Deep hole processes

[ksv86241]

Having previously posted the "surface grinding/lapping" thread, I have been convinced by my design group that the topics discussed in there were dead ends. I'm now investigating an alternative method of meeting the tolerances in question.

To re-cap, I'm trying to make a laminated superconducting magnet consisting of around 250 small iron plates clamped together. One edge of these plates constitutes the 'pole surface' and I have to keep these pole surfaces to within a 10 micron flatness across the entire 2m length of the magnet.

The method I've now come up with is as follows:
- Manufacture the plates with undersized dowel holes, with the location tolerance from hole to the pole surface controlled as tightly as possible
- Clamp the magnet plates together using separate clearance holes, so that it's essentially flexible.
- Clamp this flexible magnet down to a theoretically flat support structure, such that the poles take on the flatness profile of the support.
- Use some process (honing?) to bring the undersized dowel holes out to size, also ensuring their concentricity to each other.
- Shrink fit dowels into the holes, creating a rigid magnet assembly
- Heat treat the entire assembly, including support, to remove any stresses caused by the clamping and doweling processes
- Take the magnet off the support to wrap it with superconducting wire and pot it with resin
- Hopefully it then sits back on the support structure with the poles nice and flat, resting into the position that it was originally heat treated in
- I get a big pat on the back from the scientists and a promotion

My main question concerns the process to bring the dowel holes out to size and ensure concentricity. I like the idea of gun drilling the entire length without any pre-drilled holes, but the whole assembly is massively heavy and I'm concerned that rotating it on a gun drilling machine may be impossible. So now I'm looking at honing or something along those lines. However, I'm woefully inexperienced at these kind of processes...

Would honing be able to do this? The thing I really want to avoid is the dowel "fighting" with the clamping to the support structure, hence the desire for near-perfect concentricity. I can imagine a grinding bar, rotating eccentrically, might do the job but I've never heard of such a process being undertaken.

Does the plan as a whole sound at all feasible? Please feel free to criticize any and all aspects of it, as no one at work has the necessary experience to do so.

Many thanks,
Steve

P.S. image attached, hopefully illustrating the situation a little more clearly

 

[unclesyd]

Essentially there is only one way to finish your dowel hole and that is with the Engis Diamond Honing System. The beauty of this system is that it will give you the required hole size and not go over. The overall accuracy is outstanding even on deep holes.
I would give Engis a call and talk to an applications engineer to discuss your problem.

http://www.engis.com/single-pass-honing-bore-finishing.php

 

[MikeHalloran]

Why _three_ dowel holes?
Two will overconstrain the problem and fight each other; more is just asking for trouble.

Actually, I think the (w)hole strategy is a loser.

I think you'd have better luck lining up the primary faces and the projecting edges, then shooting something like molten zinc into the stack of punched 'dowel' holes.



Mike Halloran
Pembroke Pines, FL, USA

 

[ksv86241]

MikeHalloran,

There are three dowel holes to counteract torsional loads during the winding process. Is it really overconstrained? If the holes were drilled through each plate separately, then yes, but as they're done while the whole thing is clamped together, shouldn't it just be like drilling 3 holes into a solid lump and sticking some dowels through?

We did toy with the idea of putting in undersized dowels and filling the clearance with resin, but it would just seep out from between the plates. Maybe molten metal would be viscous enough not to though?

 

[BillPSU]

I would not pre-drill the holes. Positioning the finishing tool concentric with the pre-drilled hole will be problematic.
Tool makers for decades simply drilled and reamed the dowel holes for fixtures and driven the dowels into the assembly. The intent was to never disassemble the fixture which sounds like your intent. Using this process you can use as many dowels as you like.

 

[unclesyd]

I have to diverge a little from the above post concerning the use of tooling and dowel pins.
First is that we use 3 dowels on some large gear pumps. These dowels are the register for four shafts that have a clearance of 0.0005" ± 0.0001". The span between the outside dowel pin is 28" with the third being located on the center line of the pumps. We have numerous other precision gear pumps that use dowel pins for alignment.

The Engis diamond bore honing or hole finishing system will give outstanding results on all parameters used to qualify a bore. Unless you want to have interchangeable components the location of the dowel hole is less important than the hole geometry. This system is capable of achieving 0.000020" though I would expect less in this case due the lack of a ridged setup. We have worked from drilled holes and jig ground holes with this system. The only caveat is that it is not cheap.

 

[unclesyd]

Here are two sites that maybe of interest. Surface Finishes has some experience dealing with pure iron in this case it's the Consumate Iron from Carpenter intead of the more common ARMCO iron. I called my old boss about the two materials and he stated that he had always found the Carpenter material to be cleaner.

http://www.surfacefinishes.com/

http://www.cartech.com/ssalloysprod.aspx?id=2488