Needing help with a Profile

[ctderryberryNASG]

We have a new part with a profile call out. I have attached a rough drawing of the part with the datums labeled. The green area is the area the customer wants the profile measured. The red lines are the three areas to check the profile. We have a program we are running now that takes points (8 points per section) in that area to check the profile but I believe there is a better way of checking it. Does anyone have any insight on a better way to set up a program to check the profile. Thanks in advance.
438 Drawing.png

[louisd]

Per your use profile, it appears you are running a microval, which is a manually operated (no DCC) touch-probe (probably TP2?), correct?

-The smaller the probe you use, the less cosine error you will induce from not hitting normal to the surfaces.
-Load CAD (if you can). Use CAD to obtain nominals.
-Take more hits, but they need to be as normal to surface as possible.
Aside from this, there isn't much extra that can be done.

[ctderryberryNASG]

The CMM does have DCC. I have the 3D CAD loaded into PC-DMIS. If I am reading it right you are saying use the same program but take more than 8 hits per section?

[louisd]

Ok if you have DCC, take a butt-load of additional hits. As many hits as time permits.
If you had a scanning probe, you'd likely be taking thousands of points on that surface.
You can construct a grid, patch or UV scan using TTP if you wanted to scan a grid pattern of hits instead of just 3 rows... unless those 3 rows of hits are a customer defined method.

Also, if you have DCC and CAD, depending on surface roughness/finish, you might want to select the biggest probe you can fit. You will be confident your vectors will be accurate, so there's no concern for cosine error.

[ctderryberryNASG]

I'm using a 2by40 probe. The area I am checking is 45mm x 35mm. The part is stamped out of a coil of steel. Still think I need a larger probe?

[louisd]

No, surface finish is smooth with a stamping, 2x40 is fine.

[ctderryberryNASG]

I guess I should also mention that Datum A is a plan, Datum B is a hole and Datum C is a slot.

[ctderryberryNASG]

Yes. I have the CAD aligned with the part with the datums defined.

[louisd]

ok so why mention this? Aside from looking for tips on how to improve profile measurement, which I've provided, what else are you seeking?

[ctderryberryNASG]

That's all I am needing....an improved way to to measure that profile.

[ctderryberryNASG]

This maybe a dumb question but if datum B (the hole) is out of specification and/or datum C (the slot) is off center could it cause the profile to be out of specification as well?

[louisd]

yes, if you can prove alignment to outer edges of part for example, would pass the profile callout... You can provide that to manufacturing team, it should help them to better understand that the defect is from the datums, and not the feature being measured.

From an engineering standpoint, if B or C datum locations are the reason for the nonconforming profile dimension, and B/C are not failing on their own defined location tolerances, then the drawing is not effectively controlling the datums.

[ctderryberryNASG]

Thank you. That is what I needed.