Design department can't tolerance analysis its way out of a paper bag.


Johnny Jewel

Here is a question that, although more design-related, definitely impacts the quality field in the area of Quality Plans, measurement, and effectiveness of design documentation. First, let me describe one way the organization I work for portrays the importance of drawing dimensions.

-- Critical Dimensions -- Dimensions determined to be critical to the part's fit and function. Critical dimensions will be indicated on the print by brackets or other indicators. These dimensions must meet a Cp of 2.0 and Cpk of 1.5.

-- Major Dimensions -- Dimensions that are toleranced beyond the standard block tolerance. This would include all GD&T dimensions.!!!!!!!!! (!!!!!!!!!!! mine)

-- Minor Dimensions -- Dimensions whose tolerances are not explicitly written on the drawing (non-functional dimensions) and whose limits are defined by general tolerances or as indicated in the title block area of the print.!!!!!!!!!!!!!!(Mine, again).

-- The last 2 must pass a First Article Inspection with Majors having 5 parts measured, and 4 of the 5 must be within 63% of the tolerance range.!!!!!!!!!!(Oops, those darn!!!!!!!!!'s again!). Minors have 1 part measured, and must be in tolerance.

-- (Regarding "standard block tolerance" and "title block area", our drawings have default tolerances something like 3-place decimals tolerance is ±.005, 2-place decimals ±.010, and so on.)

1.) Has anyone seen anything similar?
2.) Does anyone know the source of this?
3.) Can you tell I am not neutral on this method?

-- I believe that the designer, thru design reviews, drawing reviews, tolerance analysis, FMEA's, and other means should decide which dimensions are significant and at what level. These dimensions must all meet the same capability minimums the company has decreed as the goal, and the tolerances can take up the slack of how important each is. By the way, the "63% IT", as it has been nicknamed, has been put forth as being a Cpk of 1.33. It is, if the process is perfectly centered, and you agree that capability can be estimated with 5 parts.

-- Aaand, there is no hierarchical nature to characteristics described with Geometric Dimensioning and Tolerancing -- it is merely the way the designer has chosen to describe the particular feature, usually because it is the best way; most descriptive and most related to the function of the part.

-- Finally, without telling you the name of my company, I will tell you we are a leader in our field -- and yet our design department can't tolerance analysis its way out of a paper bag. They work strictly with nominals in the development phases and throw in the standard ±'s with little thought. It is rare that prototype parts are even measured to see "what worked". Oh, and they also try to designate as few Critical's as possible so that meeting those pesky Process Capability targets won't hold up release of the product to the producing plant.

-- 4.) Are these attributes common?


Trusted Information Resource

I'll answer question 4 first -
I Won't say common - but standard. I have yet to see an engineer who will classify characteristics as critical, major and minor.

As for your other question, I can tell you my experience years ago working for defense contractors.

DoD drawings use to have a reference sheet which listed critical, majors and minors. Quality requirements were as follows
Critical - 100% inspection
Major - Sampling inspection performed using MIL-STD-105, Genreal Inpection Level II, .65 AQL
Minor - Sampling inspection performed using MIL-STD-105, General Inspection Level II, 2.5 AQL
Usually anything not specified as critical or major was considered minor.

I don't know if this information will help you. But without a doubt, good luck trying to get engineering to define these for you.

Top Bottom