Looking for others to road test new extrusion system

Hi There,

I am in the final prototyping stage of a new extrusion system and looking for others to road test the system. It prints flexible filament twice as quick as the manufacturers recommendations and is a vast improvement on the existing pinch type roller extrusion system as used in all other FDM style machines.

I am based in Sydney Australia and would prefer to start local.

So if you are in Sydney and want to print flex with no probs and are willing to give feed back post a reply or PM.

If there is an interest in the extruder system in general I can post more info.

Arnnie.

How is your extrusion system different? How does it work? We'd love to see pictures or diagrams.

Quote
LoboCNC
How is your extrusion system different? How does it work? We'd love to see pictures or diagrams.

I'll buy the clone on Aliexpress next week

I am a little conscious of blatant self promotion on this forum - but -.

Here is a link to a vid showing the system printing Ninja Flex through a Bowden cable. The part was sliced and printed at 80 mm / sec. and was printed for a customer who was very happy with the results.

Flex wheel Bowden Cable mod

Once I have road tested the extruder on other printers and it works I will be embarking on a crowdfunding campaign with the view of manufacturing a quality unit at an economical price. Hopefully launch in the coming months.

A.R.

Printing such soft filament through a bowden isn't easy, but it seems you went a bit far with the extruder. A while ago i designed a Bowden extruder for Ninjaflex as well and got similar speeds as you did. The key problem was always to keep the filament from buckleing after force was applied, not the application of force itself. You can find a video of the extruder here [www.youtube.com]
Good luck with the project.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Printing 3mm flexible filament isn't as difficult as trying to print 1.75 flexible filament. I've seen/heard a lot of people have had little to no issues with 3mm flexible filament in a standard extruder like the wade's extruder design or the e3d titan extruder even with a standard bowden tube. The difficulty comes in with 1.75 flexible filament in a bowden tube.

Agreed. I print in the 3mm flavor and flex PLA had no issues on my wade extruder.

Thanks for the feedback, the flex wheel will be available in 3 mm and 1.75 mm, in both Bowden cable and direct drive set ups.

Here is link to the Flex wheel printing the ‘Flexible octopus’ in 1.75 mm through a Bowden cable with the item sliced and printed at 70 mm / sec. This filament has a shore hardness of 85A (not the pretend flex that is shore hardness 90+).
Flex Printing 1.75mm 85A through a Bowden Cable

I don’t quite agree with Srek as I don’t think ‘you can go to far’ with extruder design after all I believe it is the most critical part of the printer. Why be limited by the pinch type roller. This extruder eliminates the issues with printing flexible filament and makes printing flex no more of a challenge then printing with PLA.

I use standard 4mm OD Teflon tubing with a 2 mm ID, all gaps along the filament path are filled (inside the push connectors and hot end) the force generated by the flex wheel is more than sufficient.

Here is another link to the direct drive mod printing the softest material printed to date, a crazy soft 46A ( I have posted this link on another forum thread so sorry for double dipping). This material is homemade filament where the tolerances vary widely so print speed were slowed in this print. The direct drive can print commercially available 1.75 mm 60A at 50 mm /sec. with beautiful results.
Printing 46A flex filament

A.R.

What size heater is on your hotend and how big is the block and nozzle?

I've got the Flexion extruder on an E3D V6 hotend and it is capable of running at 60mm/s, but certain Ninjaflex colors don't melt fast enough to be printed at that speed. Some are fine, but some colors just don't like to melt that fast. In my machine I've not used anything but the Ninjatek Ninjaflex 1.75 filament.

I have used many hot ends during testing and I have found the most critical factors is that you need PFTE tubing extending down towards and as close to the hot end as possible, preferably being one piece all the way back to the drive mechanism. If it is not a one piece system make sure there are no gaps along the feed path, for some hot ends I have used combinations of brass tubes slipping into PFTE tube so the filament travels unhindered. I also set my temperature at the lower end of the manufacturers recommended range as this generally provides the best finish.

As far as heater block size the first vid with Ninja Flex 85A at 80 mm / sec uses an original type Jhead heater block dimensions 17 x 17 x 10 (mm), in the 1.75 mm Bowden cable vid the heater block was 20 x 20 x 10 (mm), both preformed within spec of each other, I also use the E3D's and others generic type hot ends (you may notice different combinations in other video posts). I use standard 12 v heaters, basically whatever is supplied with the hotends.

To color a master batch only requires the addition of a small percentage of colored pellets, which should not effect the transition temperatures significantly, if you having issues there may be a variation between batches, check the filament diameter as the flexion is critical in that department and needs to be dialed accordingly. (shameless self promotion coming) The Flex wheel extruder is not sensitive to variations in filament diameter, if the filament can fit through your hot end the Flex Wheel will control and print it, you also never have to calibrate the flex wheel, just thread in your filament and the wheel does the rest.

The limiting factor with the speeds I have published to date is due to the fact I have not spent as much time on the printer build as I have on extruder drive head development, I am happy for someone with a faster calibrated machine to test the extruder for feed back (PM for details).

AR

Edited 1 time(s). Last edit at 07/28/2017 05:35PM by Arnold Lane 1000.

Guidance along the filament path is realy the top priority. In the video i linked i use PTFE tube as well. At the point of traction it has two carved out parts for the idler and the filament pulley. My newer direct extruder designs have a filament guide printed as part of the extruder body.

For a bowden extruder like yours it makes a lot of sense to use a continous bowden, i expect your design to work fast and reliably, regardles what i think about how you implemented the drive

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Just thought I would post a quick update, thanks to beta testers, we are edging closer to a release date.

In order to desiminate more information about the extrusion system I have set up a website -

www.flexwheelextruder.com

Hope for more feed back / comments and another beta tester (Australian based, sorry to the OS requests).

Edited 1 time(s). Last edit at 09/17/2017 07:51PM by Arnold Lane 1000.

---

www.flexwheelextruder.com

its a peristaltic pump for filament

Hi, just signed up to the forum whilst looking for data on Traction Extruders.
FYI, I have designed a version of a traction extruder that uses a 200T GT2 belt and geared wheel configuration.
I have it on thingyverse: [www.thingiverse.com]
It has been running well for a few months now .
(Note the "output" is "UP" in this image)

It can print with a maximum filament force of about 2.5KGf (Say 25 Newtons) (measured with my force sensor (also on thingyverse).
With my original (wade) extruders, this sort of force was enough to completely grind the filament so it would not re-feed.
With the Belt (or traction) extruder, there is no damage to the filament at all (but the motor will slip if the filament is obstructed).
in effect I can have unlimited retracts without grinding consequences.

As #Srek noted, you do need a feed system that takes the filament "off" the belt without allowing too much play. I have a very similar feed part to the illustration Srek added.

One thing I have found and would welcome discussion on is modelling the belt path. This would appear to be a non trivial exercise. I have made an excel spreadsheet (attached) that attempts to do this, but the final (exit) path on my design turned out to be 5 degrees off from the "as designed" CAD.
I have produced a modified output block to accommodate this, but would like to get the "CAD model" nearer the actual path. I suspect I need to add the belt thickness into the calculation better, but perhaps someone has a better belt path calculator?

Looking forward to discussions.
Dagnall

see
Innovative new type of extruder: the Vaeder V-struder

[forums.reprap.org]

Yeah, I've seen this type of system in several other places (like some mentioned above), I'm not sure the reprap forum is really the place for this design considering the "patent applied" at the end of your video.

Could you show a video of how it copes with 1.75mm? 3mm is considerably stiffer than 1.75mm, so 1.75mm will be the real test. I don't exactly see how this improves performance with flexibles specifically though. The problem people have with flexibles is the hysteresis they introduce through elasticity and their compliance. I don't see how changing what pushes the filament makes a particular difference when it's how you restrict the path of the filament that helps with flexibles.

Hi and thanks for the feed back.

The extrusion system is a spin off from a larger project that we have been working on which involved printing a TPE with a shore hardness of 70A through a Bowden cable at +80 mm / sec with 100% part density, i.e. the there was no tolerance for under extrusion - the volume that was meant to be extruded - had to be extruded.

As per Trakyan (I will reply in reverse order)- Yes restricting the feed path after the extruder mechanism is an essential component of printing with flex, however the problem we faced with the hob gear system was that in order to achieve desired results we were basically 'squishing' the soft filament - this caused the cylindrical profile of the filament to distort and become elliptical, trying to feed a elliptical shape into a round PTFE tube did not work, the solution was to either provide a little bit of distance between the hob gear and the bowden cable to allow the distorted shape to resume original dimensions - however this only made a cavity for filament jams. The other solution was to spread the load over a larger area - this way you could apply an amazing amount of force with out filament distortion. The other problem is that the hob gear is metal - which readily slips on polymer - the serrated teeth helped but are not 100 % efficient when dealing with the forces we were extruding at and the potential for debris to clog the hob gear was ever present. So switching from a metal component to a polymer was a much better solution.
Being able to generate more force without associated problems meant that we could overcome the additional hysteresis experienced after the 'cold end', Being able to increase speed also meant that we could 'unwind' and reapply this hysteresis more easily adding to better retraction capabilities even with very flexible material. - It basically worked 100% for our project.

Another rapid prototyping company asked if we could produce flexible items for their clients so we re calibrated the set up to see what could be achieved in this regard - which took the extrusion system in new directions- with additional fine tuning it turned out that the extruder was capable of more than we had anticipated - And in fact in could even print 1.75 mm through a 800 mm Bowden Cable at + 70 mm / sec with commercially acceptable results + print ultra flexible material if required - see video link on post dated 27 July.

We felt we had a new type of extrusion system that could help others so I wanted to network - hence my first post - Looking for other interested individuals to beta test - I am conscious of the conflicts of the open source nature of this forum and this development and have tried to be a forthright between my original intentions and crossing over into self promotions, but I thought I would follow the conversation and respond accordingly, there fore this post in response to past comments as I feel there is some value for others as well, if I am wrong please advise.

I like Dagnell53's force sensor which I shall build and mess around with, I did a very much less scientific test with video results posted here - Load testing the flex wheel extruder

The reason for patent protection is that the mechanism forms part of a larger project where we are retreading motor vehicle tires through the implementation of 3D printing techniques, this project is also wrapped up in patents - which are an unfortunate but necessary part of doing business (i.e. when seeking govt grants and investor capital etc.).

More self promotion coming ---

I am happy to discuss the retreading project further with interested parties particularly we are on the scout for advanced material specialists in the field of overmoulding TPE's to a styrene - butadiene substrate.

---

www.flexwheelextruder.com

Promotion is fine, people do it all the time on here, I wasn't criticizing you for that, just wanted to say a patented product might not be a hit on the forum.

As far as feedback on your extruder goes, from what I can see in your video, the filament does a complete 180 degree turn in the extruder around a relatively small bend radius. With flexibles this is not an issue, the larger wrap angle just means more contact areas and more grip. The problem would arise with more rigid materials like PLA and particle filled materials which are much more rigid and probably wouldn't respond too well to such as sharp bend radius and large wrap angle. Perhaps you could make the wrap angle adjustable to accommodate stiffer filaments which would have trouble bending so much.

Fair point about filament deformation and how your system addresses it.