La
Metallurgia Italiana
International Journal of the Italian Association for Metallurgy
n. 1 Gennaio 2020 Organo ufficiale dell’Associazione Italiana di Metallurgia. Rivista fondata nel 1909
La Metallurgia Italiana International Journal of the Italian Association for Metallurgy Organo ufficiale dell’Associazione Italiana di Metallurgia. House organ of AIM Italian Association for Metallurgy. Rivista fondata nel 1909
Direttore responsabile/Chief editor: Mario Cusolito Direttore vicario/Deputy director: Gianangelo Camona Comitato scientifico/Editorial panel: Livio Battezzati, Christian Bernhard, Massimiliano Bestetti, Wolfgang Bleck, Franco Bonollo, Bruno Buchmayr, Enrique Mariano Castrodeza, Emanuela Cerri, Lorella Ceschini, Mario Conserva, Vladislav Deev, Augusto Di Gianfrancesco, Bernd Kleimt, Carlo Mapelli, Jean Denis Mithieux, Marco Ormellese, Massimo Pellizzari, Giorgio Poli, Pedro Dolabella Portella, Barbara Previtali, Evgeny S. Prusov, Emilio Ramous, Roberto Roberti, Dieter Senk, Du Sichen, Karl-Hermann Tacke, Stefano Trasatti Segreteria di redazione/Editorial secretary: Valeria Scarano Comitato di redazione/Editorial committee: Federica Bassani, Gianangelo Camona, Mario Cusolito, Carlo Mapelli, Federico Mazzolari, Valeria Scarano Direzione e redazione/Editorial and executive office: AIM - Via F. Turati 8 - 20121 Milano tel. 02 76 02 11 32 - fax 02 76 02 05 51 met@aimnet.it - www.aimnet.it
siderweb LA COMMUNITY DELL’ACCIAIO
Gestione editoriale e pubblicità Publisher and marketing office: siderweb spa Via Don Milani, 5 - 25020 Flero (BS) tel. 030 25 400 06 - fax 030 25 400 41 commerciale@siderweb.com - www.siderweb.com La riproduzione degli articoli e delle illustrazioni è permessa solo citando la fonte e previa autorizzazione della Direzione della rivista. Reproduction in whole or in part of articles and images is permitted only upon receipt of required permission and provided that the source is cited. Reg. Trib. Milano n. 499 del 18/9/1948. Sped. in abb. Post. - D.L.353/2003 (conv. L. 27/02/2004 n. 46) art. 1, comma 1, DCB UD siderweb spa è iscritta al Roc con il num. 26116
La
Metallurgia Italiana
International Journal of the Italian Association for Metallurgy
n. 1 Gennaio 2020 Organo ufficiale dell’Associazione Italiana
Editoriale / Editorial
di Metallurgia. Rivista fondata nel 1909
Un 2019 positivo e un nuovo anno ricco di sfide, in primis quella ambientale
Federico Mazzolari, presidente AIM ...................................................................................... pag.4
Memorie scientifiche / Scientific papers Trattamenti Termici - Heat Treatments Bainitic transformation during the two-step Q&P process in a lean medium Mn steel containing silicon
S. Kaar, R. Schneider, D. Krizan, C. Béal, C.Sommitsch ......................................................... pag.6
n. 1 Gennaio 2020
Anno 112 - ISSN 0026-0843
High pressure gas quenching: assessment of velocity experimental measurements and steps for model validation E. Vyazmina, L. Bustamante-Valencia, V. Woimbee, G. Michel, A. Belhajria, Ch. David ................ pag.18
Influence of gas nitriding conditions on layer structure formation on greyand white-solidified cast irons (part I) A. Holst, A. Buchwalder, R. Zenker ....................................................................................... pag.28
Material-dependent aspects for the use of nitriding as a single or du-plex treatment in improving the tribological load-bearing behaviour of cast irons (part II)
indice
A. Buchwalder, N. Klose, R.Zenker ......................................................................................... pag.40
Simulation of heat treatment of Jominy specimen to improve quality of automotive gear components
E. Morgano, C. Viscardi, L. Valente ........................................................................................ pag.52
Neural networks-based prediction of hardenability of high performance carburizing steels for automotive applications
V. Colla, M. Vannucci, L. Bacchi, R. Valentini .......................................................................... pag.59
Heat treatment routes utilizing intercritical annealing of automotive medium-mn steel sheets
A. Grajcar, M. Morawiec ........................................................................................................ pag.66
Attualità industriale / Industry news New environmental-friendly developments and applications of the pulsed plasma heat treatment technology in the field of engine, transmission and power train engineering A. Gebeshuber, V. Strobl, T. Müller ..................................................................................... pag.73
Liquid Nitrocarburizing treatment on valves in internal combustion engines M. Legrand, V. Monteux .................................................................................................... pag.78
Non-destructive hardness / microstructure testing of heat-treated parts by mass production, with multiple frequency magnetic induction method A. Horsch ........................................................................................................................ pag.82
Atti e notizie / AIM news Eventi AIM / AIM events ................................................................................. pag.95 Corso Le prove non distruttive ..................................................................... pag.96 Normativa / Regulations ........................................................................... pag.97 Comitati tecnici / Study groups ...................................................................... pag.104 Eventi nazionali ed internazionali / National and international events .......... pag.105
editoriale - editorial
Un 2019 positivo e un nuovo anno ricco di sfide, in primis quella ambientale. L’anno 2019 di AIM chiude con risultati più che soddisfacenti per l’attività svolta ed il numero di iscritti. L’attività è stata intensa per quanto riguarda i Corsi e le Giornate di Studio, per un totale di 28 iniziative, a cui si aggiungono due manifestazioni di portata internazionale: la Conferenza Europea
Federico Mazzolari Presidente AIM
sui Trattamenti Termici - ECHT 2019 - tenutasi a Bardolino in giugno e la prima edizione dell’evento internazionale Metals for Road Mobility, svoltasi a Kilometro Rosso a fine novembre, che ha destato un buon interesse di settore con la partecipazione quale principale ospite del Presidente e fondatore di Brembo, Alberto Bombassei. Best seller dell’anno è stato Bardolino con 200 partecipanti, seguito dalle Giornate Nazionali sulla Corrosione e Protezione e dalla Giornata di Studio sulle leghe di nickel e superleghe sempre con oltre 100 partecipanti.
“
La grande sfida alla quale siamo chiamati a dare una risposta è quella della competitività ambientale. Federico Mazzolari
Anche per il 2020 l’attività di AIM sarà orientata alla diffusione della cultura metallurgica tramite attività ed iniziative a partire da fine gennaio. ECC 2020 a Bari, a metà giugno, la Conferenza Internazionale sulla Colata Continua, e HTDC 2020 la conferenza internazionale sul DIE CASTING ad inizio luglio saranno gli eventi chiave del prossimo anno, insieme al 27° Convegno Nazionale Trattamenti Termici a Genova in maggio ed al 38° Convegno Nazionale AIM che si terrà a Napoli dal 16 al 18 settembre, e da ultimo il Primo Meeting Internazionale sulle Materie Prime ed il Recycling a Bergamo a inizio dicembre.
La Metallurgia Italiana - gennaio 2020
pagina 4
editoriale - editorial
La grande sfida alla quale siamo chiamati a dare una risposta è quella della competitività ambientale: tutto il settore dei metalli dovrà instradarsi verso un concetto di metallurgia pulita riducendo sempre più le emissioni e più in generale l’impatto ambientale. Il problema di fondo è costituito dalla grande necessità di capitali. Le nuove direttive europee comporteranno la necessità di una massa di investimenti che dovrà essere supportata da misure di sostegno a livello europeo e nazionale.
Sotto l’aspetto dei risultati economici il 2019 è stato un anno caratterizzato da una prima metà con volumi e prezzi ancora buoni sulla scorta dello slancio ereditato dal 2018, e da una seconda progressivamente in frenata. La siderurgia in particolare e l’industria dei metalli sono correlati ai piani governativi di investimento principalmente sulle infrastrutture. L’Italia, e anche l’Europa, dovrà tenerne conto poiché la competitività di un paese è molto condizionata dal sistema logistico che deve dare garanzie di efficienza con bassi costi nella totale sicurezza e col minimo impatti ambientale.
Anno nuovo e nuovo look della nostra rivista. Da questo numero, La Metallurgia Italiana si presenta con una nuova veste editoriale più consona al formato digitale, grazie alle idee che ci hanno portato Siderweb ed i nostri collaboratori di AIM. Oltre ad alcune variazioni di impatto prettamente grafico, quali il font utilizzato e l’eliminazione della asimmetria tra pagine pari e dispari - tipica della rivista stampata - avremo la possibilità di scegliere le pagine cui accedere direttamente dall’indice con un semplice click, così come potremo accedere direttamente ai siti degli eventi AIM e delle Associazioni collegate. Le rubriche saranno riorganizzate per maggiore fruibilità. Inoltre prevediamo l’introduzione di nuove regole per gli autori, che ci consentiranno progressivamente di visualizzare foto e grafici con migliore risoluzione.
Non resta che augurare a tutti un felice 2020 insieme ad AIM!
La Metallurgia Italiana - January 2020
pagina 5
Memorie scientifiche - Trattamenti Termici
Bainitic transformation during the two-step Q&P process in a lean medium Mn steel containing silicon S. Kaar, R. Schneider, D. Krizan, C. Béal, C. Sommitsch Lean medium Mn Quenching and Partitioning (Q&P) steels with Mn contents up to 5 wt-% have recently gained a lot of interest due to their promising combination of strength and ductility. This steel group is characterized by a micro-
structure consisting of a carbon-depleted tempered martensitic (α’’) matrix and a considerable amount of retained austenite (RA) stabilized by both, C and Mn, in order to ensure a sufficient strain-induced austenite to martensite trans-
formation (TRIP-effect). Furthermore, the hard α’’ matrix contributes to the excellent performance of these steels in
bending, sheet cutting and hole expansion operations. In this contribution, the impact of the Q&P process on the transsformation behavior of a 0.2C-4.5Mn-1.5Si lean medium Mn steel is presented and discussed in detail. This includes
dilatometric experiments examining the influence of the Q&P process parameters on the phase transformations, light optical and scanning electron microstructural (LOM, SEM) investigations, as well as hardness measurements according
to Vickers. Moreover, the Q&P annealing approach was compared to the TRIP-assisted bainitic ferrite (TBF) process, in order to evaluate the influence of primary martensite (α’initial) being present in the Q&P samples at the onset of isother-
mal bainitic transformation (IBT) on the transformation kinetics. With increasing quench temperature (TQ) the amount
of α’’ significantly decreased, whereas the phase fraction of bainitic ferrite (αB) and RA continuously increased, leading to the formation of fresh martensite (α’ final) and the sharp decrease of the RA fraction at TQ exceeding 250°C. Furthermore, with increasing partitioning temperature (TP) a decreasing amount of α B, associated with an increasing amount of α’final, could be examined. The comparison of the Q&P and TBF process in terms of transformation behavior manifested a pronounced influence of the presence of α’initial, since the IBT was intensively accelerated in the Q&P samples.
KEYWORDS: MEDIUM MN STEEL - Q&P PROCESS - PHASE TRANSFORMATION BAINITE - RETAINED AUSTENITE
Introduction High demands on crash safety and environmental protection are driving factors for the application of new steel grades in the automotive industry. (1) Stringent requirements in terms of reducing CO2 emissions, improving crashworthiness and enhancing the formability of complex shaped parts are impelling motives for the steel industry to the development of Advanced High Strength Steels (AHSS)
Simone Kaar, Reinhold Schneider University of Applied Sciences Upper Austria, Austria
Daniel Krizan voestalpine Stahl GmbH, Austria
Coline Béal, Christof Sommitsch Graz University of Technology, Austria
(2,3). By the application of AHSS grades, the sheet thick-
La Metallurgia Italiana - gennaio 2020
pagina 6
Scientific papers - Heat Treatments ness can be downgauged due to their increased tensile
Rm*A80 exceeding 30,000 MPa% (18).
strength, resulting in a significant weight reduction of body structures (4). Furthermore, AHSS are characterized by an
The combination of the Q&P process and the concept of
improved ductility, which has beneficial impact on the pro-
medium Mn steels resulted in the development of lean me-
duction of complex automotive parts (5). The 1st genera-
dium Mn Q&P steels. (19) In comparison to conventional
tion AHSS includes Dual Phase (DP), Transformation-Indu-
Q&P steels, this steel grade contains an increased Mn-con-
ced-Plasticity (TRIP) and Complex Phase (CP) steels, with
tent (up to 5 wt-%), leading to a larger volume fraction of
a multiphase microstructure (6,7). These steel grades are
RA, which is chemically stabilized by C partitioning from
already in industrial application, since they are characteri-
α’ to γremain during isothermal holding in the over-ageing
zed by a good combination of strength and ductility with
region. The increased amount of RA and the presence of
tensile strengths (UTS) up to 1200 MPa and total elonga-
the hard C-depleted martensitic matrix allows for superior
tions (TE) up to 40 % (8,9). The 2nd generation AHSS, which
performance of these steels in both, deep drawing applica-
has outstanding mechanical properties with UTS*TE rea-
tions and forming operations such as bending, hole expan-
ching 60,000 MPa%, is represented by Twinning Induced
sion and sheet cutting (20).
Plasticity (TWIP), Nano-TWIP, Duplex and Triplex steels (10). However, until now, they are hardly used in industrial
By the application of the constrained carbon equilibrium
applications due to their challenging processing and high
(CCE) model proposed by Speer et al. (21), the amount of
alloying costs. (11)
RA as a function of TQ can be predicted. This simplified mo-
Currently, significant effort is put into the development of
del allows the calculation of the C-content in γ and therefo-
the 3rd generation AHSS in order to fill the property gap
re the prediction of the volume fraction of RA under three
between the 1st and 2nd generation AHSS (12). Therefore,
main assumption: (1) full C partitioning from α’initial to γremain,
recent research activities focus on the development of
while the partitioning kinetics are ignored; (2) no move-
Q&P, TBF and medium Mn steels (13,14).
ment of phase boundaries during the partitioning process; and (3) no competing reactions like αB formation or carbide
Q&P steels have a microstructure consisting of a car-
precipitation take place during the Q&P heat-treatment.
bon-depleted martensitic matrix with a substantial amount
(21)
of RA (15). Therefore, first the steel is fully austenitized, followed by subsequent quenching to a specific temperature
According to (22-27), the transformation from γ to αB du-
below the MS temperature in order to adjust the optimal
ring isothermal holding at TP has been observed in several
amounts of initial martensite (α’initial) and RA. Afterwards, the
instances.
steel is reheated to the over-ageing region, which triggers
Clarke et al. (25) and Kaar et al. (26,27) investigated a de-
the α’ tempering. During this stage the carbide precipita-
clining volume fraction of RA due to the formation of car-
tion will be significantly suppressed by the addition of Si
bide-free bainite during the C-partitioning step. However,
and/or Al. This enables the carbon partitioning from α’ into
besides the impact of the bainitic transformation on the
the remaining austenite (γremain), resulting in its appropriate
amount of RA, an influence on the RA stability was also
stabilization to room temperature (RT) upon final cooling
found (22-24).
(16,17).
This is due to the fact that the formation of carbide-free bainite contributed to the chemical stabilization of RA.
Medium Mn steels, which refer to the 3rd generation AHSS, as well, typically contain 3 – 12 wt-% Mn (12). They are
Therefore, the present contribution focusses on the investi-
characterized by a microstructure consisting of an ultrafi-
gation of the transformation behavior of a 0.2C-4.5Mn-1.5Si
ne-grained ferritic matrix with typical grain sizes below 1
steel grade. The influence of TQ and TP on the transforma-
µm and approximately 30 vol.-% RA (13). Medium Mn ste-
tion kinetics and thus the final microstructure and resulting
els have excellent mechanical properties with products of
hardness was investigated in detail. Furthermore, the Q&P
La Metallurgia Italiana - January 2020
pagina 7
Memorie scientifiche - Trattamenti Termici heat-treatment was compared to the TBF process in order
ingot of approximately 80 kg. The material was hot rolled to
to evaluate the influence of the presence of α’initial formed
a thickness of 4 mm, followed by tempering in a batch-an-
during quenching to TQ on the IBT.
nealing-like furnace at 550°C for 16h in order to provide cold rollability. Subsequently, the strip was cold rolled to a
Experimental Procedure
thickness of 1 mm. In order to investigate the transforma-
In the present investigation, a steel grade containing 0.173
tion behavior depending on TQ and TP (TB), a Bähr 805 A/D
wt-% C, 4.46 wt-% Mn, 1.47 wt-% Si and 0.027 wt-% Al was
dilatometer was used. Therefore, specimens with dimen-
investigated (Table 1). The steel was laboratory fabricated
sions of 10 x 4 x 1 mm³ were produced by wire-electrical
by medium frequency induction melting and cast into an
discharge machining.
Tab.2 - Chemical composition of the investigated steel in wt-%
Fe-C-Mn-Si
C
Mn
Si
Al
0.173
4.46
1.47
0.027
Figure 1 displays the time-temperature regimes for the
50 K/s.
Q&P (left) and the TBF (right) processes. In both cases, first the samples were fully austenitized at 900°C (Tan) for
Microstructural investigations were performed by means
120 s (tan) using a heating rate HR1 of 10 K/s. Subsequently,
of LOM using LePera etching according to (28). Additio-
for the Q&P heat-treatment the samples were quenched
nally, after electrochemical polishing, SEM was applied
to various TQ in the range of 130°C - 290°C in steps of 20°C
for higher resolution examinations.
with a cooling rate CR1 of 50 K/s. After 10s isothermal hol-
Mechanical properties were investigated by hardness te-
ding the samples were reheated with a heating rate HR2
sting according to Vickers (HV1) using an Emco Test Du-
of 20 K/s to a specific TP (350, 400 and 450°C) for 600s (tp).
raScan 20 device. SMM (29) was used to determine the
On the contrary, for the TBF cycle after full austenitiza-
volume fraction of RA, which was compared to the RA
tion, the samples were directly cooled to a TB of 350, 400
content calculated according to the CCE-model propo-
or 450°C and isothermally held for 600s (tB). Finally, all
sed by Speer et al. (21).
samples were quenched to RT using a cooling rate CR2 of
a)
b)
Fig.1 - Time-temperature regimes used for a) the Q&P and b) the TBF heat-treatments (26)
La Metallurgia Italiana - gennaio 2020
pagina 8
Scientific papers - Heat Treatments By the application of the Koistinen-Marburger (KM) equa-
MS-equations are applied twice. Initially, the MS tempera-
tion (30), the volume fraction of α’ formed during quen-
ture of the γ and the volume fraction of α’ formed during
ching to TQ was determined
quenching to a specific TQ (orange line) are calculated.
fM = 1 - e^(-0.011(MS - TQ)) [1]
The remaining amount of γ after quenching is represented by the green curve in Figure 2 a).
Here, fM describes the amount of α’ formed during quen-
Since the applied CCE-model assumes full C partitioning
ching from the γ-region to TQ. MS is the martensite start
from α’ to remaining γ during isothermal holding at TP, a
temperature, which was calculated according to Mahieu et
new MS temperature is calculated according to Mahieu.
al. (31):
Given that, the volume fraction of α’ formed during final cooling to RT is calculated (blue line), using the KM equa-
MS = 539 - 423C - 30.4Mn - 7.5Si + 30Al
[2]
tion once again. The resulting RA content as a function of TQ is shown in
Figure 2 a) shows the volume fractions of α’ and γ which
dark green. It is evident that the largest fraction of γ can be
were calculated by applying the CCE-model.
retained at that TQ where no fresh α’ is formed upon final
Figure 2 b) depicts the related time-temperature regime
cooling.
for the Q&P process. In the CCE-model both, KM- and
Fig.2 - a) Calculated phase fractions of martensite (α’) and austenite (γ) as a function of TQ and b) the related Q&P heat-treatment with the schematic microstructural evolution
Results
phase fraction could be found, since with increasing TP a smaller length expansion was observed. As a result, TP in-
Transformation Behavior
directly influenced the formation of α’final during cooling to
Figure 3 a) exemplarily shows the influence of TP on the di-
RT: Only for the Q&P samples with a TP of 350°C the forma-
latometric curves for the Q&P samples quenched to a TQ
tion of α’final could be fully prevented, whereas at higher TP
of 270°C compared to the influence of TB for the TBF sam-
of 400 and 450°C the curves confirm the formation of α’final
ples (Figure 3 b)). In the case of the Q&P regime, α’initial was
at low temperatures.
formed during quenching to TQ, since the MS temperatu-
On the contrary, for the TBF samples hardly any formation
re was approximately 310°C. During isothermal holding at
of αB during isothermal holding at TB could be observed.
TP, γ partially transformed to αB, accompanied by a length
The main transformation was the martensitic one during
expansion apparent from the dilatometer curves. Although
final cooling, indicating that the holding duration of 600 s
the amount of αB was rather low, an influence of TP on its
was definitely too short for bainitizing.
La Metallurgia Italiana - January 2020
pagina 9
Memorie scientifiche - Trattamenti Termici
Fig.3 - Dilatometric curves at different TP (TB) of a) Q&P heat-treatment (TQ =270°C) and b) TBF heat-treatment
Figure 4 gives a detail view of the influence of TP and TB on
400°C the formation of αB was not completed during the
the αB formation as a function of isothermal holding time
investigated period of time due to the slower transforma-
for the Q&P (TQ = 270°C) and TBF heat-treatment.
tion kinetics at lower temperatures.
In both cases, with increasing TP and TB the amount of αB
When comparing the two processes, it is evident that in
formed during isothermal holding decreased.
general for the Q&P samples the larger volume fraction of
However, especially for a TP of 350°C and TB of 350 and
αB was formed.
Fig.4 - Dilatation due to αB formation as a function of isothermal holding time at different TP (TB) for a) Q&P heat-treatment (TQ = 270°C) and b) TBF heat-treatment
In addition, for the Q&P steels the transformation beha-
isothermal holding increased with increasing TQ, which is
vior depending on TQ was investigated (Figure 5).
confirmed by the larger dilatation in Figure 5 b).
With increasing TQ, lower amounts of α’initial were formed
Furthermore, as already displayed in Figure 3 a) at very
due to the lower driving force for the martensite transfor-
high TQ of 270°C, the formation of α’final during final cooling
mation.
to RT was observed.
On the contrary, the volume fraction of αB formed during
La Metallurgia Italiana - gennaio 2020
pagina 10
Scientific papers - Heat Treatments
Fig.5 - a) Dilatometric curves at different TQ (TP = 400°C) and b) dilatation due to αB formation as a function of isothermal holding time at different TQ (TP = 400°C)
Microstructure
mixture of α’’ and small amounts of αB, appearing as bluish and brownish areas in the micrographs. Furthermore, RA
As already confirmed by means of dilatometry, TQ signi-
and/or α’final , represented by the white and brownish are-
ficantly influenced the phase transformation behavior,
as, could be observed in rising amounts finely distributed
and thus, the presence and the amount of the final micro-
in the matrix.
structural constituents of the investigated steel.
As confirmed by dilatometry, at a TQ of 190 and 230°C,
Figure 6 a) - c) depicts the LOM images of the Q&P sam-
the secondary phase solely consisted of RA, whereas
ples quenched to a TQ of 190, 230 and 270°C, respectively.
for the samples quenched to higher TQ the formation of
In comparison, the microstructure of the TBF sample is
α’final could not be prevented. On the contrary, the micro-
shown in Figure 6 d). All micrographs relate to the the mi-
structure of the TBF samples primarily consisted of α’final
crostructure obtained at a TP and TB of 400°C. The LOM
with small amounts of finely distributed RA islands and
images for the Q&P steels show a matrix consisting of a
only small fractions of αB.
Fig.6 - LOM images of the a) - c) Q&P samples (TQ = 190, 230 and 270°C) and d) TBF sample at a TP (TB) of 400°C (original magnification 1000x)
La Metallurgia Italiana - January 2020
pagina 11
Memorie scientifiche - Trattamenti Termici The SEM images of the Q&P samples quenched to a TQ of
independent from TQ. However, at a TQ of 270°C a sub-
190, 230 and 270°C are displayed in Figure 7 a) – c), where-
stantial amount of α’final was found in the microstructure.
as Figure 7 d) depicts the microstructure for the TBF sam-
In the present micrographs, RA appears as cavities, since
ple. By means of SEM, small amounts of cementite preci-
it was dissolved during electrochemical polishing due to
pitates with an obvious triaxial alignment, confirming the
its high C and Mn content.
presence of α’’, could be observed for the Q&P samples.
In comparison, the TBF treated sample largely consisted
Concerning the presence of αB, no significant influence of
of fine lath like martensite (α’final) with low fractions of RA
TQ was found, since the volume fraction was rather low,
and αB.
Fig.7 - SEM images of the a) - c) Q&P samples (TQ = 190, 230 and 270°C) and d) TBF sample at a TP (TB) of 400°C (original magnification 5000x)
In Figure 8, the phase fractions in the final microstructure
oling, in turn resulting in comparably low amounts of RA.
of the Q&P samples are summarized as a function of TQ for different TP. In addition, the bar charts, situated close to
In Figure 9, the amount of RA measured by means of SMM
the right rim of each diagram, depict the microstructural
is plotted in detail as a function of TQ at a TP of 350, 400 and
constitution for the steel after TBF treatment. Indepen-
450°C. Furthermore, the volume fraction of RA for the TBF
dent from T , for the Q&P samples with increasing TQ a
treated samples is shown in the form of dotted lines. In
considerable decrease of α’’ was observed, whereas the
addition, the diagram contains the calculated RA fraction
volume fraction of αB increased. However, particularly at
according to the CCE-model. This model proposed a RA
higher T , a vivid influence of TP is obvious: the increase
maximum (RAmax) of 19.8 vol.-% at a TQ of 173°C. In accor-
of TP led to a sharply decreasing fraction of αB, resulting in
dance with the model calculations, the experiments also
the formation of α’final upon cooling to RT. Therefore, the
confirmed the fundamental shape of the RA evolution as a
RA content steadily rose with increasing TQ until the on-
function of TQ. At very low TQ large amounts of α’initial were
set of the formation of α’final occured. For the TBF samples,
formed, leading to low RA fractions. With increasing TQ
it is obvious that only small amounts of αB were formed
the amount of α’initial decreased, and therefore the volu-
at all investigated temperatures as already shown by di-
me fraction of RA increased, until RAmax was achieved, fol-
latometry (Figure 4 b). Due to these low αB fractions, the
lowed by a decline in RA content. This decreasing amount
largest amount of γ transformed into α’final during final co-
of RA at higher TQ was due to the lower chemical stabiliza-
P
Q
La Metallurgia Italiana - gennaio 2020
pagina 12
Scientific papers - Heat Treatments tion of austenite and the resulting formation of α’final upon
achieved compared to the model predictions. Moreover,
cooling to RT. The experiments further show an influence
RAmax occurred for all TP, at temperatures 60 to 70°C hi-
of TP on the volume fraction of RA: by increasing TP from
gher than predicted by the CCE-model.
350 to 400°C, RAmax rose from 19.7 to 22.4 vol.-%.
It is evident that the RA contents determined for the TBF
In this context, at the TP of 350°C, the amount of RA cor-
samples were much lower compared to the Q&P samples,
related well with the model calculations. On the contrary,
ranging from 4.6 vol.-% RA at a TB of 450°C to 7.1 vol.-%
at the TP of 400°C and 450°C, a higher amount of RA was
RA at TB = 350°C.
Fig.8 - Phase fraction as a function of TQ for a TP (TB) of a) 350°C, b) 400°C and c) 450°C
Fig.9 - RA content as a function of TQ for a TP (TB) of 350, 400 and 450°C
Fig.10 - Vickers hardness as a function of TQ for a TP of 350, 400 and 450°C
Hardness
ness could be observed, especially at a TP of 450°C. For this
Figure 10 depicts the hardness according to Vickers as a
reason, in the case of the Q&P samples, the hardness was
function of TQ. In the TQ range of 130 to 230°C, both increa-
between approximately 400 and 500 HV1. In contrast, for
sing TQ and TP led to a slight decrease in hardness. However,
the TBF samples hardness was quite constant with 510 HV1
when further increasing TQ, a considerable increase in hard-
at a TB of 400°C, and 511 HV1 at 350 and 450°C, respectively.
La Metallurgia Italiana - January 2020
pagina 13
Memorie scientifiche - Trattamenti Termici Discussion
Al leads to a larger amount of αB formed during isother-
In Figure 11, the IBT for the Q&P samples (exemplarily at
mal holding at TP. According to (33) this can be explained
a TQ of 270°C) and the TBF samples is compared. The TTT
by the fact that Al shifts the T0-line to higher C-contents,
diagrams show an evident acceleration of the αB formation
enabling the formation of a larger amount of αB.
in case of the Q&P process. This is in agreement with Kaar
Furthermore, in the present case, a general impact of the
et al. (26) and Wang et al. (32), since the presence of α’initial
TP (TB) on the transformation kinetics was found: espe-
pronouncedly accelerated the IBT due to the presence of
cially at a TP (TB) of 400°C, the transformation kinetics was
geometrically necessary dislocations, acting as nucleation
evidently accelerated due to the faster C diffusion into the
sites for the αB formation. When comparing the IBT of the
remaining γ, while at 450°C a reduced driving force for the
present steel grade with that of a Fe-C-Mn-Al steel inve-
bainitic transformation seems to limit the transformation
stigated in (26), it is evident that the substitution of Si by
kinetics.
Fig.11 - TTT diagrams for a) Q&P steels (TQ = 270°C) and b) TBF steels
Furthermore, in terms of RA investigations, a considerable
αB formed during isothermal holding. This indicated that in
difference between the calculated CCE-model and expe-
the present case, RA was mainly stabilized via C-partitioning
rimental data was found. This was particularly apparent for
during the partitioning step, whereas for the Al-alloyed ste-
the amount of RA and optimal TQ, which is in agreement
el investigated in (26, 27) the RA stabilization was achieved
with results reported in literature (21,26,27,34). One reason
by both, formation of αB and C-partitioning.
for this discrepancy is the fact that the application of the
In general, in terms of RA stability, both, chemical and me-
CCE-model assumes full C-partitioning from α’initial into the
chanical stabilization have to be considered (35). However,
remaining γ. However, in the present case, the presence of
the CCE-model only takes into account the chemical stabi-
cementite was observed in the microstructure (Figure 7),
lization, which explains the significant shift of the RAmax to
which contradicts this assumption. This led to both, a lower
higher TQ than predicted. Additionally, two empirical for-
amount and chemical stability of γremain, resulting in the for-
mulas, describing the α’ kinetics and the MS temperature,
mation of α’final during cooling to RT and thus lower RA con-
are included in the CCE-model. Therefore, slight differen-
tents. Furthermore, the CCE-model does not consider the
ces in the chemical composition due to segregations could
formation of αB, associated with a decline in RA content. In
lead to deviations from the model, as well (36). Neverthe-
comparison to the Fe-C-Mn-Al steel grade investigated in
less, the CCE-model proposed by Speer et al. (21) is a vital
(26,27), the RA content was higher for the Si-alloyed steel
tool for the first estimation of applicable annealing parame-
examined in this contribution due to the lower amount of
ters in case of Q&P steels.
La Metallurgia Italiana - gennaio 2020
pagina 14
Scientific papers - Heat Treatments Conclusions
• The increase in T Q resulted in a lower volume fraction of α’ initial and slightly larger amounts of αB. Therefore, the vo-
The results of the present investigation indicate that the
lume fraction of RA increased, until that TQ where α’final
Q&P process has significant influence on the transforma-
was formed and thus RA content declined sharply. In case
tion behavior of lean medium Mn steels.
of the Q&P samples this led to a final microstructure consisting of α’’, RA and at higher TQ partially of αB and α’final.
It is of vital importance to carefully set both, T Q and T P, since they have significant impact on the microstructural
• Especially for the Q&P samples heat-treated at a low TP,
constituents and thus, certainly mechanical properties.
the comparison of the CCE-model and the experimental data showed a significant divergence in terms of RA con-
The main findings of the present contribution can be
tent. Additionally, a substantial shift of the RA-maximum
summarized as follows:
to higher TQ was observed compared to the CCE-model predictions.
• In case of the Q&P process, the formation of α’initial during quenching to TQ significantly accelerated the IBT,
• In the present case, for the Si-alloyed steel grade RA
acting in the form of vital nucleation sites.
was mainly stabilized by C-partitioning during isothermal holding at TP, whereas on the contrary for the Al-alloyed
• For both, Q&P and TBF processes, with increasing TP
steel investigated in (26, 27), the RA stabilization was per-
(TB) the IBT kinetics was accelerated, but the amount of αB
formed by both, formation of αB and simultaneous C-par-
formed during isothermal holding was decreased.
titioning.
Acknowledgements The authors sincerely acknowledge the support of the Austrian Research Promotion Agency (FFG) related to the frontrunner project No. 860188 “Upscaling of medium Mn-TRIP steels”.
La Metallurgia Italiana - January 2020
pagina 15
Memorie scientifiche - Trattamenti Termici References [1]
Kwon O, Lee K, Kim G, Chin K. New trends in advanced high strength steel - developments for automotive application. Mater. Sci. Forum. 2010:638-642:136-141.
[2]
Spenger F, Hebesberger T, Pichler A, Krempaszky C, Werner E, Doppler C. AHSS steel grades: strain hardening and damage as material design criteria. Proc. Int. Conf. on New Developments in AHSS. 2008 Jun. 15-18; Orlando, USA. p. 39-49.
[3]
Jacques P, Petein A, Harlet P. Improvement of mechanical properties through concurrent deformation and transformation: new steels for the 21st century. Proc. Int. Conf. on TRIP-aided high strength ferrous alloys. 2002 Jun. 19-21; Aachen, Germany. p. 281285.
[4]
Hashimoto K, Yamasaki M, Fujimura K, Matsui T, Izumiya K. Global CO2 recycling - novel materials and prospect for prevention of global warming and abundant energy supply. Mater. Sci. Eng. A. 1999:267:2:200-206.
[5]
Zaefferer S, Ohlert J, Bleck W. A study of microstructure, transformation mechanisms and correlation between microstructure and mechanical properties of a low alloyed TRIP steel. Acta Mater. 2004:52:2765-2778.
[6]
Matlock D, Speer J, De Moor E, Gibbs P. Recent developments in advanced high strength steels for automotive applications: an overview. JESTECH. 2012:15:1-12.
[7]
De Cooman BC. Structure-properties relationship in TRIP steels containing carbide-free bainite. Solid State Mater. Sci. 2004:8:285-303.
[8]
Hairer F, Krempaszky C, Tsipouridis P, Werner E, Satzinger K, Hebesberger T, Pichler A. Effects of heat treatment on microstructure and mechanical properties of bainitic single- and complex-phase steel. Proc. of MS&T. 2009 Oct. 25-29; Pittsburgh, USA. p. 1391-1401.
[9]
Samek L, Krizan D. Steel - Material of choice for automotive lightweight applications. Proc. Int. Conf. Metal 2012. 2012 May 23-25; Brno, Czech. p. 6-12.
[10]
De Cooman BC, Chin K, Kim J. High Mn TWIP steels for automotive applications. Chapter 6 in: New trends and developments in automotive system engineering. M. Chiaberge (ed.), DOI: 10.5772/14086.
[11]
Bracke L, Verbeken K, Kestens L, Penning J. Microstructure and texture evolution during cold rolling and annealing of a high Mn TWIP steel. Acta Mater. 2009:57:1512-1524.
[12]
Steineder K, Schneider R, Krizan D, Béal C, Sommitsch C. Microstructural evolution of two low-carbon steels with a medium manganese content. Proc. of 2. HMnS Conf. 2014 Aug. 31 – Sept. 4; Aachen, Germany. p. 351-354.
[13]
Steineder K, Schneider R, Krizan D, Béal C, Sommitsch C. On the microstructural characteristics influencing the yielding behavior of ultra-fine grained medium-Mn steels. Acta Mater. 2017:139:39-50.
[14]
De Cooman BC, Speer J. Quench and partitioning steel: A new AHSS concept for automotive anti-intrusion applications. Steel Res. Int. 2006:77:634-640.
[15]
Speer J, Matlock D, De Cooman BC, Schroth J. Carbon partitioning into austenite after martensite transformation. Acta Mater. 2003:51:2611-2622.
[16]
Speer J, Aussunção F, Matlock D, Edmonds D. The quenching and partitioning process: Background and recent progress. Mat. Res. 2005:8:4:417-423.
[17]
Edmonds D, He K, Rizzo F, De Cooman BC, Matlock D, Speer J. Quenching and partitioning martensite - A novel steel heat treatment. Mater. Sci. Eng. A. 2006:438-440:25-34.
[18]
Arlazarov A, Gouné M, Bouaziz O, Hazotte A, Petitgand G, Barges P. Evolution of microstructure and mechanical properties of medium Mn steels during double annealing. Mater. Sci. Eng. A. 2012:542:31-39.
La Metallurgia Italiana - gennaio 2020
pagina 16
Scientific papers - Heat Treatments [19]
De Moor E, Lacroix S, Clarke A, Penning J, Speer J. Effect of retained austenite stabilized via quench and partitioning on the strain hardening of martensitic steels. Metall. Mater. Trans. A. 2008:39A:2586-2595.
[20]
Huyghe P, Dépinoy S, Caruso M, Mercier D, Georges C, Malet L, Godet S. On the effect of Q&P processing on the stretch-flange-formability of 0.2C Ultra-high Strength Steel. ISIJ Int. 2018:58:1341-1350
[21]
Speer J, Streicher A, Matlock D, Rizzo F. Quenching and partitioning: a fundamentally new process to create high strength TRIP sheet microstructures. Austenite formation and decomposition. 2003:505-522.
[22]
Van Bohemen S, Santofimia M, Sietsma J. Experimental evidence for bainite formation below Ms in Fe-0.66C. Scr. Mater. 2009:58:488-491.
[23]
Somani M, Porter D, Karjalainen L, Misra R. On various aspects of decomposition of austenite in a high-silicon steel during quenching and partitioning. Metall. Mater. Trans. A. 2013:45A:1247-1257.
[24]
Kim D, Speer J, De Cooman BC. Isothermal transformation of a CMnSi steel below the Ms temperature. Metall. Mater. Trans. A. 2011:42a:1575-1585.
[25]
Clarke A, Speer J, Miller M, Hackenberg R, Edmonds D, Matlock D, Rizzo F, Clarke K, De Moor E. Carbon partitioning to austenite from martensite or bainite during the quench and partition (Q&P) process: A critical assessment. Acta Mater. 2008:56:16-22.
[26]
Kaar S, Schneider R, Krizan D, Béal C, Sommitsch C. Influence of the Quenching and Partitioning Process on the Transformation Kinetics and Hardness in a Lean Medium Manganese TRIP Steel. Metals. 2019:9:353:1-13
[27]
Kaar S, Schneider R, Krizan D, Béal C, Sommitsch C. Influence of the phase transformation behaviour on the microstructure and mechanical properties of a 4.5 wt-% Mn Q&P steel.” HTM J. Heat Treatm. Mat. 2019:74:2:70-84.
[28]
Zakerinia H, Kermanpur A, Najafizadeh A. Color metallography: a suitable method for characterization of martensite and bainite in multiphase steels. Int. J. ISSI. 2009:6:1:14-18.
[29]
Technical Report: Vatron GmbH, Restaustenit Magnetjoch – Neue Erkenntnisse und Möglichkeiten. 2009
[30]
Koistinen D, Marburger R. A general equation prescribing the extent of the austenite-martensite transformation in pure iron-carbon alloys and plain carbon steels. Acta Metall. 1959:7:59-60.
[31]
Mahieu J, Maki J, De Cooman BC, Claessens S. Phase transformation and mechanical properties of Si-free CMnAl transformation-induced plasticity-aided steel. Metall. Mater. Trans. A. 2002:33:8:2573-2580.
[32]
Wang G, Chen S, Liu C, Wang C, Zhao X, Xu W. Correlation of isothermal bainite transformation and austenite stability in quenching and partitioning steels. J. Iron Steel Res. Int. 2017:24:1095-1103.
[33]
Bhadeshia HKDH, Honeycombe R. Steels - Microstructure and properties, 3rd edition. 2006:ISBN-13: 978-0-750-68084-4.
[34]
Steineder K, Schneider R, Krizan D, Béal C, Sommitsch C. Investigation on the microstructural evolution in a medium-Mn steel (X10Mn5) after intercritical annealing. HTM J. Heat Tr. Mat. 2015:70:19-25.
[35]
Jimenez-Melero E, van Dijk N, Zhao L, Sietsma J, Offerman S, Wright J, van der Zwaag S. Martensitic transformation of individual grains in low-alloyed TRIP steels. Scr. Mater. 2007:56:421-424.
[36]
Kim S, Lee J, Barlat F, Lee MG. Transformation kinetics and density models of quenching and partitioning (Q&P) steels. Acta Mater. 2016:109:394-404.
La Metallurgia Italiana - January 2020
pagina 17
Memorie scientifiche - Trattamenti Termici
High pressure gas quenching: assessment of velocity experimental measurements and steps for model validation E. Vyazmina, L. Bustamante-Valencia, V. Woimbee, G. Michel, A. Belhajria, Ch. David
Gas quenching is a key step in the process of heat treatment of metals. To reach optimizations of the process time and quality, it is essential to correctly distribute the load inside the quenching cell. Numerical simulations using CFD (Computational Fluid Dynamics) could help us to understand the phenomena occurring into the quenching cell and identifying criteria for improving the process. However, numerical simulation must be validated versus experimental data. This paper presents the experimental and simulation works conducted in order to validate the commercial software Qobeo for the modeling of gas quenching. The computational domain has been refined, the mesh corresponds precisely to the geometries of the quenching cell and the rack. The gas flow for quenching is simulated in the atmospheric conditions for pressure and temperature. The numerical results are compared with experimental measurements of flow speed. The results obtained are very encouraging, coherent behaviours are observed in both cases. However, further improvements should be conducted for better approaching the actual quenching conditions. For modelling aspects, high pressure (up to 20 bar), high temperature (up to 900 °C) and various gases (N2, CO2, He) should be included. For experimental aspects, the uncertainty of gas flow measurement technique should be determined as well as improving the assessment of flow orientation.
KEYWORDS: GAS QUENCHING - CFD - SPEED MEASUREMENTS GAS FLOW - HEAT TREATMENT - HPGQ Context French Institute of Technology for Material, Metallurgy and Process (IRT M2P) In France, 8 French Institutes of Technology (IRT) have been created in the 2010 decade. The common target of such Institutes is allowing transfer of knowledge and development between academics and industrials in specific fields. The IRT’s sites are smartly distributed in French regions, close to the industrial activity sources as shown in Fig 1. The works presented here have been conducted by the French Institute of Technology for Material, Metallurgy and Process (IRT M2P). The IRT M2P is a partnership built up in 2013 in order to accelerate innovation, allow integration of
La Metallurgia Italiana - gennaio 2020
E. Vyazmina Air Liquide R&D, France
elena.vyazmina@airliquide.com
L. Bustamante-Valencia Air Liquide, AL France Industrie, France
lucas.bustamante-valencia@airliquide.com corresponding author
V. Woimbee Faurecia, France
valerie.woimbee@faurecia.com
G. Michel IRT M2P, France
gregory.michel@irt-m2p.fr
A. Belhajria, Ch. David Sciences Computers Consultants, France, abelhajria@scconsultants.com
pagina 18
Scientific papers - Heat Treatments new technologies and to share a technologic platform of in-
M2P with shared resources between industrials. It gathers
dustrial scale for tests.
contributions of: Air Liquide (gas industry), Faurecia (au-
The Institute is located in the east of France, close to an im-
tomotive industry), IRT M2P (supplier of the technical pla-
portant network of heat-treatment and materials industries.
tform), SCC consultants (innovative company offering its
Also, universities and technical centres of excellence are in
knowledge in fluid simulation and computer systems), ECM
the surroundings.
furnace producer.
This paper illustrates the mutual works conducted by IRT
Fig.1 - Location of French Institute of Technology. The IRT M2P is located in the east of France
Introduction
not need any post washing, the homogeneity of heat tre-
Potentialities offered by gas quenching are huge, if the
atment of the load is enhanced and the intensity of quen-
process is correctly mastered. The present work is focu-
ching can be controlled precisely, but also, the distortion
sed on optimization of gas quenching, first experimentally
of parts is dramatically reduced compared to traditional
and secondly by developing sustainable models able to
technology. Nevertheless, the parts are concentrated in a
forecast cooling rate.
small volume into the quenching cell and the technology is not as mature as oil quenching. Large potentialities of
Gas quenching and oil quenching
development exists for gas quenching, this work aims to
A key project has been launched in IRT M2P to improve
provide knowledge in this field (2).
our knowledge in advanced heat-treatments, especially quenching (1). In industries, oil quenching is largely spre-
The gas quenching technology
ad but exhibits several disadvantages: absolute necessity
During gas quenching, the parts are cooled from the treat-
to wash parts after treatment, recyclability of oil after use,
ment temperature, often approximately 900 °C, until room
detergent rejection and compliancy with ecological regu-
temperature in a given time dependent on the quenching
lations more stringent. Moreover, flexibility is weak when
intensity required. In this cell, the gas is injected at high
oil containers overcome a volume of 1 m and temperature
pressure (20 bar max), then a turbine located on the top
should be regulated precisely at different set points. The
blows the gas towards the load. The gas is heated up when
key advantages of gas quenching are: the metallic parts do
enters into contact with the hot parts. The gas is cooled
3
La Metallurgia Italiana - January 2020
pagina 19
Memorie scientifiche - Trattamenti Termici down when passes throw a heat exchanger located in the
The main difficulties are, in one hand, the prediction of gas
wall of the cell, then directed toward the turbine again. The
circulation when the metallic parts are assembled in a rack,
gas flow in the cell is shown in Fig. 2. For certain confi-
and on the other hand, the calculation of the actual heat
gurations of load, in particular the massive ones, an unidi-
exchange coefficient and the change of metallographic
rectional gas flow showed to be insufficient to guarantee
properties of metal when cooling. Also, in gas quenching
the uniformity of treatment. In order to address this issue,
parts hidden by obstacles could receive less gas flow, le-
quenching cells equipped with two turbines were design,
ading to a higher dispersion level for a given cooling rate.
in general located on both lateral sides of the cell. Some
The composition of the atmosphere used for quenching
other manufacturers design alternative flow and rotating
has a critical influence in the final characteristics of the par-
deflector to orientate the flow in all the directions (3).
ts. It has been shown that an atmosphere composed of 90 %nitrogen and 10 % helium can provide a profile of tem-
The prediction of distortion of parts submitted to heat tre-
perature decrease similar to the one of oil (4). Changing
atment and quenching is tricky whatever the quenching
the gas or the proportions of a mix, can lead to improve-
media is, oil, water polymer mixture or gas. In the case of
ments of quenching performance, so conducting tests
gas, such prediction requires a correct prediction of the
with a mixing of helium and another gases could present
heat transfer by convection from the solid towards the gas.
interesting opportunities.
Fig.2 -Industrial gas quenching facility, gas flow follows the direction of blue arrows
Quenching experimental device Furnace and quenching cell In this research, experimental tests of gas quenching
The gas quenching cell is on the front, it allows also the
have been conducted in the industrial furnace of IRT M2P
entrance and exit of the load.
shown in Fig. 3.
The pressure of gas for quenching is set to 20 bar, which is
It is divided into two chambers: the first for low pressure
the maximum pressure admissible for the chamber.
carbo-nitriding and the second for gas quenching.
The higher pressure, the faster quenching; the heat
The parts for heat treatment are ranged in a rack of around
exchange capacity is optimum.
250 kg. After the high temperature stage in the cell at the
The gas is sprayed from top to the bottom of the parts rack
back of the furnace, they are transferred automatically to
(as presented in Fig. 2).
the quenching cell.
La Metallurgia Italiana - gennaio 2020
pagina 20
Scientific papers - Heat Treatments
Fig.3 - Furnace on which test have been conducted and rack of parts
Gas flow measurements
Observations have been set by a transparent Plexiglas pla-
Experiments have been launched to get references to va-
te in place of the door, allowing to track smoke circulation.
lidate the gas flow simulation. The first trials (2) have been
A high speed camera ensures the shooting of the smoke
carried out to know qualitatively the gas flow circulation
transfer. Particular pictures are extracted from the film for
and assess visually the turbulence level. The furnace para-
expertize, see Fig. 4.
meters are set at ambient pressure and temperature.
Fig.4 - Left, quenching cell and installation for high speed camera. Right, smoke flow visualization
Fig. 5 shows honeycomb grid. On the left hand side, the
have been positioned. On the right had side, the simula-
real grid indicating the position of the gas flow meters that
tion’s honeycomb grid.
Fig.5 - Left, real honeycomb grid with the position of gas flow meters. Right, simulated honeycomb
La Metallurgia Italiana - January 2020
pagina 21
Memorie scientifiche - Trattamenti Termici Gas quenching simulation
into account by the estate law. • Modelling of turbulence with k-epsilon or other models.
Quenching simulation tool qobeo®
Heat transfer:
The numerical simulation of the gas quenching has been
• Solving of heat transfer equation.
conducted with qobeo® which is a commercial software
• Modelling of radiative phenomena.
of 3D simulation for furnaces and installations for indu-
• For liquids, calculation of boiling effects.
strial quenching. This software has been developed by
Calculation of mechanics-metallographic effects:
the reseach consortium ThosT gathering heat treatment
• Possibility of coupling results from qobeo® with FOR-
industries, the Center of Forming of Materials (Cemef of
GE® NxT (or other software).
MinesParisTech, France) and the company SCC which is
• An internal calculation model of the mechanics-metal-
responsible for the integration and commercial activities
lographic effects is ongoing in the frame of the INFINITY
(5,7).
industrial chair (9).
qobeo® is a software design to be easily used by practitio-
Regarding the complexity of modelling, in this work, a sy-
ners of heat treatment. For being used, it does not require
stematic validation of the numerical results with regards to
deep knowledge in heat transfer, fluid mechanics or com-
corresponding experiments is conducted.
putational matematics. Pre-set scenarios help the user to build up his own simulation.
Simulation domain setup
The mesh is build up with tetraedric elements (8), for gas
The simulation domain and geometry used in CFD mode-
phase simulation the immersed boundary technique is
ling is presented in the Fig. 6, left. The flow is injected into
used and for solid phase calculation, the method of fini-
the quenching cell, then blown with a turbine. The asso-
te elements is used. It allows solving a single equations
ciated flow rate is 4 m3/s, as given by the furnace manu-
system with thermo physical properties adapted to the
facturer. This flow rate corresponds to a flow velocity of
sub-domain (solid or fluid). qobeo® solves equations in
18-20 m/s. Due to the geometrical shape of the turbine the
transitory state in three dimensions for Newtonian fluids,
injected flow has an axial and azimuthal components.
as described here after. In the current simulations, a full 3D approach was chosen. Fluid mechanics:
A tetrahedral mesh of 12 million cells was chosen for the
• Resolution of the coupling speed-pressure using the
simulations, as shown in Fig. 6, right. The mesh was refi-
non compressible Navier-Stokes equations.
ned in the zone of the turbine, the flow deflector and on
• For high pressure gases, the compressibility is taken
the honeycomb grids.
Fig.6 - Left, CFD simulation domain. Right, mesh description
La Metallurgia Italiana - gennaio 2020
pagina 22
Scientific papers - Heat Treatments In Fig 6 left, the outlet boundary conditions are shown in
CFD simulation results
blue. These are open boundaries allowing the flow to leave the computational domain, it avoids any recirculation
Fig. 7 shows the main flow trajectories. The gas injected
close to the outlet. This velocity is computed and in the
via the turbine circulates inside the cell, passes through
current simulation it is approximately 150 m/s at the outlet
the honeycomb grid and goes out of the computational
of the turbine. The extra space in the cell close to the door
domain through lateral outlets.
is also included into the simulation domain. Its presence
There are recirculation regions on the left and right of the
has a strong impact on the flow due to the symmetry brea-
honeycomb grid. In industrial situations, this region corre-
king in the Y direction (rear-front of the cell).
sponds to the position of the metallic load which is quenched.
The Fig. 5 shows the grid in which the metallic parts are
Here, the recirculation points creates enhanced turbulen-
located. The load submitted to quenching consists of 4
ce zone that could help to cool down even faster the sur-
plateaus of honeycomb shape. There is a small difference
rounding pieces.
between the real geometry (on the left) and the model-
The presence of this recirculation is also found experi-
led one (on the right): there are 4 holes for real geometry,
mentally, see Fig. 4 (right frame). In the experiments, the
whereas in simulations 5 holes were imposed. We assume
smoke accumulates in the recirculation region predicted
that this difference does not induce critical error in the si-
by simulations.
mulation results.
This qualitative comparison demonstrates that qobeoÂŽ correctly reproduces the physics of the flow inside the
An unsteady approach is used. Due to very small time step
cell.
the simulations took 3 weeks using 16 processors.
Fig.7 - Flow trajectory
Fig. 8 shows the average velocity field inside the cell. The
of X axis, mainly top to down flow. The blue arrows evi-
flow passes the turbine located at the top of the cell and
dence flow in the opposite direction. Fig. 8 shows clearly
passes through the grid before being evacuated throu-
that the numerical obstacles disturbs the gas flow field, as
ghout the lateral outlets. The presence of the honeycomb
evidenced experimentally.
grid creates multiple recirculation points. The red arrows indicate the resulting flow speed in the positive direction
La Metallurgia Italiana - January 2020
pagina 23
Memorie scientifiche - Trattamenti Termici
Fig.8 - Flow field. Red arrow top towards bottom flow. Blue arrow, inverse flow
Fig. 9 presents the magnitude of the average velocity at
mogeneity between levels 4 and 1 could have a critical
the four levels of the honeycomb grid. At the top, on level
influence with regards to the quenching intensity of the
4, the velocity is higher than the ones observed in lower
metallic parts located in both grids. Assessing the effect
levels. The flow is inhomogeneous in the horizontal sli-
of such flow difference on the final mechanical properties
ce, this is due to the vorticity induced by the deflector.
of the metallic parts should be done in future works.
After passing through level 3 and 2 of the grid, the velocity distributions becomes more and more homogeneous
Analysing the global computational domain, the results
and the flow tend to be laminar. The velocity is homo-
of Fig. 9 and Fig. 8 show clearly that the flow is not sym-
geneously distributed at level 1, but the speed has been
metric in the volume of the quenching cell. Also, the ve-
divided by a factor of 10 with regards to the one of level 4.
locities become more homogeneous on the door side
The difference observed in terms of speed and flow ho-
and closer to the bottom of the cell (near to level 1).
Fig.9 - Magnitude of the average velocity at the four honeycomb grid levels
Experimental measurement (as presented in Fig. 5) and
1. Calculation of the average velocity fields in the time in-
numerical measurement points were located on level 4
terval from 15 s to 25 s.
with 17 measurement points and in levels 3, 2 and 1 with
2. Computation of the average speeds in a volume encom-
4 measurement points per level. The positioning of such
passing the measuring point. The volume corresponds to
detectors is detailed in Fig. 10. For the comparison of data
a cylinder of 2 cm diameter and 2 cm length (see Fig. 10,
from experiments and simulations, the following proce-
right).
dure was followed:
La Metallurgia Italiana - gennaio 2020
pagina 24
Scientific papers - Heat Treatments
Fig.10 - Left and centre, position of instrumentation for the measurements of flow speed. Right, measurement point
Fig. 11 and 12 present the comparison of simulation re-
• The gas speed is higher in the rear and the front of the
sults (continuous lines) and experimental measurements
rack in which no physical barriers are stablished. Never-
(dots) at different levels of the honeycomb grid. Simula-
theless, these volumes does not play a role in the cooling
ted velocities are of the same order of magnitude as the
of the metallic parts.
experimental measurements, which are very encouraging results. Some coherent results should be highlighted:
• The turbulence induced by high speed gas avoid creating “preferential” path in which the flow is laminar and
• In the center of the honeycomb grid the speed is lower
established in steady state. This is the cause of the lack
that in the sides. This is due to the physical obstacles pre-
of linearity increasing or decreasing with regards to the
sent in the middle of the grid as shown in Fig 5, left. But
speeds of Fig. 11.
also, due to the fact that the momentum of the gas provided by the turbine and the deflector is lower in the center. This effect is clearly visible in the top view of Fig 7: the farther away from the center, the faster flow.
Fig.11 - Comparison of experimental and simulated velocity at level 4. Dot: experiment, line: simulation
La Metallurgia Italiana - January 2020
pagina 25
Memorie scientifiche - Trattamenti Termici
Fig.12 - Comparison of experimental and simulated velocity at level 1, 2 and 3. Dot: experiment, line: simulation
As shown, very good experimental and numerical results
experimental result can be an averaged value over a pe-
have been obtained.
riod of time (mean value), an instantaneous value whate-
Nevertheless, for getting a more accurate comparison
ver the value is, or the value obtained in steady state (if it
and improving our analysis, it is necessary to have the
is reached).
following information on the experimental data:
The key question is then, what is the standard deviation
• Calibration procedure and uncertainty of the gas flow
for the averaged gas velocity?
detectors. • The error of the experiment produced by the repeatabi-
All this information is essential to find the error bar in the
lity of the experiment.
experiments. Without the error bar on the measuremen-
• The exact turbine flow rate and its variation in time.
ts, it is impossible to judge the quality of the experimental results. Hence, it is very complicated to perform an
As described, the turbulence in the chamber can indu-
accurate comparison with simulations.
ce a vibration in the flow speed measurement. Then, the
Conclusion and future works
experimental ones for speed distributions at the 4 levels of the honeycomb grids of the cell. Nevertheless, the
The gas quenching technology offers interesting oppor-
accuracy of experimental data used for the comparison
tunities for massive heat treatment of mechanical parts
should be improved, the uncertainty of the speed resul-
for automotive industry. This technique presents many
ts should be taken into account. Hence, improvement of
advantages such as reduction the distortion of the load
the experimental campaign is needed for more reliable
and close controlling of quenching intensity.
comparison with the simulations. It would be very beneficial to perform PIV measurements.
In order to optimise the gas quenching process, a comparison of experimental and simulation results has been
The numerical results showed important differences in
conducted. Trials have been carried out in the industrial
terms of speed and flow homogeneity between levels 4
plateform of IRT M2P. Modeling of a quenching cell have
and 1 of the rack. This could have a critical influence with
been performed using the 3D CFD software qobeo®. The
regards to the quenching intensity of the metallic parts
simulations evidenced a good agreement with measu-
located in both grids, the final properties of the parts tre-
rements qualitatively (the same flow behavior inside the
ated could present discrepancies. Assessing the effect of
cell). Encouraging results have been obtained, the nu-
such flow difference on the mechanical properties of the
merical results are of the same order of magnitude as the
metallic parts should be done in future works.
La Metallurgia Italiana - gennaio 2020
pagina 26
Scientific papers - Heat Treatments References [1]
Woimbee V. Gas quenching. Internal IRT M2P, 2017.
[2]
Woimbee V, Belhajria A. Gas quenching process: simulation can provide better understanding. IRT-M2P, in A3TS conference Bordeaux, 2018.
[3]
https://www.bmi-fours.com/products/gas-cooling-furnaces/b8-t-high-temperature.html
[4]
Bustamante Valencia L. Performance du refroidissement sous flux gazeux : choisir son gaz et les paramètres de mise en œuvre. Séminaire Bodycote - Air Liquide, France, October 2016.
[5]
E. Hachem, E. Massoni, T. Coupez, H. Digonnet. 8th. Heat Transfer Modeling Inside Industrial Furnaces. World Congress on Computational Mechanics (WCCM8) 5th European Congress on Computational Methods en Applied Sciences ans Engineerint (ECCOMAS 2008) - Venice, Italy, June 30-July 5, 2008.
[6]
E. Hachem, T. Kloczko, H. Digonnet and T. Coupez. Stabilized finite element solution to handle complex heat and fluid flows in industrial furnace using the immersed volume method. International Journal for Numerical Methods in Fluids, 2010.
[7]
https://www.scconsultants.com
[8]
G. Jannoun, J. Veysset, T. Coupez, E. Hachem. Edge-based anisotropic mesh adaptation for CFD applications. Proceedings of the 21st International Meshing Roundtable, pages 567–583, 2013.
[9]
Industrial chair industrielle INFINITY (on going) - http://chaireinfinity.fr
La Metallurgia Italiana - January 2020
pagina 27
Memorie scientifiche - Trattamenti Termici
Influence of gas nitriding conditions on layer structure formation on grey- and white-solidified cast irons (part I) A. Holst, A. Buchwalder, R. Zenker
The wear and corrosion behaviours of nitrided cast iron depend strongly on the material-specific morphology of its graphite. Furthermore, and depending on its specific form, this morphology can have either a positive or a negative effect on the potential for improvement of such behaviour. For this reason, comparative investigations were carried out to determine the influence of gas nitriding (GN) conditions on the layer structure and the properties formed for both grey-solidified (with graphite) and white-solidified (without graphite, but with ledeburitic carbides) cast iron surface layers generated by electron beam remelting (EBR). By varying the nitriding potential within the limits KN = 0.1 (no compound layers), KN = 1 and KN = 4 (with compound layers), as determined on the basis of the Lehrer diagram, the influence of different Si contents in the cast irons on the layer formation and phase composition were investigated. In addition, the influence of the morphology of the two different C-containing modifications - the stable graphite (lamellar, spheroidal) and the metastable ledeburitic carbides (secondary dendrite arm spacing: 3 µm - 7 µm) – on the formation of both the compound and diffusion layers was analysed by means of SEM investigations, hardness measurements and EBSD analysis. This allowed conclusions to be drawn about the microstructurally dependent nitriding mechanism during gas nitriding. For the metastable carbide layers, the subsequent thermal regime of the nitriding process applied (EBR+GN) may have influenced the initial microstructure. For this reason, the tempering behaviour was investigated in separate annealing tests under temperature-time regimes relevant for nitriding (540 °C to 670 °C / 4 h to 16 h). The hardness decrease induced by the coarsening of pearlitic cementite in the ledeburite was strongly dependent on the chemical composition of the cast iron used. At temperatures above 630 °C, the hardness dropped remarkably due to graphitizing. In light of these results, the nitriding parameters could be optimized.
KEYWORDS: CAST IRON - ELECTRON BEAM REMELTING - GAS NITRIDING DUPLEX SURFACE TREATMENT
Introduction Grey-solidified cast iron is interesting for numerous applications due to some of its special properties, such as good castability and associated near-net-shape com-
Anja Holst, Anja Buchwalder, Rolf Zenker
ponent manufacture, high damping capacity, high com-
TU Bergakademie Freiberg, Institute of Materials Engineering
pressive strength and emergency running properties
Freiberg, Germany
[1,2]. Alloying with silicon (Si) promotes solidification according to the stable Fe-C system. Therefore, the microstructure consists of coarse graphite particles em-
La Metallurgia Italiana - gennaio 2020
pagina 28
Scientific papers - Heat Treatments bedded in a ferritic or pearlitic matrix. As known from
particular, the role of silicon in nitriding, hypoeutectic
steel, nitriding is a common method for improving both
pearlitic and ferritic cast irons were investigated in the
the material’s wear and corrosion resistance [3]. Howe-
initial (grey-solidified) and EB remelted (white-solidi-
ver, because the graphite is not nitrided, i.e., the layer is
fied) states.
interrupted, the compound layer is inhomogeneous and
Gas nitriding was carried out with different nitriding para-
does not always fulfil its function satisfactorily.
meters (time t N, nitriding potential K N) at a constant tem-
Electron beam remelting (EBR) of cast iron surfaces car-
perature T N of 540 °C.
ried out before the nitriding treatment leads to the dissolution of near-surface graphite and, as a result of the
Experimental
process-related high cooling rate, to solidification according to the metastable Fe-Fe 3C system.
Base materials
The eutectic microstructure of white-solidified cast irons
The investigations were carried out on commercially
is referred to as ledeburite, which exhibits high hardness
available grey-solidified pearlitic (P) and ferritic (F) cast
and has been investigated extensively [4–6].
irons with lamellar (GJL), spheroidal (GJS) and vermicular
The morphology of ledeburite depends on the cooling
graphite (GJV), with the samples provided in the form of
rate. The degree of microstructural refinement can be
cast flat samples (100 × 60 × 20 mm³) by the foundry Kes-
influenced by varying the energy input during EB remel-
sler and Co. GmbH (Leipzig, Germany). Tab. 1 shows the
ting.
respective chemical composition and the carbon equi-
The higher the cooling rate, i.e., the thinner the remel-
valent (CE) calculated from it. Compared to conventio-
ted surface layers, the finer the remelted microstructure.
nal GJL-250 cast iron, small additions of chromium (Cr),
During nitriding of the ledeburitic surface layers, a clo-
copper (Cu) and titanium (Ti) served to modify the for-
sed, homogeneous compound layer is formed [7]. Such
mation of the graphite lamellae in order to improve the
combined treated surface layers (remelting + nitriding)
melting behaviour during EB remelting [10]. The nomi-
are significantly more resistant to abrasive wear and cor-
nal tensile strength was not affected. This was marked in
rosion [8,9].
the material designation by the imposed character mod
In order to understand the nitriding mechanism and, in
(GJL-250 mod).
Tab.1 - Nominal composition (Fe base) of the investigated alloys and the corresponding carbon equivalents CE (CE = wt.%C + ⅓ (wt.% Si + wt.% P))
The alloys were grouped with respect to their matrix
tic cast irons were 258 HV0.3 (P/L_2.0Si) and 288 HV0.3
microstructure (P - pearlitic, F - ferritic), graphite mor-
(P/S_2.3Si), respectively, while that of the ferritic cast
phology (L - lamellar, S - spheroidal, V - vermicular) and
iron was 167 HV0.3 (F/S_2.3Si, F/V_2.5Si).
silicon content, cf. Tab. 1. The measured mean core hardness values of the pearli-
La Metallurgia Italiana - January 2020
pagina 29
Memorie scientifiche - Trattamenti Termici Treatment parameters
were electrolytically coated with a protective Ni layer
The electron beam remelting (EBR) was carried out in
prior to embedding. To facilitate light-optical examina-
a K26-15/80 universal electron beam chamber facility
tion, the specimens were etched with 3 % Nital solution.
(pro-beam systems GmbH), which operated at a beam
The specimens investigated by scanning electron micro-
power ≤ 15 kW and an acceleration voltage ≤ 80 V. Be-
scopy were not etched, but subjected to an additional
fore remelting, the specimens were preheated in-situ
polishing step using Buehler MasterMet ™. A LEO 1530
up to a temperature of approx. 450 °C using the EB field
FE-SEM (Carl Zeiss) equipped with an HKL Channel 5
technique. The preheating procedure was necessary to
EBSD system (Oxford Instruments) was employed, along
prevent cracking of the EBR layer. The remelting beha-
with a MIRA 3-SEM (Tescan) equipped with an Apollo XP
viour of cast iron materials is mainly determined by their
detector (EDAX/Ametek). Light-optical microscopy was
graphite morphology. For this reason, the beam guidan-
used for measuring the EBR layer thickness (t EBR) as well
ce technique (multi-spot technique, meander technique,
as the compound layer thickness (CLT) after nitriding as
etc.) and/or EB parameters (acceleration voltage, focal
an average value of a minimum of 20 individual values.
position, local and global frequency) were adapted in
Secondary dendrite arm spacing (SDAS) was determined
order to produce crack- and pore-free remelted layers
by means of light-optical micrographs according to the
with surface deformation minimized as much as possi-
BDG guideline P220 (2011) [11]. This value facilitated the
ble. Once the electron beam had passed, the remelted
estimation of both the degree of microstructural refine-
surface layer solidified again almost instantaneously due
ment achieved by the solidification process and the co-
to rapid heat dissipation into the cold core material – a
oling rate. The final SDAS values mentioned were mean
phenomenon known as self-quenching.
values from 10 individual measurements. The treated
Before nitriding, the surface deformations resulting from
surfaces were characterised by Vickers hardness-depth
the EBR process were machined and ground with 1200-
profiles perpendicular to the surface using nominal loads
grit SiC paper. Gas nitriding was carried out in a chamber
of 2.94 N (HV0.3) for EBR layers and 0.49 N (HV0.05) for
furnace at a nitriding temperature of T N = 540 °C for both
nidrided layers. For EBR layers, the minimum and maxi-
different time periods (4 h, 16 h) and nitriding potentials
mum hardness values were estimated from the profiles
K N = 0.1 / 1 / 4 [atm- 1/2].
measured.
The influence of the thermal regime during nitriding on
Results and discussion
the thermal stability of the remelted microstructure was investigated through examination of the pearlitic cast
Microstructure and properties of white-solidified EB
irons. They were annealed under an argon atmosphere
remelted surface layers
at different temperatures and for different time periods,
Independently of the EBR parameters used, all cast
which are associated with both low and high thermal lo-
irons solidified in a metastable manner according to the
ads typical of conventional nitriding treatments.
Fe-Fe 3C system. While the base materials investigated
The particular annealing conditions were 540 °C/4 h,
were hypoeutectic cast alloys, the main fractions of the
540 °C/16 h, 590 °C/12 h, 630 °C/4 h and 670 °C/12 h. The pa-
so-called white-solidified microstructure were primary
rameter combination of 670 °C/12 h is unusual for conven-
austenite dendrites (which decomposed to pearlite) em-
tional nitriding, and may be considered to be an upper limit.
bedded in a eutectic matrix (Fig. 1). The eutectic matrix exhibited typical features of ledeburite, i.e., cementite
Examination methods
plates and a rod-like eutectic. While with an increasing
Cross-sectional specimens were prepared for the mi-
cooling rate (equal to SDAS↓, Fig. 1a, c), the volume
crostructural characterization. All specimens were em-
fraction of rod eutectic (austenite rods in a cementite
bedded in Struers PolyFast ®, ground and polished (final
matrix) decreased, the volume fraction of plate-shaped
step: 1 µm diamond suspension). The nitrided specimens
eutectic microstructure increased. Moreover, analysis of
La Metallurgia Italiana - gennaio 2020
pagina 30
Scientific papers - Heat Treatments the element distribution using energy dispersive X-ray
gations during solidification and the formation of another
spectroscopy (EDS) showed a strong heterogeneous
ternary Fe-C-Si phase [12,13] – which typically outlines
silicon (Si) distribution in the ledeburitic microstructu-
the edges of eutectic cementite plates – obviously led to
re. The eutectic cementite could be assumed to be Si-
the Si enrichments observed.
free. Local Si enrichments occurred along the boundary between eutectic cementite and pearlite. Si microsegre-
b) P/S_2.3Si, SDAS↑ = 5.7±0.8 µm
a) P/S_2.3Si, SDAS↓ = 3.7±0.8 µm
c) F/V_2.5Si, SDAS↓ = 3.0±0.4 µm
d) F/V_2.5Si, SDAS↑ = 5.7±0.6 µm
Fig.1 - White-solidified microstructure of remelted surface layers with small and large SDAS of pearlitic P/S_2.3Si (a+b) and ferritic F/V_2.5Si (c+d)
The cooling rate was influenced by the energy input
function of EBR layer thickness or SDAS (cf. Fig. 2a, b).
(beam current). As expected, the lower the energy input,
The hardness values measured in the remelted surface
the thinner the EBR layer thickness (t EBR) and the more
layers were in the range of 600 - 750 HV0.3. Compared to
compact the secondary dendrite arm spacing (SDAS) (Fig.
the initial hardness of the pearlitic (258 - 288 HV0.3) and
2). As a result of the preheating, the cooling rate calcula-
ferritic (167 HV0.3) base materials, the hardness values in
ted from the SDAS was ≤ 900 K/s [14]. Independent of the
the EBR layers were more than doubled (Fig. 2a) or qua-
beam guidance techniques or EBR parameters used, only
drupled (Fig. 2b) after EBR.
marginal differences in layer hardness were apparent as a a) Pearlitic cast irons
b) Ferritic cast irons
Fig.2 - Influence of the EBR layer thickness tEBR on minimum (hollow rhombus ◊) and maximum (solid rhombus ♦) layer hardness and mean secondary dendrite arm distance SDAS (solid circles ●)
During EB remelting, the spatially connected network
more widely separated manner in the microstructure –
of fine graphite lamellae of the GJL cast iron was almost
floated to the surface in the melt, and were thus not com-
completely dissolved, while the graphite morphologies
pletely melted due to their very short exposure time to
of the GJS or GJV cast irons – which were present in a
the electron beam. This graphite was removed during the
La Metallurgia Italiana - January 2020
pagina 31
Memorie scientifiche - Trattamenti Termici surface grinding that was necessary before the remelted
Due to its poor melting behaviour (pore formation, surfa-
surfaces could be nitrided. However, such a procedure
ce deformation), only EBR layers with SDAS↑ (4.9 - 6.7 µm)
can lead to a change in the local carbon content, espe-
could be provided for P/L_2.0Si.
cially in thin EBR layers. This lower carbon content led to a higher volume fraction of primary austenite (pearlite
Influence of thermal regime of nitriding on the micro-
after cooling) in the remelted microstructure and, thus,
structure of EBR layers
to somewhat lower layer hardness values (cf. Fig. 2). For further studies regarding low-temperature annealing
Starting from the initial levels of hardness after EB remel-
and nitriding, EBR layers were selected for each cast iron
ting (hollow symbols in Fig. 3), the hardness decreased
with:
continuously with increasing thermal load.
- low SDAS↓: 3.0 µm - 3.7 µm (t EBR = 0.1 mm - 0.3 mm) and - high SDAS↑: 5.4 µm - 6.7 µm (t EBR = 1.0 mm - 1.3 mm).
Fig.3 - Influence of the thermal regime during annealing on the hardness of EBR layers of pearlitic cast iron. The massive loss of hardness at 670 °C/12 h was due to graphitizing
SEM investigations of the microstructures annealed up to
3). The decrease in hardness was larger for P/L_2.0Si than
630 °C/4 h revealed that the changes in hardness (Fig. 3)
for P/S_2.3Si up to 630 °C/4 h.
were mostly related to changes in the microstructure of
The higher chromium (Cr) and manganese (Mn) contents
pearlite.
in P/L_2.0Si (cf. Tab. 1) - which are known to dissolve in
The lamellar morphology of the pearlite (Fig. 4a) was lost
cementite and, consequently, stabilize pearlitic cementi-
due to the coarsening of the pearlitic cementite (Fig. 4b).
te against coarsening during annealing [16,17] - could not
The morphology of eutectic cementite (EC) in the EBR
compensate for the lower Si content. At a thermal load of
layers was hardly affected by annealing up to a thermal
630 °C/4h, the pearlitic cementite only became coarser,
load of 630 °C/4 h.
but was largely retained (Fig. 4b).
As determined in [15], the presence of Si in the pearlitic α-iron (Fe) and along the boundary between eutectic cementite and pearlite significantly retarded the coarsening of the pearlitic cementite and stabilized its lamellar morphology. Higher silicon content hindered the coarsening of pearlitic cementite more effectively, as indicated by the different changes in hardness of P/S_2.3Si and P/L_2.0Si (Fig.
La Metallurgia Italiana - gennaio 2020
pagina 32
Scientific papers - Heat Treatments a) Annealed at 540 °C/16 h
b) Annealed at 630 °C/4 h
Fig.4 - SEM images (backscattered electron contrast) of the annealed microstructure of remelted P/L_2.0Si. The microstructural constituents, i.e., eutectic cementite (EC), pearlite (P) and ferrite (F), are labelled. Cementite appears dark in contrast, while ferrite appears bright
At the maximum thermal load of 670 °C/12 h, the hard-
Refining the microstructure (SDAS↓) by changing the
ness dropped remarkably due to graphitizing. The EBR
EBR parameters increased the amount of possible
layer of P/S_2.3Si was almost completely graphitized at
graphite nucleation sites, while the redistribution of Fe
670 °C/12 h (Fig. 5b), whereas the remelted P/L_2.0Si was
and Si was accelerated by more rapid interface diffusion.
only partially graphitized under the same conditions. The
For this reason, graphitization and ferrite grain growth
microstructure of the remelted P/L_2.0Si contained co-
were more advanced in the smaller remelted micro-
arse temper graphite as well as coarse eutectic cementi-
structures, i.e., the temper graphite particles and the fer-
te plates (Fig. 5c). In contrast to the P/S_2.3Si, the higher
rite grains were larger in the smaller remelted layers than
thermal stability of the eutectic cementite in P/L_2.0Si
in the remelted layers with larger SDAS after 670 °C/12 h
should have resulted from the lower Si content and the
(Fig. 5a,b).
higher Cr and Mn content (cf. Tab. 1).
a) P/S_2.3Si, SDAS↓ = 3.7±0.8 µm
b) P/S_2.3Si, SDAS↑ = 5.7±0.8 µm
c) P/L_2.0Si, SDAS↑ = 6.7±1.1 µm
Fig.5 - Light-optical micrographs of remelted P/S_2.3Si (a+b) and P/L_2.0Si (c) annealed at 670 °C for 12 h, where temper graphite formation occurred
In summary, the low-temperature annealing of the cast
re T N ≤ 540 °C and a nitriding time t N ≤ 16 h without chan-
irons investigated showed that the EBR layers with small
ges to their microstructures or hardness.
SDAS (≥ 3 µm) could be nitrided at a nitriding temperatu-
La Metallurgia Italiana - January 2020
pagina 33
Memorie scientifiche - Trattamenti Termici Microstructure and properties of grey- and
investigated.
white-solidified cast irons after nitriding
With the exception of P/L_2.0Si, the ledeburitic surface layers were decarburized considerably during nitriding
At the chosen nitriding temperature of 540 °C, the nitri-
without compound layer (Fig. 6a), such that after a pro-
ding potential K N was selected according to the Lehrer
cess time of 16 h, in particular, the first hardness values
diagram [18] (i.e., the phase diagram for the formation of
HV0.05 measured at a distance of 10 µm from the surface
iron nitrides, which is valid for pure iron) in such a man-
had a significantly lower value than the remaining remel-
ner that nitriding was performed in the α-Fe field (K N = 0.1,
ted and nitrided surface layer (Fig. 6c).
without compound layer), in the γ’ field (K N = 1) and in the
In the P/L_2.0Si additionally alloyed with Mn, Cr and Cu
ε field (K N = 4) for both short (t N = 4 h) and long nitriding
(Tab. 1), the eutectic cementite was not decarburized as
times (t N = 16 h).
much and no decrease in hardness was measured directly
As expected, no compound layer was formed after nitri-
below the surface (Fig. 6b,c).
ding using K N = 0.1 in the pearlitic and ferritic cast irons
a) F/V_2.5Si
b) P/L_2.0Si
c) Hardness-depth profiles, t N = 16 h
Fig.6 - SEM images (backscattered electron contrast) of remelted F/V_2.5Si (a) and P/L_2.0Si (b) with the corresponding hardness-depth profiles after gas nitriding at 540 °C/16 h, KN = 0.1.
While the near-surface pearlitic and eutectic cementite of the F/V_2.5Si was decarburized, the eutectic cementite of the P/L_2.0Si was still largely intact
The phase constituents of the compound layers of the
In simplified terms, Fe carbonitrides and Fe nitrides are
EBR layers, which were formed during nitriding with
not distinguished conceptually in the following, but are
K N = 1 and 4 [atm -1/2] for 16 h, are summarized in Tab. 2.
referred to as Fe nitrides (γ’, ε).
Tab.2 - Composition of the compound layer of the microstructural constituents (eutectic cementite; pearlite) in the remelted pearlitic and ferritic cast irons after nitriding at 540 °C for 16 h
La Metallurgia Italiana - gennaio 2020
pagina 34
Scientific papers - Heat Treatments During nitriding in the γ’ field of the Lehrer diagram
For K N = 4, only the ferritic cast iron F/V_2.5Si with the
(K N = 1), all compound layers formed on the remelted and
high Si content and the low Mn content had macrosco-
nitrided cast irons were characterized by a strong lateral
pically homogeneous ε compound layers. For the cast
inhomogeneity in their phase constitution. As shown in
irons alloyed with Mn and Cu, the nitrided pearlite was
Fig. 7, ε/ γ’ double layers were locally present in nitrided
transformed into ε and γ’ nitrides at K N = 4. Furthermore,
eutectic cementite, which was identified by EBSD phase
above the nitrided pearlite, outwardly directed Fe-nitri-
maps. The nitrided pearlitic regions of the pearlitic cast
de growth occurred at the surface (see arrows in Fig. 7).
irons alloyed with Mn and Cu were transformed to γ’. For
These surface nitrides may have been associated with the
the ferritic cast irons with low Mn contents, both γ’ nitri-
formation of pores that appeared near the surface along
des and ε nitrides were detected in the nitrided pearlitic
the eutectic cementite plates and along former α/Fe 3C (θ)
regions.
phase boundaries in pearlite.
Fig.7 - Influence of the Si content and the nitriding potential KN on the phase constitution (blue: α; yellow: θ (Fe3C);
violet: ε; green: γ’) in the compound layer of the EBR layer, measured with EBSD (black: non-indexable areas especially in nitrided pearlite due to pore formation and very fine microstructures)
a)
b)
c)
Fig.8 - SEM images of remelted F/V_2.5Si after gas nitriding at 540 °C/16 h, KN = 1 with Si nitride precipitations in the compound layer (CL) and diffusion layer (DL): (a) cross section, (b) transition CL – DL and (c) end of DL
After all nitriding treatments and independently of K N, na-
the α/Fe 3C (θ) phase boundaries and in α-Fe matrix. The
nodimensional precipitations existed in the compound
exact chemical composition of the precipitations has not
layer (CL) and the diffusion layer (DL) (Fig. 8). These pre-
yet been defined. There is a high probability that the pre-
cipitations were amorphous and only occurred where Si
cipitates were Si nitrides (Si 3N 4) as described by [19-21]
enrichments had been detected, i.e., in particular along
for binary Fe-Si alloys.
La Metallurgia Italiana - January 2020
pagina 35
Memorie scientifiche - Trattamenti Termici a) Without prior EB remelting
b) With prior EB remelting (SDAS ⁓ 6 µm)
KN = 1
KN = 4
Fig.9 - Influence of Si content, nitriding potential KN and nitriding time tN on the mean compound layer thickness (CLT) of nitrided grey-solidified (a) and white-solidified (b) cast irons
Nitriding the grey-solidified cast irons with K N ≥ 1 led to
(2.0 - 2.5% Si) of the various cast iron materials had only
the formation of a compound layer, which was inhomo-
a marginal influence on the compound layer thickness
geneous with respect to the layer thickness.
(Fig. 9b). Compared to all other cast irons, P/L-2.0Si
No compound layer was formed on the graphite due to
contained the largest amount of other nitride-forming
the absence of iron. Since the nitrogen could penetra-
alloying elements, such as Mn, Cr and Ti (cf. Tab. 1).
te particularly rapidly to greater depths along the small
This resulted in the lowest compound layer thicknes-
gap between the graphite and the matrix, the local com-
ses of all of the nitriding variants. The mean compound
pound layer thickness (CLT) varied greatly and depen-
layer thicknesses of the EBR layers for K N = 1 were gre-
ded on the size of the graphite particles located near the
ater than those of the grey-solidified surfaces, while for
surface (Fig. 9a).
K N = 4, they were lower than the mean compound layer
The absence of graphite in the white-solidified remel-
thicknesses of the grey- solidified surfaces - independent
ted surface layers resulted in dense compound layers
of the nitriding time (Fig. 9).
with a homogeneous compound layer thickness. Both the dendrite arm spacing of the EBR microstructures (SDAS↓ ~ 3.5 µm; SDAS↑ ~ 6 µm) and the Si content
a) Pearlitic cast irons
b) Ferritic cast irons
Fig.10 - Influence of Si content, nitriding potential KN and nitriding time tN on the diffusion layer thickness (DLT) of nitrided and remelted pearlitic (a) and ferritic (b) cast irons
La Metallurgia Italiana - gennaio 2020
pagina 36
Scientific papers - Heat Treatments a) P/L_2.0Si
b) P/S_2.3Si
a) F/S_2.3Si
b) F/V_2.5Si
Fig.11 -Influence of Si content, nitriding potential KN and nitriding time tN (4h: hollow symbol; 16h: solid symbol)
on the increase in hardness ΔHV0.05 (ΔHV = HVDL – HVEBU or HVBM) of nitrided cast irons in the initial (green symbols) and remelted states
The diffusion layer thicknesses (DLT) were determined
ded ferritic cast irons, the increase in nitriding potential
by means of investigating the depths at which the nano-
to K N = 4 led to a further increase in hardness (Fig. 11c,d).
dimensioned precipitations were visible under the SEM. The DLT increased with increasing nitriding time and/or
Conclusion
nitriding potential up to K N = 1 (Fig. 10). With a further
The influence of gas nitriding conditions by means of
increase in nitriding potential up to K N = 4, the diffusion
the variation of the time and nitriding potential on the
layer thickness remained almost constant (Fig. 10). Wi-
formation of nitriding layers on white-solidified (EBR)
thin the diffusion layer, the inclusion of nitrogen and the
and pearlitic and ferritic cast irons was investigated.
precipitation of special nitrides (e.g. Si, Cr, Ti nitrides
The white-solidified surface layers were generated by
etc., cf. [22]) resulted in an increase in hardness, which
electron beam remelting. By means of the targeted va-
is shown in Fig. 11 as the difference between the initial
riation of the EB process parameters, and in particular the
hardness of the EBR layer (HV EBR) or the base material
energy input (beam current), crack- and pore-free remel-
(HV BM) and the maximum hardness of the diffusion layer
ted surface layers with a layer thickness of ≤ 1.3 mm could
(HV DL). The greatest increase in hardness due to nitri-
be produced. Since the pearlitic and ferritic cast irons
ding was measured by increasing the nitriding potential
investigated were hypoeutectic alloys, the white-so-
in the range of K N = 0.1...1. Nitriding at a nitriding poten-
lidified remelted layers consisted of primary austenite
tial of K N = 4 did not lead to any significant increase in
dendrites (decomposed in pearlite) and a eutectic ma-
hardness. An analogous relationship resulted for the ni-
trix. The thinner the remelted surface layers, the smaller
trided pearlitic cast irons (Fig. 11a,b). In the case of nitri-
the secondary dendrite arm spacing (SDAS) measured.
La Metallurgia Italiana - January 2020
pagina 37
Memorie scientifiche - Trattamenti Termici Regardless of the base material of the cast iron (pearlitic
be detected by X-Ray diffraction (XRD). No Si-nitride
or ferritic), the EBR layer hardness was 600 - 750 HV0.3.
precipitations occurred in Si-free eutectic cementite.
For successful implementation of the combination tre-
In compound layer-free nitriding (K N = 0.1), the EBR
atment, the thermal stability of the EBR layer during
layer was decarburized to a considerable extent, with
subsequent nitriding was of crucial importance.
the exception of the P/L_2.0Si alloyed with Mn, Cu and
Therefore, annealing experiments were carried out under
Cr. The compound layers of the investigated cast irons
typical nitriding conditions. It was shown that Si stabili-
formed during nitriding with a nitriding potential of
zed the pearlitic and eutectic cementite against coarse-
K N = 1 exhibited a strong lateral inhomogeneity in their
ning upon low-temperature annealing. The graphitizing
phase compositions. This laterally inhomogeneous
of white-solidified surface layers at temperatures above
constitution of the compound layers was also observed
630°C, however, was promoted by increased Si content
for K N = 4, with the exception of F/V_2.5Si. With its high
and an increased degree of microstructural refinement.
Si content and low Mn content, it was only this F/V_2.5Si
The annealing tests revealed that nitriding treatment at a
ferritic cast iron that formed macroscopically homoge-
temperature of T N ≤ 540 °C for a nitriding time of tN ≤ 16 h
neous ε compound layers at K N = 4.
did not lead to any changes in either the microstructure
The microstructural refinement of the EBR layers (SDAS↓
or the hardness of the EBR layer.
~ 3.5 µm, SDAS↑ ~ 6 µm) and the Si content (2.0 - 2.5 wt.%
During nitriding, nanodimensional precipitations were
Si) of the cast irons investigated had only a marginal in-
formed in both the compound layer and the diffu-
fluence on the compound layer thickness, the diffusion
sion layer, which led to an additional increase in har-
layer thickness and the increase in hardness within the
dness. These precipitations were detected in Si-enri-
diffusion layer.
ched mi-crostructural areas, i.e., in particular along the α/Fe 3C (θ) phase boundaries and in the α-Fe matrix. From this information, it was deduced that these were silicon nitrides. Due to their amorphous nature, they could not
Acknowledgements This study was supported financially by the German Research Foundation (DFG) within the framework of the BU 2568/7-1 project. The authors wish to thank Mr. E. Siegismund (Institute of Materials Engineering, TU Bergakademie Freiberg) for conducting the nitriding experiments, Ms. G. Bittner (Institute of Materials Engineering, TU Bergakademie Freiberg) for performing hardness measurements, and M.Sc. S. Kante (Institute of Materials Science, TU Bergakademie Freiberg) for his support in conducting EBSD measurements.
La Metallurgia Italiana - gennaio 2020
pagina 38
Scientific papers - Heat Treatments References
[1]
Davis JR, editor. Cast irons. 1. print. Materials Park, OH; 1996. (ASM specialty handbook ®).
[2]
Berns H, Theisen W. Eisenwerkstoffe - Stahl und Gusseisen. 4., bearb. Aufl. Berlin: Springer; 2008.
[3]
Spies H-J. Nitrieren und Nitrocarburieren: Gefüge - Eigenschaften - kontrolliertes Nitrieren. HTM J. Heat Treatm. Mat. 2013; 68(2):86– 96.
[4]
Hillert M, Steinhäuser H. The Structure of White Cast Iron. Jernkontorets Annaler 1960; 144(7):520–59.
[5]
Rickard J, Hughes ICH. Eutectic Structure in White Cast Iron. BCIRA Journal 1961; 9:11–25.
[6]
Park JS, Verhoeven JD. Directional solidification of white cast iron. Metallurgical and Materials Transactions 1996; 27A:2328–37.
[7]
Buchwalder A, Spies H-J, Klose N, Jung A, Zenker R. Effects of Different Microstructural and Hardness Gradients Generated by Single and Combined Surface Treatments with a Nitriding Top Layer. Savannah, USA; 2016 (23rd IFHTSE Congress).
[8]
Buchwalder A, Zenker R, Fritzsch K, Rüthrich K, Nagel K. Potenziale von ausgewählten Randschichtbehandlungsverfahren sowie deren Kombinationen zur Verbesserung des Verschleiß- und Korrosionsverhaltens von Gusseisenwerkstoffen. Mat.-wiss. u. Werkstofftech. 2015; 46(6):550-62.
[9]
Buchwalder A, Spies H-J, Klose N, Hegelmann E, Zenker R. Contribution to the microstructural constitutions and properties of nitrided as-received and remelted cast irons. Nice; 2017 (24th IFHTSE Congress).
[10]
Buchwalder A, Zenker R, Rüthrich K, Nagel K, Griesbach W, Hartwig S et al. Eine neue kombinierte Randschichttechnologie für hochbeanspruchte Gusseisenwerkstoffe. HTM J. Heat Treatm. Mat. 2014; 69(3).
[11]
BDG-Informationszentrum Giesserei. Bestimmung des Dendritenarmabstandes für Gussstücke aus Aluminium-Gusslegierungen. Juli 2011. Düsseldorf; 2011 Juli 2011. Available from: URL: https://www.bdguss.de/fileadmin/content_bdguss/Der_BDG/Richtlinien/P_220.pdf.
[12]
Spinat P, Brouty C, Whuler A, Herpin P. Etude Structurale de la Phase ‘MnsSi2C’. Acta Crystallogr. 1975; (B31):541–7.
[13]
Malinochka YN, Osada NG, KovaI’chuk GZ. A new structural constituent in Fe-C-Si alloys 1961:299–305.
[14]
Holst A, Buchwalder A, Hollmann P, Zenker R. Influence of Cooling Rate on the Microstructural Features of a Remelted White-Solidified Cast Iron Surface and its Effects on Nitriding Behaviour. Journal of Materials Processing Technology 2019; 271:377–83.
[15]
Kante S, Holst A, Buchwalder A, Leineweber A. Low-temperature Annealing and Graphitizing of Low-alloy White Cast Irons Produced by Rapid Solidification Technique. Materialwiss. Werkstofftech. 2019; (50):682-95.
[16]
Zhang G-H, Chae J-Y, Kim K-H, Suh DW. Effects of Mn, Si and Cr addition on the dissolution and coarsening of pearlitic cementite during intercritical austenitization in Fe-1mass%C alloy. Materials Characterization 2013; 81:56–67.
[17]
Lv ZQ, Fu WT, Sun SH, Bai XH, Gao Y, Wang ZH et al. First-principles study on the electronic structure, magnetic properties and phase stability of alloyed cementite with Cr or Mn. Journal of Magnetism and Magnetic Materials 2011; 323(7):915–9.
[18]
Lehrer E. Über das Eisen‐Wasserstoff‐Ammoniak‐Gleichgewicht. Zeitschrift für Elektrochemie und angewandte physikalische Chemie 1930; 36:383–92.
[19]
Meka SR, Bischoff E, Rheingans B, Mittemeijer EJ. Octapod-shaped, nanosized, amorphous precipitates in a crystalline ferrite matrix. Philosophical Magazine Letters 2013; 93(4):238–45.
[20]
Mittemeijer EJ, Biglari MH, Böttger AJ, van der Pers, N. M., Sloof WG, Tichelaar FD. Amorphous precipitates in a crystalline matrix; precipitation of amorphous Si3N4 in α-Fe. Scripta Materialia 1999; 41(6):625–30.
[21]
van Landeghem HP, Gouné M, Bordère S, Danoix F, Redjaïmia A. Competitive precipitation of amorphous and crystalline silicon nitride in ferrite: Interaction between structure, morphology, and stress relaxation. Acta Materialia 2015; 93:218–34.
[22]
Mittemeijer EJ, Somers MAJ. Thermochemical Surface Engineering of Steels. Amsterdam: Woodhead Publishing is an imprint of Elsevier; 2014. (Woodhead publishing series in metals and surface engineeringnumber 62).
La Metallurgia Italiana - January 2020
pagina 39
Memorie scientifiche - Trattamenti Termici
Material-dependent aspects for the use of nitriding as a single or duplex treatment in improving the tribological load-bearing behaviour of cast irons (part II) A. Buchwalder, N. Klose, R.Zenker The author’s previously published studies proved the potential of the combined surface treatment of electron beam remelting and nitriding for improving the load-bearing behaviour of cast irons. However, further investigations indicated that a general statement could not be made regarding the suitability and benefits of the duplex treatment. This paper deals with investigations regarding the influence of cast iron base materials differing in graphite morphologies (lamellar, spheroidal), Si content and/or matrix microstructure and the nitriding conditions on the nitriding layer achieved (thickness, phase composition, hardness). Furthermore, the paper reports on the impact of these factors on the depth-dependent wear behaviour under different loading conditions. The nitriding processes for the cast iron (single treatment) and for the remelted cast iron surfaces (duplex treatment) were carried out under identical conditions. Differences in nitrogen concentration profiles and phase compositions resulted. For this reason, detailed investigations were carried out on the depth-dependent microstructural constitutions (phase composition (ε : γ`), porous zone), the hardness and the wear behaviour of the nitriding layers. This was realized by incremental layer removal using a polishing procedure. Each depth state was characterized by XRD analysis (Cu-Kα radiation), the indentation hardness, the scratch energy density and the specific wear coefficient. It was shown that in the case of the nitrided cast iron after the single treatment, the graphite morphology and the load level had the greatest impacts on wear behaviour, followed by the material and the specific nitriding layer construction. This is crucial for defining the potentials and limitations of the single and combined surface treatments with respect to the improvement of the tribological load-bearing behaviour of cast iron.
KEYWORDS: CAST IRON - EB REMELTING - NITRIDING - WEAR - DUPLEX
Introduction With nitriding (N), a wear- and corrosion-resistant surface layer can be produced on cast irons. However, the very soft graphite - specific to this material - interferes with this nitriding layer. Thus, the functionality of the surface layer applied is limited. Electron beam remelting (EBR) of the cast iron surface leads to the dissolution of
Anja Buchwalder Normann Klose Rolf Zenker TU Bergakademie Freiberg
Institute of Materials Engineering, Germany
near-surface graphite and, as a result of self-quenching,
La Metallurgia Italiana - gennaio 2020
pagina 40
Scientific papers - Heat Treatments to the formation of a white solidified microstructure
Analogously to untreated cast iron, the ‘free graphite’ re-
(graphite → ledeburitic cementite) with high hardness
sults in a low coefficient of friction µ after nitriding due
(600 HV0.3 - 750 HV0.3) [1]. During subsequent nitriding,
to its lubricating effect. Depending on the material, nitri-
a closed layer is formed [2,3]. The nitriding mechanism
ding treatment and test parameters, values of µ = 0.1...0.2
of the heterogeneous ledeburitic microstructure is very
[10]...0.3 [11] were measured in unlubricated tests.
complex due to the high Si content, which is distributed inhomogeneously (though primarily in the intercarbidic
In wear tests (pin-on-disc test, unlubricated, FN = 32.8 N, s = 3000 m),
regions) in combination with the high volume fraction of
Nicoletto et al. [10] show that the friction behaviour is
cementite.
determined more by the nitriding process while the wear
The formation of the Fe nitrides and the growth of the
behaviour is determined by the material (and by the graphi-
compound layer (CL) follow both the thermodynamics (cf.
te morphology in particular), since only the matrix micro-
Lehrer diagram [4]) and the kinetics known from the nitri-
structure is hardened and not the graphite. At low sliding
ding of unalloyed steels, though only to a limited extent.
speeds (< 0.5 m/s), the wear rate of spheroidal graphi-
On the one hand, Si inhibits the formation of Fe nitrides,
te cast iron is lower than that of lamellar graphite. When
while on the other hand, Si itself functions as a nitride
subjected to stress, the interconnected graphite lamellae
former with weak interaction [5,3], i.e., the proportion of
and their irregular shape favour the fractionation and de-
unbound silicon changes progressively during nitriding.
tachment (wear particles) of the hard nitriding layer areas
Consequently, the kinetics of Si nitride precipitation in-
immediately surrounding them. In contrast, the graphi-
fluences the kinetics of compound layer growth in the le-
te spheres remain undeformed, which indicates that the
deburitic microstructure. The exact influence of Si on Fe
harder nitriding layer matrix surrounding the soft graphite
nitride formation - and especially on the suppression of
spheres compensates for the applied contact stresses bet-
γ’ nitride formation at the expense of ε nitride - is poorly
ter (i.e., in a more elastic manner) [10].
understood [3]. One conceivable reason for this could be the low degree of solubility of Si in ε nitride. Kante et al. [6]
At higher sliding speeds ≥ 1 m/s, the graphite morphology
present an initial model of time-dependent microstructu-
has hardly any influence. In lubricated wear tests (ball-on-
ral development during the nitriding of Si-rich pearlite.
disc test; counter body (CB): Al2O3, FN = 44 N, s = 6000 m), the higher the roughness and the lower the hardness of
Along with microstructural composition, pore formation
the nitriding layer, the higher the wear rate [12].
also influences the load-bearing behaviour. The causes of pore formation can be manifold. The main cause is the
While Holst et al. [13] provide an overview of the influence
metastability of the ε nitride, which leads to the migration
of nitriding conditions on the layer formation and hardness
of nitrogen atoms into dislocations and voids which then
of cast iron, the depth-dependent layer microstructure
combine to form nitrogen molecules [7]. Furthermore, the
and the resulting tribological behaviour are presented and
hypothesis is presented that Fe diffusion along the grain
discussed in this paper. Therefore, one set of nitriding pa-
boundaries leads to an oppositely directed diffusion of
rameters (540°C, 16 h, KN = 6) was used for treating three
voids into the interior of the ε nitrides, which triggers pore
different cast irons in states either with or without electron
formation (the Kirkendall effect) [8,9].
beam (EB) remelting. The detailed investigations were carried out by analysing
The harder remelted microstructure also serves as a sup-
samples, which were polished stepwise perpendicular to
porting layer for the hard, thin compound layer. In the case
the surface and down to a defined depth. Each surface le-
of surface layers treated in this way (EBR+N), a (partially)
vel was tested and analysed. This was crucial for defining
significant improvement in the tribological and corrosi-
the potentials and limitations of the single and combined
ve load-bearing behaviour could be demonstrated when
surface treatments with respect to the improvement of the
compared to the merely nitrided cast iron (N) [2,3].
tribological load-bearing behaviour of the cast iron.
La Metallurgia Italiana - January 2020
pagina 41
Memorie scientifiche - Trattamenti Termici Experimental
tion, and in particular in their Si contents (cf. column 4 in Tab. 1), graphite morphologies (speroidal (S); lamellar
Base materials
(L)) and/or matrix microstructures (pearlite (P), ferrite (F))
The investigations were carried out on three conventio-
(Tab. 1, cf. I. in Fig. 2a-c).
nal cast irons which differed in their chemical composi-
Tab.1 - Chemical composition (GDOES), phase composition (XRD) and hardness of the substrates used as well as roughness after nitriding
Substrate
Element concentration [wt.%]
Microstructure
Hardness HV1
EN-GJ…
Fe
C
Si
Mn
P
S
Cu
Cr
Graphite
Matrix
Initial
EBR
S400
Base
3.40
2.67
0.14
0.024
0.050
0,02
0.01
S
F
156 ± 26
865 ± 16
S600
Base
3.60
2.33
0.34
0.030
0.010
0.67
0.03
S
P/F
321 ± 17
794 ± 19
L250
Base
3.20
2.11
0.08
0.024
0.010
0.76
0.17
L
P/F
238 ± 22
644 ± 14
Surface treatments
Sampling was performed by two different methods (see
Electron beam remelting (EBR) was used to generate the
Fig. 1). The focus in this article was on longitudinal mi-
ledeburitic surface layer microstructure.
crosections, which were realized by the stepwise remo-
Preheated samples (450°C (725 K)) were remelted using
val of the surface (approx. Δ2 µm) by polishing (Fig. 1b).
the meander beam deflection technique with a continuous
The thickness of the material removed was calculated by
sample feed rate of 0.5 mm/s and a beam current of 15 mA.
measuring the diagonals d1, d2 produced by Vickers har-
As a result of the rapid solidification, a ledeburitic layer
dness indentation (HV1) (Fig. 1b). Each (subsequently re-
with pearlite in the intercarbidic areas was generated ac-
moved) surface Ai+n was characterized by means of:
cording to the metastable system Fe-Fe3C.
•
Roughness measurements
There were hardly any differences in the microstructure
•
Glow
of the EB remelted layers, irrespective of the base material used. Therefore, an example of the EBR microstructu-
optical
emission
spectroscopy
(GDOES): element concentration-depth profiles •
re of S600 is shown in I. of Fig. 2d. Further information about the microstructural details is given in part I of this
discharge
Surface hardness (SH) measurements with F N = 9.81 N (HV1): mixed hardness value
•
Hardness under increasing loading conditions up to
contribution [13].
Fmax = 0.05 N (50 mN) by determining indentation har-
The deformed EBR surfaces were machined flat and
dness HIT according to DIN ISO 14577 1-3: local layer
ground (to SiC 1200) before nitriding.
hardness
The gas nitriding process was carried out at a temperature
•
of 540°C (815 K) for 16 h and with a nitriding potential of KN = 6. Characterization methods
La Metallurgia Italiana - gennaio 2020
X-ray diffraction (XRD) with Cu-Kα radiation: phase composition and quantity distribution (vol.%)
•
Light optical (LOM) and scanning electron microscopy (SEM): microstructure, geometry of the layer
pagina 42
Scientific papers - Heat Treatments
•
and wear tracks after scratch testing and pin-on-disk
abrasive/adhesive wear behaviour by calculated spe-
testing
cific wear coefficient kv= W v/(FN∙s) (W v… wear volume
Scratch testing using a Rockwell indenter with a
(measured by 3D profilometer), F N… normal load, s…
constant load of 20 N (preload: 2 N): abrasion wear
sliding distance (both were defined)) and recorded
resistance by calculating scratch energy density
friction coefficient µ
w R= F T /A R (F T… tangential force (recorded); A R…
•
scratched cross section (measured by 3D profilome-
In the diagrams presented herein, the respective sample
ter))
variant or measuring direction used is indicated by a co-
Unlubricated pin-on-disc tests (counter body: WC-
ordinate system.
Co ball (Ø 6 mm); FN = 20; s = 100 m, v = 0.15 mm/s):
Fig.1 - Variants of samples tested
Results and discussions
graphite was washed out during mechanical surface preparation, round craters remained on the nitrided surface
Characterization of the initial states after single and
of the cast iron with spheroidal graphite (II. in Fig. 2a, b).
combination treatment
For the cast iron with lamellar graphite, only small unco-
Despite identical surface preparation, the surface after ni-
vered areas were visible using SEM (II. in Fig. 2c).
triding was strongly influenced by the substrate’s micro-
For both graphite morphologies, bulging of the com-
structure (II. in Fig. 2) and, for cast irons, this is mainly the
pound layer was detected at the edges of the respective
graphite morphology (cf. Fig. 2a, b and c) or graphite size
graphite (see 3D profiles in Fig. II of 2a-c).
(cf. Fig. 2a and b). The characterization of such inhomogeneous surfaces by the determination of classical roughness parameters such as Ra, Rz, etc. is not effective. Near-surface graphite with a sufficiently large degree of expansion usually remains uncovered by a nitriding layer. Depending on the proportion of outwardly directed compound layer growth and/or the degree to which the
La Metallurgia Italiana - January 2020
pagina 43
Memorie scientifiche - Trattamenti Termici
a) S400
b) S600
c) L250
d) EBR layer (e.g. S600)
Fig.2 - Influence of the (I.) initial microstructure of the different substrates on the (II.) surface formation after nitriding (540°C, 16 h, KN = 6)
Within the compound layer and depending on the che-
examined after gas nitriding. Exceptions to this, however,
mical composition, the N concentration-depth profiles
were large graphite spheres that lay on the surface.
(GDOES) of the respective treatment states (untreated or
These spheres could also be enclosed by a pore seam. In
remelted) exhibited hardly any differences (cf. II in Fig. 3a
the combined treated surface layers (EBR+N) under the
and b). Only the nitrogen diffusion depth was significantly
selected nitriding conditions, pores could be detected
lower in the EBR state.
over almost the entire thickness of the compound layer
There was also nitriding-induced decarburization in both
in the intercarbidic regions. The latter could also be seen
treatment states, which was much more pronounced in
in the significantly higher maximum contents and depth
the case of the EBR+N state (II. in Fig. 3b), i.e. the C con-
diffusion of oxygen for the combination treatment com-
tent dropped very steeply at the transition between the
pared to the single treatment (cf. II. in Fig. 3a and b).
compound and diffusion layers (Cmin = 0.3 ma.%). For the
The solely nitrided cast iron had the typical phase com-
nitrided cast iron base material (BM+N), however, the C
position of compound layers on steel - consisting of ε
content decreased gradually (II. in Figure 3a). Further-
nitride with a maximum fraction and γ` nitride with a mi-
more, the decarburization (measured at the minimum
nimum fraction at the surface, which then counter each
C content) was significantly lower than for EBR+N. The
other by increasing or decreasing with increasing distan-
decarburization (C min = 0.4→0.8 ma.%) and the depth of
ce from the surface (I. in Fig. 3a).
the porous zone PZ (t PZ = 1.3→2.9 µm) tended to incre-
The compound layers on the remelted substrates pre-
ase with increasing nominal C content in the cast iron
dominantly contained ε nitride (60...90 vol.%) over the
(L250→S400→S600, cf. Tab. 1).
entire depth (I. in Fig. 3b). This resulted from the low
The nitrided cast iron had a pore seam which was typi-
N solubility of carbides (Fe3C), as a result of which they
cally found in the near-surface area - analogous to steels
converted preferentially into nitrides. The formation of γ`
La Metallurgia Italiana - gennaio 2020
pagina 44
Scientific papers - Heat Treatments nitride was suppressed due to the restricted carbon solu-
in the compound and diffusion layer, as was shown by
bility, and ε nitride was preferred (I. in Fig. 3b).
comparative investigations by the authors between a C15
It could be assumed that ε carbonitride was formed as a
carbon steel and a cast iron with a ferritic matrix (S400)
result of the increased carbon activity in the environment
[3].
[14].
These were mostly amorphous precipitates [15] and
In addition to the formation of Fe nitrides, the precipita-
could not be analyzed using XRD, so that they will be left
tion of Si nitrides also influenced the increase in hardness
out of subsequent discussions of the results.
a) Nitrided cast irons (BM + N)
b) Nitrided EBR layers (EBR+N)
Fig.3 - Influence of the substrate used on the depth-dependent (I.) fractions of ε and γ`nitrides (XRD Cu-Kα) and (II.) element concentration profiles (GDOES) (540°C, 16 h, KN = 6)
Investigations on the depth-dependent evolution of
of 700 HV1, the indentation depth was approx. 7 µm. With
hardness and wear behaviour
respect to the Bückle rule, it could be assumed that the
The depth-dependent hardness curves shown in Fig. 4
hardness values contained information from a depth of
resulted from the layer microstructure (phase composi-
up to approx. 50 µm. Essentially, the surface hardness of
tion and defects) described above. When interpreting the
the combined treated surfaces (EBR+N) in both the com-
results, the different measured loads (I. 9.81 N (HV1), II.
pound layer and the diffusion layer was approx. 200 HV1
0.050 N) and load directions (both vertical (I., III) and pa-
higher than on the solely nitrided cast iron (cf. I. in Fig. 4a
rallel (II.) to the surface) had to be taken into account.
and b).
The hardness measurement with a load of 9.81 N provi-
This was a reliable reflection of the significantly better
ded a composite hardness (I. in Fig. 4) as a function of
supporting effect for the compound layer in the case of
the distance beneath the surface, consisting of hardness
the much harder remelted carbidic microstructure.
parts of the compound layer and the underlying diffusion
While the hardness curves of spheroidal graphite cast
layer or substrate. For example, with an average hardness
irons were relatively comparable after both single and
La Metallurgia Italiana - January 2020
pagina 45
Memorie scientifiche - Trattamenti Termici combination treatments (cf. I. in Fig. 4a and b), those of
rous zone. In nitrided cast iron, this was formed relatively
lamellar graphite cast iron (L250) deviated significantly
uniformly up to a material- and nitriting-dependent dep-
from these and lay at hardness levels that were lower by
th.
approx. 150 HV1 (EBR+N) to 300 HV1 (L250+N). The exception was the graphite environment. Due to the The local layer hardness resulting from the microstructu-
favourable diffusion and penetration conditions for the
re (II. and III. in Fig. 4) could be determined by means of
nitrogen at the graphite/matrix gap interface (cf. [14,16]),
hardness measurements under increasing loading condi-
nitrogen could penetrate to greater depths and led to
tions, with a maximum load of 0.05 N.
pore formation.
With a maximum penetration depth of 0.5 µm, this hard-
In contrast, pore formation during nitriding of the EBR
ness value contained information from a depth of approx.
layers occurred mainly in the intercarbidic regions or at
3 µm, i.e. with the exception of the influence of pores in
the interface with the ledeburitic carbides and under the
the area near the surface, the supporting effect (for exam-
selected nitriding conditions over almost the entire dep-
ple, that provided by the hard carbides) played only a su-
th of the compound layer.
bordinate role - at least in the compound layer. This led to the influence of the layer hardness measured As already shown above, the compound layers genera-
perpendicular to the surface, which is shown in III. in Fig.
ted in the EBR layer consisted mainly of ε nitrides (cf. II.
4b.
in Fig. 3b), which led to somewhat higher average layer
While the hardness measurements carried out parallel
hardness values of 12.0 ± 0.9 GPa (II. in Fig. 4b) compared
to the surface (II. in Fig. 4) on the etched cross section
to the compound layers (cf. II. in Figure 3a) that contained
facilitated a selection of the ‘defect-free’ microstructural
more γ` nitrides on the cast iron (10.5 ± 0.8 GPa, II. in Fig.
areas - i.e., the matrix microstructure in nitrided cast iron
4a). The differences due to the chemical composition of
and the carbides in combined treated cast iron - it was
the different cast irons were rather marginal for the two
not possible to achieve this differentiation with certainty
states (BM+N; EBR+N).
in the hardness measurement perpendicular to the surface (III. in Fig. 4).
From this, it could be concluded that the large differences in the macro hardness measurements (I. in Fig. 4a)
The surfaces of the unetched ablated samples appeared
resulted from the influence of lamellar graphite and the
almost uniformly white, as the pores or graphite were
resulting strong fragmentation of the compound layer
partially lubricated during polishing. Thus, selection of
(loss of load support).
the pore-free carbides or graphite (especially with the
In the case of the combination treatment (EBR+N), the
narrow lamellae of the L250) was not possible and the
lower C content of the L250 (cf. Tab. 1) and the associated
hardness was at a lower level (III. in Fig. 4) than the hard-
lower hardness of the EBR layer influenced the suppor-
ness measurement at the cross sections (II. in Fig. 4).
ting effect, and led to a lower overall hardness of 150 HV1. In the lower, pore-free area of the compound layer, the Interestingly, the profile of the layer hardness measured
true layer hardness was measured at 11.9 ± 0.6 GPa for
parallel to the surface (II. in Fig. 4) differed significantly
the nitrided cast iron (BM+N) and at 13.6 ± 2.0 GPa for the
from that measured perpendicular to the surface (III. in
combined treated cast iron (EBR+N).
Fig. 4). This applied in particular to the significant decrease in hardness in the near-surface area resulting from the po-
La Metallurgia Italiana - gennaio 2020
pagina 46
Scientific papers - Heat Treatments
a) Nitrided cast irons (BM + N)
b) Nitrided EBR layers (EBR+N)
c) SEM image of the CL
Fig.4 -Influence of the substrate used on the depth-dependent (I.) surface hardness SH (9.81 N), the indentation hardness HIT (0.050 N) measured both (II.) parallel and (III.) perpendicular to the surface
The microstructural formation described and the re-
gential forces F T were approx. twice as high after the sin-
sulting hardness profiles influenced the tribological lo-
gle treatment than after the combination treatment.
ad-bearing behaviour, which was investigated under both
At greater depths, the scratch energy density for the com-
purely abrasive (scratch test with constant normal load)
bination treatment (EBR+N) was 2-4 times higher
and under abrasive-adhesive loading conditions (pin-on-
than for the single treatment (BM+N)(cf. I. in Fig. 5a and b).
disk test) (Fig. 5).
Failure modes generated by scratch testing with constant
On the basis of the scratch track produced by a loaded
loading were quite different on both of the nitrided sub-
(20 N) diamond cone, the so-called scratch energy den-
strates. It should be noted that comparisons between dif-
sity w R was calculated from the scratched cross-sections
ferent samples are only valid if the mechanism of failure
measured. The higher the scratch energy density in this
is the same. As a result of bending of the hard compound
measurement, the better the abrasive wear resistance.
layer into the softer cast iron, conformal through-thick-
While the nitrided cast iron within the compound layer
ness cracking occurred, both perpendicular C1 and paral-
had an almost constant scratch energy density of approx.
lel C3 (at the edges) to the scratch direction v (see LOM
21. 9 ¹ 6 N/mm² (I. in Fig. 5a), it was highly depth-depen-
image in I. in Fig. 5a). Furthermore, cracking around the
dent for the combined treated surface layers (EBR+N).
graphite C2 was observed. It was assumed that after the
In the near-surface area, the scratch energy density was
indenter passed and the scratched region was unloaded,
comparable with the single treatment (I. in Fig. 5b). Com-
the elastic deformation in the coating-substrate system
parable scratch energy densities meant that the scratched
was relaxed to some extent. However, due to the pla-
cross-sectional areas were 3-4 times larger and the tan-
stic deformation in the substrate, which resulted in the
La Metallurgia Italiana - January 2020
pagina 47
Memorie scientifiche - Trattamenti Termici formation of the visible scratch track (see LOM image in
In contrast, the combination treated surfaces did not
I. in Fig. 5a), it was not possible to completely relax the
crack or form other modes of failure (see LOM image in
substrate’s elastic deformation, and a residual stress re-
I. in Fig. 5b).
mained.
With decreasing hardness in the diffusion layer, the scra-
It is possible that the scratch energy density was influen-
tch energy density also decreased.
ced by this process, i.e., that it increased.
a) Nitrided cast irons (BM + N)
b) Nitrided EBR layers (EBR+N)
Fig.5 -Influence of the substrate used on the depth-dependent (I.) scratch energy density wR, determined by scratch testing under constant load (FN = 20 N), (II.) specific wear coefficient kV, and (III.) average friction coefficient µ, both determined by unlubricated pin-on-disk testing (FN = 20 N)
The wear behaviour under abrasive-adhesive loading
Fig. 5a). The reason for this was that the coefficient of fri-
conditions was determined not only by the hardness of
ction of the EBR+N variants (µ EBR+N = 0.78 ± 0.04, III. in Fig.
the nitriding layer but also by the coefficient of friction
5b) was essentially three times higher than that of the nitrided
µ and pore formation. The latter led to an increase in the
cast iron with spheroidal graphite (µGJS+N = 0.26 ± 0.09, III. in
specific wear coefficient kV on both substrates (cf. II. in
Fig. 5a). The diagram in section III. of Fig. 5a shows very
Fig. 5).
impressively how the friction behaviour is influenced by
If the porous zone was neglected, for the EBR+N variant a
the graphite morphology and its size at comparable hard-
specific wear coefficient of kV, EBR+N = 0.69 ± 0.25∙10 N∙m/mm³
ness. Under the given load conditions, the large graphite
(II. in Fig. 5b) averaged over all materials results, which
spheres (dS = 36 ± 12 µm) of the S400 led to the formation
would be approximately twice as high as that of the only
of a closed graphite lubricant film (see black film in wear
nitrided cast iron (kV, GJS+N = 0.31 ± 0.11∙10 N∙m/mm³, II. in
tracks in the micrographs in II. of Fig. 5a) that exhibited
-6
-6
La Metallurgia Italiana - gennaio 2020
pagina 48
Scientific papers - Heat Treatments excellent constant friction conditions (µ = 0.21 ± 0.02).
be determined:
The somewhat smaller graphite spheres of the S600 (dS
(1) The course of the N diffusion profiles (GDOES measu-
= 24.5 ± 0.4µm) led to somewhat unsteady conditions (in
rements) within the compound layer differed only margi-
terms of the formation of the graphite lubricant film) and a
nally between the two substrate variants (BM+N; EBR+N),
slightly increased coefficient of friction with a larger ran-
with the exception of the final depth reached. The diffu-
ge of fluctuation (µ= 0.32 ± 0.10). With the L250, the lubri-
sion layer was characterized by strong N enrichment due
cating effect of graphite no longer existed at all due to the
to the presence of Si.
small effective cross-sectional dimension, so that analo-
The nitriding-induced decarburization in the area of the
gous (non-lubricated) friction conditions (µ= 0.67 ± 0.04)
compound layer was much more pronounced for the
existed as in the EBR+N state, which was also reflected
combination treatment EBR+N.
in comparable specific wear coefficients (cf. L250 in II. in Fig. 5a and b).
(2) While the nitrided cast iron exhibited the typical phase
While the specific wear coefficient for the solely nitrided
structure of compound layers on steel - consisting of ε
L250 (II. in Fig. 5a) increased in the transition region of
nitride with a maximum fraction and γ` nitride with a mi-
the compound to the diffusion layer due to the drop in
nimum fraction at the surface, which then countered each
hardness (II. in Fig. 4a), it decreased successively for the
other by increasing or decreasing with increasing distan-
combination treated variant (II. in Fig. 5b).
ce from the surface - the compound layers on the remel-
As already mentioned in the scratch test, a crack network
ted substrates mainly contained ε nitride (60...90 vol.%)
formed around the soft graphite under loading conditions
over the entire depth.
(see magnified micrograph in III. in Fig. 5a). In addition to its good lubricating effect, graphite also acted as a reser-
(3) The nitrided cast iron had a pore seam typically found
voir for abrasive wear particles, i.e., it exhibited smearing
in the near-surface area, which was analogous to steels
(see magnified micrograph in III. in Fig. 5a) and, there-
after gas nitriding. Exceptions, however, included large
fore, the surface was partially hardened. When the wear
graphite spheres that lay on the surface, whereby these
coefficient was determined by the profilometric measu-
spheres could also be enclosed by a pore seam. In the
rement of the wear volume, the specific wear coefficient
combined treated surface layers (EBR+N) under the se-
was distorted (too low) for the cast iron with spheroidal
lected nitriding conditions, pores could be detected over
graphite. In contrast, the wear removal on the combined
almost the entire thickness of the compound layer in the
treated surfaces was uniform and flat (see wear tracks and
intercarbidic regions. Pores reduced the layer hardness
magnified micrograph in III. in Fig. 5b).
values measured at low levels of loading (50 mN), as did the scratch energy density as a characteristic of abrasive
Summary and conclusions
wear resistance. In contrast, the specific wear coefficient
Within the scope of this paper, comparative investiga-
- as a characteristic of the abrasive-adhesive wear resi-
tions were carried out on the tribological load-bearing
stance - increased. In all cases, this meant a deterioration
behaviour of various cast irons after nitriding BM+N (sin-
of the wear behaviour.
gle treatment) and remelting and subsequent nitriding EBR+N (combination treatment).
(4) In the layer area with low or no porosity, the scratch
Nitriding was carried out for all substrate variants at 540°C
energy density of the combination treated samples was
/ 16 h, K N = 6. The focus was on the effects of the chemi-
2-4 times higher than for the single treatment. The speci-
cal composition (particularly the C and Si contents) of the
fic wear coefficient of the solely nitrided cast iron depen-
cast irons as well as the graphite morphology (sphere,
ded strongly on the graphite morphology and the size.
lamella) on the depth-dependent phase composition in
On the one hand, the graphite served as a lubricant and,
the compound layer, the hardness characteristics and the
on the other hand, as a reservoir for abrasive particles,
wear behaviour. The following essential findings could
which resulted in completely different loading condi-
La Metallurgia Italiana - January 2020
pagina 49
Memorie scientifiche - Trattamenti Termici tions than in the case of the combination treatment. Only
The results presented show that a direct comparison of
cast iron with lamellar graphite was comparable, since
the tribological behaviour of the two nitrided substra-
the mechanisms described above did not work. Both the
tes (BM+N and EBR+N) on the basis of the determined
coefficient of friction and the depth-dependent course of
wear values is difficult. Both the stress mode (lubricated,
the specific wear coefficient were comparable with those
non-lubricated) and the resulting contact conditions as
of the combination treatment.
well as the wear mechanism were very different.
At the transition to the diffusion layer, the wear coeffi-
An essential aspect for further optimization of the nitri-
cient of the nitrided cast iron increased due to the absen-
ding process or the preferred nitriding result is the re-
ce of a supporting effect, while that of the combination
duction of pores and/or decarburization.
treatment decreased further.
Acknowledgements The authors wish to thank Dr. C. Schimpf (Institute of Materials Science, TU Bergakademie Freiberg) for his assistance in carrying out the XRD measurements and Ms. G. Bittner (Institute of Materials Engineering, TU Bergakademie Freiberg) for performing hardness measurements. Many thanks also go to Dr. H. Zimdars (Institute of Materials Engineering, TU Bergakademie Freiberg) for carrying out the nitriding experiments. Furthermore, the article contains partial results that were generated in the context of the diploma theses of Mr. B. Haas, Mr. R. Wagner and Ms. J. Fitz.
La Metallurgia Italiana - gennaio 2020
pagina 50
Scientific papers - Heat Treatments
References [1]
Holst A, Buchwalder A, Hollmann P, Zenker R. Influence of cooling rate on the microstructural fea-tures of a remelted white-solidified cast iron surface and its effects on nitriding behaviour. Journal for Materials Processing Technology 2019;271:377-383
[2]
Buchwalder A, Zenker R, Fritzsch K, Rüthrich K, Nagel K, Griesbach W. Potenziale von ausgewählten Randschichtbehandlungsverfahren sowie deren Kombinationen zur Verbesserung des Verschleiß- und Korrosionsverhaltens von Gusseisenwerkstoffen. Materials Science and Engineering Technology 2015;46:550-562
[3]
Buchwalder A, Spies HJ, Klose N, Hegelmann E, Zenker R. Contribution to the microstructural con-stitutions and properties of nitrided as-received and remelted cast irons. Proc. of 24th IFHTSE Congress, Nice, June 26-29 2017, CD
[4]
Lehrer E. Über das Eisen-Wasserstoff-Ammoniak-Gleichgewicht, Z. Elektrochem. Angew. Phys. Chem. 1930;36:383–392.
[5]
Mittemeijer EJ, Somers MAJ. Thermochemical surface engineering of steels: Improving materials performance, Elsevier, 2014.
[6]
Kante S, Leineweber A. Two-Phase and Three-Phase Crystallographic Relationships in White-Solidified and Nitrided Fe-C-Si cast iron. Acta Materialia, 2019;170:240-252.
[7]
Prenosil B. Einige neue Erkenntnisse über das Gefüge von um 600°C in der Gasatmosphäre carbonitrierter Schichten. HTM 1973;28:157-164
[8]
Matauschek J, Trenkler H. Neue Erkenntnisse über den Aufbau und die Porigkeit der beim Badnitrieren erzeugten Verbindungsschicht. HTM 1977;32:177-181.
[9]
Hoffmann R, Mittemeijer EJ, Somers MAJ. Verbindungsschichtbildung beim Nitrieren und Nitrocarburieren. HTM 1996;51:162– 169.
[10]
Nicoletto G, Tucci A, Esposito L. Sliding wear behavior of nitrided and nitrocarburized cast irons. Wear 1996;197:38-44.
[11]
Nie X, Wang L, Yao ZC, Zhang L, Cheng, F. Sliding wear behaviour of electrolytic plasma nitrided cast iron and steel. Surf Coatings Technol. 2005;200:1745-1750.
[12]
Yang Z. The Use of Nitriding to Enhance Wear Resistance of Cast Irons and 4140 Steel. Dissertation, Paper 4714, University of Windsor, 2013.
[13]
Holst A, Buchwalder A, Hollmann P, Zenker R. Influence of Gas Nitriding Conditions on Layer Structure Formation on Grey and White Solidified Cast Irons (Part I). Conf. Proc. ECHT 2019. 2019 Jun 5-7, Bardolino, Italy. Milan: AIM; 2019: paper 060
[14]
Keller K. Vergleichsuntersuchungen an gas-, bad- und ionitriertem Kugelgraphitguss und Nitrierstahl. Ph. D. Thesis. TH Aachen. 1969
[15]
Mittemeijer EJ, Biglari MH, Böttger AJ, van der Pers NM, Sloof WG, Tichelaar FD. Amorphous precipitates in a crystalline matrix, precipitation of amorphous Si3N4 in α-Fe. Scripta Materialia 1999;41:625-630.
[16]
Wiegand H, Hentze H. Dauerschwingfestigkeit und Verschleißverhalten von salzbadnitrierten Zahnrädern aus Gusseisen mit Kugelgraphit. Metalloberfläche 1959;13:238-242
La Metallurgia Italiana - January 2020
pagina 51
Memorie scientifiche - Trattamenti Termici
Simulation of heat treatment of jominy specimen to improve quality of automotive gear components E. Morgano, C. Viscardi, L. Valente Microstructure and mechanical performance play a key role for gear box reliability. According standard quality assurance procedure steel grade analysis is a fundamental base to identify correct mechanical properties achievable after heat treatment cycle. With the aim to make more robust production cycle of gear box components, samples are cut from parts and Jominy test is performed on this specimen. In case of small components like gears, specimen for Jominy test cannot be obtained and the only approach to validate process is supplier certificate obtained on steel before heat treatment. Quenchability of component can be different from data reported on nominal supplier datasheet with the risk to have failure during testing or exercise of gear box; tests on failed parts demonstrated that chemical composition variation, notwithstanding inside allowed range, drove to lower mechanical performance according microstructure not properly quenched. In order to validate heat treatment cycle, microstructure and mechanical properties, Jominy test has been virtualized into heat treatment simulation software. Real Jominy tests have been transferred inside simulation software for different steel grades used for automotive gear box application. A model to forecast microstructure and mechanical properties for Jominy specimen is built and benchmarked with experimental data. Alignment between simulation results and reality allows to transfer model definition from Jominy samples to complete real components, improving quality assurance capability for real process thanks to usage of heat treatment cycle virtualization for gear box realized with different steel grades.
KEYWORDS: HEAT TREATMENT SIMULATION - QUENCHING - MICROSTRUCTURE HARDNESS - RESIDUAL STRESS - DISTORSION - DEFORM HT
Introduction Gear box performances are based on mechanical properties that each part can achieve after manufacturing cycle and in particular after final heat treatment. Reliability of gear box assembly requests to validate each component of the assembly and respect of specification. Gears for automotive transmission are in general realized using
Enrico Morgano FCA CRF Metal Bulk, Italy
Cristian Viscardi, Lorenzo Valente Ecotre Valente Srl, Italy
hardening steel and quality of heat treatment is defined by several standards that are mandatory for automotive applications [1,2,3,4].
La Metallurgia Italiana - gennaio 2020
pagina 52
Scientific papers - Heat Treatments Steel composition of component is defined by standards
Standard size of specimen represents a problem in case
and drawings; small variations in chemical composition
of small parts because it’s not possible to get exact sam-
are allowed. These fluctuations of composition impact on
ple accordingly technical specification.
quenchability of the part and so on its final mechanical
Moreover, Jominy test is a water quench exam [5] while
performance.
gear components are in general quenched with different
To keep under control these variations, a simple but effi-
media and in this way it’s difficult to compare Jominy cur-
cient test to validate mechanical performance of a quen-
ve provided by steel mill certificate with hardness profile
ched and tempered gear in hardening steel is Jominy Test.
obtained on real component.
This methodology requests a specimen with an exact geometry (Fig. 1) and water quenching condition. Quen-
A new approach is necessary to allow proper usage of
ched specimen is tested to get a hardness profile across
Jominy Test results for components hardened with diffe-
thickness which is a curve with hardness value versus di-
rent methods and with geometry not allowing extraction
stance from quenched end.
of standardized sample.
Fig.1 - Jominy Test geometry and slice of a real specimen showing hardness test indentation
Jominy test Jominy test virtualization In order to get a reliable model to forecast performance
Jominy specimen is in steel 17 NiCrMo7 and chemical
of hardened steel gear, FEM software DEFORM-HT has
composition in reported in table 1.
been used to simulate Jominy test [6]. Tab.1 - 17 NiCrMo7 analysis 17 NiCrMo7 composition C
Ni
Cr
Mn
Si
Mo
V
Cu
Al
P
S
Nb
Ti
B
Fe
0.17
1.91
0.53
0.58
0.34
0.26
0.0027
0.16
0.015
0.01
0.03
0.002
0.001
0.0005
Rem.
Initial temperature of sample is 865 °C with full austenite microstructure; applied heat transfer coefficients for test is shown in figure 2 [7,8,9].
La Metallurgia Italiana - January 2020
pagina 53
Memorie scientifiche - Trattamenti Termici
Fig.2 - Initial condition and boundary condition applied to DEFORM-HT simulation for Jominy Test
Due to standardization, effects of specimen geometry
because it provides hardness profile result as a function
and quenching can be neglected, which is an advantage
of transformed phase fraction and its hardness. This ap-
of this test. In DEFORM-HT determination of hardness
proach is independent by geometry or quenching system
profile can be approached with 2 different methods: Jo-
because resulting hardness is coming from transforma-
miny Curve from Jominy Distance or Jominy Curve from
tion from austenite “mother phase” into different “chil-
Phase Fraction. Jominy Curve from Jominy Distance is an
dren phases” like martensite or bainite; HRC hardness
approach valid only for Jominy Test; given specimen geo-
is the result of the percentage of each phase at the end
metry and water quenching condition from standard; har-
of heat treatment. DEFORM-HT uses the TTT curves in a
dness profile result is affected only by sample chemical
dynamic way for microstructural calculation; this means
composition and grain size, inclusion and oxides. Coo-
that, despite the use of TTT is always taken into account,
ling rate achieved across thickness is driven only by con-
the kinetics of phase transformations and hence the co-
ductivity and specific heat of steel. This kind of test is very
oling rate have a fundamental role for predicting micro-
easy to carry out but it’s not flexible in case of application
structure. Microstructural transformation, like Austenite
to different quenchant media or geometry. Jominy Curve
→ Pearlite or Austenite → Bainite, is described based on a
from Phase Fraction propose to a more refined approach
TTT Diagram, like shown in Figure 3.
Fig.3 - Kinetics of microstructural transformation defined for DEFORM-HT: Austenite to Pearlite and Austenite to Bainite
La Metallurgia Italiana - gennaio 2020
pagina 54
Scientific papers - Heat Treatments Jominy test experimental results Jominy test results on sample in steel 17NiCroMo7 are reported in table 2
Tab.2 - 17NiCrMo7 Hardness profile 17 NiCrMo7 Hardness Profile [HRC] as a function of distance from surface [mm] Distance [mm]
1.5
3.0
5.0
7.0
9.0
11.0
13.0
15.0
20.0
25.0
Hardness [HRC]
44.5
44.0
41.5
37.0
33.0
30.0
29.0
27.0
24.0
23.0
Jominy test simulation results In order to validate reliability of simulation compared to reality, in particular for Fraction of Phase approach, HRC hardness profiles have been plotted in the same graph, shown in figure 4.
Fig.4 - Hardness profile of experimental results, DEFORM-HT simulation results and scatter band for H+ from Standard
Experimental results are inside H+ scatter band impo-
results; a representation of phases distribution at 90 se-
sed by EN Standard. Both Jominy Distance and Volume
conds after quench beginning is shown in Figure 5.
Fraction of phases are well aligned to experimental resul-
In every step “Mother” phase coexists with the other
ts. Jominy Distance result aligned to experimental result
“Children” Phases.
is a confirmation of correct input and correct calculation of DEFORM-HT simulation.
Other information that can be used to monitor heat treatment stage are Time between 800 °C and 500 °C that is
Regarding Volume Fraction of Phases, this result confir-
called also and it’s an index of cooling rate of material
ms that approaching hardness estimation based on tran-
[7].
sformation of single phases can allow to reach correct
La Metallurgia Italiana - January 2020
pagina 55
Memorie scientifiche - Trattamenti Termici
Fig.5 - DEFORM-HT distribution of temperature, Time 8/5, Bainite and Martensite after 90 s from water quench beginning
This method, as itâ&#x20AC;&#x2122;s not based on fixed geometry and
particular for small components that are hardened and
fixed quenching conditions as defined in standards, al-
quenched with different technology.
lows to use simulation with different geometries and in
Application to pinion gear Heat treatment cycle for gear Results obtained from validation of Jominy test has been transferred on a pinion gear in hardening steel 17NiCroMo7, shown in Figure 6.
Fig.6 - Pinion gear simulated with DEFORM-HT.Hardening Heat treatment cycle in DEFORM-HT simulation software is described in table 3
La Metallurgia Italiana - gennaio 2020
pagina 56
Scientific papers - Heat Treatments Tab.3 - Hardening heat treatment cycle 17 NiCrMo7 Pinion – Hardening Heat Treatment Cycle Duration [s]
Operation
Temperature [°C]
Carbon concentration [%]
1800
Heating Furnace
550
0
3600
Carburization Furnace
850
0.8
1200
Oil Quench
100
0
1800
Heating Furnace
225
0
3600
Air Cooling
30
0
Results obtained on pinion gear
different slice locations as well as Hardness distribution.
In this simulation, carburization stage introduces an ad-
Both pictures are captured at the end of Oil Quench ope-
ditional variable.
ration.
Carbon content is modified across gear thickness during
This simulation has been run coupling together Thermal
a heating in carburization furnace and this means that it’s
calculation, Microstructural calculation, Stress calcula-
necessary to introduce in TTT curves in DEFORM-HT
tion; as in Jominy Test, for all “Children” phases Thermal,
also dependence from Carbon content. In this way it
Microstructural and Elastic-Plastic properties are defined
will be possible to consider different TTT curve location
as a function of temperature and carbon content to obtain
in Temperature-Time diagram and different Martensite
an accurate solution.
temperatures.
Deformed gear can be exported from simulation in order to be compared with 3D scanning of experimental com-
According this modification in material input data, simu-
ponent.
lation has been run on shown pinion gear.
Simulation results are in alignment with experimental
In Figure 7 some results of simulation are shown; it is visi-
data measured on pinion gear in a range of ± 5%.
ble how Carbon content is various across thickness and at
Fig.7 - Pinion gear simulation with DEFORM-HT: Carbon content and HRC Hardness on two different slices
La Metallurgia Italiana - January 2020
pagina 57
Memorie scientifiche - Trattamenti Termici Conclusion DEFORM-HT simulation results are aligned with experi-
Fraction of Phases approach has been transferred on a
mental Jominy Test results.
pinion gear and comparison between experimental and
Simulation results of approach based on Volume
simulation data is in a range of ± 5%.
Fraction of Phases, valid for every type of heat treatment, are aligned with simulation results of Jominy Di-
Finally, DEFORM-HT predicts microstructure, residual
stance approach, valid only for Jominy Test.
austenite, mechanical properties, hardness, residual
According
this
alignment,
DEFORM-HT
Volume
stress and deformations of gear.
Acknowledgements Authors would like to thank M.Sc.Eng. Paolo Martelli for great contribution in development of this work.
References [1]
EN 10083-1, Steels for quenching and tempering - Part 1: General technical delivery conditions
[2]
EN 10083-2, Steels for quenching and tempering - Part 2: Technical delivery conditions for non alloy Steels
[3]
EN 10083-3, Steels for quenching and tempering - Part 3: Technical delivery conditions for alloy steels
[4]
EN 10083-4, Case hardening steels – Part 4: Technical delivery conditions
[5]
ISO 642-79: Steel Hardenability Test by End Quenching (Jominy test)
[6]
DEFORM User Manual, v11.3
[7]
Quenching Theory and Technology - B. Liščić, H.M. Tensi, L.C.F. Canale, G.E. Totten - (second edition) CRC Press
[8]
ASM Metals Handbook – Volume 4 (10th edition – 1990) – Heat Treating
[9]
Measurement of quenching intensity, calculation of heat transfer coefficient and global database of liquid quenchants – B. Liščić, T. Filetin – Materials Engineering 19 (2012) p.52-63
[10]
Prediction of Quench-Hardness within the Whole Volume of Axially Symmetric Workpieces of Any Shape - B. Liščić, S. Singer, B. Smoljan – Journal of ASTM International, Vol.7 No.2 – Paper ID JAI 102647
[11]
Simulation of Heat Treatment of a mold for HPDC: Cooling in gas or salt to get best performance?
A. Magistrelli, D. Mattavelli, P. Martelli, L. Valente, C. Viscardi EUROPEAN CONFERENCE ON HEAT TREATMENT 2015 & 22nd IFHTSE CONGRESS
La Metallurgia Italiana - gennaio 2020
pagina 58
Scientific papers - Heat Treatments
Neural networks-based prediction of hardenability of high performance carburizing steels for automotive applications V. Colla, M. Vannucci, L. Bacchi, R. Valentini
The new quenching processes for automotive applications, which follow the cementation stage, include the application of pressurized gas for cooling during quenching. Therefore, it is of utmost importance to have an accurate estimate of the hardenability behavior of carburizing steels, which show a higher Carbon content with respect to traditional materials. These new cooling processes also require properly designed new steels in terms of alloying contents, which ensure a proper response to heat treatment. In the present paper a neural network-based approach to the prediction of the hardenability profile is proposed, which can be applied both for the design of the steel chemistry and for assessing the suitability of the steel at the steel shop level, in order to suitable adjusting the cooling process after quenching.
KEYWORDS: LOW PRESSURE CARBURIZING - NEURAL NETWORK BASED PREDICTION - JOMINY END QUENCH TEST - TRANSMISSION GEARS
Introduction
face to the core. This is followed by a quenching process
Case hardening is a surface treatment able to confer a
and a low temperature stress relieving tempering.
component elevated resistance to wear, fatigue and in
The cementation can be carried out with different appro-
general surface damage. Moreover, this treatment gives
aches, solid, salts bath and gaseous. Between these, the
at the same time a high toughness at the core, thanks to
gaseous method is the most used because of process
the lower cooling speed, that is why it is commonly used
controllability and speed of the process itself. In this
for parts as transmission gear teeth and diesel fuel injec-
case the surface carburizing is due to carbon dissociation
tors. The process is typically applied to low carbon steels or low alloyed steels as Cr-Ni, Cr-Mo or Mn-Cr. These elements can be added to improve steel hardenability reaching a fully martensitic transformation till the core and thus sufficiently uniform mechanical properties. Case hardening is a thermochemical diffusion treatment, it consists of a first heating of the component up to austenitizing temperature where the carbon enrichment of the surface is possible and diffusion within the steel is
Valentina Colla, Marco Vannucci TeCIP Institute, Scuola Superiore Santâ&#x20AC;&#x2122;Anna, Italy
Linda Bacchi Letomec s.r.l., Italy
Renzo Valentini DICI; UniversitĂ di Pisa, Italy
enhanced by the elevated carbon gradient from the sur-
La Metallurgia Italiana - January 2020
pagina 59
Memorie scientifiche - Trattamenti Termici from a hydrocarbon (propane, acetylene or methane).
provides not only the hardness value at each point but
The innovative variant of the treatment is the low-pres-
also an estimate of the reliability of such prediction.
sure carburizing (LPC), which is carried out at very low
This approach is also exploited within the present work
pressure using hydrocarbon gases and followed by high
for the simulation.
pressure gas quench. This model was used to optimize gaseous hardening steThis process allows constancy of metallurgical characte-
els chemical composition predicting and comparing Jo-
ristics and lower deformations, surfaces without oxidi-
miny curves. Chemical composition was varied in terms
zing products and no retained austenite. Clean surfaces
of alloying elements and concentration, focusing on Mn-
lead to no need of cleaning products, resulting in lower
Cr and Mn-Cr-B steels.
cost for the treatment (1).
These steels can return a good compromise between mechanical properties and sufficient toughness, but they
Case hardening steels standards are usually very restri-
are a rational choice considering economical reason.
ctive in terms of Jominy curves, as the case of FCA internal material standard number (2).
Materials and methods Low carbon and low alloyed steels are usually used for
The Jominy End Quench Test is very expensive and time
case hardening process, to guarantee the major carbon
consuming thus a lot of models to predict hardenability
gradient between surface and core.
from the steel chemical composition have been proposed.
Considering the low carbon content, some alloy elements are introduced in the chemical composition to
The first attempts were based on traditional statistical
increase the hardenability of the steel reaching a com-
techniques, subsequently, more complex models based
plete martensitic transformation also at the very core of
on Machine Learning (ML) techniques have been em-
the component.
ployed outperforming classical approaches. In particular, Artificial Neural Networks (ANN) have been succes-
For this study, some Jominy tests were industrially per-
sfully applied by providing very satisfactory predictive
formed on various castings, considering the chemical
capabilities. In literature several approaches concerning
ranges of FCA internal standard (Table 1). Jominy End
ANN-based Jominy profile predictors have been propo-
Quench tests were carried out with reference to ASTM
sed (3,6), which predict the steel hardness for each sin-
A255 (11).
gle point of the Jominy profile, while in (7,9) a parametric approach has been proposed and refined, which predicts
Resulting hardness profiles did not match every time
each single point as part of a parametric curve, whose pa-
with desired ranges, in Figure 1 some examples were re-
rameters are correlated to the steel chemistry.
ported and the high deviation of results is evident.
A further relevant advancement is proposed in (10), where a sequantial ANN-based predictor is proposed, which
Tab.1 - Chemical composition ranges (%), balance: Fe C
Mn
Si
Cr
B
Ti
Cu
S
P
Al
0.23÷0.31
1.10÷1.40
0.10÷0.40
1.00÷1.30
≤0.0005
≤0.010
≤0.30
0.025÷ 0.040
≤0.030
0.015÷ 0.040
La Metallurgia Italiana - gennaio 2020
pagina 60
Scientific papers - Heat Treatments
Fig.1 - Jominy curves returned on castings with FCA standard chemical composition
A comparison between test results and hardness ranges
ANN-based (refer to paragraph 2). Chemical composi-
reported in FCA internal standards was made. Resulting
tion was varied within new chemical limits in accordance
hardness profile were often quite far from required ones,
with Table 2, in particular Boron was added just to impro-
thus chemical composition was optimized by means of
ve hardenability.
Tab.2 - Modified chemical composition ranges (%): balance: Fe C
Mn
Si
P
S
Cr
Ni
Mo
Cu
Al
V
Nb
Ti
B
0.23÷ 0.31
1.10÷ 1.40
0.10÷ 0.40
≤0.030
0.025÷ 0.040
≤0.40
≤0.40
≤0.10
≤0.30
0.015÷ 0.040
0.003÷ 0.006
0.001÷ 0.003
0.045÷ 0.055
0.0025÷ 0.0035
Neural networks-based prediction of the Jominy profile
prediction of the different points of the profile has been
The Jominy profile prediction is performed by exploi-
performed both on the basis of some literature results
ting a set of 2-layers standard perceptron-based Fe-
related to the influence of chemical elements on the Jo-
ed-Forward ANNs according to the hierarchical structu-
miny profile and of a preliminary statistical analysis ai-
re, which is depicted in Figure 2.
ming at highlighting the correlation between the poten-
The inputs and the number Nh of neurons in the hidden
tial input variables and the target hardness values(12,13).
layer are different for each ANN and include the most correlated chemical elements and eventually the previously predicted hardness of other points of the Jominy profile. The selection of the input variables to be used for the
La Metallurgia Italiana - January 2020
pagina 61
Memorie scientifiche - Trattamenti Termici
Fig.2 -Sequential structure of the Jominy profile predictor Table 3 shows in detail the list of the inputs which are
points of the Jominy profile are predicted in sequence
fed to each neural network for the prediction of specific
in order to suitably exploit the relations between them.
Jominy profile points: each row of the table corresponds
Due to the dependencies between single points expres-
to one point and reports both the included chemical ele-
sed in Table1, the order of prediction is not the natural
ments and the predicted hardness values fed as input to
one, but the following order, which is more fruitful for
the ANN (where Ĵ i represents the hardness value estima-
the purpose of this work, is adopted: 1-2-3-4-5-14-15-
ted at i-th distance value). Table 1 also reports the num-
13-12-11-10-9-8-7-6, where each number represents
ber of neurons in the hidden layer N h of each ANN. The
the corresponding profile point.
Tab.3 - Inputs of the ANNs of the Jominy profile predictor Point
Input
Nh
Chemical elements
Predicted hardness values
J1
C, Mn, Si
-
2
J2
C,P,S,Cu
Ĵ1
2
J3
C, Cr, S, Ni
Ĵ1, Ĵ2
3
J4
C, Cu, Mn, Mo
Ĵ2, Ĵ3
3
J5
C, B, Ti, Nb
Ĵ3, Ĵ4
4
C, B, Ti, Nb
Ĵ2, Ĵ5, Ĵ7
4
C, B, Ti, Nb
Ĵ1, Ĵ2, Ĵ5, Ĵ10
4
C, Sn, Al, V
Ĵ4, Ĵ6, Ĵ7, Ĵ9, Ĵ10
4
C, S, P
Ĵ2, Ĵ3, Ĵ5, Ĵ11
4
C, S, P, Cu
Ĵ1, Ĵ2, Ĵ4, Ĵ5, Ĵ11
3
J11
C, S, Cr, Ni
Ĵ1, Ĵ4, Ĵ5, Ĵ12, Ĵ15
3
J12
C, Mn, Ni
Ĵ1, Ĵ3, Ĵ4, Ĵ13
3
J13
C, S, Cr, Ni, Nb,Ti, B
Ĵ1, Ĵ4, Ĵ5
3
J14
C, P, S, Cr
Ĵ1, Ĵ2, Ĵ3, Ĵ5
3
J15
C, P, S, Cr
Ĵ1, Ĵ3, Ĵ5
2
J6 J7 J8 J9 J10
La Metallurgia Italiana - gennaio 2020
pagina 62
Scientific papers - Heat Treatments The ANNs are sequentially trained by means of the stan-
its general validity.
dard back-propagation algorithm by exploiting 600 randomly selected samples from an experimental dataset
Results
composed of about 1100 samples, while the other sam-
Simulation results were analyzed to evaluate the good-
ples are used for validation. This dataset was available
ness of resulting Jominy curves in comparison with desi-
before the development of the analysis described in this
red hardness defined in the FCA standard.
paper, namely none of the data which have been used in
For each Jominy position the error was calculated com-
the present work have been exploited in the training of
pared to the average value of the specification limits,
the system. For the purpose of the analysis discussed
consequently the average error, calculated on the whole
here, the ANN-based Jominy predictor has been used
curve, was estimated.
simply as a â&#x20AC;&#x153;virtual hardenability test developerâ&#x20AC;? and the
Thanks to this analysis a series of optimal and promising
goodness of the achieved results further demonstrates
chemical compositions were defined.
Fig.3 - ANN-based prediction of the Jominy curves: best results
To validate results also a series of Jominy tests were per-
Figure 3, and the comparison of calculated errors was
formed in plant. As an example, some of the curves with
reported in Table 4.
the lower average error were reported in Figure 2 and
Tab.4 - Average error estimation for Jominy hardness profiles compared to FCA reference standard Test ID
Average error (HRc) Neural network-based prediction
Average error (HRc) Industrial tests
2
3.8
2.9
12
3.1
2.5
42
3.0
3.2
43
3.6
2.8
La Metallurgia Italiana - January 2020
pagina 63
Memorie scientifiche - Trattamenti Termici
Fig.4 - Jominy results on casting with optimal chemical composition predicted by the ANN-based predictor
Conclusion
tor could be a powerful tool to avoid a large number of
The present study was a preliminary investigation with
experimental industrial tests.
the aim of optimizing production process for suitable
3) The error that occurs with the ANN-based prediction
steels for low pressure cementation. Considering the
has the same order of magnitude of experimental data di-
first results the following consideration can be gathered:
spersion on this category of steels.
1) Usual chemical composition ranges are too wide to
4) The presented results lead to the development of ste-
obtain components in accordance with severe automo-
el for the specific application as a subgroup of MnCrB5
tive reference standards, that is why a specialization is
series. These steels have the certain advantage of a gre-
required.
at economical convenience thanks to the absence of
2) A properly trained neural network-based predic-
expensive alloy elements as Nickel or Molybdenum.
References [1]
Morgano E. Trattamenti Termici Innovativi e Processi Meccanici Superficiali. Seminary at University of Pisa. 2017.
[2]
Fiat Group Automobiles standard 52414. Acciai da Cementazione Acciai al Manganese – Cromo. 2010.
[3]
Vermeulen W.G., Van Der Wolk P.J., De Weijer A.P., Van der Zwaag S. Prediction of Jominy hardness profiles of steels using Artificial Neural Networks J. Mater. Engineering, 1996, 57–63.
[4]
Dobrazanski L.A., Sitek W. Application of neural network in modelling of hardenability of constructional steels. Jour. of Mat. Proc. Technology, 1998, 78, 59–66.
[5]
Dobrazanski L.A., Sitek W. The modelling of hardenability using neural networks. Journal of Materials Processing Technology. 1999, 92–93, 8–14.
La Metallurgia Italiana - gennaio 2020
pagina 64
Scientific papers - Heat Treatments [6]
Chan B., Bibby M., Holtz N. Predicting hardness with artificial neural networks. Canadian Metallurgic Quaterly. 1995, 34, 353–356.
[7]
Colla V., Reyneri L.M., Sgarbi M. Parametric characterization of hardness profiles of steels with neuro-wavelet networks. Lecture Notes in Computer Science, 1999, 1607, 606–614.
[8]
Colla V., Reyneri L.M., Sgarbi M. Neuro-wavelet parametric characterization of Jominy profiles of steels, Integrated Computer-Aided Engineering, 2000, 7(3), 217–228.
[9]
Valentini R., Colla V., Sgarbi M., Reyneri L.M. Parametric Jominy profiles predictor based on neural networks. Revista de metalurgia, 2005, SPEC. VOL, 314–317.
[10]
Cateni S., Colla V., Vannucci M., Vannocci M. Prediction of steel hardenability and related reliability through neural networks Proc. IASTED Int. Conf. on Artificial Intelligence and Applications, AIA 2013, 2013. 169–174.
[11]
ASTM A255-10. Standard Test Method for Determining Hardenability of Steel. 2018.
[12]
Cateni, S., Colla, V., Vannucci, M. General purpose input variables extraction: A genetic algorithm-based procedure GIVE a GAP, ISDA 2009 - 9th Int. Conf. on Intelligent Systems Design and Applications. 2009, 1278–1283.
[13]
Cateni, S., Colla, V., Vannucci, M. A genetic algorithm-based approach for selecting input variables and setting relevant network parameters of a SOM-based classifier, International Journal of Simulation: Systems, Science and Technology. 2011, 12(2), 30–37.
La Metallurgia Italiana - January 2020
pagina 65
Memorie scientifiche - Trattamenti Termici
Heat treatment routes utilizing intercritical annealing of automotive medium-mn steel sheets A. Grajcar, M. Morawiec
Low-carbon medium-Mn steel with an increased aluminium content was subjected to different heat treatments using dilatometry simulations. Samples of the 0.16C-5Mn-1.5Al steel were intercritically annealed between 700 and 800°C for duration of 1 hour. These heat treatment routes were designed to enrich the austenite in carbon to stabilize retained austenite after cooling the samples to room temperature. The austenite decomposition was monitored using the dilatometry. The martensite start temperatures were calculated and determined experimentally when the Ms was higher than the room temperature. The microstructures were investigated using light and scanning electron microscopy techniques. It was found that a heat treatment temperature affects substantially a thermal stability of the austenite and other microstructural features of the multiphase microstructures. The effect of C partitioning during the intercritical annealing on the austenite stability is discussed.
KEYWORDS: INTERCRITICAL ANNEALLING - MEDIUM-MANGANESE STEEL RETAINED AUSTENITE - TRANSFORMATION KINETICS - DILATOMETRY
Introduction
steels the main route for their manufacturing is cold rol-
A new generation of steels for the automotive industry
ling after which a simple heat treatment takes place [5-
is undergoing a dynamic development because of high
8]. A typical Mn content in the cold-rolled medium-Mn
demands on various environmental, safety and fuel eco-
steels is from 7 to ca. 11%. From the industry point of
nomy aspects.
view it increases the price of finished sheet products.
These demands coming from different regulations put a
Moreover, there are various technological problems
pressure on the automotive and steel industries to ma-
related to hot-dip galvanizing, instable flow behavior,
nufacture a wide range of steels with a good combination
weldability, etc.
of mechanical properties [1-4]. The special interest is in advanced high strength steels characterized by a multiphase microstructure. Recently, the focus is on medium-manganese steels, which include from 3 to 12% Mn. The idea for these steels is to obtain a good combination of strength and ductility, together with a relatively low price related to the chemical composition and
Adam Grajcar, Mateusz Morawiec Silesian University of Technology,
Faculty of Mechanical Engineering,
Institute of Engineering Materials and Biomaterials, Gliwice, Poland
adam.grajcar@polsl.pl
manufacturing process. In case of medium-manganese
La Metallurgia Italiana - gennaio 2020
pagina 66
Scientific papers - Heat Treatments Far less attention is paid to the hot rolling process for
steel is as follows: 0.16% C, 4.7% Mn, 1.6% Al, 0.2% Mo,
this grade of steels [8-11]. One can assume that the man-
0.2% Si.
ganese addition together with carbon should increase
The steel was hot rolled and air cooled. The initial mi-
the stability of austenite and enhance the TRIP (TRan-
crostructure is martensitic due to the strong hardenabi-
sformation Induced Plasticity) effect. However, some
lity effect of manganese.
authors reported [12-14] that the increase of Mn content
The intercritical annealing was applied for hot rolled and
to the chemical composition can result in an opposite
machined samples.
effect.
The heat treatment was performed by the means of di-
The reason is that the Mn addition decreases the diffu-
latometry, which at the same time gives the information
sional enrichment of the austenite in carbon (a thermo-
on a phase transformation kinetics upon heating and
dynamic effect).
subsequent cooling.
In hot rolled steels it is also very important to know how
The samples in a shape of rods of 5 mm diameter and 10
manganese influences the austenite decomposition ki-
mm length were heated up to the desired temperatures
netics.
in a range from 700 to 800°C at a heating rate of 3°C/s,
From our preliminary tests this is clear that manganese
held for duration of 1 hour and cooled to room tempe-
shifts the ferritic and bainitic regions to very long times.
rature at the rate of 60°C/s. This heat treatment approach
In case of the presented lean-Mn steel obtaining the bai-
is presented in Fig. 1.
nitic transformation during isothermal holding at diffe-
The whole process was carried out in vacuum. After
rent temperatures was impossible even after 5 hours.
the heat treatment the samples were cut in the middle
This means that the redistribution of C to the austenite
part of the rod for the preparation for light and scanning
can not take place. Hence, the different approach of heat
electron microscopy.
treatment for this kind of hot-rolled steel was applied in
The samples were grinded with the use of 220, 500, 800
the present study.
and 1200 grinding papers and polished with the use of 3 and 1 µm diamond solution.
Experimental procedure
Next, the samples were etched using 5% Nital. The Vi-
The chemical composition of the analyzed medium-Mn
ckers hardness was measured with the load of 9.81 N.
Fig.1 - The heat treatment routes applied for the investigated steel
La Metallurgia Italiana - January 2020
pagina 67
Memorie scientifiche - Trattamenti Termici Results and discussion
temperature for the samples intercritically annealed at
The first step of the analysis was the dilatometry.
700°C.
The dilatometric results present how the phase transfor-
This is due to the different state of the austenite at the
mation kinetics changes at different intercritical annea-
intercritical region expressed by the lever rule.
ling temperatures.
At 800°C the microstructure contains a large austenite
In case of the presented results, the change of marten-
fraction but with the smallest carbon content.
sitic start temperature indicates the different stability of
Thus, the austenite does not have the sufficient thermal
austenite upon cooling to room temperature. First of all,
stability and transforms to martensite during a cooling
the martensite start temperature of the steel was calcu-
step.
lated using the mucg83 software developed by Bhade-
When the annealing temperature decreases the amount
shia [15].
of the austenite decreases because a larger volume
The calculated temperature was equal to 335°C.
fraction of ferrite is formed from the initial martensite.
This was the same value as determined by the experi-
Hence, the martensite start temperature decreases to
mental approach.
190 and 105°C, respectively for the samples held at 760
Next the dilatometric analysis was carried out to deter-
and 720°C.
mine the influence of intercritical annealing temperature
This is because the lower intercritical temperatures al-
on the M s temperature.
low for the higher carbon content to be enriched in the
The change in M s temperature indicated a change in the
remaining austenite.
stability of the austenite. This took place due to an enrichment of the austenite in carbon during ferrite formation in the intercritical region. These dilatometric results are presented in Fig. 2. It can be seen that the M s temperature decreases gradually together with the decrease of the intercritical temperature from 800 to 700°C. For example, the M s = 245°C at 800°C and is below room
Fig.2 - Dilatometric analysis during cooling to room temperature after intercritical annealing at 700-800°C
La Metallurgia Italiana - gennaio 2020
pagina 68
Scientific papers - Heat Treatments
Fig.3 - LM microstructures after intercritical annealing at: a) 800°C, b) 760°C, c) 720°C, d) 700°C
The sufficient enrichment of the austenite in carbon took place only at the temperature of 700°C. For this treatment there was no martensite start in the dilatometric curve (Fig. 2).
Fig.4 - SEM microstructures after intercritical annealing at: a) 800°C, b) 760°C, c) 720°C, d) 700°C; F - ferrite, M - martensite, RA - retained austenite
La Metallurgia Italiana - January 2020
pagina 69
Memorie scientifiche - Trattamenti Termici After the dilatometric tests, the microstructure investi-
The most important phase present in the microstructure
gation was carried out. The microstructures from the li-
is retained austenite, which fraction increases together
ght microscopy (LM) are presented in Fig. 3. It is clear
with decreasing the intercritical temperature. At 700°C
that the microstructures inherit the martensite morpho-
there is no fresh martensite in the sample (Fig. 4d). Hen-
logy. It is difficult to assess if the martensite transformed
ce, this is an optimal heat treatment temperature.
into austenite. The difference in a ferrite content can
The last step of the analysis was hardness measuremen-
only be visible in these LM micrographs. It is difficult to
ts. Six tests were carried out on each sample, after whi-
separate ferrite at 800°C (Fig. 3a). This phase is clearly
ch the average value was calculated. The results of the
seen at 760°C as thin layers between martensite laths
hardness analysis are presented in Fig. 5. The hardness
(Fig. 3b). At 720 and 700°C the amount of the ferrite is
increases at the higher annealing temperatures. This is
larger. Moreover, some ferrite grains are more granular.
in good correlation with the dilatometric and micro-
The higher fraction of the ferrite allowed to enrich the
structure investigations. At 800°C, the highest hardness
austenite in carbon.
was measured. This corresponds to the large fraction
A higher magnification in Fig. 4 allowed for more detai-
of fresh martensite formed between 245 and 20°C. The
led analysis of the microstructure. It can be seen that the
fresh martensite formed during cooling does not under-
presented laths of ferrite are very thin; some of them be-
go any tempering. This is expressed as its highest hard-
low 100 nm in width. The width of the austenite and ferri-
ness (ca. 460 HV). When the temperature of the annea-
te decreases as the intercritical temperature is reduced.
ling decreases the amount of the intercritical austenite
The separation of individual laths of austenite and ferrite
and resulting fresh martensite decrease. Moreover, the
is also stronger at the lower intercritical temperatures.
amount of retained austenite increases with the decre-
Moreover, the martensite morphology is not the same
asing temperature. At 700°C no fresh martensite is for-
in all samples. The various volume fraction of fresh mar-
med during cooling to room temperature, as the dilato-
tensite can be observed at different temperatures. Ac-
metry test shows (Fig. 2).
cording to the dilatometric analysis in a 720-800°C range
Hence, the lamellar mixture of the highest volume
some poor-enriched carbon austenite transforms into
fraction of ferrite and austenite shows the lowest hard-
martensite during cooling (it is called fresh martensite).
ness (ca. 310 V).
Fig.5 - Hardness for different intercritical annealing temperatures
La Metallurgia Italiana - gennaio 2020
pagina 70
Scientific papers - Heat Treatments Conclusions
gest ferrite fraction and resulting carbon enrichment of
Presented results showed that intercritical annealing is
the austenite. At higher temperatures triple ferrite-au-
a simple heat treatment, which can be applied for lean
stenite-fresh martensite microstructures were formed.
medium-Mn Al-alloyed steels. The dilatometric analy-
The martensite start temperatures were above room
sis indicated that it is possible to stabilize the austeni-
temperature being the reason for the fresh martensite
te during the intercritical annealing through its carbon
formation. The hardness was the highest for the higher
enrichment from the formed ferrite. The microstructu-
annealing temperature, i.e., for the corresponding lar-
res consist of different portions of lamellar ferrite, fresh
gest volume fraction of fresh martensite. The lowest
martensite and retained austenite. The ferrite fraction in-
hardness showed the sample with the duplex ferrite-au-
creased as the interctitical temperature decreased from
stenite microstructure, with the highest potential for the
800 to 700°C.
automotive sheet steel application.
The beneficial very thin duplex ferrite and austenite mi-
Further studies are needed to assess a volume fraction
crostructure was only obtained at 700°C due to the lar-
of retained austenite and its carbon content.
Acknowledgement The financial support of the National Science Center, Poland, is gratefully acknowledged, grant no. 2017/27/B/ST8/02864.
References [1]
Lehnhoff GR, Findley KO, De Cooman BC. The influence of silicon and aluminium alloying on the lattice parameter and stacking fault energy of austenitic steel. Scripta Materialia 2014;92:19-22.
[2]
Zhao JZ, Mesplont C, De Cooman B.C. Quantitative analysis of the dilatation during an isothermal decomposition of austenites. Materials Science and Engineering A 2002;332:110-6.
[3]
Steineder K, Krizan D, Schneider R, Beal C, Sommitsch C. On the microstructural characteristics influencing the yielding behavior of ultra-fine grained medium-Mn steels. Acta Materialia 2017;139:39-50.
[4]
Kang S, Speer JG, Krizan D, Matlock DK, De Moor E. Prediction of tensile properties of intercritically annealed Al-containing 0,19C-4.5Mn (wt%) TRIP steels. Materials and Design 2016;97:138-46.
[5]
Dutta A, Ponge D, Sandlobes S, Raabe D. Understanding hot vs. cold rolled medium mangenese steel deformation behavior using in situ micoscopic digital image correlation. Materials Science Forum 2018;941:198; doi:10.4028/www.scientific.net/ MSF.941.198.
[6]
Benzing JT, Kwiatkowski da Silva A, Morsdorf L, Bentley J, Ponge D, Dutta A, Han J, McBride JR, Van Leer B, Gault B, Raabe D, Witting JE. Multi-scale characterization of austenite reversion and martensite recovery in cold-rolled medium-Mn steel. Acta Materialia 2019;166:512-30.
[7]
Cheng P, Hu B, Liu SL, Guo H, Enomoto M, Shang CJ. Influence of retained austenite and Cu precipitates on the mechanical properties of a cold-rolled and intercritically annealed medium Mn steel. Materials Science and Engineering A 2019;746:41-9.
[8]
Liu C, Peng Q, Xue Z, Wang S, Yang C. Microstructure and mechanical properties of hot-rolled and cold-rolled medium-Mn trip steels. Materials 2018;11:2242; doi:10.3390/ma11112242.
La Metallurgia Italiana - January 2020
pagina 71
Memorie scientifiche - Trattamenti Termici [9]
Timokhina IB, Hodgson PD, Pereloma EV. Effect of microstructure on the stability of retained austenie in transformation-induced-plasticity steels. Metallurgical and Materials Transactions A 2004;35A:2332-41.
[10]
Liu C, Peng Q, Xue Z, Deng M, Wang S, Yang C. Microstructure-tensile properties relationship and austenite stability of a Nb-Mo micro-alloyed medium-Mn trip steel. Metals 2018;8:615; doi:10.3390/met8080615.
[11]
Qi X, Du L, Hu J, Misra RDK. Effect of austenite stability on toughness, ductility and work-hardening of medium-Mn steel. Materials Science and Technology 2018; doi:10.1080/02670836.2018.1522088.
[12]
Grajcar A, Kilarski A, Kozłowska A, Radwański K. Microstructure evolution and mechanical stability of retained austenite in thermomechanically processed medium-Mn steel. Materials 2019;12:501; doi:10.3390/ma12030501.
[13]
Grajcar A, Kilarski A, Kozłowska A. Microstructure-property relationships in thermomechanically processed medium-Mn steels with high Al content. Metals 2018;8:929; doi:10.3390/met8110929.
[14]
Haupt M, Dutta A, Ponge D, Sandlobes S, Nellessen M, Hirt G. Influence of intercritical annealing on microstructure and mechanical properties of a medium manganese steel. Procedia Engineering 2017;207:1803-8.
[15]
Bhadeshia HKDH. Thermodynamic analysis of isothermal transformation diagrams. Metal Science 1982;16:159-65.
La Metallurgia Italiana - gennaio 2020
pagina 72
Industry news - Heat Treatments
New environmental-friendly developments and applications of the pulsed plasma heat treatment technology in the field of engine, transmission and power train engineering A. Gebeshuber, V. Strobl, T. Müller
Nowadays nitrided gears have been successfully introduced for applications such as steering gears in engines, internal gears in transmissions and other not very highly loaded applications. In future, the fast growing market for top performance gears in the vehicle industry with its high load demands, is a particularly promising expansion area. Especially here is a need for reliable fatigue data from optimized material-plasmanitriding-combinations. To generate such fatigue data a step by step approach with specially designed specimens and test procedures has been chosen. This paper discusses present results and gives an outlook on the steps ahead. The first part of this investigation focuses on bending fatigue strength of the gear tooth, while other failure modes, such as pitting, will be tested in the second part of the project.
KEYWORDS: PLASMA NITRIDING - FATIGUE DATA - GEAR DESIGN RESIDUAL COMPRESSIVE STRESSES - DISTORTION
Introduction
has induced the development of new technologies in
Over the last 15 years, there has been a constant rise in
classical carburizing, such as low pressure carburizing
the amount of hard machining - mainly grinding of tooth
with gas quenching. But all these technologies require
flanks - required after carburizing. At the transmission factory of VW Kassel for example, there was an increase from 2 % in 1998 to as much as 14 % in 2005 [1]. The cost of such post heat treatment machining may thus amount up to 40% of total unit costs. The gear industry approaches this problem with diffe-
Andreas Gebeshuber, Volker Strobl, Thomas Müller Rübig GmbH&Co.KG, Wels, Austria
rent strategies. The cost saving pressure in gear production does not only affect the machining processes, but also calls for new possibilities of heat treatment, and
La Metallurgia Italiana - January 2020
pagina 73
Attualità industriale - Trattamenti Termici high temperatures and martensitic transformation as a
additional new tools and machining methods have been
metallurgical mechanism in order to increase both har-
developed, so that the availability of high strength steels
dness and fatigue life.
suitable for nitriding and the production process invol-
Both factors - high temperatures combined with fast
ved are no longer constraints upon nitriding in the gear
quenching, and volume change by martensitic transfor-
industry.
mation - lead to considerable distortion and therefore
Because of the intensive hard machining after carburi-
to an inevitable machining operation when high profile
zing required for high performance gears, and even ta-
and surface quality and narrow geometric margins need
king into account the higher alloyed steels, soft machi-
to be met.
ning of high strength steels and longer process times in nitriding, considerable savings in the overall costs can
Nitriding is a hardening process which does not use
still be achieved if the hard machining after carburizing
martensitic transformation, but precipitation hardening
is no longer necessary and can thus be dropped.
as a mechanism. Furthermore, nitriding is carried out at fairly low tem-
Investigation procedure
peratures in the range of 500 to 600°C, while fast quen-
This paper presents some results of a systematic ap-
ching in a fluid medium is not necessary.
proach to generate reliable data on plasmanitrided high
Plasma nitriding is a special form of nitriding, where the
strength steels for gear and drive application.
necessary free nitrogen is generated by a physical process - the low pressure plasma, whereas the dissocia-
Potential for plasmanitriding gears
tion of cyanates in saltbath nitriding, or of ammonium in
The potential of plasmanitriding regarding load capacity
the gas nitriding process, are chemical reactions. This
of highly stressed gears can be seen from Schlöterm-
difference in generating the free nitrogen has several ad-
ann´s investigations [2].
ditional advantages, such as
Extensive fatigue tests with gas or plasmanitrided gears
• environmental friendliness (green process)
have been conducted with different steels and material
• easy integration into production line
conditions and have been compared with the fatigue
• high accuracy and process stability
data according to DIN 3990 [3].
• low gas consumption
The conclusion of the tests is that the norm values are
• possibility of easy partial nitriding by mechani-
very low in comparison to the test results.
cal masking. A test to compare case hardened and nitrided gears was The low distortion of the nitriding process in general
conducted by H. Mallener [4].
and the above mentioned additional advantages make
This showed that nitrided gears present reduced tooth
the plasma nitriding process one of the most promising
root strength, especially in the low cycle area.
alternatives in heat treatment for high performance ge-
Another investigation concerning the load carrying ca-
ars, when low overall costs, high quality in production,
pacity of nitrided steel gears by L. Albertin and H. Winter
integration into the machining floors and pollution pre-
conducted at the Gear Research Center (FZG, Munich)
vention are an issue.
has produced similar results [5].
However, nitriding has never been used on a broader
A recent research project about bending fatigue stren-
range in the production of highly stressed gears.
gth of case hardened and nitrided gears by Prof. P. Da-
The reason is not only the nitriding process itself, but
voli [6] showed a different result. His concluding sta-
also the higher costs of the alloyed steel required, and
tement: “The damage line of the two nitriding steels is
the need of machining of teeth up to a strength level of
quite high, not far from the Wöhler line; which implies
1500 MPa and even more. In recent years, new steels and
a good overload capacity for these two nitriding steels.
La Metallurgia Italiana - gennaio 2020
pagina 74
Industry news - Heat Treatments As a final statement, the good bending load carrying ca-
core strength, surface hardness, white layer thickness,
pacity of properly heat treated nitriding steels has been
nitriding depth and residual compressive stress profiles
confirmed by the tests.“
were applied to four different high strength steels in the
By matching material selection with proper heat treat-
course of plasmanitrided processes especially develo-
ment and optimized plasmanitriding, it should be pos-
ped for such applications.
sible to obtain good fatigue values in the low cycle area as well.
According to DIN 3990-5 the tooth bending fatigue strength ranges from 270 up to 550 MPa (Fig. 1), depen-
Results
ding on material and applied surface treatment (nitriding
The scope of the investigations was the determination of
or case hardening).
fatigue behaviour of different high strength plasmanitri-
Schlötermann already showed in 1988 [2], that nitrided
ded steels.
steel can yield higher values than case-hardened steel -
During the research program, different combinations of
in contradiction to the values acc. to DIN 3990-5.
Fig.1 -Bending fatigue strength: comparison of literature data and development goal RÜBIG
This was the motivation for Rübig to start further inve-
were compared with those of case hardened specimen.
stigations with the goal of narrowing the scatter band by rising the lower band (Fig. 1).
The results show, that proper combination of steel and plasma-nitriding parameters result in bending fatigue
In a first step, during which the influence of material
values that are at least equivalent to case-hardened ste-
choice and plasma nitriding was to be studied, all fatigue
el (see Fig. 2).
tests were performed at FZG in Munich on gears with a modulus of 2,75. So it was possible to determine the tooth bending fatigue of a typical gear tooth geometry. The fatigue data
La Metallurgia Italiana - January 2020
pagina 75
Attualità industriale - Trattamenti Termici
Fig.2 -Bending fatigue strength: comparison of results RÜBIG and literature data [8]
From our investigations it is obvious that the residual
in the diffusion zone.
compressive stress profile is the dominating factor on fatigue life in all cases where surface hardness compri-
The next steps in the project are:
ses a range between 700 and 1000 HV and core strength
• Conduct fatigue tests on high performance plasmani-
between 1000 to 1600 MPa.
trided gears to evaluate flank bearing properties compared to case-hardened gears (in progress)
Just as with all the other surface hardening processes,
• Evaluation of the economic benefit due to shorter
such as carburiz-ing or induction hardening, it is the ma-
manufacturing process (pic. 3) of plasmanitrided gears
ximum value of compressive stress in the surface area
compared to case-hardened gears (in progress
and the depth of the stress profile that influences the fatigue limit positively.
The latter step is of great interest since the plasmanitriding process features very little distortion due to the
Generally, the fatigue data for nitrided steels were in the
hardening mechanism by precipitation hardening.
same range as for comparable carburized samples, with high compressive residual stresses the fatigue data are
Interim results show, that there is a high potential for the
even exceeding the values for carburized sam-ples.
substitution of case-hardening by plasmanitriding and can lead to a reduction of total costs for gear manufa-
So far we could not establish any difference in fatigue
tur-ing.:
life values that would relate to the microstructure of the white layer or to the microstructure in the diffusion zone. This may be due to the fact that all samples were, just as for the data reported in literature, nitrided with little or no white layer and with hardly any visible precipi-tation
La Metallurgia Italiana - gennaio 2020
pagina 76
Industry news - Heat Treatments
Fig.3 -Manufacturing process of gears: “case hardening” vs. “plasmanitriding”
We are convinced that upon completion of this com-
rial-plasmanitriding combinations to match maximum
prehensive investigation, we will be able to give ba-
gear load ca-pacity for a wide range of standard and spe-
sic rec-ommendations for choosing optimized mate-
cial gear applications.
References
[1]
Weißner, R., Fenstermann, J.: Anforderungen, Stand der Technik und Perspektiven von Verzahnungen im Automobilbau. Dresden; Germany: 2. Deutsch-Italienische Near-Net-Shape-Tagung, 2005.
[2]
Schlötermann, K.: Auslegung nitrierter Zahnradgetriebe, Untersuchungen zu den Auswirkungen unterschiedlicher Nitrierparameter auf den Werkstoffzustand und die Tragfähigkeit von Zahnrädern. Dissertation, RWTH Aachen, 1988.
[3]
Norm DIN 3990 Teil 5 1984: Tragfähigkeitsberechnung von Stirnrädern, Dauerfestigkeitswerte und Werkstoffqualitäten
[4]
Mallener, H., Schulz, M.: Nitrieren von Zahnrädern als Alternative für das Einsatzhärten?. HTM Z . Werkst. Wärmebeh. Fertigung 48, (1993) 3, pp. 166 – 171
[5]
Albertin, L., Frolich, R. L., Winter, H.& Höhn B.-R., Michaelis, K.: Load Carrying Capacity of Nitrided Gears. AGMA FTM; 94FTM4, 1993
[6]
Davoli, P., Boniardi, M., Longoni, C.: Bending Fatigue Resistance of Case Hardened and Nitrided Gears. München; Germany: International Conference on Gears, 2005
[7]
Trubitz, P., Kaminsky, T., Brusky, U., Kaiser, B., Adelmann, J.: Schwingfestigkeit nitrierter Proben. Vorhaben Nr.160, Lebensdauervorhersage nitrierter bauteilähnlicher Proben mit Hilfe normierter Wöhlerstreubänder – Einfluss der Randschicht. Forschungsheft FKM; Heft 197, 1995
[8]
Stenico, A.: „Werkstoffmechanische Untersuchungen zur Zahnfußtragfähigkeit einsatzgehärteter Zahnräder“, Dissertation, TU München, 2007
La Metallurgia Italiana - January 2020
pagina 77
AttualitĂ industriale - Trattamenti Termici
Liquid Nitrocarburizing treatment on valves in internal combustion engines M. Legrand, V. Monteux
In a strong position because of more than 60 yearsâ&#x20AC;&#x2122; experience in the area of thermochemical treatments, HEF Group, world leader in liquid ionic nitrocarburizing processes, has developed and optimized in that field his own treatments family. They are commercialized under the name CLIN (Controlled Liquid Ionic Nitriding) keeping as standard a commune philosophy: offering a solution in continuous development to support its customers with highest quality and robustness, still in advance in front of environment constraints. Thus, liquid nitrocaburizing became and remains an essential surface treatment solution, globally established, well known and recognized as only-one efficient solution for specific automotive components such as valves. On that specific component, the new environmental standards for a better engine efficiency and the latest constraints on surfaces technologies such as hard chromium coating, lead to new developments and designs for moving parts in internal combustion engines. These activities are the challenges of engine manufacturers. Liquid nitrocaburizing treatment is now considered as industrial mature technology of treatment of new automotive valves design, with pre-localized induction hardening, in ferrous-chromium alloys and / or nickel base alloy known for their high resistance to thermal oxidation. Furthermore, this treatment is nowadays required as a technically and economically competitive alternative to hard chrome plating treatment. The latter is strongly penalized by the restrictions imposed by the new environmental laws (REACH ...) and risk eventually being no longer competitive (high cost price).
KEYWORDS: NITRIDING TREATMENT - SALT BATH - LIQUID IONIC - ENGINE EFFICIENCY - FATIGUE RESISTANCE - ALTERNATIVE TO CHROMIUM PLATING - VALVE
Introduction In the recent years, automotive industries are facing a complete revolution in combustion engine design to face high constraints on their efficiency - CO2 emission reduction - Downsizing,
M. Legrand, V. Monteux HEF Durferrit, Andrezieux Boutheon, France
- Hybrid version development. To withstand these constraints, the valves designs
La Metallurgia Italiana - gennaio 2020
pagina 78
Industry news - Heat Treatments have been reviewed taking into consideration
- Higher stress on tip due to repetitive contact with
- Highest temperature into chamber (the valve seat
cam, tappets, follower…. Induction hardening of the
must be able to resist the corrosive effects of hot
tip become standard and obvious.
exhaust gases from the engine). New material have been successfully tested such as high alloyed iron base material or nickel base material.
INLET VALVE • • •
Larger head diameter Lower operating temperature than exhaust (500-800°C) Material choice - Low carbon martensitic steel - X45 Cr Si 9 3 (9% Cr) can be hardened for durability on tip
EXHAUST VALVE • • • •
Higher operating temperature (800-1200°C) Corrosive environment Lubrication issue Material choice - usually bi metallic valves - X45 Cr Si 9 3 (9% Cr) (can be hardened for durability on tip) on tip and stem side - Stainless steel 21% Cr on head, - Or Stainless 30% Cr 15% Ni on head - Or nickel base material on head.
Meanwhile, their standard surfaces treatments techno-
experience in the area of thermochemical treatments,
logies, such as chromium plating, used for decade are
HEF Group, world leader in liquid ionic nitro carburizing
suffering of new environmental pressure, inviting them
processes, has developed his own new formulation cal-
to consider alternatives in their new challenges.
led ARCOR® DT.
To succeed in the replacement of chromium plating on valves made of high alloyed elements, nitriding te-
This process reaches unexpected results without any
chnologies players have to adjust temperature of their
pre acid etching, with a high efficiency (short treatment
nitriding atmosphere in the lowest side of their opera-
time although low temperature), homogeneity, repro-
ting range (approx. 500-540°C to limit relieving of the
ducibility, robustness (whatever the material is) and cost
pre-localized induction hardening), keeping efficiency
efficiency (limited consumables ratio on total operating
in reduction of passive surfaces on such material (to
cost by ensuring low consumption, although low tem-
allow homogeneous and quick growing of the nitride
perature).
layers). ARCOR® DT is a low temperature nitro carburizing proARCOR DT: new formulation development
cess, operating in a temperature range (500°C to 540°C)
In a strong position because of more than 60 years’
lower than tip induction hardening temperature. Hence
La Metallurgia Italiana - January 2020
pagina 79
Attualità industriale - Trattamenti Termici re-hardening after nitriding is no more required contrary
Results and conclusion
to some others options than TUFFTRIDE, TENIFER or
ARCOR DT consists of the well-known standard se-
others nitro carburizing technologies.
quence: - Alkaline degreasing and rinsing operations,
Its efficiency comes from its special chemistry offering
- Preheating into air circulated oven
lower melting point (hence higher fluidity. Furthermore,
- Nitriding at low temperature (500°C to 540°C) for a tre-
its optimized oxo-basicity offers high nitriding rate even
atment time lower than 3 hours
at lower temperatures and even on high alloyed steel
- Cooling in industrial water,
and nickel base material.
- Washing in industrial water - Drying,
Finally, following advantages have to be pointed out
- Mechanical finishing such as blasting using thin metal-
- Cleaner valves after nitriding
lic powder.
- requiring only low pressure shot blasting after nitriding - ideally suited for hollow-stem valves, thin-stem and
In the following table, metallographic observation un-
valves with low ‘stem-bowing’ tolerance.
der optical microscopy are being given on different ma-
- greener technology.
terials.
Material
Metallographic observation under optical microscopy
Applications
for monometallic inlet valve and on tip side on bi metallic exhaust valve
X 45 Cr Si 9 3 (QT)
Nital etching, magnification x500
(9%Mn ; 4,5%Ni ; 21%Cr ; Fe balance) Used on seat side for bi metallic exhaust valve
X 50 CrMnNiNbN 21 9
Marble etching, magnification x1000
La Metallurgia Italiana - gennaio 2020
pagina 80
Industry news - Heat Treatments
(30%Ni, 15%Cr, Fe balance) Used on seat side for bi metallic exhaust valve
ECMS 3015
Nital etching, magnification x500
(30%Cr, 3%Fe, Ni balance) Used on seat side for bi metallic exhaust valve
Nimonic 80A
Marble etching, magnification x500
As a conclusion, the new formulation of Liquid Nitrocar-
- Liquid nitrocarburizing can be operated at a temperature,
burizing ARCOR DT becomes quickly the relevant solu-
which allow keeping the hardness, brought by induction
tion, showing the highest advantages for treatment on new
hardening, without acting as a stress relieving and without
generation valves:
negative consequence on produc-tion costs (limited consumables consumption; short treatment time).
- Liquid nitrocaburizing treatment improves the resistance to wear and the anti-friction qualities by in-creasing the surface hardness along the valve stem, while retaining the corrosion resistance of the material at the valve seat. - Liquid nitrocarburizing is able to treat simultaneously the different material even ones with higher al-loyed element contents.
La Metallurgia Italiana - January 2020
pagina 81
AttualitĂ industriale - Trattamenti Termici
Non-destructive hardness / microstructure testing of heat-treated parts by mass production, with multiple frequency magnetic induction method A. Horsch
Increasing quality demands, new productliability regulations, as well as international market networks force manufacturers to take special measures encompassing thefield of material testing. Nowadays, specified tolerances are extremely small. Therefore processes have to include the conducting of 100% material, structure and hardening tests on a fast, reliable and simple basis. The technology applied must be the latest state of art, it must comply with maximum safety requirements and be economical.
Introduction
stations with 8 standard test frequencies and up to 16
Today, every company carries out spot checks during
test positions have significantly increased the efficiency
manufacturing, in order to guarantee the quality of pro-
of this test method.
ducts. Within the framework of control one of the most
The field of application ranges from testing the micro-
important problems is to ensure repeatability of quality.
structure, hardening and case depth to material min-
Quality control faces major problems within this area;
gle. One of the outstanding advantages of modern eddy
these might start with supply of the material or products.
current testing systems using PMFT (Preventive Multi
Parts properties; that is structure, hardness or case dep-
Frequency Testing) is the ability to detect unexpected
th, which are determined by the heat treatment process,
faults.
are subject to variations. To ensure the correct heat treatment results, a fast and efficient test method is nee-
Magnetic test methods have been used since the be-
ded, it should be non-destructive.
ginning of the 1940s. Springer [1] and Maercks [2] have
It must be clear that heat treatment is not a standard sta-
already made their publications in 1941. Figures 1 and 2
tistical process, each heat treatment batch is a separa-
show the permeability measuring device he uses and a
te production that is not related to the previous one. In
magnetic coil sorting device.
a continuous belt furnace a quasi-statistical process is given, but here too a spot check can only provide limited information about the entire production. The heat treatment company provides only a process which can be verified by spot check. This does not say whether all
A. Horsch Arnold Horsch e.K., Remscheid, Germany
parts also comply the specification. Computer based multi-frequency eddy current testing
La Metallurgia Italiana - gennaio 2020
pagina 82
Industry news - Heat Treatments
Fig.1 - Magnetic test system from Springer, laboratory setup [1]
Fig.2 - Magnetic test system from Springer, industrial test system [1]
Both describe successful sorting tests for the sorting
checks in order to conclude the general nature of the
of differences in hardness of high-speed steel and har-
quality from the test data. Table 1 shows typical tested
dness depth differences after case hardening on gears.
parts [3].
WHY 100% TEST? Variation in product usually is in accord with statistical analysis and can be predicted or estimated. For this reason, it is sufficient to make a certain number of spot
La Metallurgia Italiana - January 2020
pagina 83
AttualitĂ industriale - Trattamenti Termici Tab.1
Application Examples of Material Testing by Means of Eddy Current Preventive Multi-Frequency Testing (PMFT) Component f. e.
Parameters to be tested
Bearing parts
case depth, structure and hardness pattern
Axle components
case depth, structure and hardness pattern
Pinion pins and axles
case depth, structure and hardness pattern
Linear guide components
case depth, structure and hardness pattern
Steering racks and the like
case depth, structure and hardness pattern
Gears
case depth, structure and hardness pattern
Cam Shafts
case depth, structure and hardness pattern
Bolts, Screws, Nuts
Structure, hardness, decarburisation
and much more
Problems develop if during manufacturing some occur-
unexpectedly. Table 2 lists possible hardening faults [3].
rences arise which are not subject to standard statistical
We also have to keep in mind that materials are not per-
distribution. First, one needs to know what may happen
fect and can have flaws nobody expected.
Tab.2
Possible hardening errors during Heat Treatment Which parameter was incorrect?
Austenitizing temperature
Austenitizing time
Quenching
In which way was it incorrect?
What are the effects of this?
too high
Overhardening, incorrect structure martensite + residual austenite
too low
Underhardening, incorrect structure martensite + bainite + ferrite
too long
Overhardening, case too high, incorrect structure martensite + residual austenite
too short
Underhardening, shallow case, incorrect structure martensite + bainite + ferrite
too fast
Incorrect structure, cracks martensite + residual austenite
too slow
Incorrect structure martensite + bainite + ferrite
formation of vapor bubbles too low
soft spots not defined Hardness too high
too short -
hardness too low
too long
hardness too high
Tempering time
La Metallurgia Italiana - gennaio 2020
pagina 84
Industry news - Heat Treatments
too slow
shallow case, misplaced case, austenitizing time too short
too fast
case too high, misplaced case, austenitizing time too long
Rate of feed by induction hardening
Damaged inductor
undefined
undefined
Mal-positioning by induction hardening
undefined
unsymmetrical hardening pattern, overheating, melting
How does eddy current testing works [4]
the transport from the hardening station to the test sta-
Eddy current testing work with coils, generators,
tion takes only a few seconds.
ac-current and ac-voltage, frequencies, field strength
Faulty parts, are immediately detected and separated,
and induction law. Eddy current test examine them for
providing an enormous savings in time and costs!
their metallurgical microstructure, thus for their me-
The energy for eddy current testing is very small, in the
chanical features like hardness, case depth or alloy.
milliwatt range. Field strength is low and permeability is
The eddy current test does not provide absolute values
in the range of the initial permeability. Test frequencies
(e.g. “56 HRC” or “2.6 mm case depth”). The eddy cur-
ranging from some Hz to some hundreds of kHz provide
rent test does detect fine differences in micro-structu-
information on undesired structures via the frequency
re with high sensitivity. In the production line, within a
dependent penetration depth of the eddy current and on
fraction of a second, a non-destructive 100 % test for
the formation of permeability. Very small electrical si-
micro structure
gnals require a very precise evaluation in order to assure
• case depth
their differentiation from ambient interferences. A small
• hardness run out,
drift from variation in temperature and high long term
• hardness pattern
stability are absolutely necessary. Digitization of input
• tensile strength
voltages immediately at the front end of the electronic
• carbon content
evaluation is of huge advantage.
• soft spots
Relative permeability [µr] is strongly affected by heat
• surface decarburisation
treatment. The amount of carbon and other alloy elements significantly influence the size and course of
is completed, and thus quick corrective reactions to any
the permeability curve. Generally, hardened (stressed)
variance from the specified structure can be realized.
structures have lower permeability than soft structures
With a suitable mechanical part handling arrangement,
(refer to graph of 0.78% C), see Figure 3.
Fig.3 - Relative permeability of different materials
La Metallurgia Italiana - January 2020
pagina 85
Attualità industriale - Trattamenti Termici Electrical conductivity [σ] is influenced only a little by
it. It is a „comparative test“. Values of OK parts (referen-
structure changes and alloy differences. Whereas, the
ce parts) which were presented to and stored in the in-
conductivity of carbon steel is < 10MS/m high-alloyed
strument beforehand are compared with the values of
chrome-nickel is about 1.3 MS/m. However, the tem-
currently produced parts.
perature coefficient is about 4 to 5 % per 10°C. Thus,
How can one get values? The alternating current i which
the temperature of the test part has influence on the test
flows through the red coil created a magnetic flux [B]
result and should be allowed to vary only negligibly (±
through the test part (grey in the sketch). The size of the
5°C).
magnetic flux and thus the size of the voltage u indu-
How can magnetic and electrical changes that correlate
ced in the blue coil is directly dependant on the electri-
with the mechanical properties be quickly and reliably
cal conductivity [σ] and on the magnetic „conductivity“
tested non-destructively?
(permeability [µr ]) of the test part in the coil, Figure 4.
Eddy current testing has proven to be well qualified for
Fig.4
The test part by ist σ and µr strongly influences the cou-
real faults never look like artificially created faults, fur-
pling between sender coil (red) and receiver coil (blue).
thermore only this one specially calibrated fault can be
Thus voltages induced in the receiver coil imply the
found. Another fault will not be found and go into pro-
structure, e.g. to verify correct hardness, case depth,
duction and will be installed e.g. in an engine.
core hardness as well as alloy. This complex signal is di-
An expected error will not be sorted out, the processes
splayed twodimensionally as vector (complex number).
must be adjusted so that this error cannot occurs. Figure 5 shows the possible distributions of OK. parts, artificial
Check against known faults
NOK. parts, real NOK parts and unexpected fault parts,
First of all, we have to determine how faults occur in the
clearly showed the not all faults will be sorted out by the
heat treatment process. In principle, they do not arise
method with expected artificial faults. The real NOK par-
as expected faults, but they always arise as unexpected
ts lift a larger spread than the artificially manufactured
faults. If an eddy current test system is set to an expected
parts, that means not all NOK parts will be sorted out.
fault, these setting master must be manufactured under
The unexpected fault will not be found.
controlled conditions. This means that the unexpected
La Metallurgia Italiana - gennaio 2020
pagina 86
Industry news - Heat Treatments
Fig.5
If permeability is tested with a single magnetic field
What occurs if another material, e.g. X40Cr13 is mixed
strength HS, test results may be unreliable as soon as
up unexpectedly and tested? The example in
other types of mixed parts become involved. Let us take
Figure 6 b shows that the permeability curve of C45
a typical mixed part test of two kinds of steel: C45 and
crosses the curve of X40Cr13 at HS (red circle).
23NiCrMo2. The largest difference in permeability is at
A differentiation with one field strength (frequency) is
HS. This test can be done with a single frequency test
not possible. Both remain mixed up in spite of the eddy
instrument set up with that field strength, because this
current test [4]. That means there is no separation.
setting provides the largest difference between both, Figure 6a.
Fig.6a
Fig.6b
Check against unknown faults
ced worldwide to test all kinds of material, including the
If we want to sort against an unexpected fault, we need
determining of microstructure and hardening characte-
a completely different approach to the testing task. We
ristics. There are numerous applications for such a sy-
need a technique which -
stem.
• has the ability, to calibrate only with OK parts • a broad spectrum of tests and many test frequencies
How it works?
• fast enough
The use of several test frequencies to induce different
• provides maximum warranty to find unexpected errors
field strengths (H1-8) in the coil means that all these dif-
Eddy current testing, especially Preventive Multi Fre-
ferences are detectable or viewable. Different structures
quency Testing (PMFT) offer this and is applicable for
from different heat treat methods (and by heat treat pro-
a large range of uses. It has been successfully introdu-
cess faults) create different permeability curves which
La Metallurgia Italiana - January 2020
pagina 87
Attualità industriale - Trattamenti Termici are detected. Thus the eddy current testing becomes
The test take place at different field strength, so to say
reliable, expected and unexpected wrong structures are
“preventively” over a larger area.
detected. Different to Figure 6a+b, Figure 7 shows the “curtains” be moved aside in order to widen the view.
Fig.7
If several frequencies of alternating current are used, a
ty OK parts only are needed to setup the instrument and
voltage vector is obtained for each frequency in the im-
to form the tolerance zones (calibration).
pedance, and there is a locus curve. Of course, other
A challenge test with NOK parts (e.g. not-hardened, in-
factors also influence the induced voltage.
correctly quenched, austenizing temperature not rea-
The position of the test part in the coil, geometric va-
ched, too short or too long tempering, annealing tem-
riances in the test part and other factors alter the recei-
perature too high or too low, etc.) can be done, but is
ved voltages.
not needed.
These factors must therefore be controlled. Neverthe-
The test system will reliably detect faulty parts with both
less, received voltage values of several OK parts will
known and unknown defects.
always vary slightly.
The multi frequency test method works reliably for all
They are subject to scattering. The vector tips of the vol-
kinds of defects which may happen during heat treat-
tages form a cloud (refer to the green dots at 4 kHz). If
ment of steel It works like a fingerprint comparison, only
the vector tips are now enclosed with an elliptic toleran-
when all parameters are 100% fit, a part will be recogni-
ce zone, the test can be reduced to a comparison of the
zed as OK, Figure 8 b [5].
vectors to inside (OK) or outside (NOK) of the tolerance zone. Testing current production with such a frequency band (eight frequencies), Figure 8 a, and the comparison with the previously created tolerance zones made with good parts has become well known in professional circles as Preventive Multi Frequency Test (PMFT). A quite wide frequency band (at least 1:1000) is used in order to detect all abnormal structures detectable by eddy current as faulty and to sort them out. Ten to twen-
La Metallurgia Italiana - gennaio 2020
pagina 88
Industry news - Heat Treatments
Fig.8a
Fig.8b
Only if all tolerance fields are matched part will be OK, if
included during calibration blows up the automatically
only one tolerance field is not matched, the part is sor-
formed tolerance zones, and the instrumentâ&#x20AC;&#x2122;s sensitivi-
ted as NOK. [5]
ty is deteriorated. Vectors of bad parts must be deleted
Careful choice of OK parts is a precondition to reliable
before the actual test is started.
testing. One or several NOK parts (red dot), Figure 9,
Fig.9
Practicle examples Used test system Modern test equipment, which are fulfilling all the described requirements are few. Only one test equipment meets the requirements of Preventive Multi Frequency Testing â&#x20AC;&#x201C; PMFT perfectly, Figure 10 a + b shows this equipment.
Fig.10a
La Metallurgia Italiana - January 2020
Fig.10b
pagina 89
Attualità industriale - Trattamenti Termici Inspection of induction hardend cam shafts
Checked for:
Forged camshafts are inductively hardened in the series.
• case depth
In this example, the camshafts are automatically tested
• structure
in a test machine in the production line Figure 11.
• hardness pattern • material mix
Fig.11
Inspection of screws
In this system, with a turntable, with newly develo-
Screws and other fasteners are extremely important hi-
ping U-coils, screws can be checked very sensitivly for
gh-security components throughout different industrys
unexpected various faults, as:
and construction industries. The failure of a screw can lead
• strength
to production stoppages or serious accidents. In the faste-
• structural analysis
ner industry is the permissible error quantity less than 1
• carburizing
ppm.
• decarburizing
For this reason, more and more testing machines will
• material mix
be equipped with Preventive Multi Frequency Testing -
• aso.
PMFT. Figures 12 a + b shows such a high speed testing
Such testing technology can be adapted at any time to exi-
system for screws, the test speed is appr. 10-15 parts/sec..
sting testing machines.
La Metallurgia Italiana - gennaio 2020
pagina 90
Industry news - Heat Treatments
Fig.12a
Fig.12b
Inspection of balls
• different heat treatments in one material
Balls ,Figure 13 a, are used everywhere in the industry,
• test speed 17 parts/sec.
not just in ball bearings. Here balls of 0.6 - 4.5mm are
must be sorted against each other, there is no mingling
tested in a flexible system.
allowed. Figure 13 b -+ c, shows the realised machine,
The goal is not to deliver wrong balls.
with 4 test channels, each test channel works with 8 test
The challenge was:
frequencies.
• 28 different materials
Fig.12a
Fig.12b
Fig.12a
Inspection Of Distance Pins
Frequency Testing – PMFT.
The typical fault of this part is a wrong hardness pattern,
The fault by Part 1 is a classical decarburization and the
Figure 14, against this fault the parts are tested.
fault by Part 2 is first also a classical decarburization, fol-
The Material is a Carbonsteel wit 0,35 % C content, tole-
lowed by a subsequent faulty carburization to reversed the
rance SHD 0,5+1mm, ≥50HRC.
decarburization.
During the series test parts were found which failed during the eddy current test. A macroscopic test with etching did not reveal any errors. A microscopic test together with a hardness depth measurement gave the hardness profiles shown in Fig. 15 and 17. Part 1, Figure 16, had a decarburisation and Part 2, Figure 18, a ferrite space within the hardness zone. This bug was certainly found with the Preventive Multi
La Metallurgia Italiana - January 2020
pagina 91
AttualitĂ industriale - Trattamenti Termici
Fig.14
Fig.15 - Hardness trace with decarburizing
Fig.16 - Bainite, Martensite and Ferrite 500:1
Fig.17 - Hardness trace with Ferrite space under surface
Fig.18 - Bainite, Martensite and Ferrite space under surface 25:1
La Metallurgia Italiana - gennaio 2020
pagina 92
Industry news - Heat Treatments Inspection of special bolts
terial,
This component, Figure 19 requires uniform hardening
• batch 1, 275-285 HB, microstructure Figure 20
in a very narrow tolerance band. The drawing tolerance is
• batch 2, 315-330 HB, microstructure Figure 21
260-340 HB, but all parts of a hardening batch must be in a
both batches are in the drawing tolerance but the spread
tolerance of 20HB.
in the heat treat batch is too high. With Preventive Multi
A high risk is a material-batch mixing that leads to different
Frequency Testing – PMFT, established a save test system
hardnesses within the drawing tolerance. In this example,
which made sure that no batch mixing or other faults are
we have two mixed material batches from the same ma-
in production.
Fig.19
Fig.20 - 275-285 HB 500:1
Fig.21 - 315-330 HB 500:1
La Metallurgia Italiana - January 2020
pagina 93
Attualità industriale - Trattamenti Termici Summary
heat treatment defects / material mingling. Thus, the
The test against unknown heat treatment defects / ma-
eddy current test is perfectly suited for a 100% test for
terial mingling by means of preventive “PMFT” eddy
medium and large quantities in order to ensure pro-
current test gives the heat treatment department a safe
duction quality.
test method with the highest possible safety. Due to the
The older testing technique of a single frequency eddy
integration in production lines, very low test costs are
current test against expected errors with artificially ma-
possible, with enormous security.
nufactured adjusting parts is not a reliable test to find
In addition to high test sensitivity for microstructural
heat treatment defects / material mingling, at all.
changes, hardness faults and case depth changes, PMFT provides a very high level of test safety for unexpected
References [1]
Dipl. Ing. H Springer, Über eine neue Praktische Möglichkeit, die Gefügeänderung bei der Wärmebehandlung von Werkzeugstählen messtechnisch zu erfassen, Härterei Technische Mitteilungen, 1941, Band 1, Seite 61-74, Germany
[2]
Dr. O. Maercks, Magnetische Prüfverfahren zur Werkstoffkontrolle und zur Einsatztiefenmessung, Härterei Technische Mitteilungen, 1941, Band 1, Seite 156-165, Germany
[3]
A. Horsch, U. Bardelmeier, Einsatz eines Wirbelstromprüfgerates mit Mehrfrequenztechnik in einer Lohnhärterei, Hanser Verlag, Harterei Technische Mittelungen 49, 1994, Seite 237-244, Germany
[4]
Herbert Baumgartner, ibg NDT-Group, Eddy Current Test, Company publication, 04.Dec.2012
[5]
A. Horsch, 100% Eddy Current Test for hardness testing and material mix, 26th Hungary Heat Treat Forum, 8-10 October 2014, Balatonfüred, Hungary
La Metallurgia Italiana - gennaio 2020
pagina 94
Atti e notizie - AIM news
Eventi AIM / AIM events
WINTER SCHOOL OF METALLURGY “METAL ADDITIVE MANUFACTURING” Corso > Trento, 26-29 gennaio TECNOLOGIA ED INNOVAZIONE NEI FORNI AD ARCO Giornata di Studio > Dalmine c/o ABB, 24 febbraio TENACITÀ E FATICA Corso > febbraio
www.aimnet.it
ANALISI CHIMICHE Corso > febbraio/marzo
Per ulteriori informazioni rivolgersi alla Segreteria AIM, e-mail: info@aimnet.it, oppure visitare il sito internet www.aimnet.it
LEGHE PER ALTA TEMPERATURA PRODOTTE CON TECNOLOGIE ADDITIVE Giornata di Studio > Firenze, c/o Baker Hughes, 17 marzo IL MONDO INDUSTRIAL: ASPETTI METALLURGICI E METODOLOGIE DI CONTROLLO Giornata di Studio > Torino c/o CNH Industrial Village, 19 marzo MACCHINA FUSORIA Corso Itinerante > 22-23-28-29 aprile - 6-7 maggio RACCORDERIA E PRESSATURA. GLI ASPETTI METALLURGICI Giornata di Studio > Suzzara c/o Intertraco Italia, aprile FAILURE ANALYSIS 11a ed. Corso > primavera-estate LO SVILUPPO TECNOLOGICO DEGLI IMPIANTI DI PRODUZIONE A CALDO DELLE VERGELLE ALTO CARBONIO E ALTO LEGATE E PER FILO SALDANTE Giornata di Studio > Lecco c/o Caleotto, primavera 27° CONVEGNO NAZIONALE TRATTAMENTI TERMICI Genova, 6-7 maggio ECCC 2020 - 10TH EUROPEAN CONTINUOUS CASTING CONFERENCE Bari, 17-19 giugno RIVESTIMENTI DECORATIVI AL SERVIZIO DELL’ESTETICA DEL PRODOTTO Giornata di Studio > Firenze, 25 giugno STAMPAGGIO: PROPRIETÀ MECCANICHE, TRATTAMENTO TERMICO E MECCANISMO DI DANNEGGIAMENTI Giornata di Studio > Ivrea c/o Confindustria Canavese, giugno PRESS & SINTER: EVOLUZIONE DI PRODOTTI, MATERIALI E TECNOLOGIE NEL SETTORE AUTOMOTIVE Giornata di Studio > Brescia, giugno HTDC 2020 - 7TH INTERNATIONAL CONFERENCE HIGH TECH DIE CASTING Vicenza, 1-3 luglio 38° CONVEGNO NAZIONALE AIM 16-18 settembre 2020, Napoli TRATTAMENTI TERMICI E MODELLAZIONE Giornata di Studio > Brescia, settembre ACCIAI AD ALTO TENORE DI CARBONIO Corso > settembre PRESSOCOLATI IN ZAMA Giornata di Studio > Brescia, autunno MATERIALI INNOVATIVI PER L’ADDITIVE MANUFACTURING Giornata di Studio > Torino c/o FCA, novembre RIVESTIMENTI - III MODULO: RIVESTIMENTI PER VIA UMIDA Corso modulare > novembre 1ST INTERNATIONAL MEETING RAW MATERIALS & RECYCLING Bergamo, 3-4 dicembre
La Metallurgia Italiana - January 2020
pagina 95
Atti e notizie - AIM news
Corso // Le Prove non Distruttive Milano, 20-21 novembre 2019
Il 20 e 21 novembre scorsi l’Associazione Italiana di Metallurgia -
(SMT) sulle prove non tradizionali in campo, Luca Bertolini (libero
AIM - ha organizzato a Milano, al Centro Congressi FAST, la quarta
professionista) su getti e fucinati, Stefano Biagioni (Trenitalia) sui
edizione del proprio corso sui controlli non distruttivi, sponsoriz-
materiali rotabili.
zato da Förster e TICAM e patrocinato da AIPnD.
Giuseppe Torsello, ex CISE, oggi TICAM, ha spiegato la possibi-
Hanno preso parte all’attività formativa circa 60 partecipanti (metà
lità di fare le prove meccaniche non distruttive. Ad arricchimento
dall’Università e metà dall’Industria) con background molto diver-
della lezione frontale, TICAM ha esposto per entrambe le giornate
si, dai completamente digiuni di PND ai pluricertificati di livello 3.
attrezzature e campioni da dimostrazione.
Per i primi il sottoscritto ha tenuto una lezione introduttiva di 45
Infine Daniele Bisi (CRF) ha affascinato l’uditorio narrando filoso-
minuti in cui ha raccontato la filosofia delle prove ed i principi, le
fia ed operatività dei CND nell’automobile ed a completamento
finalità, i vantaggi e le limitazioni dei metodi più diffusi, per tutti gli
Bernardo Barile (ArcelorMittal) ha illustrato la straordinaria capa-
altri i relatori hanno illustrato poi lo stato dell’arte dei metodi più
cità dell’industria siderurgica di raccogliere in produzione un’in-
innovativi e dei più importanti campi di applicazione.
finità di dati sulle lamiere da stampaggio e metterli a disposizione
Sui metodi sono intervenuti Umberto Papponetti (2P) con una panoramica sugli ultrasuoni di ultima generazione, Marco Castagna
dell’utilizzatore in tutte le fasi della successiva lavorazione, grazie ad una identificazione in continuo delle lamiere medesime.
(Gilardoni) sui controlli radioscopici dell’era digitale, Rainer Sailer
Il corso si è concluso con i saluti finali del sottoscritto e di una
(Förster) con una chiarissima esposizione in inglese sulle novità
folta rappresentanza del direttivo AIPND, a partire dal presidente
delle correnti indotte e Stefano Benuzzi (Tec Eurolab) sugli stu-
Ezio Tuberosa, al vicepresidente Daniele Bisi, ai consiglieri Fede-
pefacenti risultati della tomografia.
rico Raggio ed Oliviero Oldani.
Dopo uno stimolante intervento accademico del prof. Carboni (Politecnico di Milano) su efficacia diagnostica e difetto minimo rilevabile, si è passati ai settori di applicazione, dove storici docenti come Marco Casaril (OMECO) sulle saldature, Cosimo Carnovale (TenarisDalmine) sui tubi senza saldatura, Mario Cusolito (AIM) sui prodotti lunghi e Ezio Trentini (OMECO) su gru, carriponte e funi, si sono avvicendati a giovani specialisti come Igor Giroletti
La Metallurgia Italiana - gennaio 2020
Con la passione che li contraddistingue ciascuno di loro ha lanciato un messaggio molto significativo: le PND hanno bisogno di gente seria, preparata e generosa ed il mondo dell’industria e delle costruzioni in genere ha bisogno di loro per garantire la sicurezza dei lavoratori, degli utenti e della popolazione in generale.
Ezio Trentini
Coordinatore del corso assieme a Mario Cusolito
pagina 96
Atti e notizie - AIM news
Normativa / Standards in fabbrica con tubi di servizio in acciaio,
UNI CEN ISO/TS 35105:2019
AIM - UNSIDER - Norme pubblicate e
isolamento termico in poliuretano e guaina
Industrie del petrolio e del gas naturale -
progetti in inchiesta (aggiornamento
esterna in polietilene.
Operazioni artiche - Requisiti dei materiali per le operazioni artiche.
30 novembre 2019) UNI EN 253:2019
Norme pubblicate e progetti allo stu-
Tubi per teleriscaldamento - Sistemi bloc-
UNI EN ISO 35106:2019
dio (elenco)
cati monotubo per reti di acqua calda inter-
Industrie del petrolio e del gas naturale -
rate direttamente - Produzione di tubi di
Operazioni artiche - Dati oceano-meteoro-
Norme UNSIDER pubblicate da UNI
servizio in acciaio, isolamento termico in
logici, sul ghiaccio e sui fondali marini.
nel mese di novembre 2019
poliuretano e guaina esterna in polietilene. UNI EN ISO 35101:2019
UNI EN ISO 3183:2019
UNI EN 1753:2019
Industrie del petrolio e del gas naturale -
Industrie del petrolio e del gas naturale -
Magnesio e leghe di magnesio - Lingotti e
Operazioni artiche - Ambiente di lavoro.
Tubi di acciaio per i sistemi di trasporto per
getti di leghe di magnesio. UNI EN ISO 19905-3:2019
mezzo di condotte. UNI EN 17248:2019
Industrie del petrolio e del gas naturale -
UNI EN ISO 20074:2019
Sistemi di tubazioni per teleriscaldamento
Valutazione specifica del sito di unitĂ mo-
Industrie del petrolio e del gas naturale -
e raffrescamento - Termini e definizioni.
bili in mare aperto - Parte 3: UnitĂ galleggiante.
Sistemi di trasporto mediante condotte - Gestione dei rischi geologici per le con-
UNI EN 15698-2:2019
dotte terrestri.
Tubi per teleriscaldamento - Sistemi bloc-
UNI EN ISO 18647:2019
cati di tubi doppi per reti di acqua calda in-
Industrie del petrolio e del gas naturale -
UNI EN 489-1:2019
terrate - Parte 2: Raccordi e valvole isolati
Impianti di perforazione modulari per piat-
Tubazioni per teleriscaldamento - Sistemi
termicamente in fabbrica con tubi di servi-
taforme fisse in mare aperto.
bloccati di tubi singoli e doppi per reti di
zio in acciaio, isolamento termico in poliu-
acqua calda interrate direttamente - Parte
retano e una guaina esterna in polietilene.
UNI EN ISO 10893-3:2019 Controlli non distruttivi dei tubi di acciaio -
1: Ripristino del giunto esterno ed isolamento termico per reti di acqua calda se-
UNI EN 15698-1:2019
Parte 3: Controllo automatizzato mediante
condo EN 13941-1.
Tubi per teleriscaldamento - Sistemi bloc-
flusso disperso sullâ&#x20AC;&#x2122;intera superficie di tubi
cati di tubi doppi per reti di acqua calda
di acciaio ferromagnetico, senza saldatura
UNI EN 488:2019
interrate - Parte 1: Assemblaggio di tubi
e saldati (eccetto quelli ad arco sommer-
Tubi per teleriscaldamento - Sistemi bloc-
doppi isolati termicamente in fabbrica con
so), per la rilevazione di imperfezioni lon-
cati monotubo per reti di acqua calda in-
tubi di servizio in acciaio, isolamento ter-
gitudinali e/o trasversali.
terrate direttamente - Valvole in acciaio
mico in poliuretano e una guaina esterna in
isolate termicamente in fabbrica per tubi di
polietilene.
Industrie del petrolio e del gas naturale -
servizio in acciaio, isolamento termico in poliuretano e guaina esterna in polietilene.
UNI EN ISO 35103:2019
UNI EN 14419:2019
Operazioni in clima artico - Monitoraggio
Tubi per teleriscaldamento - Sistemi bloc-
ambientale.
UNI EN 448:2019
cati con tubi singoli o doppi per reti di ac-
Tubi per teleriscaldamento - Sistemi bloc-
qua calda interrate direttamente - Sistemi
UNI EN ISO 19903:2019
cati monotubo per reti di acqua calda in-
di sorveglianza.
Industrie del petrolio e del gas naturale -
terrate direttamente - Raccordi realizzati
La Metallurgia Italiana - January 2020
Strutture di calcestruzzo in mare aperto.
pagina 97
Atti e notizie - AIM news UNI EN ISO 10426-3:2019
flusso disperso sullâ&#x20AC;&#x2122;intera superficie di tubi
UNI EN 1753:2002
Industrie del petrolio e del gas naturale -
di acciaio ferromagnetico, senza saldatura
Magnesio e leghe di magnesio - Lingotti e
Cementi e materiali per la cementazione
e saldati (eccetto quelli ad arco sommer-
getti di leghe di magnesio.
dei pozzi - Parte 3: Prove della formulazio-
so), per la rilevazione di imperfezioni lon-
ne dei cementi per pozzi in acque profon-
gitudinali e/o trasversali.
de.
UNI 8810:1987 Analisi chimica dei materiali ferrosi.
UNI EN 15698-1:2009
Determinazione dello stagno negli acciai.
Norme UNSIDER ritirate da UNI
Tubazioni per teleriscaldamento - Sistemi
Metodo per spettrofotometria di assorbi-
nel mese di novembre 2019
bloccati a doppio tubo preisolati per reti di
mento atomico nella fiamma.
acqua calda interrate direttamente - Parte UNI EN ISO 3183:2018
1: Assemblaggio di tubi doppi per mezzo
UNI 8808:1987
Industrie del petrolio e del gas naturale -
di tubi di servizio di acciaio, isolamento
Analisi chimica dei materiali ferrosi.
Tubi di acciaio per i sistemi di trasporto per
termico in poliuretano e tubi di protezione
Determinazione del molibdeno negli ac-
mezzo di condotte.
esterna di polietilene.
ciai. Metodo per spettrofotometria di assorbimento atomico nella fiamma.
UNI EN 488:2016
UNI EN 489:2009
Tubazioni per teleriscaldamento - Sistemi
Tubazioni per teleriscaldamento - Sistemi
UNI 8567:1984
bloccati di tubazioni preisolate per reti di
bloccati di tubazioni preisolate per reti di
Analisi chimica dei materiali ferrosi.
acqua calda interrate direttamente - As-
acqua calda interrate direttamente - As-
Determinazione del tungsteno negli acciai.
semblaggio di valvole per tubi di servizio
semblaggio della giunzione per tubi di ser-
Metodo per spettrometria di assorbimento
di acciaio con isolamento termico di po-
vizio di acciaio con isolamento termico di
atomico nella fiamma.
liuretano e tubo di protezione esterna di
poliuretano e tubo esterno di polietilene.
polietilene.
UNI 8566:1984 UNI EN 14419:2009
Analisi chimica dei materiali ferrosi.
UNI EN 253:2016
Tubazioni per teleriscaldamento - Sistemi
Determinazione del cobalto negli acciai e
Tubazioni per teleriscaldamento - Sistemi
bloccati di tubazioni preisolate per reti di
nelle ghise. Metodo per spettrofotometria
bloccati di tubazioni preisolate per reti di
acqua calda interrate direttamente - Siste-
di assorbimento atomico nella fiamma.
acqua calda interrate direttamente - As-
mi di sorveglianza.
semblaggio di tubi di servizio di acciaio,
UNI 8363:1982
isolamento termico a base di poliuretano
UNI EN ISO 19903:2007
Analisi chimica dei materiali ferrosi.
e tubi di protezione esterna di polietilene.
Industrie del petrolio e del gas naturale -
Determinazione del calcio negli acciai.
Strutture in mare fisse di calcestruzzo.
Metodo per spettrofotometria di assorbi-
UNI EN 15698-2:2015
mento atomico nella fiamma.
Tubazioni per teleriscaldamento - Sistemi
UNI EN ISO 10426-3:2005
bloccati a doppio tubo preisolati per reti di
Industrie del petrolio e del gas naturale -
UNI 7842:1978
acqua calda interrate direttamente - Parte
Cementi e materiali per la cementazione
Analisi chimica dei materiali ferrosi.
2: Assemblaggio di raccordi e valvole per
dei pozzi - Parte 3: Prove della formulazio-
Determinazione dello zinco negli acciai.
tubi di servizio di acciaio, isolamento ter-
ne dei cementi per pozzi in acque profon-
Metodo per spettrofotometria di assorbi-
mico in poliuretano e tubi di protezione
de.
mento atomico nella fiamma.
EC 1-2004 UNI EN 1753:2002
UNI 7726:1977
UNI EN ISO 10893-3:2011
Magnesio e leghe di magnesio - Lingotti e
Analisi chimica dei materiali ferrosi. Deter-
Controlli non distruttivi dei tubi di acciaio -
getti di leghe di magnesio.
minazione dellâ&#x20AC;&#x2122; antimonio negli acciai. Me-
esterna in polietilene.
Parte 3: Controllo automatizzato mediante
La Metallurgia Italiana - gennaio 2020
todo spettrofotometrico al verde brillante.
pagina 98
Atti e notizie - AIM news UNI 7715:1977
UNI 7330:1974
UNI 6651:1970
Analisi chimica dei materiali ferrosi. Deter-
Analisi chimica dei materiali ferrosi. Deter-
Analisi chimica dei minerali di manganese.
minazione del piombo negli acciai inossi-
minazione del rame negli acciai e nelle ghi-
Determinazione del rame. Metodo colori-
dabili. Metodo per spettrofotometria di as-
se. Metodo fotometrico.
metrico.
UNI 7089:1972
UNI 6650:1970
UNI 7714:1977
Analisi chimica dei materiali ferrosi. De-
Analisi chimica dei minerali di manganese.
Analisi chimica dei materiali ferrosi.
terminazione del rame negli acciai e nelle
Determinazione del rame. Metodo gravi-
Determinazione del piombo negli acciai
ghise. Metodo elettrolitico a potenziale
metrico.
non legati. Metodo per spettrofotome-
controllato.
sorbimento atomico nella fiamma.
tria di assorbimento atomico nella fiamma (metodo previa estrazione del ferro).
UNI 6649:1970 UNI 6708:1970
Analisi chimica dei minerali di manganese.
Analisi chimica dei minerali di manganese.
Determinazione del fosforo. Metodo colo-
Determinazione del titanio.
rimetrico.
Determinazione del cobalto negli acciai.
UNI 6707:1970
UNI 6648:1970
Metodo potenziometrico.
Analisi chimica dei minerali di manganese.
Analisi chimica dei minerali di manganese.
Determinazione dell’ ossido di magnesio
Determinazione del fosforo. Metodo volu-
UNI 7337:1974
sotto forma di pirosolfato. Metodo gravi-
metrico.
Analisi chimica dei materiali ferrosi. Deter-
metrico.
UNI 7493:1975 Analisi chimica dei materiali ferrosi.
minazione dell’ alluminio negli acciai. Metodo gravimetrico.
UNI 6647:1970 UNI 6705:1970
Analisi chimica dei minerali di manganese.
Analisi chimica dei minerali di manganese.
Determinazione del fosforo. Metodo gravi-
UNI 7336:1974
Determinazione dell’ ossido di calcio.
metrico.
Analisi chimica dei materiali ferrosi.
Metodo volumetrico al permanganato.
Determinazione dell’ azoto negli acciai. Metodo spettrofotometrico.
UNI 6644:1970 UNI 6704:1970
Analisi chimica dei minerali di mangane-
Analisi chimica dei minerali di manganese.
se. Determinazione del manganese totale.
UNI 7333:1974
Determinazione dell’ossido di calcio.
Metodo volumetrico al permanganato di
Analisi chimica dei materiali ferrosi.
Metodo gravimetrico all’ ossalato.
potassio.
nelle ghise. Metodo spettrofotometrico al
UNI 6702:1970
UNI 6643:1970
fenilfluorone.
Analisi chimica dei minerali.
Analisi chimica dei minerali di mangane-
Campionamento.
se. Determinazione del manganese totale.
Determinazione dello stagno negli acciai e
UNI 7332:1974
Metodo volumetrico al bismutato.
Analisi chimica dei materiali ferrosi. De-
UNI 6653:1970
terminazione del wolframio negli acciai e
Analisi chimica dei minerali di manganese.
UNI 6642:1970
nelle ghise. Metodo spettrofotometrico al
Determinazione del piombo.
Analisi chimica dei minerali di manganese.
tiocianato di potassio.
Metodo volumetrico.
Determinazione dell’ ossido di alluminio. Metodo gravimetrico all’ ossichinolina
UNI 7331:1974
UNI 6652:1970
previa separazione dell’ alluminio sotto
Analisi chimica dei materiali ferrosi. Deter-
Analisi chimica dei minerali di manganese.
forma di fosfato.
minazione del cromo negli acciai e nelle
Determinazione del piombo. Metodo gra-
ghise. Metodo fotometrico.
vimetrico.
La Metallurgia Italiana - January 2020
pagina 99
Atti e notizie - AIM news UNI 6639:1970
terminazione del ferro totale nei minerali di
UNI 3658:1955
Analisi chimica dei minerali di manganese.
cromo. Metodo volumetrico per disgrega-
Metodi di analisi chimica dei minerali. De-
Determinazione del nichelio.
zione con perossido di sodio.
terminazione del titanio nei minerali di fer-
Metodo colorimetrico.
ro. Metodi fotometrico e colorimetrico per UNI 3893:1957
disgregazione con perossido di sodio.
UNI 6638:1970
Metodi di analisi chimica dei minerali. De-
Analisi chimica dei minerali di manganese.
terminazione del cromo nei minerali di
UNI 3657:1955 + FA119:1983
Determinazione del nichelio. Metodo gra-
cromo. Metodo volumetrico per disgrega-
Metodi di analisi chimica dei minerali. De-
vimetrico.
zione con perossido di sodio.
terminazione della silice, dell’ ossido di bario, dell’ allumina, della calce e della ma-
UNI 6636:1970
UNI 3892:1957
Analisi chimica dei minerali di manganese.
Metodi di analisi chimica dei minerali. De-
Determinazione del ferro totale.
terminazione dell’ ossido di titanio nei
UNI 3623:1955
Metodo al dicromato di potassio.
minerali di titanio. Metodo gravimetrico al
Metodi di analisi chimica dei minerali. De-
cupferron.
terminazione del fosforo nei minerali di
UNI 6634:1970
gnesia nei minerali di ferro.
ferro. Metodo alcalimetrico, per analisi
Analisi chimica dei minerali di manganese.
UNI 3870:1957
Determinazione della silice (di ossido di si-
Metodi di analisi chimica dei minerali. De-
licio). Metodo gravimetrico per insolubiliz-
terminazione del piombo, del rame e dello
UNI 3619:1955
zazione con acido cloridrico.
zinco nei minerali di ferro. Metodo elettro-
Metodi di analisi chimica dei minerali. De-
litico e gravimetrico per analisi correnti.
terminazione dell’ acqua combinata nei mi-
UNI 6631:1970
correnti.
nerali di ferro.
Analisi chimica dei minerali di manganese.
UNI 3869:1957
Campionamento del minerale caricato su
Metodi di analisi chimica dei minerali. De-
UNI 3618:1955
carri ferroviari.
terminazione del piombo, del rame e dello
Metodi di analisi chimica dei minerali. De-
zinco nei minerali di ferro. Metodo gravi-
terminazione dell’ arsenico nei minerali di
metrico ed elettrolitico.
ferro. Metodo per riduzione con cloruro
UNI 6459:1969 Analisi chimica dei materiali ferrosi. Deter-
stannoso e titolazione con iodio.
minazione del fosforo negli acciai e nelle
UNI 3665:1955
ghise. Metodo alcalimetrico.
Metodi di analisi chimica dei minerali. De-
UNI 3617:1955
terminazione degli alcali totali nei minerali
Metodi di analisi chimica dei minerali. De-
UNI 3896:1957
di ferro. Metodo gravimetrico, per analisi
terminazione dell’ arsenico nei minerali di
Metodi di analisi chimica dei minerali. De-
correnti.
ferro. Metodo gravimetrico.
UNI 3664:1955
UNI 3349:1953
Metodi di analisi chimica dei minerali. De-
Metodi di analisi chimica dei materiali me-
UNI 3895:1957
terminazione dello zolfo nei minerali di fer-
tallici. Determinazione del silicio nel fer-
Metodi di analisi chimica dei minerali. De-
ro. Metodo gravimetrico.
ro-manganese e nel manganese metallo.
terminazione della calce e della magnesia nei minerali di cromo.
terminazione della silice nei minerali di
Metodo gravimetrico.
cromo. Metodo per disgregazione con pe-
UNI 3660:1955
rossido di sodio.
Metodi di analisi chimica dei minerali. De-
UNI 3308:1953
terminazione del comportamento alla cal-
Metodi di analisi chimica dei materiali me-
cinazione all’ aria dei minerali di ferro.
tallici. Determinazione del fosforo nel fer-
UNI 3894:1957 Metodi di analisi chimica dei minerali. De-
La Metallurgia Italiana - gennaio 2020
ro-fosforo. Metodo gravimetrico.
pagina 100
Atti e notizie - AIM news UNI 3306:1953
Metodi di analisi chimica dei materiali me-
rements.
Metodi di analisi chimica dei materiali me-
tallici. Determinazione del cobalto negli
tallici. Determinazione del nichel nel fer-
acciai. Metodo ponderale all’ ossido di zin-
Progetti UNSIDER in inchiesta prEN e
ro-nichel e nel nichel metallo. Metodo
co e all’ alfa-nitroso- beta-naftolo.
ISO/DIS – dicembre 2019 prEN – progetti di norma europei.
gravimetrico.
Norme UNSIDER pubblicate da CEN e UNI 3177:1952
ISO nel mese di novembre 2019
Metodi di analisi chimica dei materiali me-
prEN ISO 15663 Petroleum, petrochemical and natural gas
tallici. Determinazione del vanadio negli
EN 14901-2:2019
industries - Life cycle costing (ISO/DIS
acciai. Metodo per riduzione con solfato
Ductile iron pipes, fittings and accesso-
15663: 2019).
ferroso e titolazione con permanganato
ries - Requirements and test methods for
potassico.
organic coatings of ductile iron fittings and
prEN 14772
accessories - Part 2: Thermoplastic acid
Flanges and their joints - Quality assurance
modified polyolefin coating (TMPO).
inspection and testing of gaskets in accor-
UNI 3106:1950 Metodi di analisi chimica dei materiali me-
dance with the series of standards EN 1514
tallici. Determinazione del nichel negli ac-
EN 14901-1:2014+A1:2019
ciai. Metodo ponderale alla dimetilgliossi-
Ductile iron pipes, fittings and accesso-
ma.
ries - Requirements and test methods for
prEN 13555
organic coatings of ductile iron fittings and
Flanges and their joints - Gasket parame-
UNI 3104:1950
accessories - Part 1: Epoxy coating (heavy
ters and test procedures relevant to the
Metodi di analisi chimica dei materiali me-
duty). .
design rules for gasketed circular flange
tallici. Determinazione del molibdeno ne-
and EN 12560.
connections
gli acciai. Metodo ponderale all’ alfa-ben-
ISO 22055:2019
zoinossima.
Switch and crossing rails.
EN 1514-2:2014/prA1:2019 Flanges and their joints - Gaskets for
UNI 2964:1949
ISO 6892-1:2019
PN-designated flanges - Part 2: Spiral
Metodi di analisi chimica dei materiali me-
Metallic materials - Tensile testing - Part 1:
wound gaskets for use with steel flanges
tallici. Determinazione del tungsteno nel
Method of test at room temperature.
ferro-tungsteno. Metodo ponderale alla cinconina.
prEN 1515-4
Progetti UNSIDER messi allo studio
Flanges and their joints - Bolting - Part 4:
dal CEN (Stage 10.99) – dicembre 2019
Selection of bolting for equipment subject
UNI 2963:1949
to
the
Pressure
Equipment
Directive
Metodi di analisi chimica dei materiali me-
prEN 10250-2 rev
tallici. Determinazione del silicio totale nel
Open die steel forgings for general engine-
ferro-silicio.
ering purposes - Part 2: Non-alloy quality
prEN 993-10
and special steels.
Methods of test for dense shaped refrac-
UNI 2960:1949
2014/68/EU.
tory products - Part 10: Determination of
Metodi di analisi chimica dei materiali me-
prEN 10250-3 rev
permanent change in dimensions on hea-
tallici. Determinazione del titanio negli
Open die steel forgings for general engine-
ting.
acciai e nelle ghise. Metodo al cupferron,
ering purposes - Part 3: Alloy special steels.
ISO/DIS – progetti di norma interna-
fotometrico e colorimetrico (assenza di tungsteno, tantalio e niobio).
prEN 10250-1 rev
zionali
Open die steel forgings for general engiUNI 2959:1949
La Metallurgia Italiana - January 2020
neering purposes - Part 1: General requi-
ISO/DIS 22605
pagina 101
Atti e notizie - AIM news Refractories - Determination of dynamic
ISO/PRF 11971
UNI EN 14901:2014
Young’s modulus(MOE) at elevated tem-
Steel and iron castings - Visual testing of
Tubi, raccordi e accessori in ghisa sferoi-
peratures by impulse excitation of vibra-
surface quality.
dale - Rivestimento epossidico (rinforzato)
tion.
dei raccordi e degli accessori in ghisa sfeISO/FDIS 9647
roidale - Requisiti e metodi di prova
ISO/DIS 19905-3
Steels - Determination of vanadium con-
Petroleum and natural gas industries - Si-
tent - Flame atomic absorption spectro-
Norme UNSIDER pubblicate da CEN e
te-specific assessment of mobile offshore
metric method (FAAS).
ISO nel mese di dicembre 2019
ISO/FDIS 4987
EN ISO 6892-1:2019
ISO/DIS 15663
Steel castings - Liquid penetrant testing.
Metallic materials - Tensile testing - Part 1:
Petroleum, petrochemical and natural gas
ISO/FDIS 4986
Method of test at room temperature (ISO
industries - Life cycle costing.
tic particle testing.
6892-1:2019)
ISO/DIS 15590-2
ISO/FDIS 439
EN 10216-2:2013+A1:2019
Petroleum and natural gas industries - Fac-
Steel and cast irons — Determination of si-
Seamless steel tubes for pressure purpo-
tory bends, fittings and flanges for pipeline
licon content — Gravimetric method.
ses - Technical delivery conditions - Part 2:
units - Part 3: Floating units.
Steel castings - Magne-
transportation systems - Part 2: Fittings.
Non-alloy and alloy steel tubes with speci-
AIM - UNSIDER - Norme pubblicate e ISO/DIS 15590-3
progetti in inchiesta (aggiornamento
Petroleum and natural gas industries - Fac-
31 dicembre 2019).
tory bends, fittings and flanges for pipeline transportation systems - Part 3: Flanges.
fied elevated temperature properties. ISO 643:2019 Steels - Micrographic determination of the
Norme pubblicate e progetti allo stu-
apparent grain size
dio (elenco). Progetti UNSIDER messi allo studio
ISO/DIS 4948.2 Classification of steel based on chemical
Norme UNSIDER pubblicate da UNI
composition.
nel mese di dicembre 2019
dal CEN (Stage 10.99) – gennaio 2020 Progetti UNSIDER in inchiesta prEN e
Progetti UNSIDER al voto FprEN e
UNI EN 14901-1:2019
ISO/DIS – gennaio 2020
ISO/FDIS - dicembre 2019
Tubi, raccordi e accessori in ghisa sferoi-
prEN – progetti di norma europei
FprEN - progetti di norma europei
dale - Requisiti e metodi di prova per rivestimenti organici di raccordi ed accessori
prEN ISO 15761
FprEN ISO 10113
in ghisa sferoidale - Parte 1: Rivestimento
Steel gate, globe and check valves for
Metallic materials - Sheet and strip - De-
epossidico (rinforzato)
sizes DN 100 and smaller, for the petro-
termination of plastic strain ratio (ISO/FDIS 10113:2019).
leum and natural gas industries (ISO/DIS EC 1-2019 UNI EN ISO 15630-3:2019
15761:2019).
Acciaio per calcestruzzo armato e calce-
ISO/FDIS - progetti di norma interna-
struzzo armato precompresso - Metodi di
prEN 14772
zionali
prova - Parte 3: Acciaio per calcestruzzo
Flanges and their joints - Quality assurance
armato precompresso
inspection and testing of gaskets in accor-
ISO/FDIS 35102
dance with the series of standards EN 1514
Petroleum and natural gas industries - Ar-
Norme UNSIDER ritirate da UNI nel
ctic operations - Escape, evacuation and
mese di dicembre 2019
rescue from offshore installations.
La Metallurgia Italiana - gennaio 2020
and EN 12560. prEN 13555
pagina 102
Atti e notizie - AIM news Flanges and their joints - Gasket parame-
FprEN ISO 10113
ters and test procedures relevant to the
Metallic materials - Sheet and strip - De-
design rules for gasketed circular flange
termination of plastic strain ratio (ISO/FDIS
connections.
10113:2019)
EN 1514-2:2014/prA1:2019
ISO/FDIS - progetti di norma interna-
Flanges and their joints - Gaskets for
zionali
PN-designated flanges - Part 2: Spiral wound gaskets for use with steel flanges.
ISO 21809-3:2016/FDAmd 1 Petroleum
and
natural
gas
industries
prEN 1515-4
- External coatings for buried or sub-
Flanges and their joints - Bolting - Part 4:
merged pipelines used in pipeline tran-
Selection of bolting for equipment subject
sportation systems - Part 3: Field joint
to
coatings - Amendment 1: Introduction of
the
Pressure
Equipment
Directive
2014/68/EU
mesh-backed coating systems.
ISO/DIS - progetti di norma internazio-
ISO/FDIS 20321
nali
Petroleum, petrochemical and natural gas industries - Safety of machineries - Powe-
ISO/DIS 15663
red elevators.
Petroleum, petrochemical and natural gas industries - Life cycle costing.
ISO/FDIS 20257-1 Installation and equipment for liquefied na-
ISO/DIS 15590-2
tural gas - Design of floating LNG installa-
Petroleum and natural gas industries - Fac-
tions - Part 1: General requirements.
tory bends, fittings and flanges for pipeline transportation systems - Part 2: Fittings.
ISO/FDIS 19959 Steels, nickel alloys and cobalt alloys in-
ISO/DIS 15590-3
vestment castings - Visual testing of surfa-
Petroleum and natural gas industries - Fac-
ce quality.
tory bends, fittings and flanges for pipeline transportation systems - Part 3: Flanges.
ISO/FDIS 4992-1 Steel castings - Ultrasonic testing - Part 1:
ISO/DIS 15177
Steel castings for general purposes.
Steel sheet, hot-rolled twin-roll cast, of commercial quality.
ISO/FDIS 4992-2 Steel castings - Ultrasonic testing - Part 2:
ISO/DIS 3087
Steel castings for highly stressed compo-
Iron ores - Determination of the moisture
nents.
content of a lot
Progetti UNSIDER al voto FprEN e ISO/FDIS - gennaio 2020 FprEN â&#x20AC;&#x201C; progetti di norma europei
La Metallurgia Italiana - January 2020
pagina 103
Atti e notizie - AIM news
Comitati tecnici / Study groups CT AMBIENTE E SICUREZZA (AS) (riunione del 4 dicembre 2019)
Consuntivo di attività svolte • Settore sicurezza: il corso itinerante “Metallurgia Sicura” si è svolto nei giorni 30 ottobre prezzo Cogne Acciai Speciali – Aosta, il giorno 6 novembre 2019 presso AFV Acciaierie Beltrame - Vicenza e il giorno 13 novembre presso Almag – Roncadelle (BS). I partecipanti al corso hanno segnalato una serie di argomenti da approfondire in prossime manifestazioni: si tratta di una quindicina di argomenti che potranno essere trattati nelle giornate di studio del 2020. Manifestazioni in corso di organizzazione • Settore Ambiente: la GdS “Le emissioni in atmosfera nel settore metallurgico” è confermata per il 5 dicembre a Verona. Iniziative future • Il premio sicurezza 2020 sarà presentato durante il 38° convegno AIM di Napoli a settembre 2020: è stato richiesto alla segreteria AIM di pubblicare il bando entro fine gennaio. • In tema di ambiente, i temi per futuri manifestazioni potrebbero essere EPD certificazione prodotto, bilancio di sostenibilità, Lyfe Cycle Assessment, Carbon Footprint ecc. Stato dell’arte e notizie • Sulla scia di precedenti conversazioni, viene presentato e discusso il caso di un grave infortunio in azienda per fornire spunti di riflessione e di miglioramento.
CT ACCIAIERIA (A) CT FORGIATURA (F)
(riunione congiunta del 11 dicembre 2019) Consuntivo di attività svolte • Si è svolto il corso “Forgiatori” organizzato dal CT Forgiatura nelle date 2-3-9-10 ottobre 1019: l’elaborazione dei questionari di gradimento è stata fatta separatamente per le prime due giornate, dedicate a fabbricazione e trattamenti termici, e singolarmente per le altre due giornate, effettuate con visite presso impianti di produzione. I risultati sono comunque positivi, e il giudizio generale delle differenti valutazioni si mantiene sempre ad alti livelli, con una percentuale di “buono” e “ottimo” che, sommate, arrivano in un caso al 96%. Anche la documentazione distribuita, solitamente un fattore critico in queste manifestazioni, ha soddisfatto i partecipanti (“ottimo” + “buono” oltre 80% e fino al 95%). Manifestazioni in corso di organizzazione • Per il 24 febbraio 2020 il CT Acciaieria sta organizzando la GdS “Tecnologia e innovazione nei forni ad arco”. L’evento si svolgerà a Dalmine; il programma è stato definito. Iniziative future Iniziative CT Acciaieria • Il programma del corso “Acciai ad alto carbonio”, organizzato da CT A, è quasi pronto; la manifestazione si svolgerà nel 2020, in primavera o in autunno, probabilmente a Verona. • Nei mesi di aprile e maggio 2020 si svolgerà il corso itinerante “Macchina Fusoria”. Sono in fase di definizione le aziende che ospiteranno l’evento e i docenti. Iniziative CT Forgiatori • In collaborazione con il CT Materiali per l’energia si pensa di organizare un corso per i valvolieri, dal titolo “Valvole e leghe speciali per le valvole”, sulla scia della GdS sulle leghe di nichel, che ha avuto ottimo successo. Una bozza di programma sarà pronta a inizio 2020. Stato dell’arte e notizie • Mapelli, presidente del CT Acciaieria, conferma che nei giorni 17-19 giugno 2020 si terrà a Bari il Convegno Europeo sulla Colata Continua ECCC 2020, organizzato dalla Segreteria AIM. Sono stati raccolti oltre 170 abstract, con buone richieste di partecipazione anche dall’Asia. Diversi sponsor ed espositori hanno manifestato interesse. • Due nuovi membri di provenienza industriale sono presenti alla riunione e vengono accettati nel CT Acciaieria.
La Metallurgia Italiana - gennaio 2020
pagina 104
Atti e notizie - AIM news
Eventi nazionali ed internazionali / National and international events
2020 March 9-11, Köln, Germany Electrical Engineering of Arc Furnaces April 20-23, Duisburg, Germany Cokemaking April 20-26, Milano, Italy Finiture Green Experience April 26-29, Köln, Germany Refractory Technology May, 4-5, Köln, Germany Hydrogen-based Reduction of Iron Ores May, 6-7, Genova, Italy 27° CONVEGNO NAZIONALE TRATTAMENTI TERMICI May 25-29, Seoul, South Korea MOLTEN - International Molten Slags, Fluxes and Salts conference June 17-19, Bari, Italy ECCC 2020 - 10th European CONTINUOUS CASTING Conference June 23-25, Rome, Italy WMM’20 - 9th International Conference Magnetism and Metallurgy July 1-3, Vicenza, Italy HTDC 2020 - 7th International Conference High Tech Die Casting (ML-P) August 24-26, Shenyang, China TMP - 6th International Conference on ThermoMechanical Processing September 8-12, Cavtat - Dubrovnik, Croatia ICTPMS - 5th International Conference on Thermal Process Modelling and Computer Simulation September 16-18, Napoli, Italy 38° CONVEGNO NAZIONALE AIM October 14-17, Milano, Italy 32.BI-MU October 18-22, BEXCO, Busan, Republic oh Korea WFC - 74th Worls Foundry Congress November, 4-5 Milano, Italy P&E (Polveri&ecoCoating)
2021 22 - 26 March 2021, Bremen, Germany 8th ECIC European Coke and Ironmaking Congress & 9th ICSTI International Conference on Science and Technology of Ironmaking
La Metallurgia Italiana - January 2020
pagina 105
Genova 2020 6-7 maggio
2
convegno nazionale
trattamenti termici
27th AIM National Conference & Exhibition on Heat Treatment
www.aimnet.it/tt.htm AIM è lieta di annunciare la 27° edizione del Convegno Nazionale Trattamenti Termici, il più autorevole ed affermato evento sui trattamenti termici a livello nazionale. Il Convegno – mostra si svolgerà nei giorni 6-8 maggio 2020 a Genova, dove si tenne la prima edizione nel 1960. Sede dell’evento saranno i Magazzini del Cotone, nel bellissimo contesto del porto antico di Genova. Maggiori informazioni saranno presto disponibili sul sito dell’evento: www.aimnet.it/tt.htm
Segreteria organizzativa Organising Secretariat
Via F. Turati 8 - Milano (Italy) Tel. +39 02 76021132 info@aimnet.it www.aimnet.it
AIM is proud to announce the 27th National Conference on Heat Treatment! The Conference &Exhibition will be held on May 6-8, 2020 at the Magazzini del Cotone, a former cotton storehouse, located in a picturesque setting right on the old harbour of Genoa. All information will be soon available at: www.aimnet.it/tt.htm
