What is correct measurement technique?

RJT · Nov 1, 2018

Had a discussion with a customer, and neither one of us had a definitive answer. Here is the scenario. Say you have a 24 inch disk, 1 inch thick, with a 2 inch bore in it. To measure parallelism of the 2 sides would you :
A) Lay it on a surface plate and move a height gage with and indicator over the part

or

B) Leave the height gage stationary and move the part under the indicator

or, it makes no difference either way?

For flatness, we always put the part up on 3 point risers (cone shaped supports all ground to exactly the same height) and check underneath with a height gage and indicator, moving the height gage.

All this assumes no CMM available.

Gordon B. Clarke · Nov 1, 2018

If we knew what the parallel requirement (tolerance?) was you'd get an answer you could use as that'll decide what's best.

How was the part even made and from what?

Richard King · Nov 1, 2018

To Please your customer I would do it his way first and record the findings and then your way. Hate to see you loose a customer because your stubborn. The answer to me is I generally move the part under a stationary indicator. It would also depend on the accuracy of the surface plate your using.

RJT · Nov 1, 2018

Grade AA plate, looking for parallelism of .002"
The customer really wasn't objecting to the procedure either way, I'm just wondering if there a correct way to do it. Part is PHT 4140, side 1 faced, turned , bored all in one holding (blank is 1.3 thick, holding on to .3 depth in stepped chuck jaws), side 2 held by OD in bored step jaws and faced to required thickness.

Eric M · Nov 1, 2018

In my opinion, you might have trouble indicating the entire surface of a 24" disc with a "typical" height gage / test indicator setup, due to the short reach. (Typically about 3 or 4 inches.) Making a longer bar might introduce some flexing you don't want. Maybe make 3 risers high enough that you can pass a surface gage with test indicator attached underneath it? Indicate the flatness of the underside, flip the part, and indicate again... YMMV

Gordon B. Clarke · Nov 1, 2018

RJT said:
Grade AA plate, looking for parallelism of .002"
The customer really wasn't objecting to the procedure either way, I'm just wondering if there a correct way to do it. Part is PHT 4140, side 1 faced, turned , bored all in one holding (blank is 1.3 thick, holding on to .3 depth in stepped chuck jaws), side 2 held by OD in bored step jaws and faced to required thickness.

Certainly more than one way but from your machining method I'd just use a micrometer 3 or 4 places around the perimeter. If the results don't vary by more than 0.002" then it's OK parallel.

art.h · Nov 1, 2018

Along with multiple micrometer measurements the use of a good straight edge to verify flatness-- no double dome or double concave, which can occur on a lathe on the flip to bring to thickness.

RJT · Nov 1, 2018

A micrometer is limited in reach, this is a 24" disk. Yes, i certainly would make a stout enough arm to overcome any deflection over a 12 inch reach. We have large / heavy enough height gages to do the job. My main question is, move the part or move the heght gage for ultimate accuracy?

PeteM · Nov 1, 2018

Deeper throat micrometers are commonly available if this is a repeat job.

If it's not a high tolerance job, I think art.h has it right. One way this can get messed up is having it machined cocked in the jaws. Micrometer readings around the periphery will quickly show that; perhaps even at the lathe and after a first cut.

Other ways it can get messed up is to have tool wear, a concave or convex cut, or warping. A straight edge (or either of your surface plate methods) should show that. Cheap CMM another option if the required tolerances are higher.

DMF_TomB · Nov 1, 2018

RJT said:
Had a discussion with a customer, and neither one of us had a definitive answer. Here is the scenario. Say you have a 24 inch disk, 1 inch thick, with a 2 inch bore in it. To measure parallelism of the 2 sides would you :
A) Lay it on a surface plate and move a height gage with and indicator over the part

or

B) Leave the height gage stationary and move the part under the indicator

or, it makes no difference either way?

For flatness, we always put the part up on 3 point risers (cone shaped supports all ground to exactly the same height) and check underneath with a height gage and indicator, moving the height gage.

All this assumes no CMM available.

.
the act of pushing a part applies force and if directly touched by hand, heat to part where touched. everyday you can setup a .00005" indicator and directly see where even a light touch or push can effect part.
.
i would normally not like to push or touch a part i am measuring. insulating where you push helps with changing temperature of part

Gordon B. Clarke · Nov 1, 2018

RJT said:
A micrometer is limited in reach, this is a 24" disk. Yes, i certainly would make a stout enough arm to overcome any deflection over a 12 inch reach. We have large / heavy enough height gages to do the job. My main question is, move the part or move the heght gage for ultimate accuracy?

Huh? I meant measure thickness at the perimeter. It is after all 1" thick. If a machinist here couldn't turn that parallel to within 0.002" (0.05mm) he wouldn't be regarded as a machinist.

I've suggested what I'd do. Everything else seems like overkill.

RJT · Nov 1, 2018

Tom B
I like your logic. Also if it's not flat it can rock and influence what you see on an indicator. Keeping the part still seems to make the most sense to me.

Illinoyance · Nov 1, 2018

Most lathes, when new, will produce a concave surface when facing.The faced side of the part will be a very shallow cone.Keep that in mind when figuring out how you are going to do the measurements.

Richard King · Nov 1, 2018

Another simple trick to see if it is flat on top of a surface plate after you wiped everything off with your hand and let the part set a bit to adjust to heat transfer is to take a plastic coated lead dead blow hammer and tap the part and see if you have a solid thud noise. The same noise all over. Only need to hit is softly on a small part that size. It won't tell you how much it's off, but will tell you if it is flat.

Mike1974 · Nov 1, 2018

move the part. if you move the indicator/height gage you could inadvertently move/change the zero on indicator (if that makes sense)...

SeymourDumore · Nov 1, 2018

RJT said:
A) Lay it on a surface plate and move a height gage with and indicator over the part

For flatness, we always put the part up on 3 point risers (cone shaped supports all ground to exactly the same height) and check underneath with a height gage and indicator, moving the height gage.

That is how I have always seen it done, with one notable addition in some extra tight instances.

After the indicator is set to the reading, it is then moved over to a height master and have it zeroed.

As you move the indicator around, from time to time you can always go back to the height master to make sure that the indicator has not been bumped in any way.

Besides, a 24" x 1" plate ain't something you'd want to be sliding around on a surface plate, and to boot, if it is set up on the three pointed references it is virtually impossible to slide anyway.

As to some of the comments about micrometer measurements for flatness on a 24" "as-turned on a lathe" plate .....

Booze Daily · Nov 1, 2018

I usually move whichever is easiest to move. If your disc has a fine finish, I’d move the height gage.

PeteM · Nov 1, 2018

SeymourDumore said:
. . . As to some of the comments about micrometer measurements for flatness on a 24" "as-turned on a lathe" plate .....

Micrometer can't tell if the plate is flat to spec. But it can tell, within about 30 seconds and on the lathe, if it's not seated properly in the jaws.

SeymourDumore · Nov 1, 2018

PeteM said:
Micrometer can't tell if the plate is flat to spec. But it can tell, within about 30 seconds and on the lathe, if it's not seated properly in the jaws.

Pete

It can be perfectly seated in the jaws, and yet still come out coned or domed.
It really is quite difficult to judge a perfect balance between good gripping force, but not too much as to bend the part.

In his situation, RJT has a 2" hole in the middle so it's easy to fit a mic in there and see if it did in fact bow ( in or out ), but let's not make a case for this being a trivial
job done on a lathe.

cyanidekid · Nov 1, 2018

depends on the size of your surface plate also, sweeping a 24" part under the indicator is hard to do on a 24 X 24" plate.

If you hold the part stationary, and sweep the indicator in an arc from a single position, it seems you can reduce the contribution of the plate error to the amplitude of the deviation.

the total will still be dependent on the flatness of the plate of corse, and squareness of the indicator base, but to the extent that the indicator base is in a somewhat more fixed trough or peak, at least it takes part of another variable (plate error) out of the measurement, at least for one quadrant at a time, which on a turned part should be diagnostic.

having said that, for .002, on a turned disk, a straightedge, feeler gauges, and a mic (or even a decent caliper) will get you there, no problem.

What is correct measurement technique?

Titanium

Banned

Diamond

Titanium

Cast Iron

Banned

Aluminum

Titanium

Diamond

Diamond

Banned

Titanium

Stainless

Diamond

Diamond

Diamond

Titanium

Diamond

Diamond

Titanium

Similar threads