What's new
What's new

Surface grinding walls of 2” square tube

Ddowns46

Plastic
Joined
Mar 1, 2018
Hi All,

I’m working on an application that uses a 2” square tube for a linear guide rail. The tubes are cold rolled and are fairly accurate right from the steel supply house, but we want to pursue additional accuracy by surface grinding the tube walls, removing minimal stock to achieve full cleanup across each face. Each tube is 2” x 1/8” wall and 32” long.

Id guess we would need to take off about .010 to .015 to achieve full cleanup.

My questions below:
- is it feasible to surface grind tubes like this? Is 1/8” wall thick enough to avoid resonance/ringing?

- what flatness would be achievable on each face? The surface would be 2” x 32”. Is .005 achievable?
 
Are you planning to heat-treat before, or restraighten after, grinding? Removing the surface layers of CR stock is likely to cause bending due to release of internal stresses.
 
Are you concerned with surface finish? You could double disk grind them but it would have a circular grind design and not the typical longitudinal grind. Double Disk (DD) grinds on both sides of the tube at the same time then grind the other sides in a 2nd operation and not 4 operations with regular surface grinding. DD helps eliminate the warping issues you see with one side grinding on a surface grinder. Another possible way would be to Blanchard (4 ops) grind them but again you would have the circular design or appearance of the grind. I know a company who could DD and Blanchard Grind them and we have a list of companies who surface grind we have a list in the Reconditioning Forum. I believe one is in CA called Shafer Grinding.
 
I think when you grind them you will releave the surface stress on the one side and you will have fits trying to keep parallel and flat Try some to see I agree with RJT 1/4 might work better but I think you may be up against a monster , invest in trying some samples and find out report back with good bad or otherwise
 
BOING ...OING....OING it went when we simply flycut a wee bit
off a side to "clean it up".

"Smooth surface" does not equate to linear accuracy.

I suggest that you bolt a hardened and ground "off the shelf" way
to the box tubing.

It will be shimmable, and alignable.
 
I would very much doubt that the mill grade material is anywhere near .005 over 32" or even 2" when you take it out of the box. So the moving around that will happen when ground, could still be better than where you start from.
First measure where you are, so you know where you are then grind, measure again, you will be better or worse. My money says it will be improved from where you are.
 
You would be far ahead to go with Thomson shafting and brackets for your rails. Trying to grind square cold rolled welded tubing square, parallel and straight is just not going to happen.

Any money you are saving on the material you going to spend several times over trying to grind them into a useable shape.
 
Like they said, grinding isn't likely to work. It will move every time you make a pass with the wheel.
 
Can you go with thicker walls?

A dozen/15 years ago i was rescraping my smaller. 6-18 grinder and needed some reference rizers/parallels. I made a set of light duty box parallels by grinding 3" OD square tube parallel and matched size on 2 sides only. The parallels are only 19" long. I ground 2 sides, then scraped the best side, then very lightly re-ground the final side.

The tube was whatever 3" outside is, the parallel sides are now 2.986 across. The wall thickness started at about .167".

Keep in mind, 2" is worlds less stiff than 3". The wall thickness is marginal. the increased length gives increased opportunity for mayhem. On top of that, a thin wall will cool and heat with local influences quite rapidly making both the grinding tricky, and the stability in use very subject to heat inputs.

I have found that to get a decent final finish (decent = good geometry, acceptable appearance) on constructions like a tube, grinding with the edge of the wheel is about the only way. Set the depth desired for clean up, then set the step-over per reversal at about .003". Coolant, of course. If you grind with the face (rim) of the wheel, the vibration will affect both the wheel and the work, and it risks need for constant dressing. If the edge of the wheel is used with minimal infeed, it can be set where vibration will be less of a factor, and the wheel will not need dressed for an extended period. I would not go too soft on the wheel - probably I or J. Running with the small step, you can probably use a 36 grit wheel (for cooler cutting) and still get a good finish. If not that, try 46. If you have to go finer, the job probably is not possible with walls that thin due to need to constantly dress and localised heating.

My experience says you could do the job. Most likely it will be time consuming. You have to ask yourself if the result will be stable enough for the app?

amt
 
0.02€.

Depends.
If the tube can be rigidly fixed and then ground, why not.

If the tube is double-side ground or maybe spot faced on supports, and the supports are flat or the whole rig is on the grinder, why not.

It is perfectly feasible to get thicker 50 mm 50x50x5 mm rectangles flat enough to run hiwin linear guides on it.
Local accuracy and flatness will be 0.02 mm easily .. and overall accuracy depends.

I made all sorts of flattened tubes, assembled assys, and welded assys with linear guides, to 0.01 mm local accuracy and max 0.03 mm local linear spot error over 1 m or about 40".
Over the past 15 years making motion-control stuff.

I would never do the same today.

Depending on endless stuff, but ..
1.
use tool steel flats on suitable near net shapes, and goops, if high temps and high max loads like a rock crusher are not a concern.
2.
use Very Heavy Flats. They are 10x cheaper than your time. Bill accordingly.
3.
make composites, or assys, if light complex strong rigid beams are needed.
Clean milled tool steel flats, solvent wash, goops, interference fits and dowel pins.

Surprisingly, 98% of stuff only needs industrial goops of not extreme cost - to be as good as welding or casting.
Often better in many ways.
4.
Outsource.
The water jet or laser, or sandblast/coaters/etc, or bulk milled steel material providers can probably provide you with net-shape material cheaper than you can buy basic stuff - perhaps delivered free, on a 2 day delay.

5.
It may be possible to get very accurate cnc dowel pin holes drilled / bored (marked) - very cheap.
Ask the provider how to do this cheaply.
Things like as-is, off-the-machine, best effort, not qualified, not guaranteed, etc...
This means the hole/feature is where the machine put it.
But the machine will put 2 holes extremely close to each other in repeatability into 2 pieces of same shape, size, material, run, temp.

Repeatability will be good while the holes might not be at an exact distance, or an exact height, in machinist terms.
Thus for beam assys the parts are great - and cheap - but not for general precision parts.

--
To the OP.
Your proposed 50 x 50 x 3 mm tubes, ground somewhat accurate, with linear guides over 30" sound like a bad idea to me.

Endless reasons and caveats.
But the engineering depends .. do You want accurate, fast, repeatable, strong, rigid, and in what order ?
Whats your goal, idea, or drawing ?
How much is size or mass a concern.
Or is cross-section say air/solar panels + wind loads the concern .. as an example.


An old Bridgeport mill table, cast iron, thick, bends a lot with 2 fingers pressure over 30".
 
Flat grinding is a bit of an artform
Your stock probably won't be grinder friendly in terms of straight and flat. If you havent done it i would be happy to help. Do you have a variable magnetic chuck? If not, this may be a real problem.

The only thing you have to worry about is if you have a grinder table long enough to fit the part. Use a 46 grit wheel or larger grit size. If you are worried about heat then grind the as diagonal as your grinder table will allow... heat really shouldn't be a problem since you will be wet grinding. Square them by holding the in a vice (or 2 depending on the length)

Heat treat if possible

This is totally possible. Not sure why people are having a hard time with it If youre looking for flatness within .005 I will say you could do .0005 without any real issue. If I read that correctly. If .005 is your target, dont even bother with grinding.

1/4 or greater is the best advices on this thread for many reasons
 
Incremental (stepping across) might be very time consuming. With having a direct drive (not belts driven)surface grinder, fair size motor (perhaps 5hp+) and a course wheel perhaps a 36H AO wheel(even a 32 might be the ticket) and having an overhead dresser. grinding to a spotter, wet, feed on the grind side small amounts of down feed, free pass on climb side, and long travel with 1" or 2" wide wheel you might down feed grind such a job… Agree you can’t burn even a little or the job will go bow.

Cold roll tube may have stresses needing a flip so one may have to evaluate that with a test grind..

What machine do you intend, what surface finish need and size allowance?

Qt: [ Id guess we would need to take off about .010 to .015 to achieve full cleanup.]
Is that .005 to .007 a side?
Qt [ Is .005 achievable?} Yes.Perhaps .002 /.001

36 or 48 x 10 or so Surface Grinder. Oil way machine best IMHO..6 on the load would save time.. (would not choose a balls way machine.)

That is what makes Blanchards work so well..they often use a courser wheel, a lot of horse power, very much solid weight to grinding forces and a dress with not lifting the wheel off the part...Agree a good grinder hand with a Blanchard would make easy work of that.

Likely if not having the grinder,experience or a grinder hand before spending big bucks I would send a number of pieces to Detroit Edge Tool, Areo or another of the local long work grinding shops..And Specify the wheel.. and the procedure ..

Yes you would send the test job to a local shop in your area.
 
Per my earlier post - I was grinding a job last week and when the down feed was about .020"- .030", the wheel ran fine with a .003 - .005 auto step over. 46I wheel, mist. The wheel formed a slight step, and then just stabilized and ran.

When i came back to do the finish pass with maybe .003 down (I forget exactly) & same step, the edge of the wheel kept loading - it was not getting enough force to break down. This was 17-4 stainless.

So i have to back up on my "small steps with small down feed". In many cases with a very small downfeed, better with a larger step over for the finish passes. It's been quite a few years since those parallels, as mentioned, but they still get used. I have not inspected recently, but better than .001" parallel and flat originally. Again, one final side was scraped before final dust-grind last side. They are very heat sensitive.

I would not use a large step-over unless only a tenth or 2 down. Any heat bows the work and blows everything. Smaller steps where all the heat is only at the corner of the wheel tend to be more reliable for me. But then you have to balance the grit, hardness, speed, and infeed. Sometimes hit it the first go, sometimes a bit more fussing or adaptation. Didn't someone above just mention this is a black art? :)

I do like 36 grit wheels a lot. Dressed to cut free and a small step they will still give a fine finish. Big steps, a little coarser.

smt
 








 
Back
Top