Roundness...do you use it?

[Sl33stak]

Is it acceptable to check for roundness (say on a valve) using the CMM (a NON-scanning head)? Or should it be done on an arbor with a dial indicator?

[bw_bob]

In my opinion a cmm is perfectly acceptable to check roundness provided you use enough points.

[Matthew D. Hoedeman]

About the only time I use roundness is for checking my results when mapping (or digitizing) details. I use it to tell me if my hits in a dowel hole (which is critical for location) were any good. If it is out-of-round by 0.0005" or more, I will check it again at a different height (they will egg-out if hammered too much).

[#2#]

I use it when necessary. Not too many roundness callouts here. I think a CMM isn't the best tool for getting diameter numbers (as stated many many times before) but, roundness is different. As long as you take a lot of hits I think it is ok.

[Paul Sarrach]

goodluck is right, I have not seen that call out in a while

[Sl33stak]

I don't have a roundness call out to check, but I did have a customer send back some valves and wnated roundness checked. They are roughly .850 in dia. and i'm gonna use a 2.5x30 mm tip, was gonna take 20 hits. Sound right?

[bw_bob]

I don't have a roundness call out to check, but I did have a customer send back some valves and wnated roundness checked. They are roughly .850 in dia. and i'm gonna use a 2.5x30 mm tip, was gonna take 20 hits. Sound right?

Me personally i use anywere from 32 to 64 hits depending on what size diameter i'm checking. Less of course if its a small diameter.

[Johnson]

I dont personally feel 20 hits is good enough for roundness. I would up it to about 60.

[#2#]

I would take an odd number of hits. If taking an even number there is a chance that you will hit all high or low spots. If the thing that made the thing round was vibrating it could have left a pattern. Taking an odd number of hits will helps ensure that you won't hit all highs or lows. At least this is whay the guy who trained me said/explained. I always take an odd number of hits on round features.

[jkb_mcg]

A TP-20 module has about .0005" variance in itself. Not sure what kind your using.

[#2#]

A TP-20 module has about .0005" variance in itself. Not sure what kind your using.

I'm not sure where you got that number. According to the Renishaw catalog a TP20 has a "Repeability of stylus change (max.)" of ±.000040in.

[BOB PARSONS]

ROUNDNESS

We use it often on precision machined parts, on ground holes you don't have
to worry about so many hits. But as with a profile the "correct" way would be
w/ an indicator on 100 % of the surface.

[Shipwreck]

if you ask a renishaw person they will tell that a TP2 has a lobbing error, there is a built in error do to the fact that the probe has inside mechanical switchs... runout is not good to check unless you have a fairly large tolerance. I am not sure but I think a TP20 has the same condition...
a TP200 would be better since it is a strain gage..

[#2#]

if you ask a renishaw person they will tell that a TP2 has a lobbing error, there is a built in error do to the fact that the probe has inside mechanical switchs... runout is not good to check unless you have a fairly large tolerance. I am not sure but I think a TP20 has the same condition...
a TP200 would be better since it is a strain gage..

Ah, but shouldn't PC-DMIS correct for the lobing when you do your tip qualification?

[Matthew D. Hoedeman]

It can only 'comp' for the lobbing (such as it can) in the direction of the touches you make when calibrating. Any touch other than those exact spots of the probe will not have the correct comp.