Root cause of harmonic oscillation in telescopic column

[WestyTea]

Hello All,

I am working on a telescopic column. The design is two square Aluminium tubes supported by internal Acetal wear blocks. There is a payload on the inner tube and it is driven up and down, the external tube is static.

I am frequently getting a 9 Hz vibration mode occurring as I drive the inner column down. As the vibration occurs at all speeds, positions and with different drives, I have managed to rule out all of the geartrain mechanics (belts, gears, leadscrew etc) and any motor interference. I also made another version with roller bearings instead of blocks and there were no vibrations, this design was not feasible for other reasons however. Different materials with better sliding properties have also been tried for the blocks.

I am confident in saying the vibration is caused by the friction forces and interaction between the tubes and the blocks.

However, I am at a loss as to precisely how or where the 9 Hz is originates. Any calcs I try give resonances many orders of magnitude higher than this. I would really appreciate some pointers or advice.

Other information
- The vibration only occurs when the payload is driven down (with gravity).
- The vibration only occurs near max payload.
- The vibration can be worse if the payload is offset, probably causing an out-of-plane twist to the inner column.
- The vibration is only in the y axis (as looking at the image below)
- There is also a 92 Hz vibration mode, I am unsure whether it is connected, but I am less worried about this for now.

All help is much appreciated

 

[GregLocock]

Shrugs. Without details of the payload, geometry of the tubes etc then it's all going to be handwavium but at a guess you have a non linear kitchen drawer sticking type problem if you only have one set of acetal blocks. 92 Hz is irrelevant.

 

[dcascap]

yeah the first mode seems irrelevant, too high with respect the 9 Hz you're getting. Is there any possibility that the max payload is generating a gap anywhere in your structure? I highly doubt it but do you see any buckling behavior in your structure or are you near buckling when you have max payload?

 

[WestyTea]

Thanks for the replies

Shrugs. Without details of the payload, geometry of the tubes etc then it's all going to be handwavium but at a guess you have a non linear kitchen drawer sticking type problem if you only have one set of acetal blocks. 92 Hz is irrelevant.

Non-linear for sure. There are a second set of smaller blocks lower down fixed to the inner tube, reducing the likelihood of racking (kitchen drawer), but doesn't rule it out completely.

yeah the first mode seems irrelevant, too high with respect the 9 Hz you're getting. Is there any possibility that the max payload is generating a gap anywhere in your structure? I highly doubt it but do you see any buckling behavior in your structure or are you near buckling when you have max payload?

An offset payload could be creating a gap for sure. We can shim the blocks closer but that increases the friction force, which can exacerbate the problem.

Are there any phrases or knowledge areas that I could research do you think? Or would you say this is more of a mechanical alignment problem rather than a vibration problem? (from a physics point of view)

 

[IRstuff]

Lots of information, yet still not enough. How tall are these columns?

 

[GregLocock]

I have seen successful estimates of stick slip problems, you need to know the axial and radial compliance of the blocks and of course the coefficients of friction, as well as all the other obvious properties of the system

 

[WestyTea]

I have seen successful estimates of stick slip problems, you need to know the axial and radial compliance of the blocks and of course the coefficients of friction, as well as all the other obvious properties of the system

Thanks I'll look into it. I took some IMU measurements yesterday and I think I may have found the root cause.
I believe it's a function of the interaction between the friction between the blocks and columns and the offset payload:

The offset payload means there is a higher friction force on one side. When the column starts to move a reaction torque is produced until the friction force is low enough for it to move. The column then moves down but the payload pushes it back over and repeat.

There are a couple of other small factors at play that I still haven't nailed down but I am going to try a roller method again to remove friction. Unfortunately only light lubrication is possible for the blocks.

Lots of information, yet still not enough. How tall are these columns?

Thanks for the reply. Sorry, but I'm not going to publish the full engineering drawings and spec of the system. The height is irrelevant to the vibration anyway (as mentioned in first post). I was hoping for guidance rather than a solution. This is Eng-tips after all

 

[IRstuff]

height is related to stiffness of the overall structure; your weight offset would cause a left/right jerkiness in motion, since one side will move before the other, and there might be some clue if the frequency changes as a function of height.

 

[WestyTea]

height is related to stiffness of the overall structure; your weight offset would cause a left/right jerkiness in motion, since one side will move before the other, and there might be some clue if the frequency changes as a function of height.

The frequency does not change with height or position. However, I wouldn't necessarily expect it to as the reaction torque arm is constant, ie the width of the column, as it is being pulled down by two belts on each face of the column. FYI the travel of the column from nested to full extension is 750 mm.

But yes, that's exactly what I think is happening. I found that the left/right jerkiness is 180° out of phase with the up/down. So as the column friction is "released", it becomes free to move down, which then makes it free to move left/right again and "stick" on the block (and repeat).