Holding +/-.0002 on lathe using Renishaw Probe

dstryr · Mar 30, 2017

I have a production part we make and the volumes just went 5x
Have 2 areas on the part that get bearings with this tolerance. We have made 1000s of these on the doosan lynx but only two of us can accurately measure and make them.

We are looking into adding a renishaw probe, rough the diameter, finish .001 larger, probe, recut.
The machine moves about .001 over the course of a day in thermal growth... very predictable

So anyone out here having good luck with this process? Looking at adding a coolant chiller at the same time as well.

*In before 2outof3 tries to sell me yet Another new machine LOL

DMF_TomB · Mar 30, 2017

dstryr said:
I have a production part we make and the volumes just went 5x
Have 2 areas on the part that get bearings with this tolerance. We have made 1000s of these on the doosan lynx but only two of us can accurately measure and make them.

We are looking into adding a renishaw probe, rough the diameter, finish .001 larger, probe, recut.
The machine moves about .001 over the course of a day in thermal growth... very predictable

So anyone out here having good luck with this process? Looking at adding a coolant chiller at the same time as well.

*In before 2outof3 tries to sell me yet Another new machine LOL

i often get differences in probe readings of .0002 to .0004", on cnc mill probe says put .0005 shim in and probe it and then it says take shim out probe again and says put shim in.
.
later i learned dont put shim in, reprobe and most times the 2nd time i dont need the shim after all.
.
lets just say i dont have too high a trust in probe readings. sometimes it is a probe problem aggravated by servo oscillation. servos can be going back and forth .0001 or more. depends on machine and what day of the week it is

dstryr · Mar 30, 2017

DMF_TomB said:
i often get differences in probe readings of .0002 to .0004", on cnc mill probe says put .0005 shim in and probe it and then it says take shim out probe again and says put shim in.
.
later i learned dont put shim in, reprobe and most times the 2nd time i dont need the shim after all.
.
lets just say i dont have too high a trust in probe readings. sometimes it is a probe problem aggravated by servo oscillation. servos can be going back and forth .0001 or more. depends on machine and what day of the week it is

You are making parts like 4' wide right? I'm turning bearing diameters that are 1/4 and 1"

dksoba · Mar 30, 2017

I've tested my probes repeatability to be no more than .0001" (minimum increment of the machine) different than the previous reading. Take this with a grain of salt though since this is when I simply repeat the probing program over and over.

The concern I'd have with the probe is that whatever causes the thermal growth would also affect the probes readings. I'd want to probe off some known reference point that is close to the spindle to quantify the thermal growth and then probe the part.

I'm all for probing, trust me, but I'd still want to have a go-no go gauge handy to check every xxx part to make sure everything is just right. Does anyone make gauge plugs that are larger than 1" to the same tolerance as deltronic gage pins?

Out of curiosity, how are these measured now? 3 point micrometer? How much is the diameter affected by tool wear and how much is it affected by thermal growth? Is this aluminum (you mentioned taking a 0.001" cut which is something I usually do only in aluminum).

Best of luck,
Matt

dstryr · Mar 30, 2017

dksoba said:
I've tested my probes repeatability to be no more than .0001" (minimum increment of the machine) different than the previous reading. Take this with a grain of salt though since this is when I simply repeat the probing program over and over.

The concern I'd have with the probe is that whatever causes the thermal growth would also affect the probes readings. I'd want to probe off some known reference point that is close to the spindle to quantify the thermal growth and then probe the part.

I'm all for probing, trust me, but I'd still want to have a go-no go gauge handy to check every xxx part to make sure everything is just right. Does anyone make gauge plugs that are larger than 1" to the same tolerance as deltronic gage pins?

Out of curiosity, how are these measured now? 3 point micrometer? How much is the diameter affected by tool wear and how much is it affected by thermal growth? Is this aluminum (you mentioned taking a 0.001" cut which is something I usually do only in aluminum).

Best of luck,
Matt

I have enough travel to probe both sides of the shaft so dont really need a reference point but yeah I figure a reference point would be good for larger stuff you cant travel on the other side of the part.

We 100% every part now and the plan would be to get the swing shift guys to run them and just 100% them in the morning. I would live with a small scrap rate if it meant better machine utilization.

We are using PCD tipped inserts so no real tool wear. Just the machine growth. Set the G53 offset to x0. in the morning and afternoon you are at .001. Next morning repeat. Has a lot to do with how long it sits between parts as well.

DMF_TomB · Mar 30, 2017

probe tip sometimes gets coolant, oil, dirt on it. as you probe the stuff can be coming off the probe tip or of course if probing a dirty part the probe can touch dirt and it can stick for awhile to probe tip.
.
just saying i have seen probe values not repeat to .0002 levels often. various reasons. like oil on you fingers as you try to clean probe tip can be leaving .0003 of skin oil sweat on probe tip. .0002 aint much. many machined surfaces have waves in surface of .0002 to .0004. we often take 3 probe readings like 0.1" apart to average readings. like looking at the 3 probe values they often vary .0003
.
you take 4 probe readings in a row of an area with slight changes in location like 0.1" i would be surprised if they all said within .0002, sometimes they do repeat but just as often they vary .0003

JNieman · Mar 30, 2017

We have an old fashioned guy here who used to tell anyone that would listen that you can't probe parts accurately because he swears he could probe and reprobe and it'd vary .001 or .002 all over the place. Our shop therefore had many renishaw probes sitting idle. I, being someone spoiled and in love with them at a previous company, argued with him all the time, in a friendly way, about how he must be crazy, had a bum probe, or it wasn't taken care of.

Eventually I gave up and got one of the people who worked under him to try it out. He couldn't believe how easy it was. He was using it on casting parts that he was machining. Then we started using it on our horizontals with a pallet pool. We use it for simple setups, locating a vise edge, whatever. We use it to check a surface in the middle of a program and then mill to clean up. Those are weldments. The main plate can vary in location up to 1/8", and we never have to mill more than .020 to get it to fully clean up. So we probe for the plate, adjust offset as necessary, then mill.

As for repeat-ability, we've never seen it change more than .0001. I didn't use it in lathes - VMCs and HMCs only. I don't see why a probe couldn't work in the lathe. Little air blast on the part before probing. Little air to clean off the probe before probing. Boom.

I think your scrap rate would be very well within acceptable ranges since you're already having to inspect 100%.

AFAIK the old guy is still relying on indicators and edge finders while several of the others in his department are in love with probes.

gustafson · Mar 30, 2017

not my wheel house, but if the drift is predictable, why not set up a go no go gauging station and measure as you go, when they drift, enter the correction in the machine, eliminate probing error.

[edit] not trying to get into the probes are/not accurate conversation, but rubber/road is the diameter of the part

JNieman · Mar 30, 2017

I think go/no-go gages for a porm .0002 diameter is going to have too much chance of operator error. If they can't be expected to run the machine in a manner that results in good parts - I wouldn't expect they could take care of, and use such gages properly either.

Red James · Mar 30, 2017

Can you say 'Indicator Micrometer' three times quickly?

dstryr · Mar 30, 2017

4500 pieces on a 5 min cycle...
Then goes on to milling on 5axis for 15 minutes

rlockwood · Mar 30, 2017

JNieman said:
We have an old fashioned guy here who used to tell anyone that would listen that you can't probe parts accurately because he swears he could probe and reprobe and it'd vary .001 or .002 all over the place. Our shop therefore had many renishaw probes sitting idle. I, being someone spoiled and in love with them at a previous company, argued with him all the time, in a friendly way, about how he must be crazy, had a bum probe, or it wasn't taken care of.

Any chance his name is Tom B?

Mike1974 · Mar 30, 2017

We have a Mazak 250msy with a renishaw probe and coolant chiller. Sorry, haven't used it other than getting it calibrated. I do know that air blast automatically comes on through the tool block when the probe comes on. Not sure if that is from the renishaw side or Mazak side...

dksoba · Mar 30, 2017

dstryr said:
Has a lot to do with how long it sits between parts as well.

I usually run a dummy run or two if a machine has been sitting for a while and I really don't want to scrap the first part (like after lunch, for example).

I guess you don't trust the swing shift guys to make offset adjustments? Are you measuring bores using a 3-point micrometer? Or maybe the 3 pin method (I could understand how only a few people in the shop could utilize this method). Probes seem to me to cost more than other options, but I could see this being worth it if this is really high quantity or if you have other jobs that you'd like to be able to probe the diameters on.

JNieman · Mar 30, 2017

rlockwood said:
Any chance his name is Tom B?

Sounded familiar, huh?

I swear it wasn't intentional. Each detail is true other than natural story-telling embellishment. The facts and figures are all right, though. I won't swear under oath about the adjectives, though.

dstryr · Mar 30, 2017

I wonder if I just built some sort of gauging feature that made it dead simple to measure and fast. We have a Zeiss CMM but the time spent measuring the part is enough to fuck up the growth cycle. Needs to be something at the machine.

They are two turned diameters.
And you are correct.... only two of us here are capable of holding that tolerance

dksoba · Mar 30, 2017

Mike1974 said:
I do know that air blast automatically comes on through the tool block when the probe comes on. Not sure if that is from the renishaw side or Mazak side...

I usually turn my filtered coolant on. It's filtered using a 5 micron bag filter, so I guess a chip that's 5 microns in diameter could sneak through, but with the coolant blasting on the probe tip it seems like I never have a problem.

Matt

dksoba · Mar 30, 2017

dstryr said:
I wonder if I just built some sort of gauging feature that made it dead simple to measure and fast. We have a Zeiss CMM but the time spent measuring the part is enough to fuck up the growth cycle. Needs to be something at the machine.

They are two turned diameters.
And you are correct.... only two of us here are capable of holding that tolerance

Deltronic gage pins or a ring gage is the first thing that comes to mind. Or a micrometer of some sort.

implmex · Mar 30, 2017

Hi dstryr:
Marposs and others make pinch type in-process gauging systems for cylindrical parts; you can often see them on cylindrical grinders doing production and holding way better than tenths.
This eliminates the thermal drift of the machine tool slide from affecting the probe position, so they're intrinsically more accurate but the downside is they are less versatile.
They might be worthwhile to look into.

The next obvious question is, what will you do next once the gauge indicates an out of tolerance condition?
Will you recut the part, will you toss out the part, will you do one or the other depending on whether the part is out of round, undersize, oversize, not cylindrical etc etc, or will you just flag all non-conforming parts and dump them in a separate bucket for rework or rejection on a separate station?

I ask, of course, because as we all know, the turning process, even on a very good machine is marginal for truly holding these kinds of tolerances, so once you can reliably gauge to this level you may find yourself with a quite large reject bucket, and no really effective way to correct non-conforming parts in-process.
As simple example; if you're two tenths up, do you feel your lathe can take another tenth per side reliably, and if not, do you simply toss your part?

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
www.vancouverwireedm.com

Vancbiker · Mar 30, 2017

I was going to advise looking at Marposs, but Implmex beat me to it. I worked with them a bit years ago on an Okuma CNC grinder and a Shigiya CNC grinder. Deadly accurate in those applications.

Holding +/-.0002 on lathe using Renishaw Probe

Diamond

Diamond

Diamond

Hot Rolled

Diamond

Diamond

Titanium

Diamond

Titanium

Stainless

Diamond

Stainless

Diamond

Hot Rolled

Titanium

Diamond

Hot Rolled

Hot Rolled

Diamond

Diamond

Similar threads