When to record values if a spec for a mfg process is set

N

NWG75

Workplace dilema; if a spec for a mfg process is set, (for instance, maximum 10 lbs pull force) is it necessary to record a value, as opposed to just P/F? The process is not validated, but is 100% verified. Obviously recording a value allows you to run stats, run SPC, etc. But if we KNOW that 10.2 lbs of force will damage a part, and 9.9 lbs won't, can we just say it passes?

Thanks as always.
 

optomist1

A Sea of Statistics
Super Moderator
generally speaking, continuous data, i.e. 2.5, 3.6, 60.9...is superior to P/F or attribute data...if affords you the opportunity for greater analysis and insight into what the process is doing...where it is heading...provides you the ability to spot negative trends BEFORE part are rejected
 

Pjservan

Involved In Discussions
If that is the case why record the values? Why not validate the process and understand what factors contribute to it so to know when to revalidate it. It looks to me that you are doing an inspection because someone thought that it was a good idea to carry out the inspection or in order to say, we do 100% inspection/testing. If you want to continue testing you can say that as long as the part meets 10.01 lbs then it passes. Why bother collecting data.
 

Ninja

Looking for Reality
Trusted Information Resource
If I collect continuous data, I can analyze it later...

No part has ever failed, so let's just test 10pcs per batch...risky
No part has ever failed, and no part has ever been above 2lbs, with 95% below 1lb. Far less risky call to reduce testing qty.

Last year, the highest was 5, then every batch since then has crept up more and more and we are now running 9.85...let's fix this before we have a whole batch fail.

Those are just two of the many things continuous data can give you toward cost savings that a P/F test cant give you.

P/F can work for the test itself...continuous can work much better in the larger picture...
 
Top Bottom