The past few weeks have been very informative; we learnt a lot.

The first week was mostly filled with research but also set some initial goals to achieve over the first couple of weeks.

  • Find the best fabric(s) to print on.
  • Discover the best print settings.
  • Discover whether there is a relation between the stretch of the fabric and the movement created by the tension.

By the second week we were printing at the AMOLF lab in Amsterdam. After these prints we collected out first findings. From these first basic prints we established print settings that produced the best results. We also discovered that our idea of using bulldog clips to secure the fabric was not the easiest way of fastening andtensioning the fabric.


Along side this we saw that prints with wide lines did not deform. The picture above shows this. Below is an example of a thinner line.


In week three our printer persisted to be a general pain and we had to repair it many times.  But the idea to try and make a hyparbolic paraboloid emerged and we drew up some experimental designs.

In week four however, we made a lot of progress with many good results. Before attemptin to print the hyparbolic paraboloid, we started with single folds. The first single fold print was too thin and lacked ridgidity. This resulted in a curve. Cool, but not the desired effect.

We decided a thicker print would improve the rigidity. However, a print too thick or wide resulted in the tension of the fabric having little to no effect. This was one of our first hyporbolic attempts.


We next looked a seperate surfaces which would retain tension and the gaps in between would create the bend.


This worked to an extent but was not very robust.


We then tried to strengthen the whole structure by adding linking strands in the gaps. This did make the form more robust but also too ridgid. So much so that the tension had no effect.

The last thing we did in the fourth week was to try and laminate the fabric with a hyparbolic shape on either side in an attempt to make them stick to the fabric better. This worked in varying degrees but as you can see in the image below th result was not very robust.

Some failure

With this last print we also found that a new fabric worked better with the PLA, forming more robust connection in places. This was a nylon pantyhose of 60 denier.


In week five we took a couple of steps backward and made several basic folds as our current strategy was not working that well. We experimented with connections, shapes, widths and different ticknesses. Where connection were narrow the tension in the fabric disappeared.


These single folds worked well so we attempted to create another laminated hyperbolic paraboloid. The findings of this shape were more or less the same as the single folds we printed, but an uneven stretch in each dircetion caused some folds to bend more than others. This in addition to the outer lines being too flexible produced a random form.

We left the paraboloid there and continued to produce many single fold experiments.

IMG_8430 (1)

  IMG_8452  IMG_8422
IMG_8454 (1)IMG_8429IMG_8419 (1)

IMG_8432     IMG_8416IMG_8431 (1)
IMG_8440 (1) IMG_8441   IMG_8445 IMG_8443IMG_8435





These experiments allowed us to test and apply our findings such as;

  • The lines should not be too thin. (weak structure causing uncontrolable results). The best thickness, depending on the size of the shape, is printing layers two layers of 0.25mm
  • For larger shapes, four layers is best.
  • The lines should not be to wide (tension will disapear). The best width was 2 mm for small shapes and 4 mm for big shapes and outlines.
  • Sucessful bends occur in the gaps between the PLA. Outlies work particularly well.
  • The fabric should not be synthetic as the PLA partially melts and destroys the fabric.
  • After much deliberation, the 40 denier nylon pantyhose worked the best.

The final thing we tested this week was single folds with different connections and combinations of shapes.

IMG_8451IMG_8442IMG_8444IMG_8447IMG_8448IMG_8411 IMG_8443 IMG_8450 IMG_8392

Our final but probably the most important finding was figuring out the height of the print bed. This determined if the PLA would stick to the fabric and ultimately a successful print. We found the best way to approach this was to adjust the bed height while the ultimaker was printing the outlines.

We have thoroughly enjoyed this project and wish we could take it further. The possibilities of application in design are very intriguing. Maybe we could all be living in 4D houses one day?