7. Computational Couture
Disciplines as programming and electronics become highly interconnected, blurring old boundaries and merging different fields of knowledge. Fashion has been already affected by this radical change. Therefore, clothes, shoes, and other accessories can now incorporate elements of hardware and software, generating a peculiar mix between fashion and computation that is incredibly fertile and inspiring. Data becomes Beauty, Interaction becomes Emotion. As a result, a new aesthetic is emerging. In this class, I was exploring computational design methods towards a new reinterpretation of cloths, garments, and accessories for fashion design, inspired by a new digital design methodology.
Design: I was investigating and working on Rhinoceros, Grasshopper
Machines: Ultimaker 2+, Ultimaker S5, BQ WitBox
Slicer Software: Ultimaker Cura
Fabrication: 3D printing the textile or 3D printing textile.
Based on my experience Textiles usually has to be flexible and to gain that either print it hard but with joints to help moving them or use flexible filament (TPU95A and Filaflex). option 2 3D Printing on fabric as it retains the flexible properties of the fabrics, doesn’t require exotic filament, and permanently incorporates the 3D elements into your garment. This method works for printing on net, tulle, lace, or other similar fabrics with holes( I used Tulle for this assignment ) (plus I experimented using Lycra and biomaterials too) .
This has been done further on varying techniques and codes and materials...
Bringing a lovely hi-tech shift in the treasury of traditional Jordanian stitches. The tradition of embroidering dresses is one of the finest art forms passed on through generations of villages in Jordan usually stitched by hand with a mono level in height (height of the thread)... I wanted to mimic the basic cell of the stitching element to produce different bigger geometries that can be as well varying in altitudes /heights to be 3D printed filaments replacing threads to a new level up
Inspiration
I was really impressed with the project that Julia Korner did where she divided the dress into segments and each one was made of basic spike element varying in height and space between the elements ... so I thought of using such an inspiration where I can transform our Jordanian traditional "Madrga/Thoub" dress into Computational/Technological 3D-Printed one :
- Having segments so can be printed on any printed considering size limitation of the printer and machine capability
- having a basic element in this case the "X" which represents the cross stitches on the traditional way of making this cultural dress and embroidery so now it can exactly replicate digitally and allows parametric design so this X can grow endlessly...
- Varying heights of the basic element/Cell "X" or distances or how they make an outer geometry of varying shapes and further parameters I'll explain below...
Of course to create these computational couture I used 3D printing.
How I started
A.1 printing with joints
so that's how I transformed my inspiration into printed embroidery like a hi-tech version of the traditional garments
I firstly printed the main cells with joints then printed them on fabric
-
Printing them with joints advantage is that you can still print PLA or any rigid filament yet be able to MOVE the full pattern as the joints gives you that range of movement and allow it (so flexibility to a certain extend depending on the joints design) … another advantage is that (well based on the joint design as well) can allow you to grow the pattern! as you don't need extra large printer to print a full outfit lets say you can print on small printers then continue to grow via joints!
-
if the joints were badly designed then GAME OVER ... would be completely useless in my opinion
-
Another point to consider is the size of the main elements VS. the size of the joints (so the smaller these elements, more repetitive they are as a result would give the full piece more movement).
-
Also you have to consider the machine accuracy so leave room for tolerance (depending on the design joint)
-
So basically the design of the joint defines to movement type and axis so how bendable the piece would end up!
A.2 Flexible TPU95A
3D printing a textile requires a filament that is flexible! Which are TPU95A and Filaflex filament that are usually marketed as flexible.
Here I used TPU95A using Ultimaker S5 3D printer
Printing with TPU is flexible but you got to be careful if you printed lots of layers might set some restriction on how flexible or bendable the piece can be!
so it works fine for few layers but might get tricky for larger prints (if amount was larger in the Z axis)
the result was flexible but I wanted more so I did another experiment below with Filaflex with same concept of stitches however more exciting patterns with the same cell "X"
cura settings
Quality
Layer height: 0.20 mm
Line Width: 0.35 mm
Shell
Wall Thickness: 0.7 mm
Top/Bottom Thickness: 0.6 mm
Infill
Infill Density: 20%
Infill Pattern]: Grid
Material
Printing Temperature: 200 °C
Build Plate Temperature: 60 °C
Speed
Print Speed: 45 mm/s
Travel
Enable Retraction: Yes
Cooling
Enable Print Cooling: Yes
Fan Speed: 100%
Support
Generate Support: unchecked
Build Plate Adhesion
Build Plate Adhesion Type: Brim
A.3 Flexible Filaflex
I used Grasshopper + Rinhoceros to achieve this My intention for the realization of the design has been to control parametrically, several aspects.
-
The width of the tip of variable form.
-
Having gradient distortion based on Jordanian embroidery geometries on traditional Clothing.
-
6 approaches digitally explored on how the gradient differs based on the customized input •chosen over a criteria• where this input is inspired from the Jordanian embroidery Patterns and these "X"s resemble the cross stitches so using grasshopper was aiming to create a lovely deconstruction gradient from each inspiration … ~where hi tech design meets tradition ✖️
3D printing a textile requires a filament that is flexible! Which are TPU95A and Filaflex filament that are usually marketed as flexible.
Here I used filaflex using BQ WitBox 3D printer
- I must admit it's one of my favorite prints from these experiments, I was satisfied how it twisted the tradition into the cool hi-tech form so matched the vision I had and was the most flexible and interesting to look at and very comfortable on the skin to wear.
cura settings
Layer height: 0.20 mm
Line Width: 0.35 mm
Shell
Wall Thickness: 0.7 mm
Top/Bottom Thickness: 0.6 mm
Infill
Infill Density: 20%
Infill Pattern]: Grid
Material
Printing Temperature: 200 °C
Build Plate Temperature: 60 °C
Speed
Print Speed: 45 mm/s
Travel
Enable Retraction: Yes
Cooling
Enable Print Cooling: Yes
Fan Speed: 100%
Support
Generate Support: unchecked
Build Plate Adhesion
Build Plate Adhesion Type: Brim
B.2.1 Flexible on Fabric
- I used flexible tulle in 2 directions with TPU95A based on grasshopper coding I did above...
For 3D printing on fabric, I used similar steps for all
-
I used small clips. I recommend you use plastic clips or small clothespins. You don’t want anything that will scratch your build plate, and you also want small clips so they won’t hit the sides of the printer when the build plates move or any sensors of the printer and end up with a short circuit! .**so please mind where you place these clips as well.
-
3D model Before you even get started, keep this in mind while designing the measurements and boundaries for the 3D printer you want to use. After baking from grasshopper, export the closed geometry as STL from rhino. I talked about this previously, but again because it's important to make sure the design won't require support material, otherwise the support material will also attach to the tulle fabric and tear when you try to cut out your piece. (It is best to avoid a design that will need it before you even get to the slicer.)
- Insert the STL file into software for slicing (I used Cura)... I sliced my model and converted it to gCode. I’ve found that 20% infill is adequate for most models. Since I used Cura for slicing, there is a feature that will allow you to put a pause in the gCode. Put the pause in after 2 to 3 layers of the design are printed. If you are new and nor familiar with this, you can do it manually by just waiting for the full layer to be printed, like counting a few, then pausing through manual control. (again, uncheck the support from the setting, uncheck Build Plate Adhesion- because if you had a brim or raft, you would tear the tulle fabric while trying to remove it and get your prints detached from the tulle) -Set up the printer and select from a USB or over network 3D model and filament type (change filament if needed). I printed 2-3 layers of my design. Place the fabric smoothly and tightly on the build plate and use numerous clips to hold the fabric taut at the middle and corners. (I learned this the hard way, but please be sure that your fabric won’t get caught in any of the moving parts of the printer and the extruder head won’t hit any of the clips while it is moving around to print. You don't want any drama! well at least not today … not anymore!)
- Resume printing (carefully watch the printer for the next couple of layers to be sure the extruder head isn’t pulling the fabric too much. Since the printer is laying down small amounts of filament it is often hard to see the first layer on the fabric, but with experience, you can detect it when printing is done! Wait for the build plate and nozzle to cool and recently I started to carefully remove my fabric (because previously I used to drag so harshly and tulle used to break as a result the separately 3D printed parts started to get detached from the tulle since its breaking and fragile... THAT'S WHY I created this piece already attached just in case tulle breaks based on my experience, but of course, at low heights, it isn't strong attachment, but better than nothing! So the edges of the "X" give extra support being attached to each other). The filament should sandwich the fabric.
So that's the thing with experience: you know how to design FOR the 3D printer and how to keep optimizing and avoiding problems by altering your design while designing, so you consider all this and that!
cura settings
Quality
Layer height: 0.20 mm
Line Width: 0.35 mm
Shell
Wall Thickness: 0.7 mm
Top/Bottom Thickness: 0.6 mm
Infill
Infill Density: 20%
Infill Pattern]: Grid
Material
Printing Temperature: 200 °C
Build Plate Temperature: 60 °C
Speed
Print Speed: 45 mm/s
Travel
Enable Retraction: Yes
Cooling
Enable Print Cooling: Yes
Fan Speed: 100%
Support
Generate Support: unchecked
Build Plate Adhesion: unchecked
this piece was flexible from the curve with the lowest height in the Z axis as there were 2 limitations to how bendable the full piece is
- The other corners were high enough in the Z axis to create some rigidity
- The X's were connected so the rigidity from the ends had passed to the X's and over all as well the connection limited the bendability
In all of my geometries no support was required and that's something to consider! make sure your design won't require support material otherwise the support material will also attach to the tulle fabric and tear when you try to cut out your piece. It is best to avoid a design that will need it before you even get to the slicer.
so make sure while designing to design something that doesn't require support! otherwise you would tear apart the tulle/fabric while tryin to remove the support and you can remove it
So basically I played with the stitching and stimulate the stitching and I was looking on how a stitch could reflect into more stiff pattern on the textile • inspired from one of the famous Traditional embroidery Motifs •Moon • “Qamar” • these "X"s resemble the cross stitches
Here I used tape instead of clips as its a mega new printer and I didn't want to risk the sensors but as I printed over night I came the next morning found it moved a little while printing in the upper Parts so tape doesn't attach the fabric firmly to the build plate especially if you have lots of height in your print (usually works for less heights)
so tape isn't the best thing you can use to stick the fabric to the bed while printing on fabric better use LOTS of clips as much as you can
Another thing to consider as mentioned previously I intended to have gradient of heights fro this print which proved on longer heights in the Z axis restricted the flexinility of the TPU and the piece
so again make sure minimize the height in the Z axis to avoid restriction of how flexible the piece of the TPU
B.2.2 Flexible on Fabric
Puff Pattern Textile - stimulating stitches
Again here approach to stimulate stitching via coding and 3D printing • inspired from the Jordanian embroidery Patterns and these "X"s resemble the cross stitches... but what differs here I was trying to create another dimension so once removed from bed it would deform as if it changes state … as if it had been stitched in the Z access or blown by air … can we call it 4D maybe?
to achieve that 2 layers were printed on the Tulle but here the X were extra light couldn't create the claw effect I wanted from their stiffness
So I went with another organization of the overall geometry of these X's the famous Traditional Pattern called Moon "QAMAR" with the X's as the cell geometry to create the Moon Motif with extra dimension
cura settings
Quality
Layer height: 0.20 mm
Line Width: 0.35 mm
Shell
Wall Thickness: 0.7 mm
Top/Bottom Thickness: 0.6 mm
Infill
Infill Density: 20%
Infill Pattern]: Grid
Material
Printing Temperature: 200 °C
Build Plate Temperature: 60 °C
Speed
Print Speed: 45 mm/s
Travel
Enable Retraction: Yes
Cooling
Enable Print Cooling: Yes
Fan Speed: 100%
Support
Generate Support: unchecked
Build Plate Adhesion: unchecked
Voronoi Pattern with outline if the traditional moon stitch
I used this pattern for the both experiments below (Printing on Lycra + Printing on Biomaterials)
B.3 on Lycra
here flexible TPU printed on flexible fabric "Lycra" so this is the most flexible piece ever (plus low height in Z Axis = allows TPU to be bendable comfortably )
B.4 PLA on Biomaterials (hibiscus bio materials)
-
Just check the temperature of the bed so it doesn't melt for gelatine based bio Materials
-
I used tape instead of small clips(as it was almost a small fast 20 minutes print on a small bio material like I didn't want to have excess bio-textile and ruin it).
-
Insert the STL file into software for slicing (I used Cura)... I sliced my model and converted it to gCode. I place the biomaterials first to the bed and directly printed on it(as the design itself was like 2-3 layers anyway only) so no pause no pre printing layers like above trials with tulle...
-
Surely as I am using biomaterials from food waste I wanted to remain with the "Environmentally Friendly Concept" that's why I used PLA.
-
PLA is what’s known as a bioplastic as it’s a product of renewable organic material such as dextrose from sugarcane and cornstarch from maize.
-
With this composition, PLA is environmentally safer as it has a shorter decomposition cycle of under 3 months in an industrial composting facility or about 6 to 12 months in a domestic compost bin. Unlike petroleum-based plastics, when PLA does decompose, and it degrades to completely non-toxic products. You can read here more about Environmentally Friendly 3D Printer Filaments
Here I printed on PLA Hibiscus Petals Biomaterial
cura settings
Quality
Layer height: 0.20 mm
Line Width: 0.35 mm
Shell
Wall Thickness: 0.7 mm
Top/Bottom Thickness: 0.6 mm
Infill
Infill Density: 20%
Infill Pattern]: Grid
Material
Printing Temperature: 200 °C
Build Plate Temperature: 60 °C
Speed
Print Speed: 45 mm/s
Travel
Enable Retraction: Yes
Cooling
Enable Print Cooling: Yes
Fan Speed: 100%
Support
Generate Support: unchecked
Build Plate Adhesion: unchecked
Photoshoot
Manikin
Here I thought I lowered the print speed just enough but ended up lowering even more manually on the printer as I was printing without support and I didn't want to risk it!
Previously
B.1 PLA on Tulle
Here again kind reminder to uncheck the Build Plate Adhesion while printing on Tulle
I learned this the hard way, as I printed via Zortrax printer while printing the PLA white spike field below (previously), and afterward, I wanted to remove the raft attached to it (as I had put the raft in the first place as it would take lots of hours to print and I didn't want to risk the print moving). So after I removed the print, I was trying to remove the raft and ended up tearing the tulle, so again I wasted the hours of printing as the spikes got detached from the tulle) so Printed again while lowering the Print Speed.
BOR short one minute video - Final presentation/ Fab academy from Batoul Al-Rashdan on Vimeo.
