Clicking on the images below will open a full size .PDF in a new tab.
Welcome back for the final time,
First of all, thank you for staying tuned this long. It has been quite a ride and we hope we have been able to give you a good (and not dreadfully boring) insight in our research. This will be our final post on the weblog, all that is left is the science fair next Tuesday.
That brings me to some very exciting news! We have succeeded in both UltiCasting and NinjaFlex printing. Next Tuesday we will have our UltiCaster running at the science fair, as well as a working prototype of the 3D printed exoskeleton for you to try out! If you are interested our process but have questions that are not answered in the weblog, we will be more than happy to share our knowledge next Tuesday.
In case you missed them, here are the details for the science fair.
Date: November 1st 2016, 12:30 – 16:00
Location: “De Tribune”, Central hall, Faculty of Industrial Design Engineering, TU Delft.
We have got a time lapse of the actual UltiCasting made in the picture above. Apart from that we don’t want to spoil to much just yet, so all that’s left for now is signing off.
Hope to see you next Tuesday flexi-friends!
Kind regards from Silicone Valley
Credits for this pun go to Jouke Verlinden
At Thursday 9.30 pm, we have concluded our first, complete, UltiCast. As our emotions of joy welled up like the silicons did in the overflow valve, we were tossed into an historical perspective that only Neil Armstrong could adeptly describe. This is a small step for one hacked UltiMaker 2+, but a huge step for UltiCast.
In all seriousness, we are stoked that we finally made UltiCasting work. Previously we have had major issues with the silicons curing inside the system, and consequently blowing up the tubing and covering our print head in silicons. This is why we redesigned the entire set up all over.
We have removed the right fan of the print head, to use that space for the nozzle. We had concluded that the mixed silicons settle faster when subjected to heat and one of the main reasons for our system to get clogged was the fact that our casting nozzle was right next to the heating element of the first extruder. Also we have once more increased the diameter of the mixing nozzle, and seeing we can now start mixing closer to the bed, we have shortened the mixing nozzle as well.
Not only the silicons proved to be though to work with, also the PVA had a will of its own. Getting the PVA to stick to the print bed proved to be a really big deal. After trying alcohol, window cleaner, vinegar and prayer, we found ye good old Pritt stick. The rest is history, we finished our first print and are waiting for the PVA to dissolve in water to see the final result. All in all, we did it, and we’ve got it on film.
To show the possibilities of FDM printing and UltiCasting soft robotics, we have looked for interesting fields in which the possibility for cheap 3D printing could prove valuable. This could be the case in the medical sector, especially with aided movement. Reducing the cost of treatment and support utilities is a goal that inspires us all, and soft robotics could, in the long term, be a big player in this progress.
This is why we have chosen to develop a 3D printed hand exoskeleton. This is in short a glove with soft robotics actuators to help move you fingers, visible in the concept drawing below. Possible applications in a completely developed product could be to assist people with arthritis, local paralysis, limited hand function or as supporting tool in rehabilitation. Our goal is not to create a finished product for next week, but rather to give a proof of concept showing the possibilities of 3D printing for soft robotics.
Good to see you again!
We are still quite unsure whether to call week 3 a success or a failure. Well at least there was some sort of progress. We didn’t quite meet last weeks goals due to some minor set backs including our UltiCast set up exploding, some burnt plastic and some broken parts. That being said, we are now ready to start UltiCasting again.
As you might have read in some of our earlier posts, we had to redesign the entire UltiCast set up to allow for bigger tubing and wider nozzles. After some cups of coffee and some flexible hours this seems to be about done. In the picture above Max is laying the final hand to the UltiCaster Mk. 2. This one is stylishly equipped with green aquarium tubing and features a new deposition system.
Due to the larger nozzle and tube size we are unable to do the static mixing inside the printer head. That is why we now mix our components before they reach the printer head, then feed them trough using a tube, and deposit them with a changeable nozzle. The advantages of this are the possibility to change nozzle diameter and the allowance of bigger tubing and nozzles. The disadvantages are that we still experience some leaking of silicons from the nozzle and we need to replace more parts after each print.
The script for the UltiCaster is almost finalized to our goal stadium. With huge assistance from our good friend, Robin Sveningson, we have managed to automate the slicing process and thereby generate a g-code containing commands for the fills operations, that can be interpreted by the Marlin firmware on the UltiMainboard. (read more about this in Methods). As soon as the hardware is ready we can start UltiCasting and kick back and relax. Hopefully.
As for the Ninjaflex FDM printing, well just look at that. We’ve never had a print that looks this promising in terms of quality so the settings should be right about perfect. So all that’s left is experimenting with some different geometries to get that real feel finger going.
Our focus now is to get our setup up and running, and prepare something cool for the science fair.
Hope to see you next week!
This article was 3D printed in the IDE Applied Labs Delft
Week 2 has been a week of breaking parts, burning capacitors and freezing Ultimakers. In short: Progress. Not only have we been able to bypass multiple “FATAL ERROR”s, as they are called, we have been able to build an actual UltiCasting setup that is ready for extensive testing next week. We are currently still doing small scale tests with water to confirm the setup but we will be able to test the actual resins and start to actually UltiCast.
While the software for the UltiCasting is almost finished and the hardware is being tested as I write this. We’ve also been making strives on the FDM printing project. Our latest prints are almost completely airtight and are already functional. We do experience some issues with the valve fitting and the filament feed that still cause minor problems but today Christmas came early when our supervisor Rob Scharff brought us a care package.
New valve geometries are already being printed to try out our new goodies as we are planning to set up a new experiment to experiment with different actuator geometries, materials and valve fittings.
Our goals for next week are to set up an experiment to measure the effectiveness of all our different tried methods and to start UltiCasting. We are ready, the silicons are ready, and our UltiCaster is almost there.
See you next week!
Ulti-broad-Casted from: IDE Applied Labs
Yesterday we have made major progress on the UltiCast project. After experiencing some setbacks with the Ultimakers we finally managed to get a second cold extruder running on the Ultimaker mainboard. Before this success our most noteworthy accomplishment was that we not only crashed the Ultimaker during testing, it also froze on startup and for some reason was completely inverted. Luckily with the help of our supervisor Lars Rossing and some tedious hours fiddling through the Ultimaker firmware in Marlin we now have a very promising set up. Our current setup will be available in the methods page.
To celebrate these successes: Here is a video of our latests FDM printed finger!
This week our focus was to get up to speed quickly. After learning what our project goals are, we immediately delved into the literature provided by our experts Rob Scharff and Lars Rossing. From this literature we quickly learned the basic principles of Soft Robotics and UltiCasting and were able to apply these on the Ultimakers.
Our first print was a test piece used to determine the accuracy of printing with flexible materials. After this we started printing basic soft actuators with the aim of making them airtight. After some trial and error this proved to be quite challenging due to the still a bit unpredictable behavior of the flexible filaments. Our progress and used settings can be found in the FDM Test Results tab.
For the UltiCast project we are programming a script that can use the provided G-Code to estimate the needed filling volume and deposit location for the casting automatically. This software will be used to drive the casting process. We are also setting up the hardware for the UltiCast process but we still have to receive the actual silicon filaments to start testing. We hope to really kick off the UltiCast project next week.
Stay flexy, stay sexy!
Ninja Flex is a highly flexible polyurethane based filament ideal for direct-drive extruders such as Ultimakers. Using this material could simplify the production of soft actuators significantly and speed up the process.
We will be investigating the properties of this material, the optimal method of printing it and the possible applications in the field of Soft Robotics. Especially for Soft Robotics it is desirable to create an airtight actuator to gain maximal control of the movements, when pressure is applied. To do so we will experiment with a lot of different printer settings and log our results in this blog.
UltiCasting, researched by Lars Rossing, is a new add on for 3D printers to speed up the printing time needed for filled structures. The basic principle involves first printing a mold and then filling this mold with a settling resin. In the video you can see the process of UltiCasting.
In the field of Soft Robotics these techniques could prove interesting. With UltiCasting it becomes possible to ‘print’ a soft actuator in a mold eliminating the manual casting process and doing so speeding up the process of creating soft actuators.
We will be using a water soluble PVA to create the casting mold which will be filled with a two component silicon fill. This will allow more complex shapes that could not be printed without support to be created on a direct drive 3D printer. If the mold quality is high the surface quality of the flexible silicon will also be very high. This is very hard to achieve in traditional FDM printing because every layer of hot filament deforms the previous layers of filament.
We will be researching the possibilities and limits of this technique and it’s applicability to the field of Soft Robotics