Incoming QC on screws 2

[limct]

My company is buying millions of screws every year. Occasionally, there were some "rejects" found on the incoming lot. My incoming QA only do samplings of 3 screws per lot and if one is found to be out of specs, the whole lot will be rejected.

For an instance, per EN ISO 2702, ST3.9 tapping screws, specs calls for core hardness 270~390 HV5, and when any one of the three readings is found out of this range, the lot is considered as reject.

I wonder if this is the correct way. Since variability is inevitable (material, operator bias, apparatus calibration and method of testing), should we consider a statistical solution by calculating the average and std deviation and judge whether the probability of failure to happen under this given scenario is acceptable.

Thank you.

Best regards,
ct

 

[PSE]

Sounds like you want to do a Gage R&R on your test equipment in order to see if the results given from its use is acceptable. The Gage R&R study will help you to find sources of variation within the gage, users, and components. There are numerous sources of information on how to select and conduct these studies.

Once you know the gage is good and the operators consistent with it's use, then you know you should have good measurement data to make decisions with.

Regards,

 

[MadMango]

You could also put this on your vendor, and have them supply you with a Certificate of Conformance (COC) to the spec you wish the hardware to meet. Then you won't have to do any inspection if you didn't want, just keep the COC on file for each lot.

Ray Reynolds
"There is no reason anyone would want a computer in their home."
Ken Olson, president, chairman and founder of Digital Equipment Corp., 1977
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[unclesyd]

If you just need a paper trail the COC or anything similar will do.
If you need to insure a quality fastener you are going to have to do some testing at your end to insure quality. It's sad to say but money rules. Doing a little work for a large fastener ditributer it's not a week goes by that we don't have some issue with quality. It is generally with the large quantities not the smaller lots. The last one was large order of A193 B7 studs and all-thread where about a third of the studs were not stamped B7 and were comingled with the stamped ones. The MTR was for one lot of material of 10,000 lbs, yet this order weighed 12,000 lbs. We are still unraveling this one.

Where I worked we had an acceptance program for all fasteners. There were two areas that required very tight controls on the fasteners, H-11 HHCS and 316/304 SS bolts and studs. Our quality control people set up an acceptance program with stores receiving where our group got the paper work and a sample of each box or lot. The purchase orders had always specified no comingling of fasteners, in other words there was traceability even before the fastener act in the US. All fasteners were held until released by the lab.

Having seen both sides of the fastener issue if you don’t have some type acceptance program at your end I think you or your represenetive should visit your supplier at irregular intervals. This is even a bigger problem at the present because of various material shortages. There is no way that I would run my end based on the paper work of an outside source without some internal controls.

 

[unp1]

How did your QA decide to sample 3 screws? Did they follow any sampling plan? If not than they should follow a sampling plan like MIL-STD-105E, ASQZ1.4/1.9. We are using MIL-STD-105E. We have more than 100 different parts. Our lots are ranging from less than 100 pieces to few millions. If we use general inspection level than our sample size becomes unmanageable for larger lots. So we use special inspection level which increases risk but reduces the sample size.

 

[limct]

Thanks everyone!
Through some friends, I too come across the MIL-STD-150E. I'll have this topic discussed soon.
Taking 3 samples for decision making is just not tied to any norm, it's our company "modus operandi".

Best regards,
ct

 

[leanne]

A sample of 3? For what lot size?

Mil-Std-105E was cancelled as part of the Perry Initiative (paperwork reduction & getting rid of military standards in favor of industry standards - another of those bright ideas by Congress - we still get contracts referencing MIL-Q-9858 or MIL-I-45208 even though both were cancelled with ISO9001 as suggested replacement). Suggested replacement is the ANSI standard.

Military standard versions of sampling plans may be downloaded free of charge online at

http://www.variation.com/techlib/standard.html

ISO documents and ANSI documents must be purchased.

A Certificate of Conformance is not generally acceptable as objective evidence in an audit (IMHO, a CofC is not worth the paper it is written on - I frequently see non-conforming material which has been shipped in with a CofC proclaiming it met the PO requirements when it did not). An auditor wants to see test data. I would not rely on a CofC to justify elimination of inpection. I would want historical data showing past inspection results that would justify skipping inspection.

 

[quadswift]

We have been here 'n' times before on this site, on specifying, levels of inspection and material certification / verifiction.
I always believe in dealing with any problem as near to the source as possible. You checking stuff to this level on receipt is not really acceptable.
You need to
- know your level of expectation of conformity
- specify the level that you expect
- specify how the supplier will verify / certify that level.
(see also 6 sigma query in this thread)
Assuming your specifying is ok, then the source of the problem is not in your receiving area, it is with the supplier.

I agree with the comment on COC.I have seen evidence of copies of cut-and-pasted docs.
However, even the 'sale of goods'acts say that, for any commodity, you are entitled to recieve what you asked for!

If you look at BS EN 10204, you can ask for various levels of 'certification'with various levels of verification. For critical purposes you can specify the levels of testing by the supplier and either witness them yourself or organise an idependant third party to witness and certify on your behalf (excellent if it is on the other side of the world).
The last thing I want to do is set up in-house systems and teams of people to check that I have got what I ordered.
Regards,
Quadswift

 

[philyang]

I agree with Quadswift, it's better you build an IQP (integrity quality plan), first ask your supplier issure certificate and demand your suppliers do intertnal quality control (ISO 9001), secondly ask a independent third party (such as SGS) perform PSI (pre shipment inspection), thirdly, do incoming material test in your end.
keep contact with your supplier and third part inspector, act prompty while promblem found is very important.