Shimming Techniques

[Dr_GMJNParticipant @dr_gmjn]

All, I'm about to have another go at shimming the column of the SX2P to get it perpendicular to the tables. I last did it back in spring, and in fact got it pretty much perfect according to my dial gauge mounted in a collet, and swung onto 4 points on the bed.

However, I've always had issues with this machine regarding chatter when side milling. Despite it working for the majority of what I've had to do to build a small steam engine, whenever I've tried side milling of any significant section, no matter what cutters or depth of cut, or techniques I've tried, the terrible noise and chatter remains.

The only reason I can think of is reduced stiffness of the column base joint: The column needed shimming front to back and side-to side, it also ended up needing an extra shim in one corner to work. All the shims ended up being pushed across the corners of the column until they touched the bolts, so a triangular contact patch – and small ones at that.

What I want is pre-cut U-shims to fit around the bolts, but I know this is going to be tricky, because the further the shims are around the bolts, the more they upset the adjustment because they're always being pushed into a wedge, not into a parallel gap (I 3tried this previously with home made U-shims). I always ended up chasing my tail – it's all perfect apart from one shim in one corner, which when you put it in place and tighten the bolts upsets the rest.

So how do you figure out the correct shims for non-parallel gaps?

Where is a good place to buy U-shims?

What I'd love to do is loosen the bolts slightly, inject a liquid shim under the column as it is, and progressively re-tighten to that. I even thought of JB weld, but unsurprisingly, the compressive strength is a fraction of that of steel, so probably more flexible than what I have now.

Of course the other option is to leave it alone, just like I wish I'd done with its gas strut, and the ML7 tailstock…

Thanks!

[Dr_GMJNParticipant @dr_gmjn]

[Tony Pratt 1Participant @tonypratt1]

Have a look at 'Tramming a milling machine with epoxy' on you tube made by Stefan Gotteswinter [August 20th 2016] it may be of use to you?

Tony

[John HinkleyParticipant @johnhinkley26699]

If you really want to squirt epoxy in the gap to tram it, you could look at Stefan Gotteswinter's YouTube video where he does just that. I wouldn't want to try it myself, but he gets it to work. I don't have anywhere near his skill level, though.

Link to Epoxy tramming

John

Slower typer than Tony!

Edited By John Hinkley on 14/12/2020 14:27:21

[Dr_GMJNParticipant @dr_gmjn]

Thanks both.

Using epoxy was really just an idle thought – Seems easy on the face of it, but is no doubt not.

I'd rather use conventional shims, but am at a loss to figure out how to make a tapered shim set from parallel steel stock.

[Rod RenshawParticipant @rodrenshaw28584]

Misread this title as shimmying techniques,

[JasonBModerator @jasonb]

One option may be to shim it as you have already done with shim in the corners. When it's OK make a note of what shim fits where and take it apart. Apply JBWeld and then replace the shims and adjust it back to the readings you want while the JBW is still soft. That way the shim will take the load but once set the JBWeld will fill any gaps and firm things up.

Myself I just have a bit of feeler gauge poked in the side of both the X3 and SX2.7

[Dr_GMJNParticipant @dr_gmjn]

Posted by JasonB on 14/12/2020 14:52:53:

One option may be to shim it as you have already done with shim in the corners. When it's OK make a note of what shim fits where and take it apart. Apply JBWeld and then replace the shims and adjust it back to the readings you want while the JBW is still soft. That way the shim will take the load but once set the JBWeld will fill any gaps and firm things up.

Myself I just have a bit of feeler gauge poked in the side of both the X3 and SX2.7

 

Thanks Jason – that's what I meant when I said loosen the bolts slightly, inject a liquid shim under the column as it is, and progressively re-tighten the bolts.

I'd really need to dismantle it to put some release agent on the surfaces before doing that anyway I guess.

Feeler gauge into the entire side would be better, but the shimming requirements of mine make it impossible.

Edited By Dr_GMJN on 14/12/2020 14:59:35

[Dr_GMJNParticipant @dr_gmjn]

Looked at the video – hmmm. I might try it, nothing to lose. Might get some of that proper shim epoxy rather than JB weld.

I have used Henkel aerospace liquid shim at work before (not a conventional application though), and it did have some residual compression when bolts were tightened. I reckon that I could probably eliminate any tiny errors by adjusting the bolt torque and checking as I go.

[old martParticipant @oldmart]

When we got the drill mill for the museum, it was a basket case, having been stored in bits under a tarpaulin in somebodys garden for three years. Everything had to be cleaned up and a lower spindle bearing was fitted. The round column was hollow and it still had a coating of foundry sand inside. I gave up trying to remove every grain of sand which kept falling off and found a large cork plug to fit in the bore. The cork had been a NOS part for a helicopter, I have no idea what it was for, but it stopped any danger of the sand grains getting into the joint between the column and the base as they were fitted together. The Taiwan made mill has never had a tram problem, luckily. I would guess that even 0.001" shim on one side would make a big difference and more worrying that the cast iron would be stressed badly by uneven clamping.

[larry phelan 1Participant @larryphelan1]

Are you guys making parts for spaceships ?

[gerry maddenParticipant @gerrymadden53711]

Dr_GMJN, I dont think you need worry too much about the elastic qualities of the epoxy.

If you imagine a 250mm bolt torqued to its maximum it will extend by approximately 1mm. That means every 1mm slice the bolt has extended by 0.004mm, or every 0.1mm slice has extended 0.000,4mm. If you now reduce the stress level to a quarter, the deflection will also reduce by a factor of 4 taking it to taking the change in thickness of that 0.1mm slice to 0.000,1mm.

Of course your epoxy shim will have a lower modulus than the bolt but is also likely to be thinner that the 0.1mm slice that I referred to, and your stresses likely to be considerably lower too. In addition, the fact that it will be in compression as a thin film will automatically make it even stiffer than the same material would appear to be under a pure unadulterated tension test. So keep it thin and all will be fine.

Gerry

[duncan webster 1Participant @duncanwebster1]

Assuming both surfaces are flat, and that the bolts are in a rectangular pattern, you should only need to shim 3 corners. The thickness of the shim opposite the one that isn't shimmed should be equal to the total of the other 2. This is probably as clear as mud, so call the corners N, S, E, W and assume N is not shimmed, then S = E+W. If this is not achieved then something is having to bend or it's not pulling down on one corner.

If you're going to back it up with epoxy, then I'd add it all round and tighten the bolts fully.

[Dr_GMJNParticipant @dr_gmjn]

Posted by larry phelan 1 on 14/12/2020 16:45:51:

Are you guys making parts for spaceships ?

No, but some of the issue is conflicting information: You read some model engineering websites or books eg Sparey, or similar, and they say words to the effect that accurate work can’t be done unless, for example the tailstock is aligned, or any number of other checks. They then go on to outline how to adjust things to perfection, using new or high quality equipment. Same with mill tramming, or gib adjustment, or…whatever. They never say ‘this is how you adjust an old/worn/cheap/hobby grade machine tool to be as good as you’re ever going to get it…

So what do you expect beginners like me, with aforementioned old/worn/cheap/hobby grade machine tools to try and do? We’re damned if we turn out dodgy work or have issues with badly adjusted (or difficult to adjust) tools, and we’re dammed if we try and fix them!

I totally get that it’s going to be a compromise, and that micron accuracy isn’t necessary, but I don’t get the criticism of struggling to get things as good as they can be. After all, how are we supposed to magically know what’s a fixable issue and what’s just ‘the way it is’?

[Tony Pratt 1Participant @tonypratt1]

Dr GMJN, this hobby/skill is very difficult to learn from a forum as you will get so much conflicting 'advice', thinking back to my apprentice days there was so much to learn but I had good teachers & I like to think a good attitude plus a mechanical aptitude, if you have the last two you will improve day by day! If you are not happy with your mill try to put it right, there are so many lessons to learn along the way & don't listen to the key board warriors who like nothing better than to criticise.

Tony

[John ReeseParticipant @johnreese12848]

Gotteswinter used a special epoxy compound for his tramming that is not available to us common folks.

[Anonymous]

Posted by Dr_GMJN on 14/12/2020 20:56:44:

So what do you expect beginners like me, with aforementioned old/worn/cheap/hobby grade machine tools to try and do?

There are many causes of chatter and I wouldn't put stiffness of the column high on the list. What happens if you take a small width of cut with a largish endmill; say 0.02mm width of cut, 10mm depth of cut and a 10mm diameter endmill in brass?

Over the years I've had all sorts of problems with poor finish and less than ideal cutting performance but they all turned out to be poor tooling and technique, nowt to do with the machines. My machines are old, worn with significant backlash and were relatively cheap but I haven't worried about making many measurements on them other than things such as tramming and tailstock alignment, I've just made parts. There are those on the forum that have turned their nose up at my work, but I don't give a XXXX!

When Sparey et al were writing the import hobby machines didn't exist, so they didn't write about them.

Andrew

[old martParticipant @oldmart]

If you know a good engineering firm that will do one offs, you could get a shim plate made. If the tilt was measured as 0.002" across the same width as the base joint, then a piece of gauge plate could be surface ground with the same angle of tilt. You could buy a suitable piece, and drill the holes yourself. Slightly longer bolts would be needed, but the mills Z height would be increased by the thickness of the plate, a useful thing on most small machines.

[mgnbukParticipant @mgnbuk]

Stefan used Diamant products IIRC. An alternative is the SKC equivalent.

Both would likely be sold to a small volume user – you may wish to be sat down when you open the quote, though.

My recollection of the SKC product was that it came in two pre-measured tins & had to be mixed and used as one lot. There is a similarly expensive release agent available to coat one face to allow for dismantling. The resulting cured compound is hard & brittle. Broadbents used this on the Broadbent Hayes Digimill under the column. They superglued O-Ring cord around the edge of the column & pulled down against this to get the column square both ways. The compound was injected through pre-drilled & tapped holes with a screw-type press / injector, moving from hole to hole as the compound made it's way into the joint. The just vacated hole was plugged with a grub screw. This method was somewhat fraught as I recall & I seem to think they went back to scraping the joint for alignment, as this ended up being quicker and more relaiable than the "time saving" method.

Boxfords just used Isopon P38 polyester filler to seat bed pedestals on to the cabinet of underdrive machines. The bed with fitted pedestals was sat onto a good bed of filler at each end on the cabinet top, then bolted down while checking for twist with a precision level. Excess filler was trimmed off with a putty knife before it had fully hardened.

Nigel B.

[Dr_GMJNParticipant @dr_gmjn]

Posted by duncan webster on 14/12/2020 19:30:21:

Assuming both surfaces are flat, and that the bolts are in a rectangular pattern, you should only need to shim 3 corners. The thickness of the shim opposite the one that isn't shimmed should be equal to the total of the other 2. This is probably as clear as mud, so call the corners N, S, E, W and assume N is not shimmed, then S = E+W. If this is not achieved then something is having to bend or it's not pulling down on one corner.

If you're going to back it up with epoxy, then I'd add it all round and tighten the bolts fully.

Does that apply to a rectangle, or just a square? Either way, I can’t see how it applies in reality, since the shims have area (which I want to maximise), and the larger that area, the more impossible it becomes to get full contact on non-parallel surfaces. The only solution for full contact is a shim that perfectly confirms to the wedge between the surfaces.

[Dr_GMJNParticipant @dr_gmjn]

What I was wondering was, is there a method of figuring out how to step (or facet) the shims to approximate the correct wedge in each corner?

[JasonBModerator @jasonb]

You could work it out in CAD, but the biggest issue will be cutting contours that are maybe 1 micron between heights

[Pete RimmerParticipant @peterimmer30576]

The time-honoured way to remedy this is to scrape the parts into alignment. What error are we talking about?

[duncan webster 1Participant @duncanwebster1]

Works for square and rectangle. As long as you get shims round the bolts, the shims are nipped without a lot of load on the bolts, and then the bolts are tightened down properly I don't think you need worry about the bits in between. However you could always add bits to the centre of the edges (if you see what I mean) these should be the average of the 2 adjacent shims, again assuming the surfaces are flat. The problem is that if these are even a tad too thick they will stop the ones under the bolts seating properly, so I'd make them a smidgeon thinner than they should be and use the epoxy, anything will do, as someone else said it is very thin so very stiff. I just did the 4 corners on my Naerok and no further trouble.

I'd get the tramming right with the head at a useful working height, it might change as the head is moved up/down

[Michael GilliganParticipant @michaelgilligan61133]

Please see my posts on this brief thread from June-2020 : **LINK**

https://www.model-engineer.co.uk/forums/postings.asp?th=166118

MichaelG.

[ Loctite 638 Appreciation Society ]

[Author]

Posts