The ElectroMotive 645 Diesel Engine By Preston Cook
I
N THREE DECADES of production following its introduction in 1965, the Electro-Motive 645 series diesel engine became one of the most successful and numerous medium speed power plants in the world. EMD 645s powered domestic and export locomotives, marine vessels ranging in size from tugboats to Great Lakes ore carriers, oil drilling rigs, and mining trucks. In power generating applications they provided peaking power to large cities, base power generation to small communities and islands, and emergency power to hospitals, nuclear plants, and aircraft carriers. Although construction details of many components were altered and improved over its 30-year production run, the fundamental design and method of construction of the 645 engine proved adequate throughout its production lifetime. EMD maintained the need for “backward interchangeability” as a primary consideration in its design programs, and as a result, many component improvements incorporated into the engines late in production could be
retrofitted to engines that were already in service. These engines were produced by the Electro-Motive Division of General Motors at their plant in La Grange, Illinois, between 1965 and the late 1990s. They were built in 8-, 12- and 16-cylinder versions with mechanical roots blowers, referred to by the builder as being “naturally aspirated.” They were also produced in 8-, 12-, 16-, and 20cylinder versions equipped with turbochargers, which provided higher power output and improved fuel economy. All used a 9¹/₁₆″ cylinder bore and 10″ piston stroke, with a “swept” displacement of 645 cubic inches per cylinder. They were 45 degree Vee two-stroke cycle uniflow scavenged diesels with intake air entering the cylinder liner through ports in the liner wall, and exhaust exiting the cylinder through four valves in the cylinder head. These engines were arranged with the cylinders in each bank directly opposed from each other, with the connecting rods sharing a crankshaft throw by way of an interlocking fork rod and blade rod arrange-
EMD’s finest in 1966, SD45 demonstrators 4351, 4352, and 4353 took the industry by storm with their unprecedented 3600 h.p. 20-cylinder 645 engines. 4352 and 4353 would be sold to the Delaware & Hudson along with No. 4354 as D&H 802, 803, and 801. 4351 went to Illinois Central as No. 7000.
40
APRIL 2010
ment. The engine normally applied to locomotive applications was “left hand” rotation, meaning that the power coupling to the generator turns counterclockwise when viewed from the generator end. However, the engine could also be built in “right hand” rotation for some marine and industrial applications. The end of the 645 engine with the
governor, water pumps, and oil pumps was considered to be the “front,” and the end with the main power coupling was the “rear”. In most locomotives this resulted in the “front” of the engine facing the “rear” of the locomotive. Looking at the engine from the generator, the “right” bank of cylinders would be on your right and the “left” bank would be on your left. Cylinder numbering began at the front of the right bank next to the governor, which was number one cylinder, progressing toward the back of the right bank, then to the front of the left bank, with the highest numbered cylinder on the rear (generator end) of the left bank.
The 645’s Successful Predecessor When the 645 engine was introduced in the mid-1960s, EMD already had an ample supply of engine experience and design talent to draw upon in the development of the new prime mover. The immediate predecessors to the 645 series diesel engine were the many versions of the EMD 567 prime mover, which had been highly successful since
its introduction in 1938. All 567s featured an 8¹/₂″ cylinder bore with 10″ stroke, with a swept displacement of 567 cubic inches per cylinder. The early versions of the 567 produced between 1938 and 1958 used engine driven Roots blowers to supply combustion air to the cylinders. The blowers were mounted on the main power coupling end of the engine, which positioned them above the main generator in locomotive applications. The original 1938 version of the engine, later referred to as the 567U due to the “U” shaped valley between the cylinder banks, was produced in 6-, 12, and 16-cylinder versions for locomotive service. A small number of 8-cylinder engines were produced for lab testing and use in a class of U.S. Coast Guard cutters. The 567U was superseded by the 567V engine in 1940, with the “V” referring to the redesigned valley between the cylinder banks that featured a “V” shape. This development was engineered to simplify welding assembly, and to eliminate cast components that had proven troublesome due
to carryover of casting sand that contributed to the formation of voids and inclusions in welding assembly. The greatly improved 567A engine was developed in 1942 for the Navy ATL program (Auxiliary Tank Landing, later called LST, Landing Ship Tank). This engine’s totally redesigned crankcase provided a water manifold in the top deck that allowed for group cooling of the cylinder exhaust risers, which were a permanently welded structure replacing the individual bolted-on parts used in the earlier engines. The 567A was introduced to locomotive use in 1943 and proved extremely durable and reliable. It was supplemented by the 567B model in 1946, a structurally similar engine that was fitted with different end covers, an integral oil strainer box assembly, and a centerline power takeoff between the roots blowers. Some locomotive types were not redesigned for the integral strainer box and continued to use separate strainers, so the 567A and 567B models remained in production concurrently through the introduction of the RAILFAN & RAILROAD
41
LEFT: Building a 645 began with the crankcase. At the left, two “C” channels have been drilled with holes to lighten them and allow combustion air to circulate; they are welded together to form a box that
567C engine in 1954. The 567C was a major redesign of the engine, using a significantly different crankcase, oil pan, and cylinder assembly. The objectives of the redesigned engine were to provide greater structural strength allowing for higher power output and to eliminate some water wetted surfaces to remove the potential for corrosion damage. EMD engineers completely redesigned the cooling system to use individual metal jumpers to take cooling water from a manifold to the cylinder liners, removing two large and troublesome lower liner seals that had been subject to hardening and leaking with age.
Turbocharging the 567 As Electro-Motive improved their prime movers, their major competitors in the railroad industry were also developing and improving their diesel engine products. The Fairbanks-Morse 12-cylinder 38D-8-¹/₈ Opposed Piston diesel engine could produce 2400 h.p. and was the subject of considerable attention among EMD managers in the early 1950s. A number of internal competitive studies were generated to evaluate the comparative economics and advantages of the F-M engine and locomotives with EMD products. However, the early experience of the railroads with the opposed piston engine led to the realization that many routine maintenance tasks on the OP were far more expensive than comparable requirements on the EMD 567. As the reality of operating OP engines set in, concerns over competition from Fairbanks-Morse faded. At the same time, the competitive pressure from the Alco 251 was a growing matter of concern to EMD management. Throughout the mid-1950s the 251 engine was proving itself to be a much better product than the predecessor 244, and Alco was busy working on 42
APRIL 2010
will comprise one bank of cylinders. RIGHT: The completed cylinder banks are fitted in a positioning jig and welded together at a 45-degree angle to create the basic crankcase structure.
improved versions with higher horsepower ratings. The 12-cylinder Alco 251 could surpass the power output of the 16-cylinder 567C, and the 16-cylinder version of the ALCO engine was correctly recognized to have development potential in excess of 3000 horsepower. Despite the excellent performance and reliability of the 567C engines, EMD’s product line would be eclipsed if they did not invest in improvements that would boost the horsepower output of their engines. To counter these increasingly powerful competitors, EMD developed the 567D engine in several configurations. They shared the same bore and stroke with the 567C, but featured improved cylinder power assemblies with larger air inlet ports and pistons designed for higher peak firing pressures. These were housed within a new and reinforced crankcase, designed to allow a maximum 2500 h.p. rating for traction in locomotive applications and a 2800 h.p. nominal maximum in power plant installations. The 567D1 model was a development of the 567C roots blower aspirated engine, with a moderate horsepower increase from 1200 up to 1350 h.p. in the 12-cylinder version, and from 1750 up to 1800 h.p. in 16 cylinders. The 567D2 used a newly developed turbocharger with no aftercoolers to provide 2000 h.p. for traction in the GP20. The 567D3 was initially offered at 2250 h.p. rating in the GP30 and 2400 h.p. in the SD24. The engine was subsequently refined to produce 2500 h.p. in the GP35 and SD35 locomotives. The turbocharger that made the increased power output possible was designed by EMD engineers in conjunction with designers from the Allison Division of General Motors, which built a variety of gas turbine engines for aviation uses. Since a two-stroke cycle diesel engine has no intake stroke to draw air into
the cylinders past a turbocharger, it was necessary to provide a means to provide a minimum amount of forced aspiration independent of engine exhaust flow and load. In two stroke cycle diesels built by two other GM divisions, Cleveland Diesel (the former Winton Engine Co.) and Detroit Diesel, this was done by combining a roots blower with a turbocharger. This compound system used the blower to provide forced air flow for starting and low load operation, with the turbocharger providing the predominant output at higher speeds and loads, and pre-compressing the air going into the roots blower. The combined roots blower and turbo system of scavenging posed many problems for application to an EMD engine, since the roots blowers and turbos would be over the top of the main generator, requiring a stack of components that would make maintenance access unnecessarily difficult. EMD chose to use a single turbocharger with a mechanical drive connected to the engine gear train through an overrunning clutch, doing the job of several components with a single piece of machinery. The turbocharger mounted in place of the roots blowers and its housing formed the rear cover for the timing gear train of the engine. In operation, the turbo was forced up to speed by the cranking and acceleration of the engine through the clutch and gear train. When sufficient r.p.m. and load were achieved to provide enough exhaust flow to cause the turbocharger to freewheel, the clutch was disengaged by the acceleration of the turbocharger. It would operate faster than the drive gear train, achieving its own balance speed that was dependent on exhaust flow and load on the engine. Buy the time the need developed for a more powerful prime mover to take the place of the 567D series, the turbocharger, its clutch, and drive system had been improved through several years of field experience.
LEFT:The partially assembled crankcase is mounted in a turning fixture, which allows the welder to position it so the welding passes can be done in a downward postion. RIGHT: The cylinder head exhaust risers are being
Developing the 645 As the 567D3A was being introduced to the market, EMD was already planning a larger and more powerful diesel engine with the potential for future growth. This was achieved by increasing the cylinder bore (consequently increasing in swept displacement per cylinder) without changing the piston stroke. The increase in cylinder bore from 8¹/₂″ in the 567 series to 9¹/₁₆″ for the 645 engines was achieved by making the cylinder liner water jacket relatively thinner without changing the liner’s external dimensions. This same method of increasing cylinder bore size while maintaining the same stroke had been used on a number of occasions by Winton and Cleveland in their twostroke cycle diesels during the 1930s and 1940s. Electro-Motive had managed to successfully compete in the engine market with the 567 prime mover from 1938 to 1958 without the need to change cylinder bore or stroke. Two decades of production featuring the same displacement prime mover was actually a relatively unusual achievement among diesel engine builders. Test versions of the 645 engine achieved public notice in 1965 with the release of the EMD’s prototype GP40 No. 433 and SD40 prototypes 434A through 434H. These austere, black testbed locomotives operated extensively in demonstration and product evaluation service on the Class One railroads from coast to coast, accumulating miles and experience, and testing the durability of the new prime mover and its supporting systems. In addition to the evaluation of the 645 engine, the experimental locomotives also allowed EMD to test and prove the combined AR10 and D14 traction alternator and companion alternator system that would be used in the new product line, and a new generation locomotive con-
installed. This area will be fitted with a top cover plate, which will form the upper deck water manifold around the risers.
trol system that used a three channel magnetic amplifier with an SCR control system designed to meet the needs of the new rotating machinery. The testing and evaluation of the prototype locomotive was the subject of a film production called Research Rides the Rails, produced as a 16mm color film that was circulated through the GM Photographic Library and was shown to civic and school groups as well as being used extensively by the EMD training center. The turbocharged 20-cylinder version of the new engine had no comparable predecessor in the 567 line. With a swept displacement of 12,900 cubic inches, the biggest 645 inhaled 10,700 cubic feet of air per minute while consuming 188 gallons of fuel per hour to produce 3600 traction horsepower at 900 r.p.m. With a weight of more than 43,000 lbs., the 20-645 was applied to only the SD45 and SDP45 locomotives, but versions of the engine would also be used in marine propulsion, oil drilling, and power generating applications. EMD built and operated four flashy metallic blue and white SD45s numbered 4351 through 4354 to demonstrate the capabilities of the 20-cylinder engines to the railroads. The 20-cylinder engine would also form the basis for a new series of housed power generating units, the MP45, which made use of some of the design features of the SD45 locomotive machinery installation. The MP45 units could be assembled in multiple unit plants to multiply their individual 2.5 megawatt power production. A group of MP45 units installed on the EMD Plant One grounds would become a central part of the ongoing engine development program, providing a convenient site for the EMD Engineering Department to do endurance testing of engine components without the need to pursue locomotives all around the country.
Assembling the Crankcase The main structural components Most of the vertical and longitudinal strength in the 645 engine crankcase is provided by the long “C” channel shaped stress sheets. These were not a standard structural shape, but were specially designed hot-rolled steel channels produced by the suppier on a proprietary set of rollers that were owned by EMD. A pair of these channels would be passed through a machining process that cut large “lightening” holes in the beam portion of the channel. The holes would provide air flow through the engine air box after the crankcase was assembled. The two channels would be welded together point to point to form a box structure. A pair of the box structures were positioned in a jig and welded at a 45 degree angle to form the stress sheet assembly for the crankcase. Two spanning plates were welded into the center of the Vee formed by the stress sheets. The peaked smaller and lighter plate near the bottom of the Vee would be the top cover for the engine main oil gallery, and the larger flat plate connecting the top edges would provide the roof of the air box and the bottom of the exhaust riser water cooling manifold in the completed engine. The completed main structural subassembly was then mounted in a turning trunnion to allow welding to be done “down-handed” for better control of the molten metal pool, and the base rails and the forged steel “A” frames that support the crankshaft bearings were welded in place under the crankcase, which was then mounted in a positioning fixture that allowed it to be held at any angle required by the welder. The crankcase was then returned to a turning fixture for installation of the cylinder head retainers and exhaust riser elbows. This was followed by the hot formed steel sheets of the RAILFAN & RAILROAD
43
LEFT: The upper water manifold area has been completed and the side sheets are installed. The crankcase is beginning to take on its final appearance. RIGHT: The carousel machine was the finest in 1960s
outer air box being welded in place, and the top deck of the center water manifold was welded to the crankcase structure and the exhaust riser outlets. Finally the crankcase was trimmed to length and the steel end sheets were welded to the assembly. Next, the crankcase assembly had to be stress relieved for up to 24 hours. During most of the time that the 645 was in production, all this work was done at Plant Two, the former Pullman Company facility located on the south side of Chicago. The complete but unmachined crank-case assemblies were then cleaned and shipped to Plant One in La Grange for final machining, line boring, and assembly.
Machining the Crankcase The crankcase assemblies flowed through the machining operations at Plant One in an order that suited the needs of production. 8-, 12-, 16-, and 20cylinder engines traveled through the same sequence of machine tools to be transformed from rough assemblies into completed crankcases. The accurate positioning of the rotating elements of the engine (the camshafts and the crankshaft) governed the machining operations. Since the camshafts had to sit atop the crankcase on a flat top deck with their location determined by machined keyways, initial machining operations dealt with the top deck area and the position of the camshafts then became a reference point for succeeding operations. Following the machining of the top deck, the crankcase was mounted in the carousel machine, the finest in 1960s machining technology. The carousel had a rotating support stage with positions for three crankcases. The stage was turned to expose the crankcase in proper sequence to several sets of machine tools that milled the underside surfaces, including machining the base rails, drilling holes for the main bearing studs, the oil standpipes, and the crankcase to oil pan 44
APRIL 2010
machining technology. One crakcase could be loaded while two others were being machined. The carousel milled and drilled the base rail and “A” frames of the unfinished crankcases.
mounting hardware. Following the machining of the base rails and drilling of the “A” frames for the main bearing studs, the “A” frames were broached with a sawtooth pattern to accept the main bearing caps, which were individual steel forgings in the 645 from its 1966 introduction. They were gradually replaced in around 1972 with steel billet main bearing caps that were hot rolled by a forming die and then cut into sections by a large bandsaw. These caps were larger, stronger, and less expensive to produce and machine than the forged caps. The billet caps were also used after 1972 in any rebuilt engines that had an overheated forged bearing cap, so it was not unusual to see an engine with both types of caps after 1972. Regardless of which type of caps were installed, once they were torqued and line bored, they became a matched component of the engine and were stamped with the engine serial number and the correct locating position on the crankcase. The engine then went to the Ingersoll mill, an enormous three-story tall boring machine that sequentially machined the cylinder bores to a 45 degree angle between the banks and in correct alignment with the center of the crankshaft main bearing bore line. The lower level of this enormous machine accepted the engine crankcase mounted on a movable carriage; the second story of the machine had a working platform, an operator’s position, and the cutting heads of the tools; and the third story housed the electric motors and pumps that powered the cutting heads and fed them with machining oil. The Ingersoll mill could handle any crankcase in the product line. The same mill would do the machining required on 8-, 12-, 16-, and 20-cylinder engine crankcases. Several generations of previous Ingersoll mills survived in the plant as part of the engine rebuilding operations. As more modern and efficient
machine tools became available, older machines were sent to the rebuild section. The tools there were used to remanufacture crankcases that had been sent in for restoration to original dimensions and tolerances. So, engines coming in for rebuilding were often serviced by the same machinery that had originally been used to construct them many years earlier. Following the completion of machining operations, crankcases would be cleaned and receive a series of visual and dimensional quality control inspections. If any discrepancies were evident, the crankcase could be taken to a precision measuring table for a thorough check using a coordinate measuring machine, which would establish the straightness and parallelism of machined surfaces and verify the location of drilled holes and other details. The oil pan was the other main structural component of the 645 engine and could be fabricated by Plant Two or by an outside supplier in the case of some specialty applications. Following machining and cleaning, the pan received similar quality control checks, and was then mated to a crankcase. The oil pan and crankcase were then machined to a matching length and were serial number stamped as a pair. After this, the drilling and tapping of gear train housing bolt holes in the end sheets was done. This assured that the gear train housings would fit flat across the two components and that the attachment bolts would be properly positioned. After a final cleaning and wash, the crankcase and oil pan set were ready to be assembled as a diesel engine.
Building the Engine On arrival at the engine line, the crankcase and oil pan were separated, with the crankcase being turned on its side for installation of the crankshaft. The main caps were installed and torqued with an hydraulic wrench, and
LEFT: After the engine was line bored, the Ingersoll mill cut the bores that would receive the cylinder liners. RIGHT: The machined and washed crankcases were then delivered to the beginning of the assembly line. They were
then the crankcase and pan were assembled with their gaskets and bolted together. The case and pan were mounted on small flanged wheel dollies that would be used to take them down the production line toward the engine test area at the end of assembly. As the case and pan were moved from stage to stage in the assembly line, the cylinder power assemblies were installed and torqued down, the front and rear gear trains were assembled, the gear train housings and turbocharger (or Roots blowers) were installed, the top deck housings and covers were bolted in place, and the exhaust manifolds were placed and torqued down. Accessory equipment including water pumps, oil pumps, the engine governor, and power takeoffs were installed near the end of the assembly line. In the final stages of assembly, the crankshaft was barred over for the first time as the gear train timing was checked and adjustments were made to the exhaust valve and fuel injector timing. The engine arrived at the end of the assembly line in an area in the back of the plant that had more than a dozen test cells running at all hours of the day. In the 1970s, EMD could produce more than five locomotives a day, and with marine, industrial, and power generating production, the engine line could sometimes produce ten or eleven engines a day. They all received a running test, thus the need for a dozen test cells to accommodate production, plus some engineering testing, and also to allow for a few cells to be occupied by engine changes and to accommodate maintenance of the test cell equipment and instrumentation.
Problems and Improvements While the 645 ultimately proved to be extremely successful, the early production had a series of difficulties that were addressed through service campaigns and rebuilding programs. Early
turned on their sides and the crankshafts were installed, followed by the main bearing caps, which were applied and torqued.
16- and 20-cylinder engines experienced scattered instances of failures of the welds between the forged crankshaft supporting “A” frames and the crankcase base rails. These failures would result in the “A” frame riding the bearing rather than supporting it, and would ultimately cause the crankshaft to fail due to lack of support. These failures usually occurred at the next to last main bearing in the engine, which handled some of the highest crankshaft torque loads but also experienced the greatest reaction to misalignment between the engine and the generator. The rear “A” frame on EMD engines was much wider and stronger than the others, so any alignment discrepancy tended to orbit through the rear main bearing and stress the next bearing and “A” frame forward. EMD initiated an improvement program for early production crankcases that sequentially increased the size of “A” frame welding whenever engines came in for rebuilding. Over several years, these modifications successfully reinforced most of the early crankcases in the field that had not experienced “A” frame weld failures. Crankcases that had failed were replaced with newer ones that featured much heavier “A” frames with a significantly larger footprint where they attached to the crankcase base rail and stress sheets. By the middle of 1972, the changes in new production 645E crankcases had essentially eliminated “A” frame weld failures. Early production 8-cylinder engines used in marine and power generating service experienced some cracking of crankcase end sheets as a result of their typically higher operating speed duty cycle and their relatively heavy camshaft counterweights. EMD reinforced the end sheets and increased the thickness of the end sheet steel plate, and also increased the size of bolts and dowels that were used to mount the camshaft bearing brackets. The end
sheet cracking issue seldom appeared in railroad engines, and these changes essentially addressed the problem. As production of the 645 engine progressed, the prime mover grew in power output, performance, and reliability. One of the most significant advances, which paved the way to a series of higher horsepower output engines, was the introduction of the “rocking pin” piston carrier and wrist pin in the 645E3A series engines applied to SD45X locomotives and the Union Pacific DDA40X fleet. Earlier production had used a radially segmented wrist pin insert bearing that was in constant contact with the pin as the rod geometry changed during the piston stroke. The “rocking pin” introduced with the 645E3A featured a three segment contact surface that shifted the load between the center and outboard segments as the rod geometry changed. This opened up a space between the unloaded surfaces and the wrist pin, allowing replenishment of the oil film during the compression and power portions of the piston stroke. The rocking pin arrangement made it possible to operate with much higher peak firing pressures and higher horsepower ratings, and was carried forward with only very minor changes into the subsequently developed variations of the 645 engines. In addition to the basic 645 models and the 645E3A series developmental engines in the SD45X and DDA40X fleets, a number of other specialty versions of the 645 engine were produced. The 567E engine was actually a 645 crankcase fitted with 567 cylinder power assemblies and was built for several years to meet the requirements of the “proven product clause” in some military contracts that required an engine with the same bore and stroke as previous production. EMD also built 645 crankcase equipped engines with 567 cylinder power assemblies for certain field replacement situations, such as when a 567 engine had been too RAILFAN & RAILROAD
45
LEFT: The cylinder power assemblies were installed at stations on the engine assembly line located in the rear of EMD’s original Plant One at La Grange. RIGHT: The engine assembly line was one of the few things at La
badly damaged by a component breakup to make it feasible to remanufacture the crankcase. In this way 645 crankcases managed as service retrofits to sneak into GP30, GP35, and SD24 applications as well as a few earlier locomotives. The 645 could also be built in specialized versions to meet the needs of marine, industrial, and military applications. These included ABS (American Bureau of Shipping) certified engines that met particular construction standards, engines for marine use designed to meet particular pitch and roll limits (the ability to operate inclined for long periods of time), engines designed to meet seismic and earthquake proof specifications, and high-shock engines for Navy aircraft carriers and tank landing ships that were designed to be survivable in high impact combat situations.
Fuel Economy Becomes an Issue Following the Arab oil embargo of the early 1970s, fuel economy had a significant influence on the development of the 645 engine and drove many changes in power assembly design during its production. EMD found that the fuel economy of the two-stroke cycle engine could be improved significantly by injecting the fuel in a relatively shorter time interval while the piston was passing top dead center. Getting the fuel into the cylinder more quickly required an increase in the diameter of the fuel injector plungers, and this in turn raised the loading of the injector rocker arm and camshaft, requiring changes in the metallurgy and design of those components. At the same time they began a series of design changes to the pistons, moving the compression ring belt upward towards the piston crown. This effectively provided a longer power stroke, since it contained the firing pressure through a greater distance on the downward piston movement before 46
APRIL 2010
Grange that was set up automotive assembly line style, with the engines moving from station to station on flanged wheel dollies running and tracks set into the floor.
the No. 1 compression ring cleared the air intake ports. Field testing of new power assembly components revealed that the durability of cylinder liners could be improved by providing an air intake port width that was the best compromise between providing breathing capacity and providing resistance to small particles in the air that could contribute to liner scuffing. The liners could also be made more resistant to scuffing through the use of laser hardening, which was first tried in the air port area and later applied to the entire upper bore of the liner. For a number of years afterwards, EMD offered a selection of three types of cylinder liners: plain cast iron bore, hardened port relief liners, and full bore hardened liners. Generally, the more expensive laser hardened liners were applied to road locomotives with the heaviest duty cycle, while plain cast iron was preferred for switchers to avoid excessive oil consumption in low load operation.
New Turbochargers One consequence of the changes in piston ring geometry and fuel injection equipment was the need to develop new and more efficient turbochargers for the “fuel economy” variations of the 645. As the engine was made more efficient through internal improvements, the heat energy available to drive the turbocharger was actually reduced under most operating conditions. This led to the need for improvements in turbo compressor impeller and exhaust turbine blade design to best match the revised flow characteristics of the redesigned engines. Two families of “fuel economy” 645 engines were developed. The 645EB series combined “rocking-pin” type cylinder power assemblies with a revised turbocharger to provide moderate fuel savings without a change in engine rat-
ings. These were built as new production and were fitted as a popular field upgrade to older engines. In the late 1970s, EMD introduced the 645EC series, which provided both fuel economy improvement and higher horsepower output. The 645EC series made use of components that were developed for the successor 645F engines. Due to its higher peak firing pressures, the 645EC package could only be applied to the newest production crankcases with the latest structural improvements to the crankshaft supporting “A” frames.
A New Crankcase The final major development of the 645 engine series was the introduction in 1977 of the 645F design crankcase. The new crankcase was developed to take advantage of the potential for greater horsepower output that was afforded by the rocking pin piston power assemblies and the larger bore fuel injectors. In order to strengthen the crankcase for higher horsepower operation, the 645F was constructed like a jigsaw puzzle, with the crankshaft supporting “A” frames forged with tabs that fit into cutouts in the engine base rails. These interlocking pieces were then welded on the inside facing the oil pan and the outside facing the engine air box to provide a much stronger structural junction than was possible with the 645E crankcase. The 645F engine was intended to operate at 954 r.p.m. in locomotive applications and was first fitted to part of the GP40X production run in 1977. The same basic set of reciprocating components were also applied to late production 645E engines with the heavy “A” frame crankcases, creating the 645E3C engine. The E3C package was not retrofitted to crankcases earlier than 1972 production, as it was thought that the service life of those engines might
LEFT: At the end of the assembly line, engines were taken to the test cell area and run for one shift under load before being installed in a locomotive or delivered to a customer. RIGHT: This view of the engine test area shows
be shortened by exposure to higher peak firing pressures. As the 1980s approached it became evident that the need for continuing improvement in fuel economy, as well as the demand for higher horsepower, would dictate the need for a new prime mover. The EMD 710 series engine was developed to meet that need, adapting the lessons learned with the 645E and 645F series production. In the 710, the piston stroke was lengthened to eleven inches from the ten inch stroke of the 645 to extract even more work from the energy of combustion in the cylinders. This required a complete redesign of the crankcase to accommodate a taller cylinder liner and a new series of crankshafts to provide the greater crankpin radius. In designing the crankshafts it was further necessary to provide larger diameter main bearing journals to maintain a suitable material overlap between the main journals and the crankshaft journals. This required a new design of main bearings and larger crankshaft supporting “A” frames.
Life After Death The introduction of the 710 locomotive engine in 1983 signaled the beginning of the end for the 645 as a commercial product. The domestic locomotive product line absorbed the majority of EMD’s engine production, with export, marine, industrial, and military applications accounting for a smaller number of engines. However, the 645 had a dedicated following, particularly in the marine industry, and it had already qualified for use in a number of specialized applications including nuclear power plants and as an emergency power generator for military vessels. As had been done with the 567E engine (a 567 built in a 645E crankcase) in the 1960s, EMD continued to produce small runs of complete 645 engines in-
the pace of production in the 1970s, when La Grange could produce up to eleven engines a day, about equally divided between domestic locomotives and other applications including marine and stationary engines.
to the 1990s, and also provided replacement 645 crankcases for roots blower and turbocharged engine applications, including a few that served as replacements for 567 engines. Through its three decades of production, and despite a few early problems with the 1960s crankcases, the 645 grew in power output from 3000 h.p. in its 16-645E3 configuration to 3500 h.p. in the 16-645F3 version. In the process, the fuel economy of the engine was improved significantly, with the final production versions generally enjoying almost a 20 per cent reduction in fuel consumption compared to their 1960s predecessors. The 645 engines powered a vast variety of memorable locomotives, including the SD45s and their derivatives, the Union Pacific DDA40X fleet, the incredibly successful Dash-2 locomotives, as well as the versatile F40PH passenger locomotives. The milestone 100 million horsepower 645 engine powered Burlington Northern (Colorado & Southern) SD40-2 No. 7854. The 645 production years also witnessed the decline of the ElectroMotive Division of General Motors from No. 1 to No. 2 in the diesel locomotive building field. The lack of interest in the 20-645 in the railroad industry following the early crankcase problems and the Arab oil embargo of the 1970s resulted in locomotive rebuilder Morrison-Knudsen developing a unique technique for converting the 20-645 crankcase into a much more useful 16-645 version. This patented process involved the torch cutting of the 20-645 into several pieces that were refitted together like a jigsaw puzzle, with intersecting “tabs” of crankcase structure providing the necessary strength and stiffness that would not have been obtainable in a simple shortening of the structure. MK Boise Locomotive Works passed
along this technique to successor MotivePower Industries. The 645 was sufficiently dominant in the railroad industry that it achieved the rather unique distinction of having to compete with a similar product, which was once again an innovation of M-K. In the mid-1990s, M-K spun off MK Rail as an independent operation, and they developed the 645FZ engine crankcase, which was constructed differently from the EMD 645F but accepted all the same principal parts. Marketing of this component was assigned to Engine Systems of Latham, New York, and construction of the crankcase design was sourced with Zgoda in Poland. MK Rail progressed to MotivePower, which eventually became a part of Wabtec. Engine Systems was subsequently sold to General Electric Transportation Systems, taking with it the 645FZ crankcase, which GE continues to produce. The popularity of the 645 has also generated interest among aftermarket parts suppliers and there are presently a number of independently produced cylinder and engine parts that are applicable to the 645 engine series. Although the 645 is no longer in production, Electro-Motive Diesel and other supplies continue to produce parts for the engine and it remains as one of the most widely used prime movers in the railroad industry, in North America as well as worldwide. Thousands of EMD 645s also continue to operate in the marine industry, where vessels typically achieve service lives of four to five decades, and it is also commonly found in power generating, oil drilling, and mining applications. There is no question that EMD’s engine development program for the 1966 product line was a resounding success that had a lasting influence on the railroad, marine and industrial engine markets. RAILFAN & RAILROAD
47