Fitting a bowed magnetic chuck

[jwmelvin]

I recently bought a B&S Micromaster 618 and have just finished cleaning it and setting it up with new oil, metering valves, and a couple VFDs. It came with a Walker permanent magnetic chuck that was very difficult to actuate and had a ton of slop in the actuating mechanism. I disassembled the chuck, cleaned it up, and made new parts for the actuating mechanism. It's back together with a little RTV gasket in the joint between the aluminum body and top plate.

The machine table looks gruesome, with pitting (a few spots 0.005-0.009" deep) and staining, but measures very flat other than the pitting (by touching the unpitted spots or with parallels on the table).

Sitting on the table, the chuck is banana shaped—high on the ends 0.008" and pivoting about its center. I mapped the surface without any tension on the clamps and it is a fairly symmetric and single-axis curvature:

If I snug the clamps with 30 ft-lb, the chuck flattens out and measures pretty flat (~0.0005-0.001") from the far left side to about 2" from the right side, and from there rises about 0.002" to the far right side.

From what I understand, one might turn the chuck upside down, clean up its bottom, then flip it over to grind the top. If I do that, I'd be removing a bunch of material from the center of the bottom and then from the ends of the top. That seems undesirable, considering the amount of material I'd have to remove. On the other hand, using the table to flatten the chuck seems undesirable, considering the distortion it must induce in the table.

I have not put the chuck on the surface plate for any checks, but since the grinder table is flat when unloaded, and the chuck is flat when clamped, it seems like the chuck is mostly a uniform thickness, just curved. If the grinder table distorts from the clamp load, that would suggest the chuck is thinner at the edges. I was concerned that I did something weird when reassembling the chuck, like nonuniform RTV thickness, but I was pretty careful to put a thin uniform layer, and a little squeezed out all around.

I appreciate any advice you can give me.

 

[michiganbuck]

Good to flat hone the bottom at least so no voids, hollows. nicks under it. You don't want to over-tighten a chuck because doing that can distort the table's scraped in underside.
I like to tighten one hand with my at elbow swing arm, not body weight and shoulder. I think the old walker chuck manual sais 12 lbs? and the push down should not be at the edge, but more toward the chuck body. The clamp tilts down or a .020 shim to set on.
The set pad should be flat honed with a big hone to be sure it has no bugs.

Some may say grind the set pad, but I don't do that unless I qualify /inspect the flatness of the ways first.

Likely someone burned the chuck with a top grinding and the stresses are still in the top.
If it bows out flat with a moderate pull that would be good enough for me for.008, much more and I would grind the chuck if the machine proved flat.
Good to have, buy a good stone and keep in its box for chuck(and the like ) only. Use the fine side for the chuck. dry fir a light skim / with oil for a figure eighs truing.
I see supposed top grinder hands hitting a chuck with a 1x 3 " stone, that is a No No in my book. * You should have your location in your stats.

https://www.amazon.com/Norton-Combination-Abrasive-Benchstone-Aluminum/dp/B0006NDPLW/ref=asc_df_B0006NDPLW?tag=bingshoppinga-20&linkCode=df0&hvadid=80814156492390&hvnetw=o&hvqmt=e&hvbmt=be&hvdev=c&hvlocint=&hvlocphy=&hvtargid=pla-4584413735926926&psc=1

 

[michiganbuck]

Grinding a burn-in a chuck top can permanently distort the chuck. If you are lucky it will be concave at the top side. it takes years, perhaps never for the stresses to come out.

 

[jwmelvin]

Good to flat hone the bottom at least so no voids, hollows. nicks under it. You don't want to over-tighten a chuck because doing that can distort the scraped in table.
...
Likely someone burned the chuck with a top grinding and the stresses are still in the top.
If it bows out flat with a moderate pull that would be good enough for me.

I will go back and stone the set pad and chuck.

It flattens with 30 ft-lbs on the clamp bolts, though I haven't tried to determine the lower limit for what will flatten it. That doesn't seem like excessive torque at all.

If someone burned it, shouldn't the chuck be concave (nonuniform thickness)? That doesn't seem consistent with it flattening out the way it does. It seems to be bowed—concave on the top—but of fairly constant thickness. I guess I should be able to measure any thickness variation with a micrometer. When I take it off to stone the set pad and chuck bottom, I'll try to investigate a bit further.

Here's my concern - to grind the chuck top, I'm supposed to tighten only one end clamp, to allow for thermal growth. But my chuck is not flat if I do that, and if I grind it flat in that configuration, it will not be flat when I snug it down. So I'm only able to grind the chuck surface when the clamps are tight. I could do that, and be extra careful not to put heat into it when grinding. Or I can try to flatten the bottom before addressing the top, so at least the shape will be stable and not depend on clamping force.

 

[michiganbuck]

To grind out a bow you plate (or set pad) check the bow and then put the hollow down with a shim under it ( middle) so it is the same as relaxed. You cool grind the top (hump in the middle) to flat.

Then you put the now flat side down and grind the (was concave)side that will hit first at the ends.

likely grind .010 off each side so not much of a grind. then mounted likely need top grind .002 to make it flat.

It is a little tricky because you need to figure out how flat is your set pad.
If the set pad is .002 high at the front you don't want to end up with double that when you mount the ground chuck..

The stresses will come a little out of the top so what you get will not be exactly what you expect.

Good first to inspect the set pad travel flatness, if you have a straight edge shimmed to zero/zero at each end and travel the top with a tenths indicator.

 

[jwmelvin]

To grind out a bow ...

likely grind .010 off each side so not much of a grind. then mounted likely need grind .002 to make it flat.

...If the set pad is .002 high at the front you don't want to end up with double that when you mount the ground chuck..

I believe the set pad is flat (and parallel to the machine ways), but I'll check it again before proceeding.

The procedure makes sense and is what I understood, thanks. My concern was that the .010" seemed like a lot to remove so I wondered about milling first. Also, since the base is aluminum, should I use a SiC wheel, or just stick to the regular wheel (currently 32A46-I)?

When I grind the chuck (bottom or top), does it seem reasonable to take 0.005" off in a pass with 0.010" stepovers? I will be using spray coolant, and that approach seemed to work okay on some initial tests I did.

 

[michiganbuck]

(If someone burned it, shouldn't the chuck be concave (nonuniform thickness)? That doesn't seem consistent with it flattening out the way it does. It seems to be bowed—concave on the top)

The hollow is not ground into a heat swelled part(the chuck), but the result of the rapid cooling shock at the burned surface in contraction (pulling together)..
I have that someplace in a government-sponsored test of heat burn warpage in grinding surface finish warping. It was something like 20 thousand tons of force applied in contraction forces, enough force to permanently bend the structure, so just removing the surface stress does not allow the part to return to the original condition.

That weight /stress is about the same as 7 Sherman tanks.
Why you never want to burn a chuck.

.

 

[michiganbuck]

I am sure Richard King would say fix the long travel to flat and square, grind the set pad, then grind both sides of the chuck, then tickle the chuck.
That is the right/best way to do it. and the machine is well worth doing it right.

Likely one could not buy a better machine for any price

 

[michiganbuck]

QT Op: (Also, since the base is aluminum,)
Aluminum is a bugger to grind. If I could mill it to about .001 /.002 or so I would mill the bottom.

Top, does it seem reasonable to take 0.005" off in a pass with 0.010" stepovers?

That seems a lot .005. Grinding a chuck you grind wet (hand spray bottle ok), have an overtravel at both ends of long travel to give a little more cooling time. don't high speed long travel, stop and feel the chuck for the least warmth, listen to the grind for any spindle slow down, dress often with a tab of masking tape to mark your down number, *as you get closer to full grind you expect more heat so become more careful.

A little washing soda in water makes an Ok coolant for chuck grinding.(not baking soda)

 

[Conrad Hoffman]

When I first read this thread I thought we were working in mm and the last digit was microns. Never seen anything so bad. Grinding 0.005" in a pass! That's nuts; it should be more like 0.0005" per pass. 30 ft-lbs! Just as nuts. Here are two docs from Walker to read:

http://www.walkermagnet.com/content/uploads/2019/05/grindinginstructions.pdf

and

 

[jwmelvin]

Never seen anything so bad.

I was certainly surprised and concerned to measure what I did. It seems like you agree with @michiganbuck (and the Walker instructions) to just flatten the bottom then flatten the top, notwithstanding the degree of distortion?

Grinding 0.005" in a pass! That's nuts; it should be more like 0.0005" per pass.

Isn't the stepover important in this question? I believe you are talking about large stepover, like 0.125" per pass, in which case the material removal rate would be slightly higher than 0.005" downfeed with 0.010" stepover. Perhaps the small-stepover approach is not appropriate for grinding a chuck (I do see that the Walker document recommends large stepover)? I'd really like to understand the tradeoff.

30 ft-lbs! Just as nuts.

It's helpful to see Walker's recommendation of 10-15 fl-lb, thank you.

 

[michiganbuck]

Here is an old PM thread about chuck griding.
I think the old walker chuck grinding sheet suggests a courser wheel than a 46.
A very open-wheel is good...having spaces between Grits.
I think Cash ( a top grinder hand) said he grids a chuck with a 32 grit wheel.
I have a corburundrum 46 I white wheel that is good for chucks but I have used 46 wheels that would burn up a chuck pretty quick.

Chuck grinding recommendations

Hi guys, I’ve got a large Jacobson surface grinder that I need to chuck grind. Before I go in and pull the wheel in to the chuck then spend half a day flattening it out... I know people love to give their opinion so let me have them! I’ve only ever done smaller grinders with 1/2 inch wheels...

www.practicalmachinist.com

 

[jwmelvin]

I think the old walker chuck grinding sheet suggests a courser wheel than a 46.

I haven’t seen wheels coarser than 46 available for sale.

Thanks for the thread cross reference. I have read a variety of discussions. I’m still not clear why small downfeed, large step over is better than the opposite.

Also, I pulled the chuck off the set pad and measured its thickness. It appears to be tapered about 0.002” along its length. That doesn’t explain the 0.008” bow that I see on the top surface. I’ll measure some on the surface plate and maybe mill the bottom a touch before returning it to the surface grinder.

 

[michiganbuck]

Re : I explained the burn-caused heat shock cooling warp (the.008) that can bend a chuck, perhaps you missed it.

I have almost 1600 words in my book in the grinding a chuck-top section.
including this rough outline in that chapter.

* Again, wet grind even if you use a hand spray bottle.
* Good to buy a large plate of steel ¾ x 6” x 12” and make this your practice grind. You can use it for a checking plate so practice with a purpose. *You might redress after .005 is taken or when the wheel seems to become loaded.

1. Set the chuck on the set pad in a position that will allow grinding the back rail and centered to the set pad. It is common to set the cuck on heavy oil or other element so to avoid rusting. Best to place the chuck where you can get the wheel off the chuck at the far side and off on the close side to be clear off the chuck. tighten the hold-down clamp nut to 12-15 lbs on the right side and 10 lbs on the left side. it is common to indicate a chuck side to make it square to the machine. Strike pencil marks Crisscross on the chuck.
2. Dress a name-brand open-wheel of perhaps 46 H to L, or 36-32 h to l. Tighten the wheel as tight as you can with one hand on the wrench and one on the wheel / Then set the wrench end on a block and make it a little tighter with two hands on the wheel. (see dressing a wheel).
3. Balancing may not be needed for the 7” wheel but for 8” and up I recommend balancing, balancing is a plus for the 7”.
4. Bring the dressed and not running wheel to hand feel the chuck at close to you at the far right and note that down feed number. Then bring the wheel down at the far away chuck center to rub and note that number. Rub very easily or you will need to re-dress the wheel.
5. Raise up and come back down to .010 above the highest number and travel the parked wheel across the chuck to see that it misses touching the entire chuck, then try .005 to sill miss the chuck.
6. fire up the spindle at let run for 5 minutes if a bearing spindle /20 minuet if a plane bearing spindle with first spark at higest chuck area/place..
7. Crank or auto-feed long travel for 5 minutes to warm up the bed.
8. With long traveling across right and left down feed .0005 till you get the first spark at the close to your standing position.
9. long travel and incremental cross ¼ to 1/3 the wheel width per pass with spraying coolant on the chuck, with over traveling off the chuck at right and left for cooling time.
10. Cross and long travel all the way to clear the chuck at the far side with cross feeds off the chuck.
11. Bring wheel back to front and with .0002 down feed cross chuck again.
12. Continue until all pencil cross lines are gone. Increasing the down fees onlyif you are confident you will not heat the chuck,
13. Grind the last .004 with the magnet turned on
14.. Pull down .00005 and take a spark out, or two.

OPS!... looks like I missed over travel off the chuck at right and left for extra cooling time..darn.
OPS! .. I did not include how tight..darn
*Yes, If any grinder hands sees something I errored or left out, do let me know
Buck

 

[Bakafish]

Given that it seems close to flat when clamped into place, and you don't want those forces to transfer into your bed, wouldn't it make sense to just make an interstitial plate out of steel? Clamp the chuck to the steel plate to pull it as close to flat as you can. Grind the base of that plate flat with the chuck attached. Then mount that plate (stress free) to your bed. You lose a little hight, but likely strengthen everything, and it is all reversible if it doesn't work.

 

[jwmelvin]

wouldn't it make sense to just make an interstitial plate out of steel?

That’s an interesting idea, though I’m not crazy about putting all that stress into the chuck. And I’d have to grind the plate flat once clamped to the chuck, so it doesn’t avoid that aspect.

It seems like the usual chuck-fitting routine will be able to address my problem, so that’s my plan. I’m just trying to make sure I understand what I’m doing, and the advice so far has really helped.

 

[Bakafish]

That’s an interesting idea, though I’m not crazy about putting all that stress into the chuck. And I’d have to grind the plate flat once clamped to the chuck, so it doesn’t avoid that aspect.

The chuck has survived so far with that stress, and machining it flat may not result in any stress relief you are hoping for if it is the top surface that is actually inducing that curvature from its natural state. You also only have to grind the base plate flat if it ends up with its own curvature, which given the base is Aluminum and you haven't chosen a plate thickness, is not something that has been established. It may be strong enough to correct the bow and not need any work at all (in which case, one could argue the table would serve just as effectively.) But if it does need to be flattened, steel is much easier to grind (and scrape) than Aluminum, so there's that. And again, it is reversible if it turns out the root cause of that curvature ends up being something you missed and correctable in the servicing you did.

But I wish you luck, wish I had a surface grinder myself. I just have a nice chuck...

 

[tdmidget]

" I disassembled the chuck, cleaned it up, and made new parts for the actuating mechanism."

How much disassembly? You may be done now if you removed the magnets.

 

[jwmelvin]

How much disassembly? You may be done now if you removed the magnets.

I have read about the effect you refer to, though my magnets seem very strong still. The chuck was a mess and needed cleaning.

 

[jwmelvin]

machining it flat may not result in any stress relief you are hoping for if it is the top surface that is actually inducing that curvature from its natural state.

Your point @Bakafish is well taken, that the chuck has internal stresses now, and will no matter what my approach (whether between the magnet plates and aluminum base, or between the whole assembly and external structure).

It seems likely to me that @michiganbuck is correct and the top surface is exactly what causes the curvature, from a former overheating event. So removing some of the top surface may (will?) change the induced curvature, but that will help me get to a straighter chuck.

That change in stress that may occur when removing some of the top surface gives me pause. It seems like it might be a good approach to incrementally address the top and bottom surfaces, to avoid chasing my tail. @michiganbuck also mentioned milling the base for bulk removal; I wonder about milling the top side until within about 0.001-.002”.

I measured the chuck on the surface plate last night and confirmed the shape I’m seeing.

Edit: I thought of one more thing. A straightening plate will necessarily be somewhat hefty. This chuck is already not light, so I’d like to avoid making it heavier. Not that I’ll be moving it much, but still…