There are a few threads on the forums already addressing PA-CF printing challenges, but it’s worth bubbling to the top again. I’ve got over 1000 hours printing on the X1C with (for the most part) great results. Bambu Slicer’s default material profiles have been excellent for PLA, PETG, ABS.
Then there’s PA-CF. I have not yet done a PA-CF print that doesn’t wrap and lift from the X1C print bed.
Things I’ve tried:
Using 0.8 nozzle
Using clean high-temp sheet WITH magigoo PA glue
Preheat chamber to 40c+
Disable all cooling for first 5 layers
In all cases, PA-CF prints lay down a good first layer (or two or three) and then start to warp/lift. This is with Bambu PA-CF and the Bambu PA-CF material profile.
I know PA-CF is generally more difficult to print, but has anyone found settings that deliver reliable prints on the X1C? Should I try lowering the default bed temps (100c)?
I used to print PA6-CF on a Raise3d. If I remember correctly bed temp was around 80c, cooling 25%, slower print speed compared to ABS but I forgot how much.
Youtube has a lot of material with tips. Look for the “Vision Miner” channel.
It’s critical to always dry the filament for at least half a day at 80C before use, even when fresh out of the sealed bag. Otherwise strange results will ensue.
Note that the CF in it is only for the color because it looks cool. It provides no additional strength or stiffness. All the mechanical and chemical properties come from the PA12 grade of the Nylon itself.
At the price point Bambu Lab sells it, it’s ground carbon dust, not “fibers”. So the question is, do you really need it?
Probably not. At least, not yet. I’m only printing with it now because 1) Bambu puts PA-CF sample spools in the box with X1C units, and 2) I may want to use it for functional parts in the future.
I probably would not have tried “PA-CF” if Bambu didn’t push it. Which is all the more reason I’m surprised it doesn’t work that well with the available 1P build plates and material profiles. ¯_(ツ)_/¯
Next up: I’ll attempt to re-dry the filament for 12+ hours and then print on a Garolite sheet. If that doesn’t work, I’m not sure what will.
Update: I did another Bambu PA-CF print, this time after 12+ hours of drying in a dehydrator, again on the High Temp sheet with Magigoo PA glue and a chamber pre-heated to 40c+.
This print got much closer to working, but still ultimately failed. The PA-CF largely remained attached to the High Temp sheet, BUT the sheet itself was pulled/warped at the edges by the print. The next test is with the Garolite sheet, but I am concerned now that the warping force of the PA-CF (with this large model) may be a problem if the magnetic build plate can’t hold the sheet flat.
For context, I’m trying to print the RepRak bracket from Printables. This model covers a large portion of the X1C’s print bed, so this may also be further pushing the limits of what the X1C can do with Nylon.
Next Update: I got a great PA-CF print using a Garolite sheet (still with dried filament and pre-heated chamber, but no glue)!
I added a Garolite sheet to the blank (“Engineering”) backside of a Bambu Cool Sheet. I went with the 255x255 version from 3DMaker Engineering. It has an adhesive backing, so it was quick and easy to attach it to the Bambu sheet.
Adding the Garolite makes the Bambu sheet about 2mm thicker, but happily, that caused no trouble for the Bambu X1C. It self-set a new z-height on the next pring and laid down a great first layer on the Garolite. And the entire print completed without any obvious warping!
Now, I did change to somewhat less taxing model (did print as much towards the edges of the sheet), but it’s great to see good results can be achieved. I’ll eventually try the RepRak bracket again, but for now, I’d say the Garolite makes a big difference printing PA-CF. Good tip and highly recommended to others.
Are you getting layer separation or just bed adhesion lift? If it’s just adhesion lift a brim may be enough to slow or arrest the lift up. Other things you can do is play with the infill in the edges that lift. Non-straight line infills tend to spread the pulling forces around and help to keep it down. Also, the amount of infill matters. 100% infill will pull up more than 10% infill. So, if strength is needed go to the trouble area in the slicer and slice those areas with low infill with the modifiers.
Definitely using a wide brim on all PA-CF prints. Before adding the Garolite sheet, the print was well stuck to the High Temp Plate, but it began to lift the sheet off the magnets! So, you’ve got some good ideas for experimenting with strategically relieving stresses.
I’m starting to have more reliable results with a garolite sheet and models that print more towards the bed center. I’ll try modifying infill for some of the harder to print models.
Ive had similar issues with pacf as well. The nylon warps & lifts the engineering plate off the magnetic bed. Im going to try printing it on glass.
The generic profile is good for the volumetric speed & cooling settings, but I had to run the orca calibrations to get the pacf im using dialled in for temp, flow & pa.
I thought id be able to print faster with a highflow nozzle, but ive found the bottleneck is cooling. 50mm/sec works well. Increasing layer height reduces adhesion too, so theres no way around slow prints with this stuff.
We exclusively print pa-cf and have ran into the same lifting issue as you. This is our first deviation from markforged printers and bambu will offer a significant cost savings for us if the bugs can be worked out.
Borrowing from MF slicer settings, i saw a huge reduction in warping when reducing bed temps to ambient or ~35C. I dont have enough time testing to determine if plate/layer adhesion will be an issue, but so far so good. Putting down the first layer at 100c and the rest at 35c does produce some elephant foot.
I am hopeful that a cool plate in combination with g10 or cf build surfaces will get this printer where it needs to be.
@hollandp363, That’s an interesting take on dealing with the warping. By starting at 100C and then dropping to 35C, personally, I would have thought would have created more issues than it helped. That’s guaranteeing a huge delta in temps (which is what causes the warping). But maybe you are seeing improvements because the temps drop to the printer’s ambient temp quickly and over less layers. So you are probably seeing a big delta in temps for about 10-15 layers but everything from then is very stable being close to the ambient temp.
Maybe a better way would be to start the printer and preheat the chamber to 45C (pretty much max temp) then print the first layer around 65-70C before dropping the remaining layers to 45-50C (at the lowest nozzle temps you can get good results with). That way the first layer should lay OK which a low temp, but the remaining layers lay at ambient temps negating the warping.
Another idea is doing it the way you do it, but with a raft. That way the raft is the item that takes the temp differential and the elephant’s foot is on that. That unfortunately means you will have no face with a pristine finish, but I would think the part should be very accurate.
I always assumed cooling the bed to that degree would lead to more adhesion problems (since we’re starting to talk PLA bed temps), but very good to know you’ve had some success in this direction. I may try @just4memike’s version with my next PA-CF print: heat soak the chamber, then 1st layer @ 70C, other layers @ 45C.
With my last PA-CF prints, while I did correct for warping with the garolite, I did notice that there was some weaker layer adhesion in the first 2 or 3 layers. When removing the prints, some of the first layer outer walls pulled away. If I go with cooler bed temps, I suppose I may need to additionally correct for that with hotter nozzle temps for the first few layers, too.
@just4memike when testing the 100c/35c bed temp settings, I paused prints after the 1st layer and allowed the bed to fully cool to 35c before resuming. My thought process was aligned with what you have stated - the temperature transition is where shrinkage occurs.
I was trying to ensure 1st layer adheasion with the higher temp. My hope is that this is not necessary with g10. Our MF X7 does not have a heated bed or chamber and lays down pa-cf beautifully on a cf plate. Of couse, this could have something to do with their proprietary materials. The solution may be to run Onyx in the bambu.
I like these guys but remember, they are running a business that doesn’t sell consumer grade printers. So, just keep that in mind when taking their advice. They definitely have a ton of knowledge and experience, so they are a good source of info, but I ask people to just understand their presence is to sell product.
In this video they seem to take issue to Bambu and other consumer grade companies bragging about a 300C hotend and a 120C bed (as well as 45-60C chambers). When compared to the machines they sell, yes it’s hard to consider those even the bottom end of “high temp”. But when compared to the market they are in, those are up and among the highest commercially available. You can buy separate components that allow higher temps, but not many complete units go above that.
All and all, they are not wrong in what they said in the video (except its arguable in that consumer grade CF’s are just for looks), but they definitely have a basis to the high temp engineering printing world and should be consumed with that in mind.
All that said, I still encourage people to watch their channel, tons of good useful info.
The next test is with the Garolite sheet, but I am concerned now that the warping force of the PA-CF (with this large model) may be a problem if the magnetic build plate can’t hold the sheet flat.
