Correctly using Position with Shadowgraph Inspection

[TopPocket]

Hi so I have some parts I wish to define. ISO please.

The requirement is that the surface of them must fit within a volume defined by a datum axis. I figured as MMR must be used I will have to use Position rather than Runout or Concentricity.

Now to inspect these I will use a Shadowgraph. They will be rotated with the datum feature in a V block. The horizonal line of the crosshairs will be used to find the highest point along the edge through the rotation. The same will be done to find the lowest point and the difference reported.

I know MMR will not be simple to apply but to be honest I expect to not have to use it except for a few edge cases. It's just there if I need it.

My question is when I define position on the diameter I usually come off the dimension, but doesn't this imply it's to the axis. How do I draw it to indicate it is to the surface or does it not matter? Will this impact whether I put a Ø on the tolerance?

 

[Burunduk]

The horizonal line of the crosshairs will be used to find the highest point along the edge through the rotation. The same will be done to find the lowest point and the difference reported.

Sounds like your intent is runout control, not position.

 

[Qcprick63]

Wouldn't TP be 1/2 the value ? R/o would be the entire value ( high to low ).

 

[Burunduk]

A position at MMR would impose a different evaluation method.

You would first need to calculate the maximum material virtual condition (MMVC) which would be the sum of the maximum allowed diameter of the measured feature and the position tolerance specified to apply at MMR. For example, if the diameter is specified as ⌀10±0.1, the MMVC in your case is 10.1+0.2=10.3.

Then, you know that the maximum radial distance from the datum to the most extreme point on the feature is 10.3/2=5.15.

A pass/fail measurement procedure could be as follows:
Set the vertical zero when horizontal crosshair line is at the datum axis. Move the horizontal crosshair line either to +5.15 (up from the datum) or -5.15 (down from the datum) and rotate the part at least one full revolution (360°), checking that no point on the feature passes beyond the crosshair line in the process.

If you want values to be reported, you would record the location of that actual extreme point and make a calculation based on the required MMVC and your measured (recorded) value. I'm unsure how this calculation should be done per ISO.

 

[TopPocket]

It does indeed feel very runouty but I didn't think you could use MMR with Runout.

Your assessment of the inspection method is very thorough. Thank you for taking the time to consider it.

The problem step is setting the shadow graph to the datum axis. The datum axis will be hard to pick out. However, seeing as the datum axis is the axis of revolution then I think I can measure the difference between the highest point reached on the top and the lowest point reach on the bottom through a full rotation. This will give me the maximum envelop the part occupies so combined with your method of define the tolerance limit should work. Right?

Unfortunately, I've actually found a further complication. The MMR would actually apply to the radius of the part, not the diameter. Do I need to call out the radius of the part on the drawing and apply MMR to that or is there another way to apply a factor of a half to the MMR on the diameter?

 

[Burunduk]

However, seeing as the datum axis is the axis of revolution then I think I can measure the difference between the highest point reached on the top and the lowest point reach on the bottom through a full rotation. This will give me the maximum envelop the part occupies so combined with your method of define the tolerance limit should work. Right?

Yes, I agree. And as long as that occupied envelope is smaller than the MMVC the feature is acceptable. That way, you don't have to set a zero on the datum axis. The distance between the locations of the extremity point at its detection (when it's all the way up there) and when it's rotated 180° (when it's all the way down there) is the envelope you compare to the MMVC.

I don't get the bit about MMR on the radius. The MMVC is a diameter, and the measured envelope is the centered-to-datum diameter the actual features occupies. Even when I mentioned moving "the horizontal crosshair line either to +5.15 (up from the datum) or -5.15 (down from the datum) and rotate the part at least one full revolution (360°)" that was because you would be setting a zero at the datum and moving half the MMVC diameter to get the crosshair to the theoretical outline of that virtual diameter.