New Machine Build Fast Cartesian 3D Printer - Page 2

# Thread: Fast Cartesian 3D Printer

1. ## Re: Fast Cartesian 3D Printer

Looking at the data again I think I've actually missed the first resonance mode, which is at about 35Hz on both beams. I just assumed it was noise at first. It is much lower amplitude than the second resonance modes around 138Hz.

Unfortunately, my kit is not really designed for very low frequency data collection so there is not a lot of resolution down at 35Hz to get a nice waterfall plot of decay.

I've already cast concrete in the beams so I can't go back and run more tests but in the future I will try to run multiple measurements and average the data to get more resolution.

Thin Tube, hollow. Note I have adjusted the vertical axis from previous plots.

Thick Tube, hollow.

I think it is important that we differentiate between 'damping' and 'loss'. If we look only at the amplitude response, added mass seems to damp the resonances, but this is only spreading the energy through time as we see in the waterfall plots. I'm not sure this is actually an increased 'loss factor'? When I think of damping what comes to mind is lossy damping that converts kinetic energy to heat, such as visco-elastic material properties. I believe this will provide a damped amplitude response as well as a faster energy decay, that would be quite desirable.

Pete, I'll have a read of those papers. Quite a lot to take in there.

2. ## Re: Fast Cartesian 3D Printer

Originally Posted by ssashton

We can see the lowest frequency resonance in each tube is actually pretty much the same frequency 138Hz (thin) vs. 128Hz (thicker). However we can see a significant difference in amplitude.
Oh sh*t, PEBCAC I meant 338Hz and 328Hz, lol.

3. ## Re: Fast Cartesian 3D Printer

Pete, a'rite. I read the E756 paper.

My test method is not strict enough to extract numbers that can be used in other situations, but I think we can calculate a loss factor for this specific example.

I'm not great with maths, so if you want to have a crack at it I believe you need the -3dB points of the main resonance. I'll provide the same for the concrete filled tubes.

I took two measurements for each sample and I've averaged them for this data.

Thin walled tube: Main resonance 340Hz. Lower -3dB 334Hz. Upper -3dB 345Hz.

Thicker walled tube: Main resonance 330Hz. Lower -3dB 323Hz. Upper -3dB 338Hz.

Originally Posted by peteeng
What is the length and wall thickness of your tubes and I'll do a modal analysis on them. length, wall thickness outside dims and corner rad? Peter
0.9m length. Wall thickness 2mm and 3mm. Outside 40mm x 40mm. Corner radius not sure.

4. ## Re: Fast Cartesian 3D Printer

I certainly could try various pulley ratios and do plan to have slotted motor mounts at least just to provide belt tension.

One thing that I like from my research into the Duet boards is the ability to do independent homing on different axis
I believe Marlin can also do this, although I've never used it. It is not something accessible from on-screen menus but if you compile your own Marlin you can uncomment
#define Z_DUAL_STEPPER_DRIVERS
#if ENABLED(Z_DUAL_STEPPER_DRIVERS)
#define Z_DUAL_ENDSTOPS
Somebody also told me about a way to run the Duet RepRap firmware on an SKR board - https://3dprintbeginner.com/duet-fir...or-skr-boards/

I am wondering about sensorless homing using stallguard.
I've never used it, but it seems like a cost saving measure for mass produced systems. Inductive switches are incredibly accurate and reliable. Better than physical switches. This is one of the main reasons I'm going to have an aluminium print surface (with a PEI sticker on top). My Creality CR10-S Pro came with a capacitor sensor, but it was too variable, effected by things like humidity and temperature.

I wonder if you should add a second Z axis motor, or change that up in some way? It seems to me like you've got a Cadillac top end, and a Yugo Z-axis.
I kind of agree, the Z axis drive is inferior to the X and Y. I do have dual Z motors and drive. Depends on the requirements though (and pockets not infinitely deep). Gotta remeber that in a 3D printer the Z axis does not move fast and does not need to handle cutting forces. During printing the Z axis will 'Z hop' by about 0.2mm while the X and Y do significant rapids, just to avoid collisions. Repeatability and accuracy are also important of course, but I find my CR10-S Pro has pretty decent Z accuracy with T8 leadscrews. The load on the leadscrews will be pretty static, just the weight of the plattern. I did think hard about using profile linear rails on Z, but they need support which then needs to be accurately integrated in the frame x4. I thought about in the future I might want to use it as a mill but.. I already have a bigger CNC router so unlikely really.

GT2 is a further revision of HTD, designed to give less backlash
Thanks! I struggled to find short belts with decent width, but I did just find some so that's great.

The thing I'm struggling with right now, is if I want to use the industrial Yamaha belt driven linear stage..
I'm considering if down the road I may wish to use this machine for more than 3d printing, and if it would be worth it therefore to go with all ballscrews.
I've not much experience to offer here. I have seen chain driven CNC routers cut ali. I guess a good belt would probably fine too if you find the best feeds and speeds. It's not the tech, but how well it is implemented that matters. Can we see a picture of this linear stage? Like that enginearding porn!

How good are you at welding and fabricating?
Ahhh hahaha. I'm absolutely a noob at welding. I got a 150 amp mig with C02 about 6 months ago. Grinder and paint make me the welder I ain't.

I was thinking I'd mill the top surface of the tube flat, then tack weld it to the side panels to get it aligned. Then fully weld. Maybe I should fill at least the top beams with concrete before doing face-off and tack weld. I can't fir the assembled 3D printer frame on my router to face off the top beams as it's too tall.

I bolted my CNC router at first. It had low frequency wobbles. Later I welded it and that was better. Bolting between certain sections with a damper material may be a good idea, say between gantry and rail block. I'm not sure it's good to have lossy fixings on the frame itself that is the endstop for vibration in the system.. yeah I dunno.

The 1mm steel plate....have you considered using something with better insulating properties? And perhaps putting some doors on the front and a top on this machine?
print materials like nylon. At this point I'm not sure how far I can go without adding water cooling for the extruder and steppers. That will be future progress I suppose.

you would have leveling knobs in the four corners to adjust the actual print bed.
You are right, I intended to add that and forgot. Once level I hope it will stay that way as I will not use sprung mounts for the Z axis. However I do need to add some kind of adjustment at each corner.

5. ## Re: Fast Cartesian 3D Printer

Originally Posted by ssashton
Can we see a picture of this linear stage?
I just grabbed this pic off EBay seller 80servo, so credit to him for the pic. His is already taken apart so you can see everything. I have the exact same one.

Originally Posted by ssashton
Ahhh hahaha. I'm absolutely a noob at welding. I got a 150 amp mig with C02 about 6 months ago. Grinder and paint make me the welder I ain't.

I was thinking I'd mill the top surface of the tube flat, then tack weld it to the side panels to get it aligned. Then fully weld. Maybe I should fill at least the top beams with concrete before doing face-off and tack weld. I can't fir the assembled 3D printer frame on my router to face off the top beams as it's too tall.

I bolted my CNC router at first. It had low frequency wobbles. Later I welded it and that was better. Bolting between certain sections with a damper material may be a good idea, say between gantry and rail block. I'm not sure it's good to have lossy fixings on the frame itself that is the endstop for vibration in the system.. yeah I dunno.
I don't know how well concrete will react inside a tube that's being welded. I think it will cause some big problems.

If you mill the top surface flat, it's just going to distort again when you weld the rest of the stuff on there. Although if all you did was tack weld this together, it should be strong enough with just the tacks, and you'd get less distortion.

1mm steel plate...yeah, you're going to have some problems trying to weld this without blowing through the material. TIG is for sheet metal like this. I wouldn't even try it. LOL.

I'd suggest getting some scrap pieces of steel of varying thicknesses and practicing.

It would be interesting to see what kind of welder you have and how well it performs. I have an old school 220V miller stick welder. When I say old school, it's probably 30 years old. Still works. I like it. And it's simple (no gas required). It's alot harder to do vertical welds though, than with a mig.

Originally Posted by ssashton
print materials like nylon. At this point I'm not sure how far I can go without adding water cooling for the extruder and steppers. That will be future progress I suppose.
Actually some printers that are geared towards nylon don't even have a heated bed. Higher temp extruder, but print on unheated garolite. The markforged mark two, for example, doesn't have a heated bed. And even though it is enclosed, I don't think there's any kind of dedicated chamber heating or temperature control either.

Steppers are rated for max temp. Most are good for around 100 deg C if I'm not mistaken.

Yes, the printers that do really high temp stuff like ultim, have to provide motor cooling.

But even for things like ABS, is an enclosed chamber not a big help?

Correct me if I'm wrong, but with the exception of the really high end industrial printers, I haven't seen even one enclosed chamber 3D printer with built in stepper motor cooling.

6. ## Re: Fast Cartesian 3D Printer

Hi Simon - I modelled a 50x50x5mm 900mm long SHS steel. One open, one filled with concrete and ran a modal analysis. You've published your sizes so will rejig around those. I "hung" the tube from one end but modes should be independent of how its hung. The filled tube had the lowest freq at 30.5htz then the open tube 34htz then the filled at 232 and then the open at 262htz. So filling does not change the freq remarkably. My guess is that the tube stiffness in this case is dominant. I used 50GPa for the concrete stiffness and 2700kg/m3 for density.

The question is "is it damper?" and you seem to have the equip to figure that out. Regards Peter

The lower freq is a cantilever wobble mode, like a tuning fork tyne. This will be damped physically by the way it is supported, the tube at both ends. If you hang the tube from one end this mode may show up better.

7. ## Re: Fast Cartesian 3D Printer

Hi Sash - I modelled the 40x40 and had a play with restraints. The very low modes depend on how you hold it. If "free" the first mode is 317htz in the thick tube, then 324htz in the thin. This is the banana mode.The thin tube bends across the diagonal which is interesting didn't expect that. To get the low modes the end has tobe held in some way to allow it to wobble like a spring board. Then the first mode is 50htz for the thick and 51.6htz for the thin tube. The banana shape kicks in at 309htz for thick and 315htz for thin. Movies would be good but this site can't load up MP4 files. Peter

8. ## Re: Fast Cartesian 3D Printer

Thanks for the details. I didn't think I had any fixture dense enough to not affect the measurements, hence I went for free hanging. However, thinking about it I could have hung the beams vertically from only one end, or even bolted to the concrete floor outside my workshop. Inside workshop is wooden structure with floating floor.

Never mid, I think we will get enough data to see if concrete filling adds lossy damping or only mass.

I will carry forward with the tests, I just want to be sure the concrete is fully cured and then waiting for a spare moment.

I'm just going to post some comparisons between my CNC router stepper drives and the Trinamic ones I bought for this 3D printer. I think it might be interesting: https://www.cnczone.com/forums/stepp...05186-cnc.html

9. ## Re: Fast Cartesian 3D Printer

The results are in an interessssting. Compare with post #17

Thin walled tube, filled:

Thick(er) walled tube, filled:

Thin walled tube, filled:

Thick(er) walled tube, filled:

So what do I make of it? Seems like the filling gave a significant increase in mass and has reduced the initial amplitude of the resonance in both tubes to a similar level. Both are quite improved, by over -10dB (actually that's HUGE, 3dB is a halving of amplitude!) the thicker walled tube maintaining a slightly higher frequency of resonance than the thin one.

When we look at decay rate, we see a really noticeable difference! The thin walled tube decays much faster than the thick(er) filled tube. Both are improved over the decay rate of the hollow tubes.

I'd hypothesise this is because the thicker walled tube carrys more of the stress / strain in the steel, while the thinner walled tube passes more of the load to the concrete. Does the latex (SBR) concrete mixture offer lossy damping? Yes, clearly it does.

HOWEVER!!!

Here we are looking at the 2nd resonance really, not the 1st one. As noted in the earlier discussion, a lower frequency, but also lower amplitude (at least in my test setup) resonance exists around 35Hz.

Thin walled tube, filled:

Thick(er) walled tube, filled:

Here we see that the filled thin wall tube has improved greatly (over hollow version), yet the thick(er) walled tube has shown only a small improvement. It seems that at the lowest frequencies the mass and stiffness contribute more to the material behaviour than visco-elastic damping properties. The thick(er) walled tube still maintains the lead, just.

This isn't really very surprising. Visco-elastic damping is more effective at higher frequencies than low. This is where constrained-layer damping methods are purported to maintain much greater effectiveness.

So... Which is better?

I'm not sure. I see only a small difference in the low frequency resonance on the two filled tubes (thicker tube 2dB better) while the 2nd resonance in the filled thin walled tube is much faster decaying. I can't see much info about the lowest frequency decay in this data. The thin walled tube does look cleaner all the way down the frequency range, but that may just be a knock-on effect from that big 2nd resonance.

In conclusion I would say 'it depends what frequency your excitation is at'. Over 100Hz? Choose thin walled tube with concrete filling. Definitely under 100Hz? Just aim for maximum mass and stiffness, you won't easily manage to get lossy damping down that low.

I'd very much like to try a thin walled tube with concrete filling under compressive pre-load. In buildings they cure concrete around steel bars stretched under tons of load. When the concrete is set, the steel is released from it's load and it contracts on the concrete putting it under a massive compressive pre-load. I imagine this would get the best of the visco-elastic damping at higher frequencies, while also gaining great stiffness and mass at the low frequencies. Trouble is, that isn't so easy as pouring in to an existing structure.

That is a hydraulic jack pulling the steel cables under tension to load the concrete (after casting in that case).

10. ## Re: Fast Cartesian 3D Printer

Hi Sash - The reason to prestress concrete is to keep it in compression when a tension load is placed on it as concrete is weak in tension. So internal stress goes from compression to less compression vs compression to tension. I don't know if it affects its damping. Peter

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•