Accurite Measurement Of A Long Shaft

[Ski]

I have to shorten a long shaft and want to get it as close as possible. My question is what have others made to make a measurement. Shaft is about 36 inches long. I will need a accurite end to end and also one end to face of a shoulder it will have. I have a couple ideas but thought I would see what has already been engineered. Ski

 

[Uglydog]

While an accurate bench center is best.
I've used my lathe. However, you gotta be sure the lathe is dialed in.
All depends on your tolerances.
Remember that round is a different measurement from straight.

Daryl
MN

 

[JimDawson]

Short of having a really long caliper, the best way might be to use a tape measure. The last Stanley I checked was better than 0.005 in 20 inches. Clamp a straight edge at the 1 inch mark, and measure away, don't forget to subtract the inch.

You could also clamp a caliper to the tape measure in this case at the 33 inch mark then measure from there, should get you pretty durn close. if you are careful in the positioning of the straight edge and the caliper.

I'm sure there will be more and better ideas forthcoming.
.
.

 

[RJSakowski]

You didn't say what accuracy is required. I have a 24 inch Starrett that can measure to .001" but when I get past that length, I'm up the creek. One possible method would be to face lengths of threaded rod to lengths you can measure and stack them with coupling nuts. Leave the last one a little short and measure the difference with the depth feature on your calipers. Not exactly tool room procedure but better than a yardstick.

One thing to be aware of at that kind of length is thermal expansion. The length will be dependent on its temperature and just the process of picking it up can change the measurement.

Bob

 

[Ski]

My Lagun mill has a 42 inch table. The x axis screw is pretty worn and I bid low on a longer new one with nut (54") on ebay and won it. New from lagun is high dollar. Heck,I think the nut is over 300 and the shaft is 750. Don't quote me but I think I got a quote 3 years ago for 1000 for both. The original one in my machine is 2 piece with a pin. My goal is to match the original as close as possible. I am not sure what degree accuracy I should shoot for other than matching the original. I have 250 bucks in the new screw and nut. The shafts are soft enough to file. Ski

 

[Ski]

My thoughts were to weld up something for a 1 time use. Flat U shaped flat stock jig,filed to fit.

 

[RJSakowski]

Ah, a different problem. Can you just use your old shaft as a comparison? You should be able to get close enough. You mentioned a pin. How is that incorporated? If possible, allow for the use of shims to adjust clearances. That would probably make the new shaft a hair longer. The shim pack would take up the difference.

 

[Ski]

I have not taken the pin out yet. I am thinking a taper pin but not sure as I did not measure the ends. It appeared as a solid pin viewing from the ends of the pin. There is slop inthe pin as well. There are shims.

 

[Tony Wells]

With the shafts side by side and blocked together for linear position, you can use a depth mic from one to the other for comparison. I would expect to get match to within 0.001/0.002 with this method, as long as you use some good, stable surface to put them on. Is your lathe long enough lay them in the ways? You could also slide the old one in one of the grooves, set up a indicator and mag base and do a comparison.

 

[Ski]

Thanks Gents, Good food for thought. My lathe is a 5 foot bed 13" South Bend. The screw fits thru the head stock and I made a 4 screw spider for the left side of the headstock so I can dial both sides in. I have an old nut that I split and faced and plan to screw 1/2 on each side to get a good anchor point onto the screw so I don't damage it.