[GME]

Probably.

Could also be wear on the gear bushing. Maybe conical wear towards the ends which lets the gear "cant" but difficult to feel as the middle portion is still tight.

In terms of sand, you could drill small (1/4 or so) holes and funnel the sand without gantry disassembly. I would expect it will dampen the resonance nicely.

The expanding foam will definitely not do anything at all. It doesn't have the mass, the structural strength or the adhesion to have any effect.

Thanks, David. Gear wear is a possibility. I'm still struggling to get my arms around the intricacies of resonance. More particularly, how it can show up at 5 imp and be mostly gone at 7 ipm. It's an observable fact, but it strikes me as counterintuitive.

Drilling holes in the gantry might work, and drilling holes was something I considered Getting the lower cavity to fill would be interesting. Remember, I am using a 40mm X 160mm profile and there are upper and lower cavities to fill. To complicate matters even more, I would need to drill holes in the center of the 40mm width to get the greatest amount of fill. My X axis pinion passes right over where the holes would be drilled. The pinion only has a few thousands clearance. That could make sealing off the holes a little more challenging. Maybe silicone sealant and a razor blade to shave it off level with the extrusion surface. As much as I hate the idea of removing the gantry, it may be the best way to do it Cap one end, stand the extrusion on end, fill it up. and cap the other end. The worst part of would be removing the Z axis assembly, although removing the cable track, cabling, and coolant lines would come in at a close second.

I definitely agree about foam's lack of mass and structural strength, but have you ever used expanding foam, or at least the stuff that sprays out of a can for sealing around windows and doors? Sticks like hell and it's pretty much impossible to get off of anything it touches. The first time I used it, I didn't know (real men don't read all of the instructions, ya know). What a mess! Good to know that it would be a waste of time. I would only get one shot with foam. Once in, it would be practically impossible to remove, and would take off the table.

The more I think about it, the more attractive avoiding 5-6 ipm becomes. Maybe I'll talk myself into servos.

Gary

[wmgeorge]

I am trying to understand how a little noise at certain IPM is going to make a difference, or enough to spend any amount of money or time on? If it was me and its not, I would just move on and get the project done.

[davida1234]

I would drill like depicted in the pic. Inside the channel. Maybe every
10" or so. That area does not affect much the capacity of the profile.

Fill up, run a bit to settle, top off again, repeat until it doesn't settle
anymore. Seal with something removable like PlastiDip.

Your foam sticks only to porous or rough surfaces. Removal is a mess
because it breaks down into tiny foam pieces rather than coming off
in one piece due to low strength of the foam.

80/20 has a slick coating. Even 3M 5200 (white, 48 hours), probably
the strongest and most tenacious caulking glue of them all, has limited
hold on 80/20 profiles.

[GME]

I would drill like depicted in the pic. Inside the channel. Maybe every
10" or so. That area does not affect much the capacity of the profile.

Fill up, run a bit to settle, top off again, repeat until it doesn't settle
anymore. Seal with something removable like PlastiDip.

I see where you're coming from, David. Unfortunately, your example does not represent the cavities in a 40 x 160mm extrusion, which is what I'm using

Here's a pic of the 40 x 160:




Note that this extrusion has 4 slots/side vs 3, and by drilling where you suggested, you can't fill the extrusion cavity all the way to the top. There are also 2 other cavities between the 2nd & 3rd slots. A partial fill may work (or not), but will be sand moving around during cutting. Hopefully, what I was describing makes a little more sense, now that you see the profile I used.

Your foam sticks only to porous or rough surfaces. Removal is a mess
because it breaks down into tiny foam pieces rather than coming off
in one piece due to low strength of the foam.

80/20 has a slick coating. Even 3M 5200 (white, 48 hours), probably
the strongest and most tenacious caulking glue of them all, has limited
hold on 80/20 profiles.

I have never tried to foam with 80/20. I'll take your word for the foam not wanting to stick to it. If It won't work for dampening, I won't bother trying it. Sand is a better bet.

Gary

[GME]

I am trying to understand how a little noise at certain IPM is going to make a difference, or enough to spend any amount of money or time on? If it was me and its not, I would just move on and get the project done.

Bill, I want the best possible outcome from this build. That's it in a nutshell.

The problem is not going to stop me from moving forward with the build, or using the machine. I am cutting the counterbores and holes for mounting in the BB base layer this evening. I'm doing it on the CNC. I still have the gussets to cut, although I now doubt they will help much, if any. They will add rigidity to the frame, but the problem doesn't appear centered in the frame. Can't hurt to try, though.

Anyway, I am moving on. There may be some half-steps back to remediate the resonance, but this issue isn't shutting down forward movement. Also, I haven't spent an additional dime at this point. Unless I were go to servos, which wouldn't be a bad thing without this problem, Without going to servos, I doubt I could spend more than $20 or so on remediation.

Gary.

[peteeng]

Hi Gary - If the Ganty tube is vibrating then do the following. Get a large heavy G clamp and clamp it to the gantry. Run at the vib speed. Then move it along a bit and do same. What you are looking for is a midpoint between nodes. At a node there will be no difference as a node does not vibrate. Another methods is to sprinkle some dust on the gantry and it will dance and accumulate at the nodes. Then you can clamp midway between the nodes. Automotive parts retail sell damping tape. This is heavy mastic tape that you may have seen on the inside of vans. This is a much better approach than filling. Personally I would not do any filling or anything major as a punt. I'd keep working at it until I understood and correctly diagnosed the problem. Keep poking with the screwdriver. If you put it lightly on a place that vibrates you will feel it. Then that place needs clamping, damping, damping tape or extra support so it can't vibrate. Peter

[GME]

Hi Gary - If the Ganty tube is vibrating then do the following. Get a large heavy G clamp and clamp it to the gantry. Run at the vib speed. Then move it along a bit and do same. What you are looking for is a midpoint between nodes. At a node there will be no difference as a node does not vibrate. Another methods is to sprinkle some dust on the gantry and it will dance and accumulate at the nodes. Then you can clamp midway between the nodes. Automotive parts retail sell damping tape. This is heavy mastic tape that you may have seen on the inside of vans. This is a much better approach than filling. Personally I would not do any filling or anything major as a punt. I'd keep working at it until I understood and correctly diagnosed the problem. Keep poking with the screwdriver. If you put it lightly on a place that vibrates you will feel it. Then that place needs clamping, damping, damping tape or extra support so it can't vibrate. Peter

Thanks, Peter. If i put the metal rod (or screwdriver) "lightly" against the gantry, it jumps all over the place. I had to apply pressure just to be able to make some sort of comparison from one place to another. Again, the frame members exhibited considerably less vibration.

I'm not sure I understand what you suggest with the clamp. I should clamp, listen, move, and repeat until I find spots (nodes) where there is no vibration? This sentence confused me: "At a node there will be no difference as a node does not vibrate." "[n]o difference" suggests the vibration would be the same at a node, but that doesn't make sense. If there was "no difference" a node would be like other spots and be undetectable. So, it appears I am looking for spots "nodes" with no vibration. Correct?

I've never noticed the tape you are talking about. I checked two national auto parts chains with stores nearby. They don't carry it. I checked Amazon and found it there. OMG. This stuff is pricey. A 3" wide roll x 36 yds long x 15 mils thick (5 mils of polymer and 10 mils of alum. backing) costs $261.62 (U.S.) for 3M 2552. There were other listing for $450 rolls, but you have to buy 2 at that price. You didn't mention how thick it needs to be.

Heck, I already have a large bag of clean, dry sand, so I'm set (were I to go that way). Just curious, but why is damping tape a better option than filling the extrusion with sand? Sure, there is some deflection from the weight of the sand, but it's under 0.001". On the plus side, sand reputedly has excellent dampening properties. If I wanted to drop some bucks, I could even mix sand with epoxy (I have about 3/4 gallon of West System crying out to be used.

When it comes to using sand, I'm just being lazy. I can't handle the gantry with the Z axis in place, so I'd have to remove it. I'd also have to deal with the cable chain and all it contains. Overall, a big PIA. :violin:

All kidding aside, I am looking at problem at only 5 and 6 inch/minute (127-152.4mm/minute). I may hit that low a speed during cutting, but for only an instant. Is there a downside to just homing at 7 ipm, or even 10 ipm and avoiding the lower speed? I'll never intentionally try to run a toolpath at those speeds. As much as I tend to go for as close to perfection as I am able, a straightforward cost-benefit analysis seems to favor avoiding the offending speed.

Gary

[peteeng]

Hi Gary - at 5ipm or 127mm/min I'd ignore it as well. But if you want to fix it then first you have to understand the problem. A vibration often can be solved just by putting your finger on the correct spot literally. Adding sand or foam is just a stab in the dark unless you can absolutely say this is the problem. Can you post images of the gantry please, so I can see its construction. Plus its thickness. I'll look back through the thread. Also what steps per min is 5ipm? so I'd like steps per min and microsteps per min please to understand what the excitation frequency is. Peter

Edit - so your using 40x160 can you get the suppliers link to the cad of it? If so and if you tell me the length I can run a modal analysis and tell you want frequencies it will vibrate at. If this agrees with the 5ipm excitation freq then we have some evidence. Or publish the dxf of the section here or the step file.

Peter

see node - https://en.wikipedia.org/wiki/Node_(physics) if you clamp a node you don't change the vibration. If you clamp between nodes you stop or change the vibration. There are global and local vibrations. The first global vib is the gantry moving like a banana with nodes at its ends. (like a wobble board) Then it has a half wave, with a node at its middle. This would be in the thin direction. Does the gantry move more in the up down direction or the back fwd direction? Then there are local vibs. The channels vibrate or the internal bits vibrate. These would be at high freqs as they are small.

https://en.wikipedia.org/wiki/Wobble_board

Another thing done is to isolate the motor from the structure somehow. Sometimes a plastic plate between the motor and the mount will do it. Perhaps clamp a piece of steel or al along the gantry if possible to add some mass. Photos of your motor mount area?

edit- I looked at the photos in the thread. Is this the current Y drive? If so the vibration has two paths into the structure. Thru the R&P or through the pivot pin. Try removing one side and driving on one side only. Just to check its not a something to do with how the two are coupled elastically. Or if its one motor perhaps? Then you can try tightening the pivot pin if possible. nd does the piosition of the Z assembly make any difference? If its in the middle or the sides? Cheers Peter

[ger21]

I've never noticed the tape you are talking about. I checked two national auto parts chains with stores nearby. They don't carry it. I checked Amazon and found it there. OMG. This stuff is pricey. A 3" wide roll x 36 yds long x 15 mils thick (5 mils of polymer and 10 mils of alum. backing) costs $261.62 (U.S.) for 3M 2552. There were other listing for $450 rolls, but you have to buy 2 at that price. You didn't mention how thick it needs to be.

I never thought of this, but a much cheaper option is the stuff they sell for car audio.

https://www.amazon.com/Noico-deadeni...dp/B01KZ5X7KO/

[davida1234]

I never thought of this, but a much cheaper option is the stuff they sell for car audio.

Or truck bed liners. I think they call it RhinoSkin or something.

However, car audio stuff is for dampening sound waves and
liners for physical protection with limited mass.

My guess is that sand would still work better due to higher mass.

[ger21]

If using sand, I've heard that oiled sand is much better, as it acts like one large mass.

[davida1234]

If using sand, I've heard that oiled sand is much better, as it acts like one large mass.

Interesting. Could very well be true.

[davida1234]

Here's a pic of the 40 x 160

If your gantry profile is a 40x160, that implies that it is four times as tall as it is wide.

Your pic is twice as tall as it is wide?

[GME]

If your gantry profile is a 40x160, that implies that it is four times as tall as it is wide.

Your pic is twice as tall as it is wide?

I misstated the size. The profile is 80mm X 160mm, or twice as tall as it is wide (2 slots on the top and bottom, and 4 slots on the perpendicular faces). This isn't the first time I've made this mistake. I guess it's because the extrusion is part of the 40mm series and I keep picking up the 40mm number. Sorry for the confusion.

Gary

[GME]

Hi Gary - at 5ipm or 127mm/min I'd ignore it as well. But if you want to fix it then first you have to understand the problem. A vibration often can be solved just by putting your finger on the correct spot literally. Adding sand or foam is just a stab in the dark unless you can absolutely say this is the problem. Can you post images of the gantry please, so I can see its construction. Plus its thickness. I'll look back through the thread. Also what steps per min is 5ipm? so I'd like steps per min and microsteps per min please to understand what the excitation frequency is. Peter

Edit - so your using 40x160 can you get the suppliers link to the cad of it? If so and if you tell me the length I can run a modal analysis and tell you want frequencies it will vibrate at. If this agrees with the 5ipm excitation freq then we have some evidence. Or publish the dxf of the section here or the step file.

Peter

see node - https://en.wikipedia.org/wiki/Node_(physics) if you clamp a node you don't change the vibration. If you clamp between nodes you stop or change the vibration. There are global and local vibrations. The first global vib is the gantry moving like a banana with nodes at its ends. (like a wobble board) Then it has a half wave, with a node at its middle. This would be in the thin direction. Does the gantry move more in the up down direction or the back fwd direction? Then there are local vibs. The channels vibrate or the internal bits vibrate. These would be at high freqs as they are small.

https://en.wikipedia.org/wiki/Wobble_board

Another thing done is to isolate the motor from the structure somehow. Sometimes a plastic plate between the motor and the mount will do it. Perhaps clamp a piece of steel or al along the gantry if possible to add some mass. Photos of your motor mount area?

edit- I looked at the photos in the thread. Is this the current Y drive? If so the vibration has two paths into the structure. Thru the R&P or through the pivot pin. Try removing one side and driving on one side only. Just to check its not a something to do with how the two are coupled elastically. Or if its one motor perhaps? Then you can try tightening the pivot pin if possible. nd does the piosition of the Z assembly make any difference? If its in the middle or the sides? Cheers Peter

Peter,

I misreported on the extrusion profile. It's 80mm x 160mm. The 80/20 CAD library has 2d DFX files, but not 3d. No .step files. The available formats are SLDPRT, SLDLFP, 2D_DXF, IPT, and SKP. Will any of these work?

Thank you for the link about nodes.

The vibration, before I added the side braces was both up and down and front to back. I think mostly front to back, but I can't be sure. After adding the side braces, It seems more front to back. I say that because an item placed for observation didn't seem to dance up and down so much, but did move around. I recognize that my observations may not accurately reflect what was really going on. I'm out of my depth on this issue. The placement of Z doesn't seem to make a difference, as far as I can tell, but I wasn't looking for a correlation.

Looks like much to check. Unfortunately, I don't have any experience with observing vibrations in the ways you have expertly outlined. I wish I had someone looking over my shoulder to point out things like, "see, when you do A at this point, you are observing X, and when you do B at that point, you are observing Y." So far, I haven't been doing a very good job of connecting the dots. Hopefully, I do better with a little more practice. Thank you for your patience. Much appreciated.

You asked about steps/microsteps per minute at 5 ipm. The native resolution of my Gecko 203Vs is 10X microstepping. As far as I can tell, they can't be set to run at full steps. Anyway, 10x microsteps per minute at 5 imp = 10,191.355. 1/10th that would = 1,019.1355? full steps.

Much left to do.

Thank you so much for your help.

Gary

[GME]

I never thought of this, but a much cheaper option is the stuff they sell for car audio.

https://www.amazon.com/Noico-deadeni...dp/B01KZ5X7KO/

Or truck bed liners. I think they call it RhinoSkin or something.

However, car audio stuff is for dampening sound waves and
liners for physical protection with limited mass.

My guess is that sand would still work better due to higher mass.

If using sand, I've heard that oiled sand is much better, as it acts like one large mass.

Good information.

I've not heard about oiled sand. But then, all this about dampening is new territory for me. I understand the some of the basic concepts, but that's about it.

One would need to made a good, high quality gasket to be sure no oil settles and leaks out, and gets all over everything. I suspect the oil might also contribute to dampening, even if only in a small way.

Thanks, guys.

Gary

[GME]

Or truck bed liners. I think they call it RhinoSkin or something.

Rhino Linings has a whole line of spray on bed liners. You would want to be darned sure it would work before spraying it on. If it didn't, you'd probably be shopping for a new extrusion. You would also have to be pretty selective where you sprayed it, given the linear rails, sensor flags, etc, which are mounted to the gantry. Rhino Linings doesn't sell RhinoSkin, which appears to be some sort of spray-on skin care product.

Gary

[ger21]

The sprayed on bedliner in my truck is thin and hard, and I don't think it would provide any damping.

[wmgeorge]

If it was mine, and its not I have some assorted round steel bars in my pile of steel inventory. If the ends of the extrusions are open I would just slide some of those inside. All your doing is experimenting with adding mass to change the resonant frequency as i understand. Yes I understand if it did work and the bars were in permanently you would need to hold in place to prevent more noise. Once again.... just brainstorming.

[GME]

If it was mine, and its not I have some assorted round steel bars in my pile of steel inventory. If the ends of the extrusions are open I would just slide some of those inside. All your doing is experimenting with adding mass to change the resonant frequency as i understand. Yes I understand if it did work and the bars were in permanently you would need to hold in place to prevent more noise. Once again.... just brainstorming.

An interesting idea, Bill. Unfortunately, I don't have much steel laying around, and what I do have (some square tubing) is too large to fit in the cavities. Keep the ideas coming.

Gary