Machining A Vise Square

[dkwflight]

Hi
Since clamping pressure is the issue, think about how the pressure is applied.
I am thinking the vice needs to be clamped on strips to limit clamping pressures on the edges near the clamps,
You might consider milling out the centre section of the bottom so the clamping pressure is eliminated from the center section. Not much metal needs to be removed, just a few thou.

 

[ronzo]

Surface grinding the bottom of the vise and the floor of the vise flat and parallel is the best option. Other wise if it was mine this is what I would do


Sent from my iPad using Tapatalk

 

[mekanix48]

Had a similar problem with the vice I bought here in the UK; I could measure on the 'Y' axis & found a it was out .007" over 4" on a 2-3-4- block & out on the 'X' axis .005"... not good! it was mounted on a swivel base so the first thing I checked out was; were the top & bottom faces of the swivel base parallel to ea. other... not! Obviously the manufacturer had just done a quick set up before milling the faces without checking for any discrepancy & milled over the faces leaving 'tramlines' as the finish. I ran the base over some emery cloth, 400 grit, on a plate glass flat, & revealed that the base was high across approx 2/3 of the base. I proceeded to machine .005" off across the base to flatten off as much as possible, reassembled the vice & rechecked with 2-3-4 block in the jaws again... run out down to .003" on 'Y' axis, better but not good enough. Both the mating faces on the swivel were checked (blued ) & revealed that the vice base was high in the centre compared to the outer diameter... a skim of .002" was completed & faces checked again... a much better contact pattern. All was reassembled & checked again... .002" out on 'Y' axis & .0015" on the 'X' axis, checking the DTI reading on tightening & slackening the base revealed a .002" flex front to back, obviously stress factors in the cast base are having a fun time... so what to do? I ended up deciding to get the best I could considering it's not a premier no lift jaw type of vice & ended up placing some .002" aluminium fizzy drink can shim under the back side of the vice base & tightened up, ending up with a discrepancy of .001"( over 4") front to back & .001" & on the 'X' axis over 6" parallel.

I eliminated as much jaw lift as I could by removing the bolts from the clamp plates under the vice body, replaced with studs & fitted Nylock nuts & snugged up so that the body was neatly snug up against the slides when opening & closing the vice. I took this into account together with inbuilt stress/ flex as part of the reason for the continued discrepancy overall. Considering that most retailers, UK & US, get their stock from china & these types of vices do leave a lot to be desired I reckon I've got as close as I can get so decided to call it a day... maybe Father Christmas will leave me a nice no lift jaw precision milling vice this year... .

George.

 

[jmh8743]

skipmeister has a point. automotive shops mill heads. they must be flat. they use big surface grinders.

 

[lpeedin]

Somebody asked previously if I know for sure that my table is flat. I am questioning that now. I built a dual dial indicator tramming aid and got it installed in the spindle last night. I was very surprised how much things varied from one end of my table to another. In certain parts of the table, there is a .006" variance over the 5.5" span between the two indicators. Other areas it is dead flat, some places only .001" or .002". This was in both axiis. (sp?) I am not sure what my next steps are. For now, I am just going to make some chips, smile, and be happy I have a lathe and a mill.

 

[John Hasler]

Some of that might be the table, some might be the ways. Need to check with a surface plate to tell.

 

[Crock]

I think I am just going to buy a better vise.

But then you wouldn't have the pleasure of telling people how much effort you put into making your vise work right. After all we are machinists and that is what makes tweaking something and see it come into spec so enjoyable. But then I'm retired from tool making and have a mill, grinder, and lathe but I don't have to be perfectly precise on my stuff. But it's still enjoyable.

 

[lpeedin]

Some of that might be the table, some might be the ways. Need to check with a surface plate to tell.

Agreed, hard to tell for sure what is what. I took the whole table apart over the weekend and was surprised by the results with my micrometer. The table itself is within .0003" tolerance on thickness at each end, with the front being approx. .0008" thicker than the rear. Then, the middle is approx. .0015" thinner than at each end. Since I don't have a surface plate, all I can check is thickness, not actual straightness / flatness.

I then checked the saddle and it was within .0003" all the way around, but I discovered an issue here. After cleaning everything thoroughly, and lightly stoning the bottom of the saddle as the scraping pattern on there was very aggressive in a couple spots, I placed it on the base ways. Three of the four corners sit flat, with one corner, the front right when looking at the machine rising slightly. In fact, I can get a .006" feeler gauge under with essentially no rubbing. For some reason I can't find my .007", so I checked with an .oo8" and it goes in about 3/8". It appears that my saddle is slightly twisted / warped.

So, after finding that, I put it all back together. Then I took a 12" piece of 1.5" x .5" flat bar stock and put it in the vise and proceeded to make a skim cut on each side, just enough to clean up the face. I checked it after the cut and again, I can't check flatness, only relative thickness. I was within .0012" over 11.5", with approx. 7" of that being absolutely dead on, then all of the .0015" variance occuring over the last 4.5" or so. I think what is happening is that when the weight of the table shifts from the left side to the right side, the saddle is able to settle in the corner with the gap, causing the table to tilt slightly. I would definitely like to get the saddle surface ground to clean up the bottom face before messing with anything else.

 

[lpeedin]

I should also add that what tipped me off is that with the table removed, I measured the distance from the top of the saddle down to the top of the base ways and that front right corner was about .003" higher than the other corners.

 

[John Hasler]

My Grizzly table had similar problems, but worse. I was able to scrape it in well enough to be usable using a piece of plate glass as a surface plate.