Surface grinding/lapping

[ksv86241]

I'm designing a laminated superconducting magnet which is 2m long, made up of around 250 individual plates. The required flatness of the pole pieces across the length is 10 microns. Our original idea was to join all the magnet plates together using dowels, controlling the location of the dowel holes to the pole surface very accurately. The flatness from pole to pole would depend on the hole location tolerance and larger scale errors would be taken out using shims. However, we've not been able to achieve the necessary tolerances on the plates; there just seems to be too much tolerance build up with the location and straightness of the hole, flatness of the poles surface etc. The best pole to pole flatness we've achieved is about 30 microns.

To me, the obvious solution is to assemble the magnet, then do a post-assembly machining operation i.e. surface grinding or lapping. However I'm having trouble finding out whether such an operation is possible at the lengths and tolerances in question. Obviously surface plates can be made to the sub-10 micron flatness level but would that process translate to this application?

Apologies if the above is poorly conveyed, it's a Friday afternoon...

Thanks,
Steve

 

[saberblue]

It is not going to be easy finding a shop set up to do this kind of work. However what I would do is contact the manufactures of 4 or 5 different types of laps and ask them if their equipment could do the job -- they may have lab equipment and do an occasional job if asked properly -- or ask them to send you to the job shop that is most qualified to do the job --

My experience is that lapping is as much an art as it is a science so good luck

 

[unclesyd]

I would get in touch with like American Grinding and discuss you problem. The will not be able to get to your sated tolerance but very close.
There are some large gantry polishing machines out there that could diamond lap to your requirements.

http://www.americangrinding.com/divisions/grinding/

 

[KENAT]

Maybe, talk to semi conductor/data storage equipment makers or related industries. They do some crazy stuff on lapping etc.

http://www.veeco.com/optium-lapping-dicing-systems/index.aspx

for example.

I'm not saying they can do something as big as you list but maybe you can learn something.

Posting guidelines faq731-376

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

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

 

[BillPSU]

Lapping on a grey iron or ductile iron surface can be performed down .000035" within a 8" diameter. This was achieved on a hydraulic charge pump body. The problem you will have will be the size of the part. A 2 meter size part will require a huge lapping machine and I would be concerned about the part rigidity being able to maintain the flatness in a position not identical to the lapping process. Another concern would be the material being lapped. The lapping machines I have experience with have a cast iron platen which has abrasive charged to its surface. If you part is too soft the abrasive particles could actually transfer into the work piece.

 

[MikeHalloran]

Suppose you secure your superconducting magnet in its normal position and (somehow, magically) lap the pole pieces into alignment within 10 microns. Just for purposes of argument, let's assume that it can be done.

Now, invert the magnet, still on its normal mounting means, on some theoretical infinitely rigid support.
Are the pole piece faces still going to be aligned within 10 microns?
I.e., how much is the assembly going to deflect under the influence of its own weight?



Mike Halloran
Pembroke Pines, FL, USA

 

[unclesyd]

mikeHalloran,
I was thinking the same thing concerning the high flux and keep the field aligned. Using a Blanchard and Mattison grinders he should be able to get the faces parallel and flatness to 0.0005"

ksv86241

If it is possible to get to the 0.0005" flatness and parallel with a the appropriate setup diamond lapping should get you the required precision.

I would contact someone with the National Magnetic Lab at Florida State University.

http://www.magnet.fsu.edu/search/personnel/getprofile.aspx?fname=alexander&lname=gurevich

I would also contact someone at the Mirror Lab at the University of Arizona

http://mirrorlab.as.arizona.edu/LBT.php?navi=1cast

 

[ksv86241]

Thanks to all for the replies, some hugely useful info there.

BillPSU: That's a really good point about the material. We'll be using as close to pure iron as possible, so transfer of particles may be a problem. I'll have to follow that one up.

MikeHalloran: The magnet will be supported at regular intervals from a massive beam. Any deflection under it's own weight should be able to be corrected with shims. I think it's really the flatness from each pole to its neighbour that is critical?

Once again, thanks for the contacts and links!

Steve

 

[MiketheEngineer]

Kind of agree w/ MikeH... Once had a mfg gloat how flat his piece was...Then I put my hands on it. Within ten seconds, it was out of bounds.

 

[MikeHalloran]

If the magnet's own weight can distort it out of spec, what is the applied field going to do to it?



Mike Halloran
Pembroke Pines, FL, USA

 

[unclesyd]

ksv86241,
Can you comeback with the overall dimension's with the require precision.

What is the allowed radius on the periphery of the point of interest surface?

what is the procedure for measuring the flatness of the surface?

 

[ksv86241]

unclesyd,

Obviously a drawing would be ideal in the circumstances, but I'm only a graduate, so a little wary of giving away anything I shouldn't...

There are two magnets facing each other, 2m long and the surfaces in question are 80mm wide. The magnets are made up of alternating poles and coils and it is the height of the poles that is critical to the field profile.

I'm not sure I understand what you mean by the peripheral radius? I believe the outer edge of the surface is largely irrelevant to the field profile, so hopefully it shouldn't matter.

I'm afraid I don't know the name of the system we'll be using to measure the flatness, but it works by using a pendulum to measure the direction of gravity with respect to the surface in question. I believe it's accurate to less than a micron per metre. Apologies for my ignorance on the subject.

Our other solution, asides dowel location and post-assembly operation, is to lay the plates on a flat surface and insert a dowel into a clearance hole, which will then be filled with resin. I'm not entirely convinced by the long term stability of resin under the loads in question, hence my preference for post-assembly machining.

 

[unclesyd]

The reason ask, probably not the best description, was for one possible way to achieve your required flatness. You dimension is too narrow for my idea, You have to be real careful not to roll an edge when grinding or lapping, diamond being the exception. When grinding or lapping with conventional abrasives we pug holes and put the part in a frame to mitigate the rolled edge.

From your description of two opposing magnets brings into play that the face and bottom of both magnets would involve some pretty high precision. This would require both magnets to be machined an ground together.

I talked to a colleague of mine and he stated that he could constantly grind to 0.0002.5" on parallel sides and lap to quarter light band on one surface. We also discussed the situation where you are using ARMCO iron as we work with hardened tool steel. We used a 60" Blanchard and a 48" diamond lapping machine for our work. We also discussed the Mattison grinder and our consensus was that with care you could beat 0.0005"

 

[unclesyd]

Hope you are still checking.

I've thought about several potential approaches to your problem. The first would be to talk to someone like Standridge Granite with your requirements. As they produce Laboratory Grade Granite Surface plates they may be able to point you to someone who makes Cast Iron Surface Plates with similar tolerances.

Another approach might be to contact someone who makes CI work holding tombstones.

The last one is to contact Lafayette Grinding as they can handle large objects and finish with good precision.

http://www.standridgegranite.com/surface.htm#8

http://www.lafayettegrinding.com/index.htm

Off the wall:
If you could get your part close to size by grinding one of the Granite Finishers might be able to bring it to size. The only problem I see is that you would be finishing an edge.