Lathe Leveling Problem

[epanzella]

I'm trying to level my Grizly G4003G which I just put on the bases last night. I bought a machinsts level with .0002/10 inch accuracy. After I level the headstock it moves when I level the tailstock. One is three graduations one way and the other is about the same in the other direction. It says .0002 on on 10 inches on the level. I've been chasing my tail for an hour trying to get this and it's just not happening. This level is so sensitive the bubble shifts if I get a flashlight too close to it. So, how level is level? .001? .0001? Do I have to wait to let the machine settle onto onto the leveling pads? I'm leveling across the V-ways. Is this Ok or do I have to rig parallels to level across the flats................................................................................................................................................................................................................................................

 

[epanzella]

I was laying in bed last night and it hit me like a thunderbolt. Being as my lathe is bolted solidly to two independent bases, if one of the bases is 90 degrees to the floor and the other is 89 degrees to the floor, they will act like two levers and twist the bed. I could go up and down with the levelers forever and it would never fix that problem. I went down the basement at midnight, put a block against the tailstock base, and gave it a light hit with a sledge. Poof, the lathe bed is now level to better than .0002. My machine sits on eight leveling pads. I wonder how you would fix this if the machine was bolted to the floor?

 

[Ray C]

Guys, guys, guys...

Level to horizontal is meaningless. The machine needs to be aligned so it will cut straight and that has nothing to do with it being level.

Please search this list and look for threads about lathe alignment. Also look-up "2 point method" and "Rollie Method".

You can also look at my thread called "YALB Yet Another Lathe Bench" which gives background information and leads-up to pointers to the proper methods.


Ray

 

[epanzella]

Guys, guys, guys...

Level to horizontal is meaningless. The machine needs to be aligned so it will cut straight and that has nothing to do with it being level.

Please search this list and look for threads about lathe alignment. Also look-up "2 point method" and "Rollie Method".

You can also look at my thread called "YALB Yet Another Lathe Bench" which gives background information and leads-up to pointers to the proper methods.


Ray

I'm familiar with the 2 point method but I don't have power to the lathe yet. I think having the ways true to each other with no twist between the headstock and tailstock are important. Leveling them both front to back seems to be the easiest way to accomplish this. I definitely was not going to leave an observable twist in the bed of two to four thousandths if I could help it. Was I wrong in making the bed straight?

 

[Ray C]

I used to think the same way but finally shed my stubborn beliefs. Over a 30" distance, a few thousands of twist will hurt positively, absolutely nothing! Metal can almost flex that much under it's own weight. Even if it were twisted much more than that, it would just flex right back. If you look at the yield and working strengths of the sizes of pieces of metal at-play, you are orders of magnitude within the metal's ability to restore from deformation. Now, if you set 10 or 20 pallets of lead bricks on your bed, that might be a different story.

You did no harm by leveling it but, it will be a mircale if it cuts straight that way. It will be totally pure/chance luck if it does. Bolt it to a solid table or put pedastils under it and adjust the legs or levelers until it cuts straight. Done!

I did this on mine and it's only "approximately" level and each leg of the bed is slightly different. I've cut shafts 26" long and had total variance end-to-end and everywhere in-between at about 0.0006. I suspect if I waited long enough for the shaft to evenly cool off, it would have been even better than that.


Ray

I'm familiar with the 2 point method but I don't have power to the lathe yet. I think having the ways true to each other with no twist between the headstock and tailstock are important. Leveling them both front to back seems to be the easiest way to accomplish this. I definitely was not going to leave an observable twist in the bed of two to four thousandths if I could help it. Was I wrong in making the bed straight?

 

[Tony Wells]

Being that this is a new machine, getting it level with the world is a simple way to ensure that it will be a straight as it was on the machine that ground the bed. Leveling across the top of the vees is not quite proper, though. If you look closely, you'll probably see that they aren't ground. Normally, parallels or 1-2-3 blocks are placed as needed to use the flat ways that were ground to be together.

Ray's points are well taken, but apply in a more realistic way to a used, worn machine. We all know that the end result is what tells us if everything is aligned. If this machine is new, then it should be very well aligned if it is leveled properly. They are two different things. Getting the twist out of a new machine is imperative. Then and only then can a cutting test be effectively used to gage a machine's accuracy and appropriate measures taken as needed, if needed.

 

[epanzella]

Being that this is a new machine, getting it level with the world is a simple way to ensure that it will be a straight as it was on the machine that ground the bed. Leveling across the top of the vees is not quite proper, though. If you look closely, you'll probably see that they aren't ground. Normally, parallels or 1-2-3 blocks are placed as needed to use the flat ways that were ground to be together.

Ray's points are well taken, but apply in a more realistic way to a used, worn machine. We all know that the end result is what tells us if everything is aligned. If this machine is new, then it should be very well aligned if it is leveled properly. They are two different things. Getting the twist out of a new machine is imperative. Then and only then can a cutting test be effectively used to gage a machine's accuracy and appropriate measures taken as needed, if needed.

The lathe is advertised as having hardened and ground V-ways. The surface looks ground and the seam for the gap bed is all but invisible. In my original post I asked if I could level across the V's. I have a depth mike but right now I don't have anything both precise enough and long enough to span both V's to measure down to the flat ways. I'll see if I can scare up some blocks or parallels. I just getting back into this after being away for 20 years. (and I wasn't that experienced 20 years ago either)

 

[Ray C]

Hi...

I didn't intend for this to apply to just well worn machines...

When a bed is ground (and I'll ask if I can share the pictures of the PM machines being made) they rough-out the feet then, bolt it to a work station which is rolling cart anything but perfectly level. It passes through several stages of operations (drilling, tapping etc). Toward the end, it is induction hardened then ground. After that back it goes onto industrial rolling benches that are anything but perfectl flat where many more things such as mounting the head, attaching all the brackets etc -and all of that work is done. Point is, all the secondary work that was done will forever change the shape (very slightly) of the overall machine. Once it is fully assembled, they bolt it to a level table and check various tolerances. Those tolearances need to fall within certain limits but, they are no guarantee it will cut perfectly straight. In order for that to happen, the factors from which the tolerances were derived must be tweaked to as close to dead-on as possible.

The point being, I highly doubt a freshly assembled machine bolted to an infinitely rigid and perfectly flat table stands a chance of cutting perfectly straight. Just by looking at the variables involved from a purely mathematical view of manufacturing statistics, each and every machine will need to be fine-tuned when it gets home. If for any reason, once it's strapped to a pallate and bounced acrosss a couple continents, it will never be the same as it was the day it was made.

As for leveling it... It think it's prudent to place it on level, even ground and try to level it so the drain pan drains properly. After that, the legs etc need to be tweaked by fractions of an inch to make it cut properly.

Ray

Being that this is a new machine, getting it level with the world is a simple way to ensure that it will be a straight as it was on the machine that ground the bed. Leveling across the top of the vees is not quite proper, though. If you look closely, you'll probably see that they aren't ground. Normally, parallels or 1-2-3 blocks are placed as needed to use the flat ways that were ground to be together.

Ray's points are well taken, but apply in a more realistic way to a used, worn machine. We all know that the end result is what tells us if everything is aligned. If this machine is new, then it should be very well aligned if it is leveled properly. They are two different things. Getting the twist out of a new machine is imperative. Then and only then can a cutting test be effectively used to gage a machine's accuracy and appropriate measures taken as needed, if needed.

 

[epanzella]

A followup;
Got my G4003G breathing finally. I got both ends of the bed true to each other within .0002. I did the 2 point test with the rings about 6 inches apart and was within a .0001 or two. One caviat; The instructions I read on the 2 point test said to use stock 1.250 or greater. I only had one inch stuff and had to turn .030 off it to true it up. Does the smaller stock somehow mask any bed twist or is the 1.250 recommended only to prevent spring? My last three cuts were .005, .003, .003. It's funny how some things stay with you. It's been 20 years since I had a lathe and I still keep reaching with my left hand to turn the machine on or off. This one has the on/off lever on the carriage to my right. I can't believe how clumsy I feel running this machine and 75% of the stuff is in the same place as my old one.
Thanks
Ed P

 

[darkzero]

the 1.250 recommended only to prevent spring

Yes, that is the reason.