On extruding reliable filament (Page 1 of 2)

225C?! Holy moly what gave you that idea?!

The specifications for that brand of ABS, it says melt and nozzle temperature are between 205 and 245.
Could you explain where this 25°C difference (205-180) is coming from ? and if I should deduce 25°C from every specified melt temperature for other materials, like PLA ingeo 4043D (210°C).

(took some time to design a custom funnel and hopper)

I tried again at 180°C this time. The bubbles disappeared and the length of the good/straight filament increased, but it's far from enough. I was able several times to extrude around 3,5m of straight filament (and one time 8m) but sooner or later the filament tangles just after the nozzle, gets blocked on the copper pipe and I end up with a pile of noodles.
Any wind blow can ruin the filament before it cools. But even if an enclosure is made to prevent that, the way the filament falls, turns, piles up on the floor leads sooner or later to tangling at the nozzle. I still think that you can't make reliable filament out of the filastruder. Without some spooling mechanism it's a plug and hope for the best device.

P.S. extrusion speed: 27,3cm/min
filament diameter: 1,7345 ±0,035mm

Just saw the pictures in the OP. The filament guide block is way too far from the nozzle face. Did you follow the instructions on filament guide block placement? It says in the instructions that the top of copper guide is to be 1.4 inches from the nozzle face. You have at least 2.5 inches of space, maybe 3.

27.3cm/min is right where it should be. 75cm off the ground is a little low - shoot for 100 to 120cm for best results, with nothing the filament could hit on the way down.

The horizontal distance (distance along a line parallel to the ground) should be 1.4in, or about 3.5cm.

Cooling the nozzle is indeed important. The idea is that you keep the melt temperature higher than the nozzle/die temperature (the pro extruders do this also). The pro extruders usually don't need a nozzle fan because 80% of their heat comes from barrel shear, so theres a natural temperature gradient between the barrel and nozzle/die. Most of the Filastruder's heat comes from the heater, so we create the thermal gradient with a cooling fan.

Really Tim we need a phablet titled "Why I did this" to help us understand why in the world you had the fan cooling the nozzle.

That was going to be one of my first modifications to see if I could improve production.

Ralph

"How about just trusting the work of myself and all the beta testers, who collectively had over 1,000 hours of runtime during the development stage? "

I know, I know sorry, it just didn't seem logical to be cooling the nozzle.

I don't have any understanding of the extrusion process since my Filastruder is sitting waiting on my 3Dprinter.

I have read the 58 page sticky but I really do not remember any specific points.

Documentation helps to drive sales but so far you seem to be doing well, hopefully it will continue.

Ralph

Experimenting is great. I am all for it. I am VERY far from being an expert in the field.

What is bothersome is when people don't follow the instructions, then complain their "filastruder" isn't working. Trouble is, they haven't built a filastruder at all - more like a (personsnamehere)struder.

Moral of the story: build as outlined in instructions, you should have a Filastruder that works well.

Edit: Another example of something like this is the shaft coupling. Some people might think it is a product of being dumb/lazy/etc. I know a fair bit about shaft couplings. Story time!

During my undergraduate studies, I was asked to do some work for a local power plant. This plant was having trouble with their boiler feed pump. This is the pump that feeds the steam generator on a combined cycle powerplant. It was 1,000HP, and the motor and pump were each about the size of a Honda Civic.

(That is just a picture I grabbed off the internet.)

Anyway, this plant was destroying their shaft couplings. These couplings had to withstand around 1200 ft-lbs of torque, and were comprised of dinner plate sized discs. Like this:

They had 8 boiler feed pumps in service, and were destroying roughly one every 2-3 weeks. That tripped the unit offline and resulted in a loss of 200MW of generation for 48 hours or so. Management was not happy.

I walked over to the marketing coordinator's office and asked to borrow their DSLR. I cranked the shutter speed to the max, set it in burst mode, and held the camera under the coupling's cover. In hindsight, it was really dumb and unsafe - if the coupling came apart at that moment I'd probably be missing a hand today. I walked inside and looked through the pictures. The discs were deforming only in one quadrant, which told me it was an alignment issue, not a overload issue. (if aligned properly but overloaded, the discs will deform in all 4 quadrants). The misalignment was fatigue cycling the metal disc packs, leading to premature failure. Some of the pumps exhibited no flex in their couplings. As the couplings detoriated, I could see individual discs failing, and warn management that a failure was imminent. The next time one failed, I helped revise QC/QA processes on the laser-based alignment between the feed pump and motor. No more failures after that.

Back to Filastruder: The shaft coupling was chosen with these goals in mind:
- allow for radial, axial, and angular misalignment
- mechanical fuse functionality (the washer will shear on overload)
- easy to source a replacement if lost or damaged ($1.99 at Lowe's)
- cost: a Lovejoy (or similar) coupling capable of handling the torque would have been the most expensive item on the BOM, by a good bit.

In hindsight, the second goal didn't work too well. The washers ended up having +/-30% difference in thickness, resulting in unpredictable shear loads. Post-KS, I'll probably just use a stainless steel pin instead, and forego the mechanical fuse capability. Maybe a PTC self-resetting fuse on motor current.

Heh, yeah, the user can be blamed for not reading the manual.
The thing with this particular product is that the general approximation and lack of information in the instructions as well as the low technicality of the kit lets people think (rightfully or not) that it didn't require an engineer to design it and that it will work as well if not better if some modifications are made.
A manual can always be improved. If something is very important it should be flagged. Most things might be important, you'll say, but there are points of frequent confusion, we found one with the fan and the nozzle. The (uncommon) height of the table should definitely be added.

There's nothing the filament can hit on its way down, right? No table legs or table supports?

Painters tape sticks to polymers. There's a reason it is used on print beds. Use aluminum foil or nothing.

The board looks a little long relative to the nozzle face. Try cutting an inch off.

One other thing to try - extrude at 170-175C.

Now re: the fan cooling the nozzle question above.

I went through a lot of the pictures people have posted of their extruder some working great most with problems and see that there is a lot a varieties of placing the insulation around the nozzle. Some have the nozzle protruding by a 1/2" others have it kinda flush and some have it completely wrapped.

Tim in another thread said:

          "You should tilt the fan away from the nozzle 10-15 degrees, and make sure the insulation is slid forward, flush with             the nozzle. That should take care of your temperature problems."

If the insulation is slid forward flush with the nozzle how much of a temperature gradient would you get? You'd get some cooling of the nozzle, I suppose, but not as much as if you had a 1/2" exposed.

This kinda stuff is what made me think I'd have to experiment with the fan placement and then of course Ian posted a picture of his vertical extruder without a fan.

I sure would like to see "This is what is recommended and this is the reason why" noted in a pamphlet.

Obviously there is confusion that was not clarified in the sticky.

As far as:

           "I'd much rather spend any free time i have on R&D than creating a (massive) post on each design decision, when most of that information has already been documented in the sticky."

Spoken like a true engineer, I spent a good part of my professional career fighting with engineers over documentation, especially easily read informative documentation, I usually lost.

Ralph

The instructions are vague about the insulation placement too even if it matters.

Putting the filastruder vertically was what I was going to try next - have you got a direct link to that thread ?

------

In the previous run I forgot to put back my hot end enclosure. Now I put it back, I didn't cut the board shorter but I put aluminium foil at its edge, I cleared the floor and I moved the insulation so it's flush with the nozzle face. The result is not one tangling until I stopped the machine (and a matching temperature). Several tens of meters of straight filament (I didn't measure). I think that the enclosure is a real game changer.

Then I switched to the supplied PLA. I poured some pellets in the hopper after it was empty of ABS and after some times I set the temperature to 160°C.
I also got many meters of straight filament but... it was very thin and very fast:
extrusion speed: 70,8cm/min
filament diameter: 1,305 ±0,085mm
(I made sure that I measured pure PLA not an ABS mix. Until I get masterbatch I'll use some coloured plastic to mark the coming of a new polymer - anyone doing this ?)