Skip to content

8. Soft robotics

Checklist for the week

test

Research, Inspiration & Resources

Since this week's topic was completly unlike anything that i had read about or seen before, i spent the week looking at works of others. Here is a list of some of the people and resources that inspired and guided me this week:

  1. Adriana Cabrer's entire YouTube Channel, but especially this tutorial on soft robotics

  2. Montserrat Ciges's final project at Fabricademy and her explorations in her Soft Robotics Assignment

  3. Saskia Helinska's MINDBLOWING body of work and explorations in her Fabricademy Final Project and at Fab Lab

  4. Adrian Torres's 3D models and explorations helped pave the way for our experiments this week

  5. This article with a lot of helpful details by Autodesk Instructables

  6. The entire YouTube chanel of Soft Robotics Toolkit but especially this tutorial on making a Pneumatic Wrist Brace

  7. These two tutorials on Open Soft Machines about making an acrylic mold with lasercutting and silicone bending actuators were my main point of refernce this week, whenever i felt lost about the next steps

  8. This research paper on Pneumatic Wrist Brace Fabrication Guide was very interesting to see real world applications of this week's learnings

  9. This research paper on Soft Robotic Grippers with its timeline of mapping the advancements, explorations and pushing the envelope projects in real world use cases was also very interesting

  10. Lastly, this past lecture called From Soft robotics to biobotics was an intersting watch to see what else could i try to make in the future

Softwares, Tools & Materials

SOFTWARES:
  1. Autodesk AutoCAD
  2. Adobe Photoshop
TOOLS:
  1. Lasercutter
  2. 3D printer
  3. Iron
  4. Blowing tools like syringe, balloon pump, keyboard dust blower, blood pressure inflator, metal straws
  5. Kitchen scale
MATERIALS:
  1. Heat Transfer Vinyl
  2. Baking Paper
  3. Ironing cloth
  4. 2mm Acrylic Sheet
  5. Ecoflex 00-30
  6. Transparent plastic cups
  7. Gloves
  8. Stirrers
  9. Feviquick
  10. Cutting Mat
  11. Cutter
  12. Paper Tape
  13. Empty containers
  14. Plastic sheet for workspace
  15. Round tip paint brush

Work Flow:

3D Printing Existing Designs

To begin with, it was recommended to us that we first start by trying to work with existing models. So we selected the 2 models we wer most keen on understnding the mechanism of for this week.

We used the designs shared with us for a 4 Leg Gripper and a Wendy Mold, and set these for 3D Printing. Once both these models were printed, we tried to examine and understand how these would inflate.

test

Arm Brace Design explorations for Pneumatic Activation using Heat Transfer Vinyl

Now that we had analysed the 3D prints, we got a better idea to imagining how these designs worked. So me and Shefali set out to whip up a few designs to test, inspired by the body of work of Saskia

Printing the Design

We collected references for the shapes we wanted to inflate and converted them into drawing for printing. And since we were over-enthusiastic with experimenting, we selected a LOT of shapes to play with.

TIP: take 2 sets of printouts of the design! One to use for Baking sheet and one to use for Heat activated Vinyl.

test

Cutting Baking Paper

Once we had the printout of the shapes, we overlayed the baking paper on one set of the prints and carefully started cutting the baking paper. Here we are cutting the air chanels through which air will enter and inflate.

TIP: ensure the tails are nice and long when cutting these designs

This was a quick and dirty process and it was reccomended to us to not be super precise with this. We created around 20 shapes!!

test

Cutting Vinyl Sheets

We used the second set of prints to cut the overall outer form of these shapes on 2 sets of vinyl sheets (since the 2 surfaces will stick to each other). We also kept these print outs on top to identify the designs, while we were cutting.

test

Inserting Baking Paper

Once the 2 sets of mirrored vinyl sheets are cut, we carefully place the baking sheet cutouts, ensuring the tails are well placed.

test

test

Heat Activating

Now that ALL the pieces are ready, it's time to get ironing. We setup an ironing cloth below to create a smooth surface to iron on.

Honestly, the vinyl could take a lot more heat than i expected!! But the trick was to iron BOTH sides and let it cool down, for a smooth finish.

This is how we did it

Inflating

Once all the pieces had cooled down, we used a Keyboard dust blower to inflate them.

test

test

This is what they looked like

We realised that we weren't seeing as drastic of a movement as we expected, and got in touch with our Local mentors. Rico gave us a tip to try cutting the excess vinyl and test inflating. This worked like a charm we saw much better movement from our pieces!

test

The trimmed viny worked really well!

Building Arm Brace Design

In the Global as well as Local mentoring sessions, we were asked to look at inspirations from nature. And so we set out to looking at images of examples of flora and fauna that woould inspire us to come up with abstrct shapes for this week.

Designing

Having experimented with Vinyl shapes, i knew i wanted to create a design with an air chanel with diagonal axis to get a more radical movement. So i looked at similar drawings inspired by the Octopus, Pufferfish and Starfish, and created a version inspired by them.

I worked on AutoCAD to make the final design

test

Lasercutting

Once the design was ready, i created a file for laser cutting, which actually needs a bit of calculation and foresight.

Since we had 2mm Acryllic lying around, we made our layer calcutions based off of it.

There is a Top layer, a Middle spine layer and a Bottom layer to create the moulds for casting the Ecoflex. I lasercut 2 pieces each of the shapes, that i would stick together, to allow for enough height for the cast.

test

Assembling

Now that the pieces were cut, we began assembling our designs to make them ready for casting. We used feviquick to join the various acryllic surfaces. Knowing that we were also going to be pouring a liquid into these moulds, we paid extra attention to sealing the edges with feviquick, in order to ensure no leakage happened.

We stuck 2 pieces of the Top layer to 1 piece of the Bottom layer. So three 2mm acrylic layers. Inside this, i carefully placed the Spine layer and stuck it, along with sealing the edges. This was one part of the mould for casting. The second part of the mould will use the remaining 1 Bottom layer, on which the other part will be cast.

test

Ecoflex Pouring

We had 4 mould ready for pouring, 2 3D printed moulds and 1 mould design each between me and Shefali. We were ready to pour and cast the Ecoflex.

It was also a good decision to do this together, because Ecoflex is exteremly expensive, and the mixture has a use time of 45 mins, so anything that isn't used, needs to be disposed of. SO two sets of hands meausring, mixing and documenting was an ideal setup and worked realy smoothly for us.

Ecoflex Part A & Ecoflex Part B need to be in 1:1 Proportion in the mix at all times

Since we were working together, we mixed the Ecoflex to be able to cover all the 4 moulds. And setup a timer of 45 mins to complete ALL the pouring.

test

3D Printed Molds

Wendy Mould

We started with the Wendy Mould. We mixed:

Part A: 39g + Part B: 39g = Total 78g

We then kept stirring for 1 min on the timer, as was instructed on ALL the turotials and the box intructions. However, we ensured that we didn't form bubbles in the misture.

After 1 min, we carefully poured the mixture in the 3D printed mould. In order to create the top, we poured a thin coat onto a waste acrylic piece, whose size was bigger than the mould.

test

Gripper Mould

We ended our pouring session with the Gripper Mould. We mixed:

Part A: 24g + Part B: 24g = Total 49g

As perviously done, we kept stirring for 1 min on the timer and then poured the mixture into the mould, slow and stead stream ensuring there are no bubbles formed during the pouring a well.

We poured a thin coat in a waste container top, whose size was big enough to fit the base shape.

test

Laser Cut Molds

For our individual moulds, we realised we would be needing more mixture than the 3D moulds and made a conjucture for the total quantities accordingly.

For our moulds together, we mixed:

Part A: 53g + Part B: 53g = Total 107g

We stirred the mixture at a steady pace for 1 min to ensure there are no bubbles. And we pour the mixture at a steady thin stream to ensure there are no bubbles in the cast.

We poured the mixture on the acrylic mould pieces and gave it a few taps to remove any bubbles that might have sneaked through. The remaining mixture was used to create the thin layer on the second acryllic sheet, to create the top for the mould.

test

De-molding

Once the Ecoflex mixtures were poured, we moved all the pieces to their final resting place, within the 45min resting time.

We let ALL the poured moulds to cure for 24 hours.

We first stared by removing the cast in the 3D printed moulds. In the beginning we struggled a lot, because we were worried we would tear it, BUT the silicone cast was quite resilient.

With a lot of wrangling on the gooey textured surfcae, we were successful in de-molding them.

This is us trying to delicately wiggle out the cast silicone from the 3D molds

The Lasercut Acryllic casts were way simpler!

Sticking the Molds together

Now that we had all the indivudual pieces, we needed to stick the tops and the bottoms, to create sandwich pieces, through which the air could pass, once sealed.

We mixed a really small batch of Ecoflex and used a thin painbrush to apply it to the individual surfaces. And then we stuck the pieces together.

We let this cure in a corner for 5 hours

test

test

Inflating the Final Molds

Once the sandwiched pieces had cured, we took the learning we had from the Heat transfer Vinyl Inflatables and trimmed the edges of our silicone casts.

test

We first tested inflating the 3D printed silicone mould casts. We created a little puncture for airflow and began pumping air into the moulds.

At first we tried inflating with a syringe, beacuse we wanted to be carefull to not tear up the cast with too much air flow. BUT the silicone was super resiliant! So we shifted to using the Key board Dust blower.

As we pumped air, we saw them poof up, but not compeletely. Something was OFF!

test test

We tried inflating the other pieces and realised that we had made a mistake. We had glued the ENTIRE surface, where as we were only supposed to glue the SIDES!! We had never seen this detail in any of the tutorials and wished someone had flagged this.

test

TIP for the FUTURE: Do NOT stick the entire surfaces together with Silicone. ONLY stick the SIDES!!

As demoralising as it was, having put in soo many hours into this week's exercise, we atleast had Partial Inflation, and realised details we hope future students don't need to make. It has been a huge learning.

Inflating 3D Printed Molds

We atarted with inflating the 4 Legged gripper

And then tested the Wendy mold

Inflating Laser Cut Molds

And this was the half-successfull inflation of the Acrylic Mold Design

Learnings

IDEAL Work Flow:

  1. Explore and play with available designs in Heat Activated Vinyl
  2. Once there is clarity on how these shapes work, draw up the Initial Design Draft for Self Design
  3. Test this Initial Design in Heat Activated Vinyl
  4. Incorporate the learnings from the previous step and update the drawing
  5. Make the laser cut drawing and hope for the best

Setting up Heat Activated Vinyl

I had never worked with Heat Activated Vinyl before. So just before i was about to begin with it, i ended up tearing a part of the sheet to test the heat settings. Only then did i realise that the vinyl has a peel off layer and acts like a sticker on that side.

Secondly, we used a home iron used for ironing clothes. You need to KEEP a Baking Sheet BETWEEN the vinyl and the hot surface. Else, the Vinly WILL STICK to the iron and ruin it. I never knew that Baking Paper is THAT resiliant as a material!

During the test run, i had put the iron on super hot and the vinyl got super fragile. So for the initial explorations, i put the iron on a less hot setting. However, THAT was a MISTAKE that we realised when we inflated the designs. My tip would be to KEEP the IRON HOT. Once the vinyl cools down, it will be fine.

Another learning we had was how important the airway chanel is in each design. Ensure that the Baking Paper design has an inflow airway chanel.

Also, while inflating the vinyl, it WILL NOT inflate to it best at the start. BE PATIENT and keep pumping air a few times. Once all the airchannels have had some stretch, the vinyl will start actuating properly.

A tip here is to inflate the vinyl on both sides of the baking paper. We found best results once both sides had air pass put through it a few times, for the final movement to work properly.

Material Selection for Laser Cutting

Acryllic was the most accessible material and it worked really well. However, we didn't find any details on the thickness of the materials suggested.

Based on our experience, we used 2mm sheets and found that the ideal width of the mould works well with this selection, not too thick and not too thin.

Preparing the Laser Cut Drawings

Since we didn't find details of the preferable material thickness, we took our chances with setting up our drawing. After the First attempt, we learnt through trial and error, and the drawing uploaded here have the relevant designs prepped and composed for 2mm acryllic sheets.

2 base layers: 1 for the base with the design + 1 for the base cover on the other side

3 boundary layers: 2 for attaching to the base with the design (for the perfect height) + 1 for putting on top of the base cover on teh other side

2 design layers: 1 for use + 1 for backup

Handling the Molds

We had a silicon leak happen on both the 3D printed moulds. Luckily we had kept a plastic sheet on our table before we started working, and and A4 sheet below each mold we were pouring into. It helped us indentify where exactly the leaks were coming from, and also helped during cleanup.

For the Lasercut molds, we needed a LOT of Feviquick, more than we had expected. And the glue took a while to dry as well. So we figured out 10 mins time slots to work on the mold. It was also important to seal ALL the EDGES of the various layers to ensure there is NO LEAK.

Sandwiching the Cast before inflating

MAJOR TAKEAWAY: Do NOT stick the two surfaces together!! Only STICK THE EDGES to each other.

As obvious as this seems in retrospect, and despite having this clarity during the Heat transfer Vinyl Inflation, when you are the one working with a quick dry material that has a shelf life of 45 mins to work with, sometimes it is difficult to put two and two together.

My hope is that no one else needs to make this mistake after learning from our boo-boo.

Fabrication Files

Model for Lasercutting

3D Print file for Wendy Mold

3D Print file for 4 Leg Gripper

Heat Activated Vinyl Design Shapes

Personal Design Lasercutting files for 2mm Acrylic


People to thank for this week

Shefali: for being my constant comapanion and an extra set of hands on this assisnment, be it sourcing the materials, or getting the work done!

This week in emojis: 🀩😲🀌πŸ₯ΊπŸ«‘πŸ€πŸ§‘β€πŸ³πŸ€³πŸ§‘β€πŸ’»πŸ˜