cast iron t-slot set-up table

[Old Mech]

Greetings-

 

I have a small quandry. I am setting up a cast iton t-slot set-up table in the shop. It is 4 tables bolted and keyed, ground flat, 12' by 19', set into the floor with 72 anchors, into a slab that is joined to 6 30' deep pilons into the ground. Heavy stuff. I am in the process of truing and leveling this table using a Hammar Laser Level.

 

This table had been re-worked several years ago, and set up in its present position by an installer. After several people have re-leveled it, and about a year since its last check, it has gone a bit out of whack.  We use it to set up machine frames for building to get them as true as we can before moving them to the construction bay. I have been assigned to make it perfect. I think that is because I am now one of the "old guys" who know all the secrets. (I don't know everything)  Problem is, there are no other "old guys" around to ask.

 

When I got to the job, there was an overall .010 sag in the center. I loosened up all the anchors and let the baby float and relax. Then I shot all 4 corners, got an average, and set them all to a common zero. Then I started in the center, brought it up to level, and began setting the anchors in a circular pattern to keep things evened out. When I got done, the center was .005 high. I put it down, and the next row went up. It started to act like a bowl of jello, so I just left it set for the weekend.

 

What I would like to know is: Is there a preferred pattern to setting the anchors to 0 so that I do not end up chasing around in circles? Should I have gone diagonally and criss-crossed over the center instead of a circle?  How can this much iron move around that much? I am familar with leveling machines and using the tools (30+y), but this is a mystery to me. A good answer or a link to material to find out would be so kind.

[DSW]

Your problem may lie in your foundation. The slab my sit on pylons, but the area in between may be under supported. I've jacked out numerous floors and slabs to find the sub base underneath has settled, or in some cases moves due to seasons or moisture. Some soils expand when moist and contract when dry. Freeze/thaw lifts and heaves things, probably not you issue, but temp swings can also make changes.  They put expansion joints in concrete floors for reasons.

 

To me a 6" slab is "thin". 6" is the minimum we used to pour for residential driveways. On poor subsoil or when heavy traffic was expected, we'd go no less than 8". My driveway sits on 36" of crushed stone and then had 8" of 3500 PSI concrete with a mat of #4 on 12" centers and #4 4x4 wire  mats in it on 2 1/2" chairs.. I've cut out industrial floors more than 18" to 30"  thick with double and triple mats of #8's and larger rebar in them. Then you hit the footings...

 

A good test to see if a floor moves is simply to drive something heavy on it. Check certain spots with your laser and mark them for reference. Then load up the forklift and drive over by your reference marks. If they change, guess what your floor moved no matter how tough you think it is. Movement due to moisture in sub soil is harder to determine, same with temperature changes. Usually you need to take a series of readings when conditions are known and compare them to each other.

 

 

Good luck.

[Old Mech]

Thank you DSW. While I do not know the exact construction of the floor, I do know that 100 years ago it was swamp. I will try your test and keep water and soil in mind. I will also try and find out more about how much concrete there is.

[Black Frog]

I have not had experience with exactly what you're doing, but when tightening or tensioning something with an array of fasteners, I was always taught the parent rule is to criss-cross the pattern of tightening.

 

If you go in a circular pattern, you end up applying lots of tension/stress to one side before the other side gets any application of force.

This does not lend itself to keeping things consistent through the entire package....

 

Don't know if that helps any, but wish you the best of luck.

[Dodge]

Greetings-

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

 

When I got to the job, there was an overall .010 sag in the center. I loosened up all the anchors and let the baby float and relax. Then I shot all 4 corners, got an average, and set them all to a common zero. Then I started in the center, brought it up to level, and began setting the anchors in a circular pattern to keep things evened out. When I got done, the center was .005 high. I put it down, and the next row went up. It started to act like a bowl of jello, so I just left it set for the weekend.

 

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

 

I tend to agree with Black Frog. Criss-cross is the standard or the industry in tightening. From wheel lug nuts to large machinery. But I take it this table isn't being set simply in a typical 1/4" tol blacksmith shop. I used to work in a company manufacturing structures from traffic 10' light poles to 150' power transmission poles and our tightest tolerance was plus or minus .03" or 1/32". Having trouble wrapping my mind around plus or minus .005" or 5/10,000" LOL

Anyway, I'd try criss-cross tightening and shim where required.

 

(Is tin foil thinner than five ten thousandths of an inch?? :huh: )

[Judson Yaggy]

What is the tolerance of the laser level you are using?  I'm not familiar with that brand, the ones I have used +/- .005" in 100 feet or so is well within tolerance.  What is the ground under the laser level doing over the course of your work day?   If you are not already doing so you can shoot a benchmark on a known immovable object and keep checking it during the workday to see if the laser is shifting.  Any trucks or trains driving past your building while you are trying to dial in the floor?

 

EDIT

 

Ok, googled the laser, the laser tolerance is probably not the problem.  

[yahoo2]

Having trouble wrapping my mind around plus or minus .005" or 5/10,000"

gone one order of magnitude too far Dodge, its 5 Thou (or mil) or 127 microns in Metric.

 

I would suggest digging up the original certificate that came with the table install and see exactly what specs it was milled to in the first place. You are not going to flatten a table that was never that flat, especially if there was some tension in the legs when it was milled.

 

I had a look in one of my books, the finest milled surface without blueing and scraping for a 19' long base plate  would be DIN 876/I, it is listed as 0.00269" tolerance  876/II is 0.00539"  876/III is 0.01078"

 

Best thing I can suggest is not to go for the flat number but instead creep up on it and aim for a tolerance, perhaps shift a few high and low areas 1 or 2 thou on the adjusters then re-measure. my gut feeling is that it is more predictable when you work against gravity ie, lifting the hollows rather than lowering the high points for the final tweaks but I don't have any paperwork to back that up.

[Old Mech]

Good Morning Gentlemen-

 

Wow. You guys ARE good. Here is an update:

 

The table is anchored to an 18" slab that has a lot of steel in it. Rebar, not netting. This slab is 24' x 24', and is isolated from the rest of the flooring and the building. This slab is also attached to six 30' Deep Pylons screwed into the ground and filled with concrete. The specs left by the installer show that the original tolerance was .0003" Flatness. I did not get that good this time.

 

It is, in fact, very similar to Jello. I found that criss-crossing was better than circular, although I have not yet decided on a pattern. I ended up getting the table in rough (+/-.010), and then I went at it piece by piece. I tweaked the corners, and then the middle, and then criss-crossed all over the place, checking, checking, checking. If I moved an anchor, say, .002, I found that sometimes the one next to it would move, and sometimes it would like skip over to another row. So, by very carefully and slowly tweaking the worst of the worst, I gradually got the dang thing within +/-.003 from one end to the other. I shall now write up a MAP and detail my work so that next time I can jump in and kick right off. I will also order some better tools. I also found out that the lazer had to be checked constantly, because the base would sometimes go off; so I checked the levels on the Laser every time I walked by it. Then I found out that the fan I was using was also playing havoc with the readouts, so I had to make sure I had a nice coating of oil on the table to resist the sweat dripping off my nose. :D

 

Thanks alot to all of you, your suggestions were more than worth the time.