WORKSHOP 1: COMPUTING MATTER Tutor: Mustafa Al Sayed Apostolos Despotidis Morgan Graboski Andrius Laurinaitis Shimu Wang Aya Riad 2015
CONTENTS GENERATION 1: STITCHING METHODS GENERATION 2: PATTERN GENERATION 3: PULLING POINTS TOPOLOGICAL RESEARCH GENERATION 4: TOPOLOGICAL STUDY #1 GENERATION 4.1: TOPOLOGICAL STUDY #2 GENERATION 4.2: TOPOLOGICAL STUDY #3 GENERATION 4.3: TOPOLOGICAL STUDY #4
1
GENERATION 1:
STITCHING METHODS
FLAT FABRIC PATTERN
CAST IN FRAME
FLIPPED CAST, OUTSIDE OF FRAME
FABRIC STRETCHED IN FRAME
GENERATION 1: STITCHING METHODS For this cast we experimented with multiple stitch types: dots, lines, and pinches. We stretched the fabric from each of the four corners and used two opposing corners to pour into.
0.75m
0.50m
PULLING POINTS
FLOW BY GRAVITY FLOW BY PRESSURE 0.75m
0.75m
0.50m
PLASTER FLOW
Since we only poured into 2 of the 4 corners, the main flow of plaster was down from the corners into the center and bottom of the fabric pillow. Once the lower portion was sufficiently filled, the plaster flowed up into the other two corners.
0.50m
PLASTER POURING POINTS
1
GENERATION 1: STITCHING METHODS Analysis Due to the random placement of stitches, a high level of control was not achieved. The plaster bulged out in the areas with less stitching. Also, because we only poured into 2 corners, the two which were not used for pouring did not fill up completely. A more dense, uniform pattern approach will control the plaster more. The most effective stitches for controlling the flow are dots and lines through both layers of fabric.
LINE PINCH
POINT STITCH
STITCHES USED:
1
2
GENERATION 2: PATTERN
PATTERN: BEFORE SEWING
STRETCHED IN FRAME
AFTER CAST OUTSIDE OF FRAME
PATTERN: AFTER SEWING
GENERATION 2: PATTERN 0.50m
In contrast to the first cast, we used a dense pattern for the second cast. We used the pinch stitch in an inter-woven pattern. Each corner was used as a pulling and pouring point.
0.50m
FLOW BY GRAVITY
PULLING POINTS
FLOW BY PRESSURE 0.50m
0.50m
PLASTER FLOW
POURING POINTS
Analysis Although we poured from each corner, the stitches were too narrow and close together for the plaster to flow freely into the entire pattern. As a result, the flow of the plaster was impeded and most of the fabric was not filled with plaster at all. Scaling this pattern up would result in a more successful cast.
CROSS-STITCH
2
3
GENERATION 3: PULLING POINTS
STRETCHED IN FRAME
AFTER CAST, IN FRAME
AFTER CAST, OUTSIDE FRAME
GENERATION 3: PULLING POINTS
0.50m
The flat pattern for this cast was a simple pillow with three apertures cut out from it. The edges were sewn with openings at the corners. The rest of the stitching pattern was executed by hand while the fabric was stretched in the frame. A uniform dot pattern of stitches was applied after multiple pulling points had been established. As we learned with the previous casts, the pattern must be dense enough to control the flow of plaster but not so dense as to stop the flow completely.
0.50m
PULLING POINTS 0.50m
0.50m
0.50m
0.75m
POURING POINTS
PLASTER FLOW
FLOW BY GRAVITY FLOW BY PRESSURE
3
GENERATION 3: PULLING POINTS Analysis A good level of control resulted from the point pattern we used. There was no unwanted bulging of plaster. While the outcome was successful, the form created does not stand vertically (as it was cast,) on it’s own.
POINT STITCH
STITCHES USED:
PULLING EFFECTS
3
TOPOLOGICAL RESEARCH: SHELL STRUCTURES
We arrived at the shell typology early on and decided to pursue it further and in more complex ways. We decided on a three-pointed hyper-curved shell, which is composed of surfaces spanning in two directions resting on three vertices. The load distribution of this typology embeds it with an optical effect of “lightness and radiating.� (Moussavi,
Farshid, and Daniel Lopez. 2009. The Function Of Form. Barcelona: Actar. p.416)
TOPOLOGICAL RESEARCH: SPACE FRAMES
A space frame is a three-dimensional framework which behaves as one component to withstand loads from any direction. This concept would add lateral strength to our shell typology, as space frames can function in flat applications as well as curved.
4
GENERATION 4:
TOPOLOGICAL STUDY #1
TRIANGLE A
Triangle A
GENERATION 4: TOPOLOGICAL EXPLORATIONS The basis of our pattern is an isosceles triangle divided into parts. The veins become the flow of the plaster, and so the edge veins are the widest because the most compressive support will be needed in these areas. A smaller, proportionally similar triangle connects to the large triangle at the corners and at the midpoints of the main “arteries.�
TRIANGLE B
Triangle B
4
B over A, overlay scheme and connection points
TRIANGLE B OUSTIDE CONNECTION
INSIDE CONNECTIONS
CONNECTION POINTS
Side View Connection Scheme
TRIANGLE A OUSTIDE CONNECTION
4
FLAT FABRIC PATTERN
PULLING POINTS
STRETCHED IN FRAME
MAIN FLOW Pouring - Constructing the Skeleton POURING POINTS
SECONDARY FLOW COLLECTION POINTS
PLASTER FLOW
FLOW BY GRAVITY FLOW BY PRESSURE
MAIN FLOW MAIN FLOW SECONDARY SECONDARY FLOW FLOW COLLECTION POINTS POINTS COLLECTION
4
POST CAST IN FRAME
POST CAST OUTSIDE OF FRAME
POST CAST OUTSIDE OF FRAME
4
GENERATION 4: TOPOLOGICAL EXPLORATIONS Analysis No control points were used in the pattern and as a result the nodes created where arteries and veins met bulged with plaster. Because of the scale of the cast, this was not a serious problem, but scaling up requires more control in these areas. Because of this bulging, the intermediate members became folded. Again, because of the scale, this was not a serious problem but will need to be addressed later. The size of the veins was strong enough for this scale, and will likewise need to be increased for scaling up. Additionally, the three corners which became the “feet� of the object when it was flipped, were not strong enough and cracked.
4
4.1
1
GENERATION 4.1:
TOPOLOGICAL STUDY #2
375
65
375
0 65 0
70
50
375
70
TRIANGULATION: Triangle A and B
1300
GENERATION 4.1: TOPOLOGICAL STUDY #2
43
251
The pattern here is similar to the previous cast, but scaled up and an equilateral triangle. Each side of the big triangle measured 87 cm when laid flat. The same network of veins and arteries is used, with more attention paid to differences in the sizes of each. The main arteries and edge veins measured 7 cm, while secondary veins were 5 cm. We introduced control points at the nodes which occurred within the triangle in the form of point stitches. The smaller triangle, also equilateral, employs a scaled down version of the veins and arteries. The arteries are 5 cm and the veins are 3cm. The connections between the two layers occur at the three corners and at the midpoints of the arteries.
251
5 43 5
50
30
251
50
870
4.1
FLAT FABRIC PATTERN
STRETCHED IN FRAME
CORNER TO CORNER MEMBER TO MEMBER
4.1
DIGITAL CONSTRUCTION
MODELING P
PROCEDURE
4.1
DIGITAL SIMULATION
FAILURE POINTS
FAILURE POINTS
4.1
CONTROL POINTS
POST CAST IN FRAME
POURING POINTS
PULLING POINTS
PLASTER FLOW
FLOW BY GRAVITY FLOW BY PRESSURE
4.1
ANALYSIS
SIMULATION VS. PHYSICAL FOLDING
GENERATION 4.1: TOPOLOGICAL STUDY #2 Analysis Due to their weight and lack of control, the boundary edges pulled inwards causing folding to occur in the intermediate members. These members needed to be pulled away from the center of the model. Additionally, they needed to have control points to prevent over-bulging of plaster. Since the cast failed and was not completely filled with plaster, the feet still need to be resolved at this scale.
FAILURE POINTS
4.1
4.2
2
GENERATION 4.2:
TOPOLOGICAL STUDY #3
32
375
5 32
375
5 32 32 5
375
5
40
40
25
TRIANGLE A
1300
GENERATION 4.2: TOPOLOGICAL STUDY #3
43
251
The pattern size is identical to the previous cast with alterations made to control the flow of plaster. The edge veins are 4 cm wide and the inner veins are 2.5 cm. For the smaller triangle, the edge veins are 3 cm and the inner veins are 2 cm. Due to the failure of the previous cast, control points and lines were added to the nodes at the edges of the triangle and the center nodes have an increased number of control points. Additionally, extra fabric has been added in the form of a sock over each of the three corners to add additional material at the feet of the object. To prevent folding in intermediate members, the edges were pulled during casting =.
251
5 43 5
30
20
251
30
TRIANGLE B
870
4.2
PRE-CASTING
B over A, overlay scheme and connection points CORNER TO CORNER MEMBER TO MEMBER
FLAT FABRIC PATTERN
4.2
CONTROL POINTS
STRETCHED IN FRAME
4.2
DIGITAL CONSTRUCTION
DIGITAL SIMULATIONS
4.2
POURING POINTS
POST CAST, IN FRAME
AFTER CAST, OUTSIDE OF FRAME
PULLING POINTS
STRETCHED IN FRAME
4.2
POST-CASTING
PLASTER FLOW
FLOW BY GRAVITY FLOW BY PRESSURE
4.2
PULLING POINTS: DURING CASTING
VIEW FROM UNDERNEATH
4.2
3D PATTERN CONSTRUCTION
LARGE TRIANGLE: EDGES
LARGE TRIANGLE: MAIN ARTERIES
SMALL TRIANGLE: EDGES
SMALL TRIANGLE: MAIN ARTERIES
LARGE TRIANGLE: SECONDARY VEINS
LARGE TRIANGLE: COMPOSITE PATTERN
SMALL TRIANGLE: SECONDARY ARTERIES
SMALL TRIANGLE: COMPOSITE PATTERN
4.2
3D PATTERN CONSTRUCTION
LARGE TRIANGLE: SECONDARY VEINS
LARGE TRIANGLE: COMPOSITE PATTERN
LARGE TRIANGLE: CONNECTION POINTS
INTERCONNECTIONS
SMALL TRIANGLE: COMPOSITE PATTERN
SMALL TRIANGLE: CONNECTION POINTS
SMALL TRIANGLE: SECONDARY VEINS
INTERCONNECTIONS AND POINTS
4.2
ANALYSIS
FOLDED MEMBERS CRACKS
FAILURE POINTS
GENERATION 4.2: TOPOLOGICAL STUDY #3 Analysis While there was a good level of control with the flow of plaster, the pulling points made during the casting process did not prevent folding in the intermediate veins. More pre-tensioning of the model is necessary to fix this. This should be done before casting and also sooner during the casting process to prevent cracking. The extensions we made at the feet of the model worked fairly well in creating sturdy feet, however, they are not a perfect solution. Additionally, cracks appeared in several members post-casting. We think this is due to the folded members. To prevent this, member to member connections between the two layers will be avoided.
WEAK MEMBERS While the model was poured it was hanging in tension, but when flipped it must function in compression. The members shown in red here are members which failed when put into compressionthese members “want� to be in tension when the model is upright, however plaster does not allow for it. This is why cracks appeared in these areas.
4.2
4.2
4.3
GENERATION 4.3:
TOPOLOGICAL STUDY #4
5
16
2 .5
16
2 .5
24
3 .7
5
24
3 .7
5
40
3 .7
15
24
281.5
281.5
5
187.5
3 .7
375
24
TRIANGLE A
1300
25
40
GENERATION 4.3: TOPOLOGICAL STUDY #4
43
251
For this cast we used a different triangulation in the 2D patterning; we created more nodes for interconnections between the two shells/layers. In order to create more lateral strength, the two layers are interconnected by a series of intermediate tubes between nodes on each triangle and the midpoints of the main arteries on the large triangle. We also pre-tensioned the fabric before casting to prevent folding in the intermediate members. In an attempt to improve the feet of the model, rigid rings were used as the pouring and pulling points.
251
5 43 5
30
20
251
30
TRIANGLE B
870
4.3
PRE-CASTING
FLAT FABRIC PATTERN
CORNER TO CORNER MEMBER TO MEMBER
B over A, overlay scheme and connection points
FLAT INNER MEMBERS
SIMULATED INNER MEMBERS
NODE TO NODE
4.3
CONNECTIONS
A POINTS
A POINTS
A TO AB CONNECTIONS
A TO A CONNECTIONS
A AND AB POINTS
BC POINTS
BC POINTS
C TO BC CONNECTIONS
AB TO BC CONNECTIONS
C POINTS
CONNECTIONS BETWEEN LAYERS
ALL CONNECTIONS A POINTS AB POINTS BC POINTS C POINTS
4.3
CROSS-SECTION THROUGH CONNECTIONS
PRECAST-CAST COMPARISON
4.3
PLASTER MOVEMENT IN THE CAST
POURING POINTS
PLASTE
ER DIVISION POINTS
COLLECTION POINTS
4.3
DIGITAL CONSTRUCTION
Second
Connec
main
ary Layer
ting layer
layer
4.3
DIGITAL SIMULATION
4.3
FLAT FABRIC PATTERN
STRETCHED IN FRAME
AFTER CAST
For this cast, we used rigid edges at the pouring points in order to control the feet of the model more.
PULLING POINTS
POURING POINTS
PULLING POINTS: DURING CASTING
4.3
PLASTER FLOW
FLOW BY GRAVITY FLOW BY PRESSURE
4.3
CONTROL POINTS
4.3
CONNECTIONS BETWEEN LAYERS
4.3
3D PATTERN CONSTRUCTION
LARGE TRIANGLE: EDGES
LARGE TRIANGLE: MAIN ARTERIES
SMALL TRIANGLE: EDGES
SMALL TRIANGLE: MAIN ARTERIES
LARGE TRIANGLE: SECONDARY VEINS
LARGE TRIANGLE: COMPOSITE PATTERN
SMALL TRIANGLE: SECONDARY VEINS
SMALL TRIANGLE: COMPOSITE PATTERN
4.3
3D PATTERN CONSTRUCTION
LARGE TRIANGLE: SECONDARY VEINS
LARGE TRIANGLE: CONNECTION POINTS LARGE TRIANGLE: COMPOSITE PATTERN
INTERCONNECTIONS
SMALL TRIANGLE: CONNECTION POINTS SMALL TRIANGLE: COMPOSITE PATTERN
SMALL TRIANGLE: SECONDARY VEINS
INTERCONNECTIONS AND POINTS
4.3
4.3
ANALYSIS
TWISTED MEMBERS
SLANTED FEET
GENERATION 4.3: TOPOLOGICAL STUDY #4 Analysis While this cast was able to improve some of the issues with the previous model, there were still some problems. Some of the intermediate members became twisted, weakening them. Additionally, the rigid boundaries used for the feet in this cast were not held horizontally during pouring, causing them to become slanted. What we learned with this cast is that the upper layer filled up through the pressure of the bottom layer filling and pushing plaster up through the intermediate connections. In the previous cast we tried to fill this layer separately, but because the space-frame members introduced in this model became vertical with gravity, they filled the upper layer.
POSSIBLE PROPAGATIONS
4.3