vibrations translated from floor thru a Kinematic mount

[vono]

I have a very very sensitive x-ray microscope that weighs 4600 lbs. this device is designed to be move in and out of position. It is currently sitting on a 4 point kinematic mount with 4- 1" dia stainless steel balls on every corner. ( 3 corners drop into corresponding vee block and one corner sits on a flat plate) To move this device we manually crank up the ball mounts and the frame drops onto wheels so it can be moved.
The issue is that the device does not drop back into position accurately. It needs several raise and lower attempts to get it back into the vee blocks. I see there are tungsten carbide " canoe spheres" that might improve the repetitive accuracy.
but
A point was brought up that these "Canoe Spheres"

http://www.precisionballs.com/All_Vee_Blocks.php#canoe

add a larger overall footprint ( contact surface) to the device and it could potentially amplify known vibration from the floor into this device. IS THIS POSSIBLE??? will a larger footprint potentially cause more vibration?
thanks for the input ...

 

[KENAT]

So, it would appear part of the problem may be that you don't actually have a kinematic mount.

You have 7 points of constraint not 6. The 3 ball & V blocks give you all 6 constraints - the extra flat is superfluous or will stop one of the other mounts acting properly. Of course if your device isn't stiff then it may sag into all the points but the repeatably will be impacted.

Assuming your device is squareish, and the C of G is moderately symmetric then it seems your whole set up will tend to be unstable i.e. rocking on a very small level.

Also you may have issues with the balls sticking if your surfaces aren't smooth etc. and if the Herzian contact is high enough you may be damaging them.

The website you link has some excellent practical technical information such as:

https://www.precisionballs.com/Micro_Inch_Positioning_with_Kinematic_Components.php

I suggest reading over the technical info on that site very thoroughly and if you still aren't clear the last time I had a problem I was able to call them and get put in touch with their founder who has forgotten more about kinematic mounts than most of us will ever know. This was about 3 years ago though, not sure if he's still involved with the company as he was a little advanced in years.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[IRstuff]

> It's very likely that your balls are too small. if you assume that each contact is 1/10th inch square, you get 453MPa of contact pressure. you may want to point an inspection scope at the vees to see if they've been damaged
> we had a much lighter system, about 40 lb, on 3/8th inch balls and wound up dimpling the vees when we torqued down the bottom and top
> seems to me that if you're able to move the position by raising and lowering, then the mechanism that doing the raising and lowering might be driving the repeatability also.
> kenat's comment about sag should be taken seriously. Even steel bends, and keeping something that weights 4600 pounds stiff is challenging
> the canoe spheres seem to be a plausible way of getting a effect of larger balls without consuming the commensurate volume. wish we had thought of that 20 years ago.




TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[vono]

Thank you for the thoughts and information. Yes I know this frame is not on a true Kinematic mount but it is what I have to live with at this time. I suspect there are several reasons for the lack of repeatability ( under size ball point, wobbling in the post that supports the balls, frame twisting, dimpled or worn vee blocks etc.) further testing is needed to determine if I am getting any rocking motion from the frame its self or other sources inside the device. ( not so easy task to get to).
I will be moving the device in a few months to a new location and I want to put some thought into possibly putting the device on a rail system that can be locked down. Again the question comes up ...will locking the device to the floor add vibration?

thanks

 

[Strong]

"...will locking the device to the floor add vibration?"
No, if the stiffness and damping properties of the floor does not change, then locking would not make a change to the vibrations in vertical direction. Lateral vibrations may change from locking, but that would depend whether the lateral floor vibrations are a potential source or not.

Walt

 

[KENAT]

Are you going to have a seismic survey of your new location done?

Any potential for improving things such as moving to basement with thick foundations and or tied into bedrock or equivalent pedestal etc.?

How about changing the mount to true kinematic - even if not optimal - as you move?

I.e. split one of the V blocks so one side acts on on ball and the over side of the V acts on another. i.e. 2 'V's' and 2 opposed slopes. The force vectors may not play out quite as well but might be good enough - or at lease might be better than 7 point pseudo kinematic mount.

Would give you a chance to reconsider ball & groove size and finish. Also to play tricks with the material and even things like a 'gothic arch' rather than simple V.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[vono]

The new location that I need to move the device into is not negotiable. There will be vibration testing done next week so I will have a baseline to reference from. If I just increase the ball and groove size will this help? I do notice some wear on the vee grooves so now is the time for me to improve this design. I do not have the space to make this a 3 point mount. If I could make a new frame then I could do this but time and money do not allow this. As I said before once this device settled into position it was okay...but the repeatability was an issue. The CG is in the center.

I have attached a photo of the vee block and ball.
thanks

 

[KENAT]

Don't just make the balls bigger - do the math and work out how much bigger, what shape & angle the 'V' blocks can be and what the respective materials need to be, surface finish required etc.

Hertzian contact stress isn't that complex - even I worked it out eventually.

Do you have facility for balls in all four corners?

If so then as I mentioned before you can split one of the V blocks into 2 'ramps' so that all 4 corners are supported. See below image 2 blue 'v blocks' 2 green 'ramps' giving total of 6 points of contact if you have 4 balls, one in each corner of mating part.

You may need to consider the geometry for optimizing the force vectors etc.

Do some research or at least Googling don't rely on us to give you all the answers.

http://www.precisionballs.com/tech_papers.php

Or if you have budget the below folks may be able to do the hard math for you (even if not budget they have some FAQs).

http://www.g2-engineering.com/technology-FAQ

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[KENAT]

I was doing some searching on a kinematic mount I may need to design and came across the below link:

http://www.precisionballs.com/split_vee_block.php

Points out that the spherical ball element(s) touching the split V need to be tangential to both surfaces for true kinematic/to avoid various issues. I.e ideally it would be one very big ball or truncated ball. Failing that portions of the nominally single ball.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​