Surface profile questions

[d.evans]

I have a workpiece with a large area "hogged out" to form sort of a swimming-pool shaped feature set... Several planes, cone-shaped tapered corners, and cyl-shaped corners...

There is one callout for the surface profile. It is referenced RFS to the main datum framework [.010|A|B|C].

What I've done is for each "feature" that comprises this surface, I measured them as I always do, using auto-planes, auto-cyl for corner radii, auto-cones for tapered corner radii, and 3D lines constructed from the intersections of the planes involved. After all the individual features were measured, I constructed a "Constructed Set Feature" using all the above measured features. Then I dimensioned that one feature against the callout (+/-.005 form and location) and it comes back .0005" out tol.

Am I going about this the right way?

[kbotta]

I prefer to use individual points and report the 'T' val.
KB

[Matthew D. Hoedeman]

OR, measure with vector points and construct a set with those. I would not try to do it with planes, cylinders, cones and the like, just single points. You can take enough points on a cylinder shaped feature so that you COULD construct the cylinder, the same with the cone, ditto the plane, ditto the line, but I would only use the individual points for the profile callout.

[george frick]

1 thing i noticed first off is this is an example of a unilateral profile tolerance. do you notice the phantom lines on the inside of the pocket. this means the tolerance applies to the inside of the pocket only. in other words your tolerance is .010 to the inside of the pocket only not +/-.005. see if this helps. as far as the way you checked the part it seems to be correct to me. just redimension to the inside .010 and you should be ok.

[BSchwark]

Ditto on the points for profile ....................


BS

[george frick]

i agree using points unilaterally though.

[d.evans]

I checked with the engineer on that one, and he intended it to be +/-.005. We both went back and looked at the y14.5 1994 spec and it had a drawing with two arrows pointing to both ends of a unilateral callout tolerance zone. He's only got one arrow, which to him, means bilateral.

[kbotta]

Pg 164 illustrates this very well (asme Y14.5) top of fig. 6-11
And I believe your eng would be correct.
kev

[george frick]

interpret the dwg

I checked with the engineer on that one, and he intended it to be +/-.005. We both went back and looked at the y14.5 1994 spec and it had a drawing with two arrows pointing to both ends of a unilateral callout tolerance zone. He's only got one arrow, which to him, means bilateral.

the machinist probably interpreted as unilateral (to the inside of the pocket) as shown on the print. if you are showing .0005 o/t he machined it right on nomimal .0055. LOOK AT SECTION
A-A clearly unilateral

[d.evans]

How far apart should I space the V points? I am currently spacing them approx .5" apart. I'll be here all day typing in X, Y locations. Is there a quicker "automatic" way of doing theses Vector Points?

[kbotta]

Just because the arrow is pointing there, does not make it a valid unilateral. I would agree that it is 'unclear' as to the intent - with the chain line and all. In any case (bi-lateral, or unilateral) I believe the designer should have been more clear.
And just think. Soon this will be different (circle U...Maybe).

KB

[kbotta]

As many as you need to accuratley define the area (I know it is a canned response - but the right one...)
Kev

[d.evans]

Yes it looked clear to me, but I asked anyway. I was surprised the answer he gave me.

[kbotta]

Which begs the case "why the chain line"?
Is it a part outline (does not look like it to me...but it's your part).

He should fix that. After I 'zoomed' in on it, the arrow is directly pointing to the chain line. Very unclear.
KB

[d.evans]

I think he was surprised that I asked. I don't think he respects the job I do very much. In fact, he has dubbed my CMM "the NCR Machine" (NCR=Non-Conformance Report). I think I caught him off guard a little, and he HAD to be right.