Using SPC to determine overall Thread Quality

[oldqamanager]

Does anyone know of which kind of chart i can use to measure overall thread quality throughout my manufacturing facility.

We run about 30 different gauges, each of these gauges has its own threads ( pitch, minor and major diatemeters). My inspectors measure these parts and i do have SPC data for each gauge.

I have been asked if i can take the results of all of my inspections to roll them into one chart. Different part numbers, each part number with a different thread measurement.

How can i do this??

Thanks for your help.

Frank

 

[bobdoering]

Before you try to make a "data cloud" chart of thread characteristics (more of a fog), let me ask you what operations are you using to generate the threads?

 

[oldqamanager]

we are running the threads on a cnc machine.
and when these threads are assembled onto gauges, the final step in inspection
is to check the threads on the assembled gauge with a Johnson Gauge.

we have very few failures.

 

[bobdoering]

we are running the threads on a cnc machine.
and when these threads are assembled onto gauges, the final step in inspection
is to check the threads on the assembled gauge with a Johnson Gauge.

So far so good.

How many parts are run per part number set-up?

we have very few failures.

I would expect that would be the case. So, what I am trying to do is determine the correct methodology for the most benefit. It sounds like you want a report card chart that covers everything. That should be your last choice.

 

[Bev D]

It sounds like you want a report card chart that covers everything. That should be your last choice.

it would be my last choice too, but it sounds liek your management is asking for somethign exactly like this?

if so, a control chart of the reate of out of control events is probably the best approach. you could use a simple c chart and trend the number of OOC events per day.

but it may be more productive to enquire - in a positive helpful way - waht the requestor hopes to learn from such a chart. often teh requestor knwos what they want to know but doesn't know the best way to get that information.

 

[bobdoering]

Unfortunately, management usually has even less of an idea of charting methodology - therefore creating a wasted effort that further reinforces to some people that SPC is just a paper chase.

To jump to an endpoint, single pointed threads are a case of thread wear - so it is non-normal and traditional charts have no meaning as far as process control.

But, it could be further tempered with short run - which is why I asked how many parts. Short run in machining has an issue with warm-up. Any time a CNC is running and has not achieved steady state, it is not in a state of control - warm-up itself is a special cause. Often, if runs are short enough, it is never in control due to warm-up. It depends. But, that does not mean they are making out of spec parts, and the proper charting methodology still reinforces they are making good parts, and ultimately will help minimize overcontrol.

So, I really am trying to guide an SPC charting effort away from rubber stamping report card charts, because long term they are just shredder fodder, and undermine any real SPC implementation.

 

[oldqamanager]

Yes, Management, LOL.

We run variable size lots, however, i would say that the typical or average size run is approx. 150 pcs.

I appreciate your help. You are both correct! I was keeping this data on individual charts, but management as usual doesnt have time to review them, or listen to me, they want a "cloud chart" as you describe it. LOL

Frank

 

[oldqamanager]

Thank you very much BeV I appreciate the information
As one Taurus to another, ;-)

Thanks

 

[Bev D]

yeah it's a stupid request, but I've found that gently helping them down the path and providing them what they really want is more effective than what little Bev says (in my head most of the time) which is "what the hell are you thinking"?

 

[oldqamanager]

LOL, okay, thank you,
Ill give that a shot!!!!!