Shimming Techniques

[Dr_GMJNParticipant @dr_gmjn]

Posted by Steviegtr on 18/12/2020 17:27:30:

Yes i thought it would be awkward to do. I do hope you find a way. When you mentioned JB weld or someone did. Have you looked at the different version they do. When i bought some ,there was different strengths. Also many years ago i had some 2 part resin that claimed to as hard as cast iron, it was called BELZONA.

Steve.

Thanks Steve – I think I got the lower strength version. I'll have to check.

[HOWARDTParticipant @howardt]

Where I used to work we always used u shaped shims 5mm thick and machined them to suit. Easier than fiddling with shim steel.

[mgnbukParticipant @mgnbuk]

the upper and lower faces of the joint aren't parallel (which they're not)

The one or both of the castings will flex when you pull the bolts down.

What is the machine mounted on ?

Nigel B.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dr_GMJN on 18/12/2020 19:47:24:

[…]

It comes down to the impossible task of shimming a tapered gap with parallel shims. Hence the initial question about liquid shim.

[…]

.

And hence my observation that I have used Loctite 638 very successfully in such circumstances.

… Now that I have seen the four-pad arrangement, I am even more confident that it would do the job nicely.

MichaelG.

[Dr_GMJNParticipant @dr_gmjn]

The mill is mounted on a bench I made from fence posts and other spare bits of wood:

The top is two sheets of plywood glued and screwed together, giving a total of 35 mm depth. The mill base is mounted on an additional a plinth of 25 mm ply – mainly to raise the y-axis handweel to a practical height above the bench surface. The mill base is bolted through the entire thickness of the three layers of ply. Not sure why the photo rotated itself, but there you go:

Basically it's solid, and I don't see this as a primary issue anyway: it's a small footprint mill with a relatively stiff base. You could probably not even bolt it down and it would be fine. It's the column-base joint that's going to be the issue, not the base to table.

Re. the Loctite 638. I'm sure it would be OK, but why use retainer primarily designed for high shear-strength joints in cylindrical components in this application? Its *compressive* strength doesn't even seem to be listed on the Henkel datasheet. Surely I'd be better off using something like JB Weld, that at least has the primary mechanical property I'm looking for listed on its data sheet – and it's pretty high at that?

I assume that JB weld could be applied to the gap after allowing it to cure somewhat, so that it doesn't run everywhere, then tighten in two stages as I proposed previously?

Thanks all.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dr_GMJN on 19/12/2020 00:28:11:

.

Re. the Loctite 638. I'm sure it would be OK, but why use retainer primarily designed for high shear-strength joints in cylindrical components in this application? Its *compressive* strength doesn't even seem to be listed on the Henkel datasheet. Surely I'd be better off using something like JB Weld, that at least has the primary mechanical property I'm looking for listed on its data sheet – and it's pretty high at that?
 

.

I used 638 because we had it available … and because I had predicted from its known properties, that it should work.

… the sharing of knowledge and experience seems appropriate to the forum, but I am quite content to leave you to find your own method.

MichaelG.

.

Quoting from my post on 16th : 

Please note that my successful use of 638 [in its original formulation] as a ‘flat’ shim is clearly outside the product’s design-envelope … 

 

Edited By Michael Gilligan on 19/12/2020 08:22:40

[Ron LadenParticipant @ronladen17547]

Just a thought but have you considered that even if you could get the column/base connection completely rigid that the chatter problem is coming from somewhere else. Could it be flex in the column itself, the head to column connection, the cut speed, feed, depth etc..?

I have only experienced chatter on my SX2P when I have got the set up wrong with the right settings it is fine.

With regards to the base/column fixing you know the shim sizes and positions to achieve alignment plus you know that you have a tapered gap across the shims. I think Jason's suggestion is a good one in applying epoxy and tighten the bolts to achieve alignment with the epoxy filling the tapered gaps, sounds good to me.

[Martin ConnellyParticipant @martinconnelly55370]

The department I was in at work often had to shim parts with anything from peelable shims to nominal 50mm wedge shims. parts were joined with your basic machine screws up to hydraulic (Pilgrim for example) nuts. The only real contact area between parts was in a ring around the hole about twice the diameter of the hole. You could tell this was the case because when parts were separated after some time the parts of the faces away from the holes would have signs of corrosion. We used to mount 50 tonne alternators on 4 pairs of 2 off Ø150mm screw adjustable machine supports with spherical bearing faces. I'll try to dig out a photo.

I think extra contact points between the column and the base would complicate shimming without adding to the stiffness of the joint.

Martin C

Edited By Martin Connelly on 19/12/2020 09:46:19

[Dr_GMJNParticipant @dr_gmjn]

Posted by Michael Gilligan on 19/12/2020 08:13:57:

Posted by Dr_GMJN on 19/12/2020 00:28:11:

.

Re. the Loctite 638. I'm sure it would be OK, but why use retainer primarily designed for high shear-strength joints in cylindrical components in this application? Its *compressive* strength doesn't even seem to be listed on the Henkel datasheet. Surely I'd be better off using something like JB Weld, that at least has the primary mechanical property I'm looking for listed on its data sheet – and it's pretty high at that?

.

I used 638 because we had it available … and because I had predicted from its known properties, that it should work.

… the sharing of knowledge and experience seems appropriate to the forum, but I am quite content to leave you to find your own method.

MichaelG.

.

Quoting from my post on 16th :

Please note that my successful use of 638 [in its original formulation] as a ‘flat’ shim is clearly outside the product’s design-envelope …

Edited By Michael Gilligan on 19/12/2020 08:22:40

Thanks Michael,

Yes, I read your note before re. it being outside it’s intended use. That’s really why I was asking further. I’ve got some JB Weld already, so if I was to buy an alternative it would be something more intended for liquid shimming.

[Martin ConnellyParticipant @martinconnelly55370]

Alternator on Vibracon machine mounts. Vibracon technical page

Martin C

Edited By Martin Connelly on 19/12/2020 09:52:04

[Dr_GMJNParticipant @dr_gmjn]

Posted by Ron Laden on 19/12/2020 08:42:30:

Just a thought but have you considered that even if you could get the column/base connection completely rigid that the chatter problem is coming from somewhere else. Could it be flex in the column itself, the head to column connection, the cut speed, feed, depth etc..?

I have only experienced chatter on my SX2P when I have got the set up wrong with the right settings it is fine.

With regards to the base/column fixing you know the shim sizes and positions to achieve alignment plus you know that you have a tapered gap across the shims. I think Jason's suggestion is a good one in applying epoxy and tighten the bolts to achieve alignment with the epoxy filling the tapered gaps, sounds good to me.

Thanks Ron.

Is your SX2P column shimmed? If so, what’s the contact area like?

I have considered that the chatter is caused by other things, but you may remember my early posts back in May/June where I was trying to figure out the issue. I tried all suggestions from tooling, to materials, to ensuring axes are locked, to adjusting gibs, to climb/conventional milling. I tried everything suggested but the issue remains. The column joint has always been at the back of my mind, because I know it’s not great.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dr_GMJN on 19/12/2020 09:45:03:

Posted by Michael Gilligan on 19/12/2020 08:13:57:

[…]

 

Thanks Michael,

Yes, I read your note before re. it being outside it’s intended use. That’s really why I was asking further. I’ve got some JB Weld already, so if I was to buy an alternative it would be something more intended for liquid shimming.

.

If your difficulty is that… “Its *compressive* strength doesn't even seem to be listed on the Henkel datasheet.” then, to a first reasonable approximation, you could use the figures for Perspex.

MichaelG.

.

Edit: 

Maybe Duncan Webster could be persuaded to do the sums for a material thickness of [say] 0.1mm

Edited By Michael Gilligan on 19/12/2020 10:07:42

[Dr_GMJNParticipant @dr_gmjn]

Posted by Martin Connelly on 19/12/2020 09:08:40:

The department I was in at work often had to shim parts with anything from peelable shims to nominal 50mm wedge shims. parts were joined with your basic machine screws up to hydraulic (Pilgrim for example) nuts. The only real contact area between parts was in a ring around the hole about twice the diameter of the hole. You could tell this was the case because when parts were separated after some time the parts of the faces away from the holes would have signs of corrosion. We used to mount 50 tonne alternators on 4 pairs of 2 off Ø150mm screw adjustable machine supports with spherical bearing faces. I'll try to dig out a photo.

I think extra contact points between the column and the base would complicate shimming without adding to the stiffness of the joint.

Martin C

Edited By Martin Connelly on 19/12/2020 09:46:19

Thanks Martin.

I hear what your saying about adding area not making a difference to stiffness, but I’m thinking in this case I’m not really adding area, rather I’m trying to get back to something approaching a minimum design area.

If we assume the joint was actually designed rather than guessed (I know, I know), say it will still perform with 50% original contact area due to shimming – I’m currently nowhere near that 50%.

I always get back to thinking along these lines: taken to an extreme of assuming contact area doesn’t affect stiffness, it would mean that you could have four contacts of, say 4 mm^2. Would that really be as stiff as the same joint with four contact patches of c. 700 mm^2 area? OK it’s an extreme example, but at what point is the area not an extreme example.

The way I see it I need to Maximise the contact area while being able to maintain perpendicularity with the table.

[HopperParticipant @hopper]

Posted by Dr_GMJN on 18/12/2020 15:02:47:

Thanks Hopper.

What you describe is pretty much exactly what I've got now. The width of the "L" shims is about 4 mm or 5 mm. Any more than that and it clashes with the bolts.

Each bolt looks to be within a machined rectangular pad of about 30 mm x 25 mm:

I

I can't put shims half way between the bolt centres, because it would be in the unmachined pocketed parts of the base.

If you imagine only the outer two edges of each pad being shimmed to a maximum of 5 mm from the edges, that's my current setup.

To me, it looks precarious, and that is why I wanted to get significantly more contact area.

Edited By Dr_GMJN on 18/12/2020 15:03:12

Now I have seen what we are talking about, I have to wonder what makes you think that very flimsy casting won't pull into shape and sit nice and flat on larger shims once you tighten those bolts down?

We are talking only a few thou of variation across each shim, if you make each shim the size of one of those four pads, with a U cutout for the bolt . Each one of those bolts will exert up to about a ton and a half of clamping force. That's about ten times the force needed to pull that little casting into shape to match the surfaces of the shims.

Just whack the large flat shims in there that give you correct column alignment with the bolts tightened and carry on milling.

KISS — Keep It Simple, Sam.

[Ron LadenParticipant @ronladen17547]

Yes I shimmed mine with cut pieces of feeler gauge and I guess I was lucky in that I have only needed to shim the two front pads.

My column was leaning forward and I found that a pair of 0.005" shims corrected that error, the shims (20mm long) where fitted along the Y on the outside of the bolts. I then found that the column was leaning left in X so changed the right hand shim to 0.002" which was too much and went for 0.003" this put both the X and Y quite close. You will know that tramming with shims is a lot of trial and error and I finally arrived at 0.003" front right and 0.005" front left but in a L shape running down the side in Y and across the front in X. The two rear pads are not shimmed at all.

Out of interest I just checked to see if things have settled and I have as near as damn 0.001" in Y across 135mm and 0.0015" across 175mm in X, I,m quite happy with that for the work I do.

Ron

[Dr_GMJNParticipant @dr_gmjn]

Hopper, the whole reason for the thread is that when shims of reasonable area are fitted and the bolts tightened, the column deflects out of true. Much trial and error showed that the only reliable method required shims of marginal width. This is because it's a gap that's tapering diagonally, and the lowest corner is always going to press into its mating surface while leaving the opposite corner of its pad in fresh air.

[Dr_GMJNParticipant @dr_gmjn]

Posted by Ron Laden on 19/12/2020 10:51:08:

Yes I shimmed mine with cut pieces of feeler gauge and I guess I was lucky in that I have only needed to shim the two front pads.

My column was leaning forward and I found that a pair of 0.005" shims corrected that error, the shims (20mm long) where fitted along the Y on the outside of the bolts. I then found that the column was leaning left in X so changed the right hand shim to 0.002" which was too much and went for 0.003" this put both the X and Y quite close. You will know that tramming with shims is a lot of trial and error and I finally arrived at 0.003" front right and 0.005" front left but in a L shape running down the side in Y and across the front in X. The two rear pads are not shimmed at all.

Out of interest I just checked to see if things have settled and I have as near as damn 0.001" in Y across 135mm and 0.0015" across 175mm in X, I,m quite happy with that for the work I do.

Ron

Thanks Ron, yes shimming two adjacent corners was fine – to take out error in one axis. As soon as I try to correct for the other axis, that's when things get messy.

I know its very easy in theory, and adding another full shim to the third corner should be trivial with a bit of trial and error, but there was just no way it was having it with shims of reasonable area.

[SteviegtrParticipant @steviegtr]

Here have a look at this site. This is the company i mentioned. They do liquid metal. They have a place in Harrogate.

Belzona

belzona polymetrics Claro Road. Harrogate. HG14AY. Tel 01423 567641.

Steve.

[Martin ConnellyParticipant @martinconnelly55370]

Get some spherical washer sets for the four bolts, shim them individually with u shaped shims and that will sort out any taper in the joints. You will raise the column by the thickness of the spherical washer sets but get the best joint stiffness possible. The limit on stiffness is the base casting and column fabrication not the contact area between the two parts. Shimming only sorts out the orientation of the parts and cannot be used to increase stiffness. Poor methods may introduce a slack joint but good methods cannot increase stiffness beyond the that of the design.

Mount the column with 3 spherical washers and no shims. Measure the gap for the fourth spherical washer. If it is larger than the remaining spherical washer then shim it to fit the fourth space. If it is less then shim one of the fitted washers to allow the fourth one to fit. You then have a starting point to measure and shim the column for a solid joint with no soft foot problems.

Martin C

[Dr_GMJNParticipant @dr_gmjn]

Martin – I'm saying that the shims I've ended up fitting will probably have reduced the joint stiffness – I'm not hoping to increase it beyond what it would be for full contact, just wanting to get back to something like it was without shims.

Interesting idea about the spherical washers. I guess they'd have to be slightly wider ID than the bolts, since they have to move horizontally a small amount in order to orientate themselves parallel with each other? Maybe standard clearance would be enough for the amounts we're taking about. I will investigate sizes.

Thanks.

[Martin ConnellyParticipant @martinconnelly55370]

This is a spherical washer set, it's just an example of what I am talking about, you would need some to suit the bolts you have.

Ø13 bore spherical washer set.

Martin C

[mgnbukParticipant @mgnbuk]

Basically it's solid it could be better, and I don't see this as a primary ssue anyway it could be a contributary factor: it's a small footprint mill with a relatively stiff flimsy base.

Fixed that for you.

You could probably not even bolt it down and it would be fine

Doubtful. Bolting down to a substantial foundation is a pre-requisite for many heavy machine tools to perform as designed – why do you think a lightweight hobby mill would be any different ?

It's the column-base joint that's going to be the issue, not the base to table is an assumption

If the plywood base has no bracing underneath, it could be that the mill is just sat on a diaphragm. If there is no brace, adding a 3"x2" stiffener fore & aft under the centreline of the mill base (long side down), glued & screwed could help, as could replacing the top raising section of plywood with a pressed concrete paving slab & bolting through the lot. Plywood is a lot more flexible than concrete.

Don't under estimate the value of a solid foundation – it makes a big difference to large machine tools with far more rigid castings than a hobby milling machine. If you dig around, you will find reports of users of similar style hobby machines to yours resorting to adding a substantial bracket from the top of the column to the building wall to stiffen up the machines.

Maybe look at some other aspects before fixating on complicated "work-arounds" for less-than-perfect construction aspects that I suspect will not yield the improvements in performance that you desire ?

Nigel B.

[Dave HalfordParticipant @davehalford22513]

You said this in your OP!

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.

All I would have done is put the dial gauge on the fourth bolt tab, read the extra deflection given by the tightened bolt and increased the shim by the same amount .

Then try a cut and see the difference rather than just assume it will be rubbish.

[duncan webster 1Participant @duncanwebster1]

This is a sort of mind dump, no sums involved yet. A preloaded bolted joint with the faces in contact acts as tho the joint was not there, solid metal, until the load applied exceeds the preload. All of this preload is applied via the underside of the bolt. I've no idea what size the bolts are, but if they are M8 that works out at (13^2-8^2)*pi/4 = 82 sqmm (not allowing for a clearance hole which will make it smaller). If they are M10 it comes to 148 sq mm. As long as the shim area is greater than the bolt head area I think it should be OK, so at 5mm wide they need to be 16 or 29mm long, which ties in with the photo.

As Gerry Madden has said a thin shim will not affect the overall rigidity by very much at all. A 0.01" thick shim 1 sq in has the same spring rate as a piece of steel 100 sq " and 1" long. I'm still not happy that the shims are not under the bolts, doesn't feel right!

[Dr_GMJNParticipant @dr_gmjn]

Duncan, if we’re talking about preload in the joint itself, and if I understood what you wrote correctly, the preload is not confined to the area under the bolt head. It depends on the joint, but load falls to zero some distance from the bolt axis – can’t remember the figures but it’s something like 2x to 4x bolt radius.

[Author]

Posts