Tensile made critical characteristic, but Cpk? High rate of rejection in a part

S

Salman

We are trying to resolve a very high rate of rejection in a part. We know that tensile strength of a small ring used in the part needs to be maintained within spec.
We have pursuaded our supplier to send us tensile certs with every shipment. In order to include this requirement somewhere in our system, our (Design) Engineer made the tensile on drawing to be a critical dimension. (Even though criticals are mentioned on drawing to mention criticality, and not to get certs).
Now our supplier says that since we have made tensile a critical dimension, they are bound to do Cpk on tensile, every time they run the parts in furnace ... which is not possible (min. 30 readings would be required).

My understanding was that even though tensile is critical dimension on drawing, we do not need them to do Cpk every time, rather they only need to do it once during PPAP. Sadly (for me), this suggestion was knocked down by the supplier's Quality representative. He said that as per AIAG manual, every characterisitic with critical on it, HAS to have Cpk done everytime it is run.

Unforunately, I don't have AIAG manual with me right now. And he does seem right in this statement. Still I think it is weird that we (customer) are BOUNDED by some manual to have Cpk done on a dimension, EVEN IF WE DO NOT NEED IT.

What do you guys think about my suggestion to the supplier?

---------

By the way, in the end I pursuaded the Design Engineer to take off the critical from tensile, because critical is put on drawing for functioning, and not merely to get certs. He then added a note on the drawing saying something like tensile certs are to be accompanied with each shipment.
 
B

Bill Ryan - 2007

Wow :eek:

To start - you, as the customer, are not bounded by any manual. The B3 and some tier 1s "override" the manual because the three of them can't/won't agree on what they feel is necessary for a quality product.

There is NO requirement in "the manual" for "Cpk studies" being done on a critical characteristic. The only place that would be called out is in the Customer Specific Requirements (no matter the tier).

I don't know what material the ring is made of, but when we need to tensile test a zinc or aluminum die casting, there is a ton of work done on processing parameters so a destructive sampling can be performed to validate the process setup. There are subsequent samplings to assure the setup parameters haven't shifted undesireably, as well as, to minimize the risk of scrapping a lot of product on long runs.

Not trying to be a smart:ca: , but why are you letting your supplier dictate to you? I guess I can see where your supplier is coming from, though. In my mind, the tensile cert with each shipment (whatever that is and does) could be handled in the P.O., Supplier Agreement contract, or whatever, and not on the print. I have always raised an eyebrow :rolleyes: when a customer decides something is "critical" after we've been in production for a year or ten.
 

Jim Wynne

Leader
Admin
Bill Ryan said:
Wow :eek:

To start - you, as the customer, are not bounded by any manual. The B3 and some tier 1s "override" the manual because the three of them can't/won't agree on what they feel is necessary for a quality product.

There is NO requirement in "the manual" for "Cpk studies" being done on a critical characteristic. The only place that would be called out is in the Customer Specific Requirements (no matter the tier).

I don't know what material the ring is made of, but when we need to tensile test a zinc or aluminum die casting, there is a ton of work done on processing parameters so a destructive sampling can be performed to validate the process setup. There are subsequent samplings to assure the setup parameters haven't shifted undesireably, as well as, to minimize the risk of scrapping a lot of product on long runs.

Not trying to be a smart:ca: , but why are you letting your supplier dictate to you? I guess I can see where your supplier is coming from, though. In my mind, the tensile cert with each shipment (whatever that is and does) could be handled in the P.O., Supplier Agreement contract, or whatever, and not on the print. I have always raised an eyebrow :rolleyes: when a customer decides something is "critical" after we've been in production for a year or ten.

Good response:applause:

The only thing I would add is that there is a PPAP manual requirement for "...process capability or performance to be determined to be acceptable prior to submission for all Special Characteristics..."* but there's plenty of wiggle room in satisfying the requirement.
And one more thing: What you want the supplier to do, as Bill suggests, is control the process parameters that affect tensile strength. The supplier should be able to demonstrate an understanding of the relevant variables and that the variables are controlled to the extent that continual destructive testing isn't necessary.
*AIAG PPAP manual, 4th Edition, page 7
 

Wes Bucey

Prophet of Profit
Bill Ryan seems to have covered the Cpk issue pretty fairly.

In my experience, heat treating parts can be affected by whether they are merely tossed in a pan or individually racked before putting them in the oven. Physical location within the oven chamber can also affect the heat treatment. Racking is normally only done for parts which need to avoid "dings and dents." Heat treating is done in batches which can have variables of temperature, time, (and in large operations) whether the batches are done in the same oven.

The requirement for a "critical characteristic" of tensile strength can probably be met by judicious sampling (for destructive tensile strength tests) from various locations within the pan or from various locations on the rack. A Cpk centered around the mean is probably not necessary, since the requirement is met regardless whether the parts are ALL at the bottom of the tolerance range or at the top.

The problem is compounded by the unhappy fact that all folks do not mean the same thing when they use the term "critical characteristic," despite the classic definition. If this were MY problem, I would get the customer engineer and the heat treater metallurgist in the same room (or on a conference call) and hash out exactly what is necessary to meet the customer requirement. (Remember about the word "ASSUME!")
 
K

Kevin H

Both Bill and Wes are bringing out some good points. Additionally, if the small ring is quenched and tempered steel, Rockwell hardness often correlates well with tensile strength, is a much less destructive test, and is much cheaper than tensile testing. You could potentially require that representative parts from the top and bottom of the corners of trays and the center be tested for Rockwell hardness to check for uniformity of heat treatment.

Or if more of a continuous process, from the start, middle, and end of a run.

I'd seriously check/discuss with the supplier their heat treating operations - have they done a temperature uniformity survey of their furnaces, are they controlling the temperature of their quench media? What about controls on atmosphere, assuming they are using one?
 
S

Salman

Bill Ryan said:
Wow :eek:

To start - you, as the customer, are not bounded by any manual. The B3 and some tier 1s "override" the manual because the three of them can't/won't agree on what they feel is necessary for a quality product.
It was my feeling too. Thanks for clarifying.

Not trying to be a smart:ca: , but why are you letting your supplier dictate to you?
Well, its the combination of being a very old supplier and a young customer quality engineer. But they can't dictate. It's just that I was confused at that time.

I have always raised an eyebrow :rolleyes: when a customer decides something is "critical" after we've been in production for a year or ten.
Bad design. It was a new design, and is clearly a failed one.

Jim Wynne:
What you want the supplier to do, as Bill suggests, is control the process parameters that affect tensile strength.
We are pretty confident that the supplier can control the process parameters. They have made changes in their process, and are very cooperative ... even when we are modifying the design every quarter.

Wes Bucey:
In my experience, heat treating parts can be affected by whether they are merely tossed in a pan or individually racked before putting them in the oven.
The good thing about our case is that our supplier uses a continuous process for heat treatment. The rings fall on a conveyor belt, and the belt moves through the oven.

Kevin H:
Additionally, if the small ring is quenched and tempered steel, Rockwell hardness often correlates well with tensile strength, is a much less destructive test, and is much cheaper than tensile testing
The problem is the small geometry of the ring. The maximum internal diameter is 0.413". Don't remember the cross-sectional diameter. But it is small enough that taking hardness value does not give reasonable confidence.

----

Thanks guys for your excellent replies. It cleared many of my doubts.
 

Wes Bucey

Prophet of Profit
I'm curious. What kind of ring? a retaining ring like a Truarc or Rotoclip? Does the supplier make the ring as well as heat treat it? Whose design, yours? your customer's? your supplier's?

I'm very familiar with the processes for designing, stamping, and heat treating the Truarc-type ring, less familiar with the spiral ring (looks like two turns of a Slinky!) About the only thing that can screw up the Truarc process is an unannounced change in the BTUs of the fuel for heating the furnaces. Everything else is pretty cut and dried, plus they have practical, functional tests for such rings which deliver same info as tensile test at much lower cost.

For the customer to be making demands in confirming tensile strength may be an indicator that the design for the ring in place exceeds the functional limits of the original ring design causing failures wrongly attributed to the ring. Tell me more. I may be able to help you find and correct root cause. Sometimes the "error proofing" requires a very minor design change.
 
S

Salman

It is not retaining ring like the one used in pistons. It is made out of a wire, and looks like the letter 'C'. It is a very low-tech part, and that is a problem because we are asking a low-tech part to save us from potential warranty issues from our customer.

The supplier starts from a coil of wire, and end up with a small C-shape ring. Heat treatment follows the ring-cutting process. That is, each ring after being cut falls on a conveyor, which goes through a furnace maintained at 275 F.

The problem that supplier had earlier was that they were using a different furance, whose temperature could not be maintained, and looks like they did not have very good means of checking temperature either. Now they have changed the furance, and hopefully, there will be consistency in tensile strength (and hardness) from now on.

Since no one is listening from our company (hopefully), let me tell you guys that it is a pathetic design. I just had a training session on trade-secret/copyright stuff, and so I would not like to go in much detail. But I can say that there are many variables to be perfect before the part can function properly. Ring is just one of those variables.
Our guys have been working on this same issue for a couple of years, and have been getting warranty claims from our own customers, and yet still we have not been able to resolve it. Shoots our monthly PPM right through the roof. I for one get the feeling that even after all this, we have not been able to identify all variables responsible.

Anyway, we are planning to do a DoE soon. We hope it will give us better understanding. sigh.
 
Last edited:

Wes Bucey

Prophet of Profit
Salman,
I've sent you a private message about this topic. It could be worthwhile for you to follow up on that message without violating trade secrets.
 

Hershal

Metrologist-Auditor
Trusted Information Resource
You know.....one possible solution is to have an accredited test lab run tensile tests on a periodic basis under the ASTM or other applicable standard(s) and track that to make sure that you have consistently.....independent and possibly less money.

Just a thought.

Hershal
 
Top Bottom