Difficult profile to probe

[[Er...]]

I have a profile inside a bore that is .05mm from A to B along a cylindrical surface. It is a plane to a 45 degree cone to another plane, where the distance from plane to plane is .21mm and connected with a .127mm radius on each end of the cone. They are asking me to come up with a way to check this, and currently all i have is a 1mm probe on our Contura G2 CMM. I'm not even confident a .3mm probe with be small enough. Any ideas or suggestions?

[[Cl...]]

Section (EDM) the part and check it with an overlay.

[[Mi...]]

Facsimile/reprorubber and optical comparator?

[[Cl...]]

We do have a MarSurf that works well for stuff like that.

[[Je...]]

.

Eric,

I agree that optical inspection may be the most feasible choice. If destructive testing is allowable, then sectioning the part, ideally with an EDM, and then using a an optical system like an O-Inspect could do the job. You could ask if this would meet print or if you need to capture the entire 360° area of the controlled feature. I don't see an "all around" symbol, so technically it's just at the depth of the print section view.

Outside of CT scanning (Zeiss Metrotom) or feather probing, which Zeiss does not offer, I don't see another option. I'm not optimistic that facsimile would be successful in simultaneously capturing the datums and controlled feature. And even if you could probe the datums and controlled features with your Contura, the various RDS angles required would blow your uncertainty budget to pieces for a .05mm tolerance.

I'd love to hear updates on how this assignment turns out. Very challenging indeed.


Jeff Frodermann
Meier Tool & Engineering
Anoka, Minnesota

.

[[Er...]]

One way to proof this would be have a wire EDM cut a rectangular block. Preferably with half of a cross section of this shape. That way you could try to "reach in" with your probes. Then try to measure it.

[[Jo...]]

I would open the Curve feature and section the area and scan as one continuous line between A and B, then report profile of a Line, if allowable.

[[Ro...]]

I would do what Joe said, just scan Free form line(s) from a to b.
Or maybe create circular section cuts on the low surface, the high surface, and right in the middle of your 45º feature and combine them into 1 Free Form feature.

[[Je...]]

.


Agreed. A cautious approach would be to make sure profile of a line is acceptable over profile of a surface.

Avoid using the shop bandsaw to section the part. The heat and forces from the saw would distort the part.


Jeff Frodermann
Meier Tool & Engineering
Anoka, Minnesota


.

[[Er...]]

Thanks to all for your replies. Unfortunately EDM or destructive testing for inspection will not be possible. The part has 4 of these sections, and if I recall correctly this customer has 100% inspection requirement on anything .05mm or less. My first initial approach will be to scan a line working from point B to point A. I'm just not sure how well points will measure along the 45 deg surface with the depth of the 2 planes so shallow. I'll have to do in 3-4 places for surface profile, I don't think a single line will satisfy the customer as a surface profile. I'll update this more when we actually have a physical part machined. But not have any training with Curve, this will be one of the biggest challenges yet. Not to mention the customer is planning to come visit and watch the machining/inspection process as well.
I didn't even mention that there is a bore on the opposite end of Datum -D- that has to be in position within .025mm. With Datum D less than an 1" long, we fear i will be dealing with projection error with this as well. This will be interesting to say the least.

[[Er...]]

Please sign in to view this quote.

Clark, we do have a marsuf for surface finish. What model works well for checking profile like this? how would you set this up?

[[Je...]]

.

Eric,

I'm signed up for curve & freeform courses in February. I have been working with curves for six months now, and I have learned a lot since I started. We have a ZeissCare agreement, so I generate a support ticket almost every week to ask curve questions. My questions are about specific measurement problems, but the amazing Zeiss application specialists usually give me 10-15 minutes of general curve & freeform instruction per phone call. I have them remote into my computer, which helps a lot.

Anyway, you are welcome to call, email or text me anytime if novice advice on curves would be useful (can provide contact info via pm if necessary).

About the projection issue you mentioned, I would highly recommend constraining Datum D to Datum C. There are a few different ways in the software to accomplish this. You could run this type of constrained datum setup by the customer, but it appears to follow the form and fit of the part. Although position is an acceptable control for the feature you mentioned, I think runout would be more appropriate and also provide information on the orientation of the bore.


Jeff Frodermann
Meier Tool & Engineering
Anoka, Minnesota

.

[[Er...]]

Please sign in to view this quote.

To help with the projection error I would scan circles and lines on -D-.

[[Er...]]

Please sign in to view this quote.

We have zeiss care as well so maybe I need to make more calls as well. I appreciate you offering the help, I may indeed reach out to you. Thank you.