Proper way to check perpendicular when can't loc on datum 1

[Dave1bart]

I recently recieved a messsage from QC about a perpendicular check they were having trouble with.
This turns out to be an ASME Y14.5 issue. The print pointed to the threads and said perpendicular to a datum face within .0005 (in geometric block form). Our inspection has a Johnson gage locator stand and can locate on the threads. They can then sweep the face with an indicator. The problem is, this is backwards from the ASME callout, but we often have to check backwards. But this creates a problem because the threads are .450 long and the face is 8 inch dia. So instead of checking over the .450 length we are checking over an 8 inch length which magnifies any error. Does anyone know of any document that supports (or debunks) my position that they should be checking to a tolerance of .0088 when indicating the 8 inch face instead of indicating the thread as geometric block says?

 

[smurali1]

Dear Dave1bart,

1. When the perpendicularity is called out in thread you should be checking on the axis formed by thread when the part is located against locating surface. Your current checking method seems to be not right.

2. Generally the perpendicularity of threads are measured closer to thread starting length and not on projected length. If designer's intention is projected length, he should specify with a modifer (p).

Thanks

 

[Dave1bart]

Item 1. You are correct, it is not the correct way to check it and that is what creates the checking problem. I have seen a lot of inspectors check backwards and assume there is a one-to-one conversions of tolerance. Unless to the tolerance is in degrees, perpendicular is not one-to-one when checked backwards.
Here is an example. Suppose you have a part similar to a 1-2-3 gage block. And supose the designer used a geometric tolerance of perpendicular .001 pointing to the "1" side and referencing the "2" side. The correct interpretation is to set the "2" side on a surface plate and indicate the "1" side using an Indi-square. When you do this you only indicate over the 1 inch distance. Suppose you get a reading of .0002.
If you were to check backwards locating on the "1" side then you would check over a 2 inch distance thereby doubling the error reading to .0004 (ignoring surface irregularities).
My question is is there any document that describes the adjustments needed or is checking backwards just not allowed?
Item 2. I agree

 

[smurali1]

Dear Dave1bart,

Sorry for late reply.
But could not find any documents to justify reverse calculations. May be we should measure in the intended way only.

Thanks

 

[Dave1bart]

I agree that we should check the correct way but, I cannot find any documents that say we can't check backwards.
This issue keeps coming up here. I do print checking here and we issue in-process sketches where we show the location surface and what to machine at this step. I just got into an arguement because the other Engineer called out runout by attaching the runout to the location surface! When you do this, this reads "hold locator to machined dia". Well, you are not machining the locator so this is not logical. Additionally many geometrics depend on the indicated surface be of a certain length. If you reverse them, you are changing the desiger's or manufacturing engineer's intent.
I need some documented help. I am doing nothing but making everyone mad at me for telling them they are wrong. Can anyone help!?