guideway surfaces regrind

[gerry maddenParticipant @gerrymadden53711]

[gerry maddenParticipant @gerrymadden53711]

My microscope X-Y table guideways might need a regrind. They have a little wear but it's the brinelling and corrosion pitting that's the real problem as this affects the smoothness of operation.

I think the regrind would need to be a precision job with no more than say 5um taken off each of the surfaces of the 45 deg. flanks. At the same time I cannot risk having any part of the 250mm length of the guide not cleaned up or that might introduce some other problems.

Do you think this is a feasible operation ?

If so does anyone happen to know anyone within say 50 miles of Milton Keynes that might be capable of this ?

The reason I need to keep cut depth to a max of 5um is because I have to compensate for the machining with some oversize balls to avoid introducing play into the system. I think the largest oversize 1/2inch balls are about +10um.

Your thoughts are welcome.

[blowlampParticipant @blowlamp]

Have you considered lapping the surfaces?

Mill some copper/aluminium to 90 degrees, charge it with abrasive paste and give it a go is what I'd try first.

Martin.

[Michael GilliganParticipant @michaelgilligan61133]

What an interesting problem, Gerry !!

… Those are mighty big balls for a microscope [presumably the OMT]

A quick trawl of t’internet found TE-BAR : **LINK**

TE-BAR Slideway Grinding

More than 50 miles from MK, but it looks like they would have the capability.

… Worth asking for a quote, if only as a baseline

MichaelG.

[Ian PParticipant @ianp]

Regrinding the whole face seems a bit OTT as there is only line contact. Why not get bigger or smaller balls?

OK, I dont know the the construction but I would have thought smaller balls and packing to correct the height should be doable.

Ian P

I made the mistake of replying before reading your whole posting and seeing the the oversize ball bit. I was thinking of a change of ball diameter sufficient to move the running area away from the worn track

Edited By Ian P on 13/10/2020 23:15:26

[Ian PParticipant @ianp]

May or may not be relevant but I made the mistake of thinking the faces of typical engineering Vee blocks were a 90 degree included angle.

Your guides might be 45 degrees but the exact numerical angle value is not critical for operation so if you do regrind best to check first.

Ian P

[Tony Pratt 1Participant @tonypratt1]

You state 'cut depth to a max of 5um', which is .005 mm so not a grinding job as it would be nigh on impossible to set up & remove such a small amount but as previously stated lapping would be you best bet.

Tony

[David ColwillParticipant @davidcolwill19261]

Posted by blowlamp on 13/10/2020 22:08:07:

Have you considered lapping the surfaces?

My first thought too!

You can also use 3M lapping film which is self adhesive. This can be stuck to a piece of ground steel and comes in various grades.

Regards.

David.

[gerry maddenParticipant @gerrymadden53711]

Thanks all for your comments.

Michael, your link was an interesting one – how do you find these things so quickly ?! Anyhow I've sent them a mail and asked for a quote. I think it will be expensive, but as you say its a baseline.

Tony – As I thought about it more last night I concluded that the best way might be to drop the bars into V-blocks and skim each flank using a surface grinder. (The side of the wheel should clear the other flank due to the undercut at the root.) Assuming the guides are already precision machined, most of the error 'could' be in the heights of the V-blocks so some shimming might be required here to get the guide perfectly parallel with the table. I'm by no means an expert in such a field so I would be very interested in your thoughts as to why 0.005mm would be a challenge.

Lapping proposers – yes this went through my mind too. I was worried by two things. Firstly straightness is important. Unless the lap is equally straight, as long and rigid as the guideway itself, would I be at risk of producing small undulations along its length? My second concern was to do with the width of the surface. Its only about 5mm wide so I was thinking the lap would be a little unstable and i would probably end up converting the flat into a camber. Have I got the wrong idea here ? Sorry Blowlamp – I have just re-read your post and I see you are proposing a 90 deg block, presumably to do both flanks at the same time. Yes this solves my stability concern.

Ian P – unfortunately the guides are dowelled and bolted to flat surfaces so it wouldn't be easy to allow for a smaller or larger ball without a lot of argy bargy.

[Tony Pratt 1Participant @tonypratt1]

Posted by gerry madden on 14/10/2020 19:33:46:

Tony – As I thought about it more last night I concluded that the best way might be to drop the bars into V-blocks and skim each flank using a surface grinder. (The side of the wheel should clear the other flank due to the undercut at the root.) Assuming the guides are already precision machined, most of the error 'could' be in the heights of the V-blocks so some shimming might be required here to get the guide perfectly parallel with the table. I'm by no means an expert in such a field so I would be very interested in your thoughts as to why 0.005mm would be a challenge.

With respect I don’t think you realise how small .005 mm [.000197”] is in reality, in practical terms it will be extremely difficult to set up a grinding machine to take off evenly .005 mm, the parts are very unlikely to be flat within said .005 mm, there will be twist, distortion etc. etc.

Reworking a part to remove such a small amount is not really grinder territory but is very doable using a lap, how you measure the metal removal is open to conjecture. You would have to rely on the existing part to guide the lap.

Tony

[gerry maddenParticipant @gerrymadden53711]

UPDATE …..

Tony I think you were right. I did some flatness checks with a micron DI on a surface plate and although the guides were surprisingly accurately ground it was quite obvious a 5um cut would leave some areas not cleaned up. This was the last thing I wanted. I therefore used the corners of a small V-block and some carbide papers to clean up the surfaces.

I didn't manage to remove the brinells completely (which I measured to be about 5um deep) but removed enough metal to take away the rims around the craters, and the corrosion pitting and staining. This should make the things smoother.

As I need to make new ball-spacer bars, I'll produce the new ones with the ball spacing slightly different to the originals. This will ensure that all the balls don't fall into what remains of ruts at the same time, hopefully further improving the smoothness of travel.

Thanks all for your inputs.

[old martParticipant @oldmart]

Those guideways look exactly like the ones used at work for machines. You may be able to get new ones and save a lot of bother.

[gerry maddenParticipant @gerrymadden53711]

I probably couldn't afford new ones but I would be interested to know who might sell such things. Do you have any names old mart ?

[old martParticipant @oldmart]

Sorry, I didn't have much to do with them actually. In the fitting shop we had some surplace new ones given us which had been cut in half and they made excellent parallels. I think they were German make. There are lots available on the unmentionable website which might be identical in dimensions to yours.

[Nigel Graham 2Participant @nigelgraham2]

Interesting problem…

I would say this is "two-thirds" of the reciprocity method for making matched trios. Two-thirds because the V-block here is a tool, not one of the work-pieces.

With this, 3 identical work-pieces are involved, A, B and C . A is lapped to B, B to C and A to C until all three are mutually flat, and they will be flat to other reference surfaces within reason; without needing a "fourth-party" tool. I believe this is the way the first surface-plates, optical flats etc, were made.

It needs 3 because using just two work-pieces reciprocally risks making one convex to a matching concave brother. In this instance, using the Vee-block avoided that hazard.

I am not sure it would be as safe to use a block like that for finishing parts from milling, rather than servicing existing ground parts needing only very light trueing. That is because having to cut more metal risks very small differences in profile, tiny but possibly significant in an application like microscopy.

What may work for a V-guide, and I have not experimented with it, is a reciprocal lapping of each long face in turn against each of those on the other guide.

'

Incidentally I encountered this A-B-C approach first not in mechanical engineering, but the basis of a standard method for calibrating trios of identical hydrophones of types that can be driven to emit sound as well as detecting it, their normal microphone action. As I recall though, it still needs their other electrical characteristics determining by precision measuring equipment, and certainly needs calibrated amplifiers and voltmeters etc., so is not as absolute as the metal-finishing analogy.

[Author]

Posts