LR4 "Twist Test"

Okay, so I proposed a “Twist Test” in Doug’s thread, but I kind of figure that it needs its own thread to get more feedback and see what is actually what.

Hypothesis: securing the braces and/or strut to the X tubes will produce a resistance to twisting forces in the gantry beam. Additionally reinforcement of the strut to tube should serve to reduce deflection of the beam in Z, though given that the horizontal strut cannot be connected to the tube, no additional resistance to deflection on the Y axis should be expected.

Methodology:

By lifting the toe of the XMin side YZ plate with the core at a preset position, and Y also consistent, or Y belts disconnected, we should be able to measure a difference in the amount of lift between a non-reinforced, and a reinforced gantry on the same machine at the X Max side. Ideally, any lift of the toe at X Min will produce a comesurate lift at X Max.

Corroborating evidence: currently, on my 32" beam, I am able to lift the X Min side toe by several centimeters before the X Max toe lifts from the rail. I have not yet reinforced my gantry.

My proposed reinforcement will be use of silicone caulking between the face strut plate and the X rail tubes, on the theory that the weaker link (compared to screws in the braces) over a larger area will produce equal or greater reinforcement, and have the ability to be removed if this introduces an unwanted twist in the beam. This also allows a further reinforcement of the screws later if further reinforcement is desired.

My expectation is a significant reduction in the amount of twist that the weight of the beam will naturally allow.

Alternate testing: Since Doug has managed to insert screws onto his X rails, he can perform this test in the “after” state as-is. I would suggest that removal of the screws would be very nearly the “before” state, and would validate, but not quantify any increase in twist rejection, since additional friction as the holes in the tubes reach the edges of corresponding holes in the braces is to be expected.

6 Likes


Okay, this is a starting point. This is a piece of scrap that I used to get the toe wheel 11mm from the table. Any further and the wheels lift from the rail on the X Max side. Both heels remain on the table. I suppose that the heel lifting from the table would also indicate a limit to the beam twisting.

I am apparently out of usable silicone caulking. I do not think that JB Weld or similar is necessary, silicone will adhere more than well enough, and remain removable with a razor knife if necessary.

2 Likes

I think this is an awesome way to test @SupraGuy and I can’t wait to see what your results are. I will do my best to remember and run these same test on mine before and after the screw additions

2 Likes


Okay, now I need to let the silicone cure. Supposed to be 24 hours, I think, but should be OK to clean up the excess in the morning.

This stuff says that it remains flexible, but my experience is that just means that vibrations won’t crack it. This is a plus over something more rigid like JB Weld, since I expect the machine to vibrate. I also don’t have high expectations of JB Weld to span gaps well. The gap is small comparatively, but larger than the cracks that I have good experiences with it previously.

5 Likes

Dan, thank you for figuring out a reproducible way to test flex, and sharing what you discover.

Given your thoughts on JB Weld, I might try Polyurethane Construction Adhesive based on AI advice too?

2 Likes

Have fun ever using the core removal trick where you slide the rail out. :stuck_out_tongue:

But I think it’s pretty neat you are trying it. :slight_smile:

1 Like

So Ryan wants to know if the sheet metal screws do anything also!

2 Likes

At present, I’m not confident enough in my ability to make the holes in the tube line up straight. Doug has managed it, which is great, and I hope that he does a similar test to see how much they help, as I outlined on his thread and copied into post #1.

I figure, however that I’ll probably have pretty good luck drilling the beam once the silicone is cured, lol. It’s dry enough that I can clean up the excess this morning, but I haven’t touched it yet.

4 Likes

I am curious to see if your results change from the silicone. When I drilled the beta I used the same method that doug did with the self centering bit. But I believe this time I will go with a different plan…

https://a.co/d/fr7VYV6

A transfer punch set like this should be able to line up on the hole in the 3d printed part just right and leave a perfectly centered punch mark for later drilling. Then I will pre drill with an appropriate drill bit before using these screws that I picked up…

https://a.co/d/3QB9R0z

3 Likes

you linked the punch set twice…

2 Likes

:man_facepalming: I fixed it… I think :rofl:

2 Likes

I will try to do this test hopefully either tonight or tomorrow.

1 Like

Well, the silicone isnt fully cured yet, but…


Before the lift was 10.88mm, after, 4.75mm. From this, I think it’s reasonable to say that the silicone on the front brace has a significant impact on the ability of the beam to resist twisting forces. I don’t know if further curing of the silicone will increase the resistance, it is still very soft compared to the stuff I caulked my shower with (same product, but different tube.)

I doubt the actual product makes much difference, of course, but this is what I used. My strut plate is 1/4" Melamine/MDF.

5 Likes

That is less than half the previous amount. I would say that is pretty significant improvement. Now I wonder how that will translate into actual cutting.

I suspect I will never know, as the limits I hit were the ability of the machine to remain on the Y rail. When I tried cutting too fast and too deep, I managed to push the X Max YZ plate off of the rail before I noticed any other issues. Still, twist was a known weak point in the LR3, so I don’t see how an improvement in this area isn’t a good thing on the LR4.

1 Like

oh wow. I have not had that happen at all. I did see some flex in the beam when pushing it REAL hard but that was mostly before we moved the one brace screw up somewhere between RC1 and RC2. I didn’t notice it as much after that. For sure not like the LR3.

In order to do the test in a way that hopefully matches your conditions, help me on these questions:

Are you doing this with a router installed in the core?

Where are you positioning the core for the test?

You mentioned doing it with the belts off which would be very easy for me right now because I don’t have my belts on yet. I would have to go ahead and finish building my core and install it.

1 Like

Harbor freight has this too! https://www.harborfreight.com/12-piece-cold-chisel-and-punch-set-66440.html

1 Like

Yes. I have punches like that. But the transfer punch set will be sure that you are dead center. Use the correct size for the size of the hole and it will be dead center every time. I have a set already I just have to find them LOL.

2 Likes

As it happened, I just used the machine as it was, so yes, there is a Makita RT0701C in the core, and it is 12.7mm asway from the home position. The extra weight there keeps the X Min heel from lifting, I think. I left the Y belts on for this, and it is 300mm away from the Y home position. While removing the Y belts would remove some of the extra forces in question, I kind of wanted to see the real world forces as it would be when in use.

3 Likes