What's new
What's new

survey of surface grinder

dian

Titanium
Joined
Feb 22, 2010
Location
ch
this is a "haarmann" surface grinder. the company is obsolete. it has hydraulic traverse and electro-mecanical cross feed. it has a flat way in front and a v-way in the back. it was capable of grinding "parallel shaped" parts to under 0.0004" over 8", but i was getting a taper of the same magnitude over a shorter distance cross wise. the chuck is 12"x6". i decided i want more precision out of it. i have made a lot of measurements and would like to ask for advise on how to proceed.

1. the table top is 14"x5.5". it is scraped. it is essentially flat, being high in the middle by maybe 0.0002" traverse wise. this was measured with a 400 mm beveled straight edge (din 00) using the light gap method and also a 0.0002" shim (pull test). should i scrape that out?

2. the traverse ways are worn convex. this makes sence. as i dont have a parallel of the lenght and precision needed i measured it again with the straight edge aligned to within 0.0004", interrogating it with a large radius indicator tip. the result is in fig.1. i have a flat area of about 9" traverse wise and then the straight edge rises by about 0.0008" on the ends. however in the area marked by the arrows the table kicks up by over 0.002". (the front and rear measurements are not really consistent, because i have to move the table.)

3. the cross ways are worn convex also. this doesnt make sence, since there is little overhang unless a lot of large parts were ground on the grinder. as can be seen in fig. 2 the area about 1" from the edges is very flat, then it rises rapidly by 0.0004". this was measured with a 200 mm parallel adjusting for its inaccuracy.

currently im not planning to attack the ways, i can make do with the 9"x10" area of flatness. however the table kicking up that much in one spot is a mystery. first a visual and tactile inspection of the portruding ways yields nothing remakable. there are some scores but they are smooth. but more important, the ways are fully retracted at that point so what can make them jump? some mechanical interferrence? even if the hydraulic rod was bent it cant have a hump in it like that.

btw, i dont have a helper at tis time, so im not taking the table off. and i have an additional needle valve to reduce oil flow, so the ways are oily but nothing is running out now.

Edit: there is another funny thing. when the table moves to the left zero is at point B. when to the right its at A. (fig. 1). when i stop the table between these points it drops by about 0.0002" in about 2 second. it stays stable if i stop at any other location. an additional twist is that measuring directly on the table the "kick up area" is narrower but is clearly about 1" to the right. no idea what to think of that. am i chasing some kind of oil bubble? first i thought of moving the chuck to the left to avoid it, but with its location as the straight edge shows that not possible any more.
 

Attachments

  • DSC00556.jpg
    DSC00556.jpg
    83.3 KB · Views: 272
  • 0 182.jpg
    0 182.jpg
    89.8 KB · Views: 230
  • 0 032.jpg
    0 032.jpg
    91.4 KB · Views: 210
  • 0 030.jpg
    0 030.jpg
    89.1 KB · Views: 229
Last edited:
You can rarely make logic of a grinders wear with not knowing its history. Perhaps a right-handed guy was blowing an air hose this way or that, and completely different with a left-handed guy. That is a nice-looking machine and will worth planning to re scrape it. Photos look like it is not very clean and it need be spotless clean before making any plans.
Sometimes one can make hard wool spacers and then lift one table side at a time, then with it raised slide it off onto a high bench.
This book may be worthy of purchase (?).
https://www.amazon.com/Learning-Scr...s&sr=1-1-dd5817a1-1ba7-46c2-8996-f96e7b0f409c
 
I see it is sitting on wood blocks? Many grinders sit on 3 points...I would first check that. You said your using a beveled straight edge? You mean a prism with out a camel back? You need to use a camel-back as the prism straight edges bend into a concave or over a convex. I suspect this is happening and something or all the ways are high in the middle. The top of the table is scraped? That could also be an issue. Blue up the table top and hinge a camel back on it. With a camel back check all the ways. When you have a helper flip the table over and set it on 3 points at 30% and check the hinge of the V's and flat. it should hinge be at 30% from both ends.

Lay a parallel on the table top and first hinge it and see if it is flat. If not slide in feeler gages under it at 30% from the ends and make your tests. Do that in cross feed too. Your using a proved gage (parallel that is .0002" or better) to measure instead measuring an unproven way or table top. Picture one, is that the V way? It's not scraped? A tip: when posting pictures, just above the pic's type "Left to Right" ( L to R) and write there what were looking at instead of going back to the body of the post to figure it out.

If you have a short precision level, set it on the table centered and shim under it to make it read level, then slowly feed it and watch the bubble. Another thing is check the bottom of the mag, chuck. Blue up your surface plate and blue up the chuck. If that is good I would "grind" the table top a table and not scrape it. Then set the chuck on and only tighten the left hand tee bolt and leave the right side snug. We can figure it out and when your done .0001" should read all over.

If you need a 12" camel back I can sell you a casting. I have a sale on them now.
 
Looks like a really nice machine. Don't forget stupid stuff like rubbing way shields and anything mechanical that can hit. Not an issue for yours, but my small grinder drives the table with a gear. At one spot a taper pin in the gear was lightly hitting the rack, causing a bump. I think I had a couple other problems, none of which were actually the ways. Make the room as quiet as possible and listen to the action.
 
the machine is as clean as can be without stripping the paint. you can see the beveled straightedge in the picture. it is very precise, better than 400 nm (violet light) and it doesnt move, as it sits on the table. i performed other measurements and did some bluing up, but i would like to find out first if anybody has any idea where the hiccup could come from. so again: the table jumps up when the table ways have disappeared 2" into the saddle (or whatever its called on a grinder). can anybody think of a situation that might cause this? im almost expecting something stupid to show itself, as conrad said.

btw, i have removed the table before and there was nothing unusual, but maybe i was not paying enough attention.

edit: anything flat hinges in the middle of the table top and the biax will get rid of the 0.0002"hump in a couple minutes.
 
I didn't see the parallel, sorry. Mount a indicator under the bottom of table on right side and indicate the flat as it feeds to the right. The indicator shows plus on left side of table top, but that could mean the table is dropping on right and it is lifting the table. I am trouble shooting this and have seen all sorts of issues on surface grinders over the years. I have had a Myford table drag on clearance surfaces because as I scraped it down if fell. I had to use some Dykem layout fluid painted on the surfaces to find the spot. As I said if you ground the table top like every new machine builder does you would probably cure the problem. I have never seen a grinder table top scraped. You ask for advice and Im giving it as I have been rebuilding surface grinders for 50+ years and have taught classes in several new machine builders. I hold .0001" easy on small grinder like that.
 
yes, the swarf is really solid after a lot of cleaning. the ways are still a bit dark from the remnants of some molly additive some po thought he should put into the system. it comes out only slowly. and the picture of the grinder on wooden blocks is when i got it a few years ago.

richard, i just did what you said and a 0.01 mm indicator is rock solid everywhere incl. the spot where table jumps. i had to use a long reach test indicator on the left side because of the oil tub. the table top is scraped as well as the bottom of the chuck. i thought i would scrape the table top rather than grinding it because of the convex ways. but on second thought i might grind it, as i only have to take off in the middle, where the ways are flat.

edit: the numbers on the chuck are meaningless. look at the drawing.
 

Attachments

  • 0 033.jpg
    0 033.jpg
    84.3 KB · Views: 111
We already went over the grinding vs. scraping of the chuck surface in another thread. It is a mistake to scrape the top of the chuck. The top of the actual machine table and the bottom of the mag chuck might be ok, but not necessary I think.
 
yes, i remember, you said:

"Both surfaces when ground on the first surface will end up being the same shape as the ways, doesn't matter if the part pulls down or not - machine travel is machine travel. If you put a flat-scraped mag chuck on the machine rather than one ground in place on the machine, whatever geometric error the ways have will show up as a parallelism error in the part. Even with the bottom held perfectly flat, the machine and thus the work would still travel in an arc or whatever shape it has while grinding the top. When the chuck is ground in place on the machine, it has that error already ground into the top surface, so when you put your part on the chuck it will pull down into the same shape that is being ground on top, producing a parallel workpiece.

Trouble arises with very thick workpieces that can't be pulled down tight against a non-flat magnet. Since wear on most grinders is usually pretty minor, this usually only happens on a massively worn out machine though. (Alternatively also when grinding with very low magnet force). In that case you can get double the geometric error present in the ways, as you grind an arc (or whatever shape) in the first surface then flip it and grind another in the second surface."

(surface ground part low in the middle) (post #29)

but that was not really conclusive (getting double error), right? but if i were to grind some curve into the table top, how do i put the chuck on, grind the same curve into the bottom? thats why i figure these two surfaces need to be flat.
 
i have found out what is causing the table to to kick up in that spot (or at least par of it). its two things. first the side pressure of the right rod is perfectly able to push the table up when retracted, especially without the chuck now. no idea what i can do about it, because the left rod is perfectly centered, probably enlarge the right slot. second (yes richard, you were right) the underside of the table is hitting the hydraulic cylinder* in that spot. so the table has to come off.

Edit:* actually its hitting the bracket.
 

Attachments

  • 0 034.jpg
    0 034.jpg
    92.6 KB · Views: 110
  • 0 035.jpg
    0 035.jpg
    89 KB · Views: 106
Last edited:
Im glad you found the problem and you can figure it out now. You are correct the feed rod should not rub on the end bracket and just float. On many machines the end of the rod is threaded and you have a nylon lock-nut with washer on each side of the bracket that are snug and can float and move up and down. You will probably have to use a die grinder and open up the bracket hole. Your correct about the curvature of the table travel and grinding the chuck and parts. I use the example of a how a cylindrical grinder grinds with single point contact. But if you scraped the machine straight ground the table top and chuck top then all should be OK and that won't happen.

I found this booklet on another brand of grinder and on page 19 they tell how to grind the magnet chuck. Different machine but the principal is the same. http://vintagemachinery.org/pubs/257/18421.pdf

One last check after you grind the chuck with mag off set a hand scraped straight edge or your sweet parallel on it and hinge it and it should hinge 30% from both ends of the chuck. That will show if the curvature issue is gone.

It is nice seeing these issues resolved. That is whats so cool about this forum. :-)
 
thank you.

the table is off, its probably over 100 lbs, so the wife struggled quite a bit. the ways are completely worn, i thought they had scraping on them but i obviously mixed it up with the od grinder. they are 21" and 30" long. a quick approximate check yields:

sadle flat 0.0008" convex
saddle v 0.0008" convex
table flat 0.0020" concave
table v 0.0012" concave

no, i wont be scraping them. i have no means to move the table around and i only have a 24" camelback anyway.
 

Attachments

  • 0 036.jpg
    0 036.jpg
    90.1 KB · Views: 105
  • 0 041.jpg
    0 041.jpg
    94.6 KB · Views: 96
  • 0 038.jpg
    0 038.jpg
    93.5 KB · Views: 98
  • 0 039.jpg
    0 039.jpg
    93.8 KB · Views: 97
  • 0 040.jpg
    0 040.jpg
    92.5 KB · Views: 92
That's bad news....so sorry. I hate to say it, but the saddle ways are probably worse. Many times they scrape the base ways convex if the table travels over the end of the saddle to compensate for unsupported sag.

Be sure you set the table on 3 points when you scrape the ways. 2 points under the ways and one centered where the feed rod centers in the table at 30%.
 
the idea occured to me to have the table (and the top) ground and then use it to scrape the saddle. well, the quote is $2000 if i bring it in.
 
the idea occured to me to have the table (and the top) ground and then use it to scrape the saddle. well, the quote is $2000 if i bring it in.

May need long travel and cross to make it right.

Care to make it ground closest to existing can save a lot of scraping time...

Otherwise you would be better just scraping the whole thing.

You dont grind the same off the flat and the Vs but calculate that to drop down, the take will be diferent from each to keep top level.
 
Or you can do what Ballen did on his Studer Cyl/ grinder (see above in sticky above). Cut new oil pockets in it by giving it 2 square cut scrapes in both directions, lower 1/3 of the middle of the table so it hits on the outside 1/3's . I thought you scraped? Depending on how accurate you need to grind parts you maybe fine if you cut the ways, lower the area that was hitting the paint, grind the table top and chuck and see what happens. All your labor and test it before spending the $2000.00 plus transportation costs.
 
Or you can do what Ballen did on his Studer Cyl/ grinder (see above in sticky above).

In my case, in addition to the square cut scraping of new oil pockets, I needed to fix the swivel table (it had been incorrectly ground at one end, was off by ~150 microns!) and fix the height and alignment of the workhead, tailstock and grinding head. But once I had done all of those things, the machine geometry was still good, meaning that the machine grinds straight and round to the precision that I can measure (one or two microns). I think this might be because the damage to the ways was not severe (the original scraping marks were still visible almost everywhere) and because a cylindrical grinder is quite insensitive to vertical errors and only sensitive to "straightness" errors in the long V-way. In contrast, a surface grinder is first-order sensitive to the vertical errors, so less forgiving than a cylindrical grinder, which is only sensitive to these vertical errors at second order.
 
Last edited:
so i put the table back on, it runs smoothly now by hand.

4. the table top and the underside of the chuck are scraped and now convex by about 5µ. i plan on relieving the middle (1/3 in lenght and width) on both by scraping. any thoughts? maybe they should be convex, so they mate when the chuck is torqued down? probably not.
 
Flatten the ways to .00005" per 12" and then lower the short side middle .0005" to .001". You will have to measure the table travel so the table does not fall into the relief. I would not scrape the table top or match fit it the chuck. I would grind both with coolant or air / water mist. Oh before you scrape the table ways, scrape the table top or grind it to relieve the work hardened surface where the chuck sets. If you read the Do-All grinder manual it tells you how to tighten the chuck down. tighten the left side and leave the right side snug. All grinders work on the same basics. If your trying to hold tight tolerances you have to scrape the saddle and table on 3 points, not sitting cockeyed on a work bench. Set it on a work bench on 3 points.
 








 
Back
Top