In need of help improving cold saw repeatabilty

[rkremser]

Currently we have a Kalamazoo cnc cold saw. We are looking to get repeatability within .005". Currently we are struggling around the ball park of .015" the feed system is using an pneumatically driven hydraulic cylinder. (air is moving old into the cylinder) on the indexer I had setup a dial indicator and it is repeating within .002" from each end of the travel. Could the blade be causing this much variance or what else would be suggested to look into. The material being cut ranges from 1"-4" stainless thin wall tubing. Thanks
Saw model can be seen here.
http://www.kalamazoo-saw.com/Product...sAutomatic.htm

[rkremser]

Anything at all? It seems the jaws of the saw vice are also out of square from the saw blade itself. This is causing some of the variations that have been seen but I'm lost as to how to accurately square the jaws to the blade. I have gone through the process of squaring heads on mills but that was much easier with having the moving axis to check against.

[HuFlungDung]

If this saw is capable of miter cuts, it should be able to be adjusted for squareness relative to the vise around a vertical axis. If the locking detente (at 90°) is out, you might have to fiddle with that. Check the squareness of the cut on the end of the tube after taking a cut. It might be necessary to adjust the height of your support stands to ensure a square cut.

Indicate the saw blade for true running around the spindle, just to check the condition of the arbor flange. If there is no wobble ( minimal wobble) to the blade you will at least then be able to use a precision square against the blade to check vise squareness. Blades are usually hollow ground, so allow for that. If you have a large enough precision square, you can remove the guard off the saw, and check for simultaneous square contact at both top and bottom or front and rear of the blade.

Thin wall tubing like that may be reasonably tough and could easily damage the sharp corners of the blade teeth. If the blade gets dulled, it will wander very easily. Make sure the rpm is correct for the blade diameter used, and use flood coolant for stainless.

They usually recommend that 2 or 3 teeth be in the cut at a time, so you need to have a suitable pitch blade available for the wall thickness of the cut. While the saw will cut thin wall very rapidly after it breaks through the top, the feedrate should be moderate for the downfeed as it seems to me when I really 'push' the saw, is when I get wandering in the cut.

[rkremser]

The saw is one capable of miter cuts and does not have a good locking position at 90. The large T-square was going to be the first attempt but i was wondering if there would be a more appropriate method. As for the saw blades we have several out for sharpening to see if that will indeed help and the feed was actually pushing quite quickly. We will have to play around a bit more to get it where we want but thanks for the input.

[gus]

I have just this saw in the non ferrous version. It is 14 years old and i can get +- .01 without much attention, and it was good for 5 when new.


Exactly what is the nature of your error?

[rkremser]

The cut lengths seem to be varying up to +-.02 which is much to large for our purposes. One thing we have found is the jaws are significantly out of square of each other but it seems there isn't' a good way to square them. The jaws on the indexer are a few degrees out of alinement from the jaws at the saw blade, this ends up causing the material to catch when being indexed. (which most likely is the cause for the large variations) Is there some easy way that i have missed to help align these aside from using shims between the jaw faces and the supports?