Anyone had warping or deflection issues with larger Trident frames?

[datageek]

About to adventure into building the Trident. I see some comments from time to time of people needing to reinforce the frame due to expansion at high temperatures.

Has anyone experienced this in a noticeable way???

Thinking if this is really an issue, it may be worth building the outer frame (the cube) from 2040 to make it stiff and have twice the screw points. Looking at the build every thing will still work, with out any modifications. Instead of building all with 2020 and then buying more materials to stiffen it.

Keen to hear experience of people with 300+ build plates and if they have had to stiffen the frame.

Thank you.

 

[brendanm720]

This is definitely a thing, but it is mostly experienced in the gantry with the frame and the rails expanding at different rates and amounts. It gets worse the bigger the printer gets... I'm seeing a bit on my 250, and I know most folks with bigger printers add backers to combat this. Heat soaking the printer and then doing a home/bed tram (QGL or Z-Tilt)/bed mesh while everything is hot helps too.

WhoppingPochard has done all of the testing and math and it's available here if you want to check it out: https://mods.vorondesign.com/detail/ewDI1Cntz7urtuq3Cm9wGQ

As far as sizing up to 4040/4020 or something different than the stock extrusions, a bigger/beefier frame usually only needed if you scale bigger than 350 or want to add some insulation for higher chamber temps/lower room temps. The Doomcube folks use 20-4040 and 20-4020 extrusions to allow for lots of room between the gantry and the panels to install insulation and additional panels inside for passive thermal boost (think double-paned windows with insulation). There is a big chonky trident with a sidepack being built, and there are numberous V2s out there that have been scaled beyond spec that use bigger extrusions.

 

[datageek]

@brendanm720

Brilliant thank you that article makes alot of sense with the major cause being the different metals.

I would assume then, using Carbon Fibre or Steel square tube on the Y rails would reduce the issue in that direction.

If the flex is mostly coming from the aluminium 2020. Then would you expect if you used a 2040 on the Y rails this would also reduce deflection. Annoying part with that as a solution is I would need to modify the STL of the components to fit.

Think some steel tube may be easier solution over all.

Thanks again for the link to the article. Very helpful.

 

[brendanm720]

If you can find tubing made from the same steel as your rails, that would be ideal, but I think that may be cost prohibitive. If the steel is different, there would still be a bimetallic expansion effect to some degree.

I'm not sure how CF would react but it could work.

 

[Sandman50]

This being an issue really depends on why you want your printer.

If you are doing cosplay or toys, or stuff like that, then probably not.

If you are doing large parts for engineering prototypes, or investment casting, or something like that, then sure, it can be a problem.

The issue is only really a problem for large prints. The bow won't really show up on smaller ones. At least the margin for error is inside the capability of extruding a plastic thread with consumer grade steppers, slides, etc.

 

[ChrisA]

About to adventure into building the Trident. I see some comments from time to time of people needing to reinforce the frame due to expansion at high temperatures.

Has anyone experienced this in a noticeable way???

Thinking if this is really an issue, it may be worth building the outer frame (the cube) from 2040 to make it stiff and have twice the screw points. Looking at the build every thing will still work, with out any modifications. Instead of building all with 2020 and then buying more materials to stiffen it.

Keen to hear experience of people with 300+ build plates and if they have had to stiffen the frame.

Thank you.

I'm lookimng at doing exactly this, using some 2040 and also the same rounded corner 4040 as the doom cube. What I suspect is that warping is not the issue but rather vibration. Newton's law about reaction forces from the granty to the frame. I think the frame needs to be much more massive than the moving parts.

Many peope have posted input shaper graphs that typically show VERY large and steep resonance in these Voron printers. Then they use software to deal with it. Why not kill the resonance as best you can FIRST?

One experiment I want to do that I've not seen anyone do yet is to place more than one accelerometer on the printer. Place one on the print head but also the other parts and look to see what is moving relative to what. This is how it is typically done in industry, you'd use several accelerometer channels. The questions this could answer are: Is the frame moving? Is it moving as a solid unit or are parts of the frame out of phase with other parts of the frame? In fact the whole concept of "phase" can not be measured unless you have at least two channels.

Making the frame more rigid might help but the other "'fix" is damping. Could printed ABS parts do this? I don't think anyone has looked into this yet.

All that said, I expect the belt is the biggest "spring" in the printer. Moving to a 9mm belt might be more effective then moving to 2040 extrusions. But you can't kknow unless you can measure it.

I'm building a V0.2 now, so it will be so,me time until I can get around to this.

In the end, one thing is always true: You can not fix what you can not measure. It does not make sense to suggest solutions until you have a way to measure the problem.