MBD and MIN/MAX Dimensions

[EDCSRPM]

I've read many threads touching on the use of Min and Max for product definition. There seem to be two camps, one that see no application value, and one that see it as useful in limited circumstances. Personally, I see the application on Min and Max dimensions as very necessary part of complete product definition. The min and max provide a way of conveying a refinement of other toleranced dimensions while avoiding a double dimensioning situation. Just like GD&T refinements that set up a correlation of available tolerance with the as produced feature of size, I see the min and max fulfilling the same for toleranced dimensions.

Take the following example, an axisymmetric part consisting of two cylindrical journals. Each cylindrical portion would have a toleranced radius suited for a specific interface fit. The step from one cylinder to the next would be stacked by combining then feature of size limits with any form controls for centrality. What if there were an engineering requirement for a maximum step. This maximum requirement being less than the stack. Would we limit one or both journal limits to something tighter than the interface fit, or normal manufacturing process capability in order to achieve the desired step? Would we add a third unilateral tolerance to the step?

Many have suggested that the answer is to provide a bilateral tolerance zone, but wouldn't this result in double dimensioning and an interpretation issue relative to inspection acceptance?

It seems to me that the goal is to communicate to the manufacture that they must control the variation between the first and second operation. If the first operation favors the minimum side of then journal tolerance band then the second operation must follow on order to achieve the maximum limit.

Thoughts?

 

[3DDave]

Picture?

I really don't know the reason against 'double' dimensioning. For certain there is double tolerancing, applying multiple tolerances with respect to multiple feature control frames, so why not double dimension?

My guess is that manufacturing wants to push design to make a one-shot process to manufacture the part, but the drawing is for inspection and either the part meets all the requirements or it doesn't. If the drawing is such that no part can meet the requirement, that's a problem that is independent of dimensioning technique. More important, if the drawing describes a part that is not useable that is a bigger problem than having extra dimensions (not acceptable) or extra tolerances (completely acceptable.)

I've gotten pushback on assembly dimensions. If I have two pieces welded end-to-end, I've been forced to remove an overall length dimension. This despite the weld causing the length to be greater or smaller than the sum of the original lengths, depending on weld technique. The kicker? I would make the parts longer and have the weldment finish machined, but the shop feels it is too costly to run the parts through the machine shop twice. Result - much more expensive design and more parts to account for the much greater variation that is not controlled at the weld operation. It's ok. The contract is fixed-price.