Trying to make a precise linear system right now. The brackets holding the motor to the aluminum extrusions is 3d printed and the component which prevents the ball screw housing from rotating by attatching it to two linear rods next to it (basically a 12mm thick plate).
I am trying to find a suitable material which is highly stiff (to prevent inaccuracies due to flex with the motor turning) and will not deform under the rotational force of the motor and screw over time. The system is not moving anything like a cnc spindle but rather a laser optics system so apart from its own load, there won’t be heat or wear to worry about. I also need a material to hold the aluminum extrusion’s ends together because I couldn’t find a profile that fit my need so im slapping together a few 2020’s and 2040’s
I am currently debating on using:
PLA-CF for the ballscrew to Linear rod mounting plate
PLA-CF for the part where the motor screws onto, and this will be attached to the main bracket
PETG-CF for the part which attaches to the end of the aluminum extrusion holding the extrusions in the correct shape and the PLA-CF mounting plate for the motor will be screwed onto this with 4 m5 screws
I’d prefer if the material is AMS compatible so I can print supports
I’ve also considered PAHT-CF for the motor bracket and aluminum extrusion end parts but that might be overkill and will it make that much of a difference once it gets wet? I’m not in a humid environment but I still dont want to dehydrate my parts every once and a while.
As with anything related to 3D printing, material is only have the battle. Design has more influence on strength, durability and survivability.
Do you have a concept of what the device will look like that you can share? Size? Weight?
What are the environmental conditions that it will have to survive? Can we assume indoors? Are there temperature extremes such as contact with a hot motor?
Once the parts are extruded, moisture absorption is not an issue. However, you didn’t state how you’re mounting the pieces. Were you relying on self tapping screws? Are you embedding nuts or thread inserts into the material? Or are you tapping threads into the material? Or are you using simple through-hole nuts and bolts?
If you’re looking for precision, you’ll want to use as much metal as possible. Also, whatever filament you chose, you’ll want to use solid walls and not rely on infill. I posted an example of that the other day which will show how to get strength out of infinite walls and zero infill.
For my material, my preference is PC because of its rigidity when printed at 100% wall thickness, but one could achieve similar results with any of the filaments you mentioned. I happened to have a fondness for PC because it produces ABS like results without the fumes, albeit, it will never win any beauty contests because there will usually be some zits, loops and other items that may need to be cut off. It is also a more of a functional filament because it’s never fully opaque and therefore the limited colors it comes in give almost a hard candy appearance due to translucent properties of the filament, but the results as far as strength and regidity speak for themselves.
Also, don’t forget that layer orientation will have more influence over flex and strength than anything you can possibly design.
I was experimenting with a linear motion system as you described. I printed everything in PLA, because that’s all I have, it feels sturdy, although there is a barely perceptible wiggle, which I blame on the lm8uu bearings I used. The trick to achieve sturdy builds with PLA is to make the parts really thick, and use 3 or more inner walls in the slicer.
I agree with Olias, you should go full metal if your aim is ultra precision. You can take a look at some ready made linear motion systems, I saw a couple on Aliexpress for ~$50. Finally, I would also recommend to use closed loop stepper motors.
heres an example, its printed out of Nylon 6.6 with no fiber reinforcement. I’ve been able to get pretty good accuracy with only plastic parts holding the components together but the nylon warping after printing is a bit of a struggle which i’m trying to solve.
My completed system will have one of these on each side as a Y stage and a X stage which is just mounted on top of the plate which ill put heat set insert threads into later.
The additional load mounted on the X stage will be <0.5kg and the system will move at max 20mm/s
I’ve narrowed it down to PAHT-CF or PLA-CF so ill be able to print with multi material supports but I was wondering which one would be more rigid and less prone to creep/deformation over time from the rotational stresses
Sadly getting custom parts machined is not an option because it would cost more than an entire printer and i’m willing to do some compromise on precision. I also need the system to be custom.
EDIT: Been eyeing the new PPA-CF filament but I want to wait and see the reviews on it, also how flexible is PETG-CF exactly? Will it deform as much over time like PLA-CF will?
Thanks for posting the clear screengrabs. That’s enormously helpful and saves a lot of back and forth.
Based upon my own experience, I would more likely persue PC filament as an option because of it’s rigidity. But as I stated initially the filaments you mentioned above would likely work just fine as a basis for precision. However, the question of durability over time is the one that you may want to explore.
Also, the design hints that there won’t be a lot of torsional forces on the device in what your model currently shows in the Z-axis. But there are two things you may want to explore in your design.
I would consider one of two approaches. The first is adding simple 45 degree chamfers to the boundary areas. Noted here:
However, it is in the use of 100% walls in place of infill that I found to be the best trick. This creates a solid near-molded plastic strength. I’ve built some pretty solid clamps using this method. The combination of adding chamfers and solid walls combined made for extraordinary durable product.
Then there is approach 2. You could take the offset vertical posts and try your hand at making them inserts that could then be printed horizontally. To me, that requires a lot of work but assuming you are working on your own CAD model, this should be relatively easy to do using Boolean intersections. This would give you the best of both worlds because you could then print each part in it’s own orientation, thus taking advantage of layer lines. Then simply use acrylic cement to “weld” the two parts together for a bond that is stronger than the surrounding plastic. I’ve used this technique with great success.
Thanks for all the help, Ive decided to go with PAHT-CF (nylon 12) for everything because of easier printability than its Nylon 6 counterpart and I don’t feel like switching between different filament types for different components. Ill look into gluing the pegs separately but it should be fine because screws into the extrusion ends will hold most of the stress. I can’t do the 45 degree junctions though because the top part sits flush with the extrusions.
Ill post some results later once I reprint the parts get a dial gauge to test for repeated accuracy of the plastic parts. Shooting for repeatable micron range rn with closed loop feedback
also, for future readers who are trying to decide between pa/paht cf or pla cf, I did some research and while PLA is super good for specific super rigid applications, it is unsuitable if you have through hole screws attatching it to anything because the pla will crack and deform with the pressure of the screw washers on it
