What's new
What's new

shimming less than 0.0004"

DMF_TomB

Diamond
Joined
Dec 13, 2008
Location
Rochester, NY, USA
had a job yesterday i had to recut some rail surfaces .00035" off with a twist. the casting
near as i can tell was twisted when the opposite side was machinned similar to what cold rolled steel does when machinned.
.
part is about 40" long (200lbs) and supported on 3 points. since i usually use .001" and .0015"
shim stock or feeler gage material i had to think about how do i shim at support points
0f .00035, .0003, and .00025" ??
.
i have plastic onion skin of about .0004" but it is flimsy and hard to put into a small gap.
so i figure i would try 3 each .001" shims at each of the 3 supports and a screw jack at 4th
corner. i would vary hold down bolt tension and use screw jack to squeeze the .00025 to
.00035" difference. obviously the more shims used the more they can vary when tightened
due to small amounts of oil and dirt still on shim material even if only .0001" on each shim.
.
it worked in that i squeezed over a .0002" by varying bolt torque and using a screw jack.
i was just wondering how others shim parts with less than .0004" difference between spots ?
obviously if i had a larger amount of shims they could squeeze too much when they are
tightened and be unstable. as it was the shims i used, it did vary .0002" after 10 minutes
and i had to measure and readjust just before machinning part.
.
i have heard of the make a crayon mark and put part on and measure and the crayon mark
having thickness does act as a shim but i did not want to keep lifting a 200 lbs casting
up to keep making multiple crayon marks and measuring so many times. i use a .0001"
test indicator and magnetic base stuck to the spindle of a big cnc gantry mill for
measurements. as it is the cnc machine servos moved back and forth .0003" and take
some time to stop moving after a Z axis adjustment to get a stable reading.
.
i am just asking, does anybody have a better way of shimming heavy castings of less than
a .0004" difference other that use multiple shims and varying tightening tension ???
 
would it be ridiculous to grind blocks to size? grind 3 pads the same size, bolt the piece down and measure the error, then grind the one you need to the few tenths.
 
would it be ridiculous to grind blocks to size? grind 3 pads the same size, bolt the piece down and measure the error, then grind the one you need to the few tenths.

.
there are side bolts and if shimmed even 0.1" they they would not go in the holes. basically 3 bolts pulling
down and 3 bolts on side pulling to stops horizontally
.
i suppose i could have .005 shims ground to .0047 to get a .0003" difference but no surface
grinder easily close by. i would have to ask to get premade but i am guessing they will say it is
not needed and do it the same way it has always been done by varying bolt tension.
......... but i will ask if i can get some .005" shims ground to .0048 and .0046, etc sizes. i will
have to mark size on them and save them or they will get lost with the little pieces of shims
we have in a pile.
........... if the main support blocks were machined .0003 in the machine they would have to be replaced
with 3 each the same size when done. plus lifting casting clear off the fixture to get at support pads
is a pain. it is easier to stick shims in a gap from jacking up casting .010" to install shim.
.
maybe just having .005" shim stock ground to .0048, .0047, etc sizes would be easiest method if boss says
it is ok. it can be hard at some places changing ways to do stuff if, they want to do it the way they always
do it........ there are places that make custom sizes of shim material. but my boss is the kind to complain
when i get a new set of allen wrenches (hex keys) from the tool crib. you might be surprised at how
cheap some places are.
 
thinking about it, maybe the easiest thing is to take .005 shiim stock thats only 1/2" wide and
making a magic marker mark on it and sanding it off with 100 grit sand paper and measure shim
.
often each time you do this if it removes .0001"........... i know sounds kind of crude but if it only
takes 10 to 20 seconds of time to make .0050" shim measure .0048" maybe thats the easiest
way to do it. if it costs $100 to make custom surface ground shims that are only used for 10 minutes
or me taking less than 5 minutes to hand sand and measure maybe thats what my boss would want.
....... i can always just ask what they think is the best way. either way i check with a .0001 indicator
after shimming part to confirm everything is good. (dirt has a way of knocking things off .0005" easily)
.
it is just a new learning experience for me trying to shim parts to less than 0.0005" tolerances
 
Need a really thin shim? Dykem. It can be built up, too. Just don't get too crazy with clamping pressure.
.
thats why i mentioned a grinder trick of making a mark with a crayon marker that leaves a wax line. but
thickness is difficult to control. plus lifting a 200 lbs part to "adjust" marks would be difficult. plus it
is not the most stable setup as the crayon wax can crush flatter from pressure over time.
.
i am kind of thinking sanding a .005" feeler gage shim is easiest. like how long can it take to sand .0002" off ??
 
maybe im missing something, but i have 0.010 mm and 0.005 mm shim stock (stainless spring steel) in my drawer.
 
One time a hobbyist jeweler showed me a cute little hand cranked 3-roller mill that he used to infuse patterns into various metals. For example, he'd place a tree leaf in between two strips of brass sheet, and roll it through the mill and it would emboss the vein structure and texture of the leaf into the brass, while slightly reducing the thickness of the strips.

Seeing as how that was so quick and simple, maybe you could buy or build a simple little device like that to custom roll your own shim stock.
 
As I recall, the cellophane wrapper on cigarette packages is .0002 thick. I bought a roll of the stock some thirty years ago for work and used it extensively for insulating two surfaces that are clamped together but must have good heat transfer.
 
I have used silver leaf and gold leaf to shim to tenths on motion picture film camera mirrors. That's standard procedure as the work was always inside those tolerances.
 
You could beat on one end of a piece of shim stock with a hammer until you have made it into a tapered wedge.

I have made crude wedges for welding fit up by flattening the end of an aluminum welding rod with a hammer.
 
rail surface

Sounds like a scraping job, and with that amount to remove ... A pretty quick job...
.
these are linear rail surface 2 each about 1" wide and 40" long and up tight within .003" to locating edges or surface at 90 degree.
.
shimming and remachinning to .0003" flatness takes about a hour. they got rid of most of the hand scrapers as they always took far, far too long to do. it is done with cnc milling machines for the past 2 decades now.
.
this is a classic case where when one side was final machined (opposite side warped) the part curled or banana shaped when unbolted .0008" over 40" bow with a .00035" twist across 12" width
 
Even tightening the workholding nuts a tiny bit will move that much. Or for such a large part its warmer on one end than the other.

Maybe you can put a thin malleable material like a thin copper sheet on the bottom of the fixture points. So you can clamp down on those areas and it wll push those points lower as the material deforms and the part digs into it a bit more.
 
Even tightening the workholding nuts a tiny bit will move that much. Or for such a large part its warmer on one end than the other.

Maybe you can put a thin malleable material like a thin copper sheet on the bottom of the fixture points. So you can clamp down on those areas and it wll push those points lower as the material deforms and the part digs into it a bit more.
.
what i tried and works is put 3 each .001" shims on under each support. the more shims the more squeeze
change as bolt torque is varied. but putting say 10 each .001" shims under each of 3 supports it squeeze changes too much and can be unstable over time. casting temperature is stable and when machinning only .001" is removed. you cannot detect any heat change although i am sure it might change 1 or 2 degree F
.
basically a casting that was final machinned on one side the part warped or curled and twisted the other side. this is similar to when cold rolled steel is machinned on one side and vise opened part is often not flat any more. 40" long part curled .0008" and twisted .00035"
.
nothing like explaining this to a boss or engineer who has no machinning experience. they look at you like you are crazy. same people have a hard time understanding metal changes length when its temperature changes. or understanding a part can sag from its own length like a 20 foot piece of 1/2" dia steel rod
 
.
what i tried and works is put 3 each .001" shims on under each support. the more shims the more squeeze
change as bolt torque is varied. but putting say 10 each .001" shims under each of 3 supports it squeeze changes too much and can be unstable over time. casting temperature is stable and when machinning only .001" is removed. you cannot detect any heat change although i am sure it might change 1 or 2 degree F
.
basically a casting that was final machinned on one side the part warped or curled and twisted the other side. this is similar to when cold rolled steel is machinned on one side and vise opened part is often not flat any more. 40" long part curled .0008" and twisted .00035"
.
nothing like explaining this to a boss or engineer who has no machinning experience. they look at you like you are crazy. same people have a hard time understanding metal changes length when its temperature changes. or understanding a part can sag from its own length like a 20 foot piece of 1/2" dia steel rod

Yea thats a good way. Shim it up with compressible malleable material as support and crank away as needed to push the part down more as it deforms the shim. Only problem is rigidity if machining forces were high it can also knock the part out of alignment by deforming the shim material some more.

Yeah, for a casting, have to machine out the skin first and the part can change dimensions because of residual stress. Not a problem for lathing a casting though as its symmetric on all sides normally.

There are always a few of those who make it up the higher echelon with poor understanding of the underlying but is otherwise just a supervisor/management type with a lot of words but not as much game. A good engineer or boss should understand that though. And if they do and still press on things, it means they're only really interested in completion timeline.
 








 
Back
Top