Rubber 3D printing – Makerbot Replicator 3DP Rubber Filament Mod

3DP “rubber” filament mod… or rubber-like fused filament fabrication (FFF)…

Hi and welcome, below you can see my work to modify the Makerbot Replicator (dual) so that it can print a rubberlike conductive and flexible TPU material.  BTW my electrically conductive TPU used below can now be purchased here:

1.75mm / 2.85mm

Also see the dedicated homepage:

rubber 3d printing - Be your own flexible filament "rubber" 3D printing professional




So why did I want to print rubber? Well it´s sort of the missing link;) Now I can make: Dampers/Chock absorbers, Tires, Tank treads, Bellows, V-Belts, Timing/Synchronous Belts (or any other type to transfer power), Couplings, Grommets, Seals etc :-)


Step 1:

I have developed a suitable TPU filament, I named it PI-ETPU 95 Carbon Black, see some material data below:

PI-ETPU 95 Carbon Black  (Typical values shown, print method might alter these values)
Diameter ø 1,75 mm Aprox (always measure before use)
Density 1,30 g/cm³
Tensile Strength 15 MPa ISO 527
Yield Strength 15 MPa ISO 527, 10x4mm part (not FDM)
Elongation at Break 250 % 10x4mm part (not FDM)
Volume resistivity <300 Ωcm PRE021, 10x4mm solid part
 Surface resistance  <200.000  Ωcm  IEC 61340-2-3, ANSI/ESD. STM 11.11, 400µm solid thick sheet
Hardness 95 Shore A ISO 868 / D-2240
Color Carbon black
Recomended Temperature range 190-230 °C Depending on 3D printer, nozzle and print speed
Self-ignition Temperature >340 °C Aprox
Pre-drying 90°C for 1 hour Only needed for optimal performance if the print surface starts to show small bubbles


See some test prints further down on this page…



Step 2:

I designed a new “filament feeder assembly” for the Replicator, I call it PI Drive Block (original design can not handle soft high friction filament at higher speeds).

My first version of the PI Drive Block works really great for rubber, later I will upload the STL files to but you can download it below now (It will not handle the standard plastic filament and it is designed for the Left extruder). The next generation will handle both rubber & plastic within the same footprint as the original but might be even more difficult to get accurate enough in the Makerbot (still work in progress).

If the demand is big enough I might even invest in tools to make real moulded parts to get the best performance/tolerances/reliability (yes I have the knowledge and connections needed).


You could also try some other ball bearing versions like but the lack of guiding below the ball bearing is not good…


Step 3:

To get everything fully automated the 1/4″ feed-tube is replaced (going from the print head to the back) with a PTFE version tube. I also made an low friction spool holder suspension. Since the rubber is not as stiff as the plastic it can not turn the original spool holder without risking bad flow from the print head, with the original spool holder and feeder tube hand-turning is a must…

The new spool holder design works incredibly good, almost zero friction and also much faster/easier to unload/load a new spool :-)  I designed it for glass marbles (aprox ø16mm) since most people have them easily available. Yes I will upload to later… The parts need to be “glued” with acetone to get the best result.


Download the PI Spool Holder One-Piece





Test prints:

TEST 1. A test-print using PI-ETPU 95 Carbon Black, size 30x33x8mm, 4mm thick walls (layers=0.27mm, fill=100%, temp=220°C, feed=30mm/s, cold table):

See movie below: To avoid warpage I use the trick to add PVA-glue (in Swedish: Trälim ;o) to the surface to get a good “stickiness”

See more tips on avoiding warpage here



TEST 2. Another test-print using PI-ETPU 95 Carbon Black, this time the center piece of this 3D printed ø8 mm axle coupling (layers=0.34mm, fill=100%, temp=220°C, feed=35mm/s, cold table):



TEST 3. Size T5 synchronous/timing belt (and simple tank tread belt, later used in the quickly designed caricature WWI toy tank for my son: Renault FT-17) test-prints using PI-ETPU 95 Carbon Black, 1mm thick, 4mm wide, layers=0.34mm, fill=100%, temp=220°C, feed=35mm/s, cold table (Pulleys in ABS) As you can see the first print is a bit messy since the filament diameter settings was wrong. The 95 shore hardness secures a rather good traction power, it feels willing to transfer some power without stretching too much: 

2013-10-01, PI-ETPU 95 Carbon Black successful extrusion

As you can see in the text below it has been problematic to find a skilled extruder company. I got tired of the mediocre tolerances they could deliver and said to myself – how hard can it be?

Ok so I purchased a small lab/hobby extruder, that one did not work either but I managed to design a new setup to get it right :-) and now I have several meters of filament within 0.05mm tolerance and a perfect smooth/glossy surface! I felt very satisfied when I finally hit the bed early this morning :-)  Stay tuned…

PI-ETPU 95 Carbon Black successful extrusion

2013-10-14, Perfect print result

Recently I have been making TPU “rubber” 1.75 filament myself since the different extruder companies I contacted could not reach the tolerances required. I have found a good technique and are now building a dedicated extruder that I can easily duplicate so that I can get a “extruder farm” up and running. All that takes time and for the fun of it I did a test print today, as you can see the result is amazingly good :)

See movies:


2014-03-10, Update

Sorry for the slow progress but now things are moving again. I´m running my 2:nd generation extruder and I made 2 rolls of filament this weekend. I measured the 2:nd roll´s filament diameter rather frequently (digital calliper) and the entire length was between 1.71 and 1.75 (+-0.02mm for approximately 0.9kg of TPU filament). Even if the calliper is not the most accurate method I´m still satisfied with that :-)

I have started 4 market trials to get some end user feedback on the material, 1 trial in Germany and 3 trials in Sweden. They are experienced and well known actors in the 3D-printing world. the feedback so far is great :-)

Now I´m designing the 3:d generation extruder and that version will be made in bigger volumes to really test the design.

Kind Regards /Thomas



2015-01-07, Update

Now things are moving again. I´ve run 5kg of 2.85mm filament this weekend with great tolerances :-) anyone that would like to buy samples can contact me here (also for 1.75mm)

Kind Regards /Thomas

The conductive TPU filament is for sale. Go here to find out more:



14 responses to “Rubber 3D printing – Makerbot Replicator 3DP Rubber Filament Mod”

  1. Asad says:

    Hi Thomas,
    What Printer are you using for this material
    What grade TPU are you using (I mean according to hardness like TPU 98A or TPU 65A etc)


    • palmiga says:

      Hi Asad, I am using it in a Makerbot Replicator and a MBot Grid II (both modified to get a very reliable and precise operation, was working ok most of the time before though). The best 3D printers for printing soft materials has a PTFE tube from the feeder wheel down to the warm nozzle.
      As you can see in the table above the hardness is 95 Shore A.


  2. Chris says:

    Hi Thomas,

    I just stumbled upon this site today and this is exactly what i am looking for. A question or three if you don’t mind. How does this hold up (after it is printed into a part) in the heat(melting Temp)? what is the wear resistance? Also (after looking at the Octo you printed, very nice by the way) what is that tolerance? 100, 200, or 300 micron? Any idea on time to complete? And once this is converted is there a way to still be able to print other filaments?

    Thanks for taking the time to read,

    • palmiga says:

      Hi Chris, glad you stumbled your way here. The bond is very strong but as for all materials the print method itself creates air pockets and thus the material is not solid strait through, thus it is not to be compared with a molded solid part. It’s hard for me to say anything sophisticated in regards to wear resistance, I think it is good but I did not do any special tests or anything. Any suggestions on valid tests? I do not remember the layer thickness I used (if that is what you are asking), I guess I used 0.2mm or something…
      I now have 3 beta-testers of the material (they like it;) but I am still working on the extruder design. It is giving nice tolerances but I need to make some changes so that it´s easier to manufacture lots of them :-) I am not promising a SR-date here since things out of my control are happening right now that is slowing down this project. I do have both a manufacturing partner and a good sales channel already now, only waiting for me to release :-)
      The conversion for the extruder etc is beneficial for all materials (not the currently downloadable fixed-ball bearing version though). Hope that was helpful for you.

      Kind Regards

  3. Duke says:

    Hello, i would like to compliment to you for the excellent work you are doing. I’m very interested in Printing termo plastic elastomer TPE/TPU.
    Unfortunately it is hard to find 1.75 mm TPU filament :( your project is absolutely amazing to me (also for the new extruder design).

    I’m interested in buying some filament (and i could also evaluate to buy the extruder set for the makerbot replicator 2x, as the quality of your work looks very compelling!)

    Do you have any target date for the release of the filament?

    Is it possible to have some samples already?

    • palmiga says:

      Hi Duke, thanks for your kind words :-) I am not surprised it’s hard to find 1.75mm TPU… it’s not easy to make within tolerances in larger volumes;) I’m struggling myself to find the skilled enough extruder company, been through 3 companies now and got many rolls of filament from them that is not up to spec fully. It’s possible to sort sections of filament for good 3D printing, but close is no cigar… And that’s why I don’t want to offer samples just yet. BUT I am not giving up! the 3D printing with this TPU rubber material works too good:-)

      Kind Regards

  4. Carl-Axel Schoug Västerås says:

    Hej. Är intresserad av att köpa flexibel plast för att göra larvband. Har en Rep 2X.


    • palmiga says:

      Hej Carl-Axel, skoj att du är intresserad. Jag återkommer när materialet finns tillgängligt. Det bästa är om du anmäler dig till nyhetsbrevet “3D printable rubber” till höger på denna sidan.
      Med vänlig hälsning,

      Hi Carl-Axel, nice that you are interested in buying this material. I will let you know when it is available. The best would be that you subscribe to the “3D printable rubber” Newsletter on the right side of this page.
      Kind regards

  5. Don Jewett says:

    I am starting a project that will involve designing and building a few models made of a rubberized material, to create a “chinese finger trap” which is something that fits on cylindrical shape, and then when the cyl is withdrawn, the object’s friction with the cyl leads to a lengthening of the trap which causes the DIAMETER of the trap to become SMALLER, and this in turn increases the friction. It would be desirable if the trap were thin, but would have some “rings” on the outer surface to give the gripping effect. There might be some knobs on the inside as well.

    Here are my question: 1. Does your material change its properties say by heating more or less?
    2. Do the properties change as a function of thickness?
    3. How thin could this material be if it were laid down just as an open cylinder?
    4. Do any of these answers change with the different diameters that you have (1.75, 3)?
    5. Would you be able to make other sizes of filament if that effects the resulting model propoerties?
    6. What 3D printer does you device fit?

    Incidentally, I started to answer your questions, but the WebSite told me that I had already entered
    answers, which I hadn’t. I am planning to get grant funding for the project, and so would expect to
    pay commercial prices for your products.


    Don J.

    • palmiga says:

      Dear Don, fun that your asking so many questions and thanks for letting me know the voting is not working, please try that again.

      Let me start by letting you all know that I will not be able to know how your individual use will turn out in the end, the variables are endless and you will simply need to test yourselves. That said I will do some guesswork…

      1. All materials behave a bit different with different temperatures and I would guess a too high temp would start to brake down the material resulting in a stiffer material, only a guess;)
      2. Of course, thicker means stiffer.
      3. I did some successful test with only one layer in width, using the 0,4mm nozzle, turns out great and soft. This might not work with a part too high since the part is maybe flexing and the nozzle will miss the next layer.
      4-5. The actual material will not change significanty with different diam filament but it’s a science in regards of tolerances on the finished part, I will not get into that…
      6. The filament will fit any printer that already has or has been updated with low friction feeding. I have done modifications to a Makerbot printer, but I cannot see why other printers would not be possible. In the future I might make parts for some others as well, time and demand will tell…

      Hope your satisfied with these answers.


  6. KenB says:

    I am interested in this material. Is it bio-safe?
    Where are you located?


  7. palmiga says:

    Please comment… Maybe you want some other materials shaped as filament?