Help with PP/PE warping

[Reinier1994]

Hello guys,

I'm working at a company that has a lot of PP and PE waste, so they like to start 3d printing with their leftover material. I know these plastics are not the best materials to print, but that's the challenge. I have to find a way to make it work, so their waste gets reused.
The current setup is:

ABB IRB welding robot (from 2003)
Leister Weldplast S2 hand extruder. The extruder is connected to the robot I/O-board, and only has an on/off filament feed control. It's impossible to adjust temperature/filament speed while the robot is doing it's routine.
Vacuum table.

Right now I'm experimenting with HDPE filament, but the problem we get is that there's so much plastic, the internal forces get so high that a lot of warping occurs. I have attached some photos of the current setup and the problems. In the long run, the extruder will be fed with either granulate, or extruded filament out of granulate.

higher res images:

Huidige opstelling.jpgWarping.jpg

As you can see, some products turn out nicely, but a lot of them suffer from too much warping, to make it a valuable addition to the production plant. The following parameters have been tested.

- Filament feed speed
- Layer height
- Layer width
- Filament temperature

A lot of warping has been solved by using a vacuum table, on which a PP plate is secured with clamps on the edges. I had some thoughts on solving this issue, and I'd like to hear some suggestions and ideas from you guys.

- Heated pad, or a heat circulation. It's impossible to build a heated room, since the robot head cannot withstand a constant temperature of above 70C.
- Fluid cooling (oil/water) inside the workpiece. Let's say you're building a case, and after a few layers you start filling it with enough cooling solvant, so the piece cools down, except for the top 1-3 layers (so the next layer is hot enough to attach).
- Warm fluid inside the workpiece. Since it's not possible to make a heat room, could it be possible to fill the piece gradually with fluid between 50-120C, and when it's finished, cycle fluid through the piece while gradually lowering the temperature?
- Cool air gun. Make a cold air flow behind the extruder, so it cools down faster (but not too fast so the molecular structure is affected).
- Smaller layer width. By adjusting the extruder nozzle (filament is either 3 or 4 mm) it's possible to lay a smaller line. This means there's less material and hopefully less internal temperature differential.

Thank you for your time and hopefully I can work something out and test your ideas!

[curious aardvark]

That is totally amazing !
Just love the way you're using a robot arm as a 3d printer :-)

You don't want to cool it while printing, you want to keep it hot to cut down on warping and then cool it all at once fairly gradually.

Ordinarily a print would not be watertight so the water method would not work.

That said, with you using such large beads, there should be a good watertight seal between layers. So if you can work out a system to use hot liquid inside a hollow shape - it ought to work.

Alternatively, and as you are using a robot arm not a rod mounted extruder.
You could build the model inside a watertight container. Ie: have the plate and model in a topless box and the arm can reach over the side and print inside the box.
That way you can fill the box with the hot liquid and it should act like a heated enclosed print space.

As long as you can work out a pump system to keep the liquid just below the current layer of print and keep the liquid hot, and be able to empty the box after the print has finished and the liquid has naturally cooled. It should work.

[Sebastian Finke]

Printing PP and PE... where to start...

Your prints will warp unfortunately. PP has shrinkage value of approx 1.8% and PE anywhere from 2-3% depending on the grade. I can only suggest using a very heavy heavy raft and make sure your layer adhesion is spot-on. There's a reason why these materials aren't popular for 3D printing. Both are easily recyclable though so if you aren't doing any injection moulding yourself you might be able to flog it off to someone that is.

[Mjolinor]

A container filled with a gas would be better than a liquid. Something like CO2 or Argon that is heavy will stay in an open topped box even when the gas is warm and it will stay warm easier than a liquid. Using a gas you would only need to heat the bed in order to warm all the gas as convection currents are sufficient, if the bed is not sufficient then a heated container would certainly do. You would have to watch the temperature of the head though if it were completely in the container.

If you wanted to get really enthusiastic you could have very cold air coming out from just below the head like they do in hot volatile liquid wash baths to prevent loss of the volatile liquid through evaporation.

If all else fails just give the whole set up to me, I would dearly love one of those arms. I keep looking for a cheap one on ebay but they are always too much money.

[Reinier1994]

Sorry for the late reply. Thank you all for your insights and information. I've started printing and I'd like you to see some results so you maybe can share some tips on how to improve the print quality.

Biggest issues right now:

- Corner warping/shrinking. PE-HD shrinks a lot, and on straights the welding gun can just push the raised layer a little bit, and improve adhesion. In corners the product falls inwards, resulting in a lower height and circumference. This results in the next layer being laid down aside, so there's no adhesion, and every next layer gets even worse.
- Bed adhesion. This is not a huge issue, but the first layers need to be pre heated by a heat gun. Warping isn't a big issue, the vacuum table actually holds down the ground plate the entire time.
- Print speed and filament feed. Since the project's goal is to recycle PE-HD waste, it's not a solution to lower the print speed for a better quality. There has to be a way to print at around 20-30 mm/s with the best quality possible.

My settings I have been playing with right now are:
- Air temp (of the welding gun blower) : 180 C . Since the air doesn't really do anything, usually the air gun heats the welding surfaces, but the previous layer is always hot enough to bond with the next one. This is the lowest possible setting without disabling the wlder.
- Extruder temp: 200-240 C. I've been trying to get it as low as possible, so if I turn up the welding speed, it doesn't collapse due to the high heat.
- Layer height. I start with a layer height of 2mm, with the first one being pushed on the bed (25% first layer height) and ending at 3 mm. The next process has layers of 4 mm, starting at 3 mm.
- Extrusion speed. Since it's a welding gun, I cannot give an exact number, but it's around 1,0 kg/h.
- Filament width is 4mm, and the extruder has a nozzle of 8mm. Since this is a welding gun, I have no access to smaller nozzles. I think a smaller nozzle will result in a significantly better result. I want a small extrusion, being pushed to the side so the material is as curved as possible on the sides, so it can release it's heat.
See this album for a lot of pictures:

http://i.sli.mg/a/sluwww

As you can see, the first finished product (the tub) went pretty well. This design was meant to be wide in the middle, and smaller on the top and bottom. Only downside is that the print speed was 10mm/s so it takes up to 6 hours to finish. This is not a valuable result :P.

As for now I'm not in the position to buy a heated bed or chamber, but I can ask to mill a welding shoe with a smaller nozzle.
Thanks everyone in advance for your replies and insights.

[Reinier1994]

I have been experimenting with infills, and I'd like to try a brim aswell, since the biggest problem right now is warping in corners. Since the first layer adhesion is fine, and the vacuum table pulls the baseplate straight during the entire print, the corners always want to go inside. This causes huge problems after 5-10 layers, since the radius of the corner goes down by 30-50%. Every next layer falls down and makes the print shrink even more.

Right now I've tried three infills. My conclusion is that infill in the middle and on straights is a bit useless, so I'm focussing most on corner strengthening. Since the PE extruder is a simple on/off device, I cannot work with travel points, so my pattern needs to be continuous in the printing process. Here you can see some of the results:

http://imgur.com/a/JTGEh

I'm pretty sure I want to (for starters) decrease the huge nozzle width of 8mm. I''d also like to see if you guys have an idea which type of infill or outline adjustments could make a bettter result. Thanks for the help!

[curious aardvark]

these days I pretty much just use a triangular infill. Seems to stress the printer less than any other and offers all round reinforcement and fairly low infill levels.
That's through simplify3d.