Improper Heat Treatment?

[Lescombes]

Hello everyone,

*For the purposes of anonymity I am going to be deliberately vague on some details, and I appreciate your understanding in advance.*

I work for an aircraft manufacturer, and I have recently discovered that we may not, in my opinion (and I stress the in my opinion bit), have the required material traceability on an aircraft structural item.

This item is an aluminium part that is supplied to us in a semi-finished state. The design data for this part stipulates that its material conform to a MIL spec that includes the letters QQ and A, however the Supplier's certificate of conformance makes no mention of this spec, only lists the required chemical composition of the family of alloy and does not identify a material temper. I've had reason to contact the Supplier on sort-of related matter and just "threw in" a question about material conformity - and the response I received further furrowed my brow.

I have exercised my duty of care and informed the relevant departments and I expect that in the near future I will be asked what can be done about aircraft that may be in-service with one of these suspect parts. An example of the part was subjected to ultimate load testing during aircraft certification, however the part has not been tested to destruction that I know of. My first thought was that I could require an inspection of any part's hardness and conductivity, however research has led me to a paper indicating that there is very weak or no correlation between hardness or conductivity and material tensile strength for this family of aluminium alloy (which is in no way exotic, for the record).

Short of destructively testing a part from each batch ever supplied or replacing everything in service, is/are there any different ways of tackling this situation that I haven't thought of yet?

Thank you,
Lescombes

 

[tbuelna]

If the engineering documentation you must comply with for manufacture of this component requires raw material traceability based on a QQ-A aluminum spec, then that's what you need to comply with. QQ-A aluminum material is not mil spec, it is federal spec, which is slightly different. If your PO required raw material and heat treat certs, then the vendor's CofC should have confirmed the product delivered complied with this requirement. If your vendor did not supply the required documentation, then you need to take the issue up with them.

 

[rb1957]

what does your QA/QC department say about this ?

I guess you don't do receiving inspection (like most in the business) ... sigh.

The part is completed on-site. What is inspected then ? just dimensional changes ?
I'd expect the finished part is defined as "make from XYZ part" so the material definition may be hidden there. possibly you heat treat it, in which case the definition would be a QQ-A spec and it would matter less the definition of the in-process component (the more definition the better, but ...).

another day in paradise, or is paradise one day closer ?

 

[WKTaylor]

Lescombes...

"You deserve to get what You asked for. But if You didn't ask for it don't expect to get it."

Your PO should have been crystal clear regarding raw material certification [material Certs] and any follow-up inspections and destructive tests... which should be part of the delivery paperwork.

I have had reputable vendors forget to send [omit] key paperwork... such as certifications, destructive tests, NDI, etc... with the delivery PW... but thy always follow through immediately when they learn-of their error, since they track everything and finding copies of the lost documents was not a problem. This documentation is life/death for acft parts and materials companies and their ISO/AS and NADCAP aerospace business certifications.

However, small businesses and non-NADCAP companies have failed to meet this obligation time and again... and it can cause hell down-stream. AND the pressure to 'accept the parts' AS-IS, or with further 'testing', is enormous especially when delivery is time-critical. This happened to some critical engine mount parts from a sub-tier processor [heat treater, I think] to our prime-vendor: the E-mount parts ended-up being scrapped at great expense to the prime vendor and my company's reputation with our customer, which was in desperate need for the parts.

IF there is a honest-working ethics [business integrity] group within Your company, then they need to be informed. At very last resort, I believe that the FAA as a hotline for reporting fraudulent [bogus] parts. Unfortunately if this slips through, then the entire 'quality' system at the vendor and/or within Your company may be corrupt. This happened at the first company I worked for: an internal QA whistleblower notified corporate which opened and investigation on-site; when the story emerged, the buyers and QA agents involved were fired immediately; and then the incident was grudgingly reported to the FAA and law enforcement... which ratcheted-up the tension dramatically. people went to jail and were forced to pay 'enormous' fines [for the late 1970s]. This was my first/most critical aerospace ethics lesson!




Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[verymadmac]

So how to prove something is of sufficient strength when its gone out the door.
I would approach it this way (assuming at this point the supplier has become non-cooperative)

Rough out the cost of replacing the parts, workible solutions are anything that cost less than this.

So did the work to complete the part in house make any usable data, places to look would be
Concessions relating to this part (often extra stress analysis is buried in them).`
CNC / automated assembly used may have tool force etc logging as part of the maintenance system (you
might be able to collate to the material condition).
Did any of these parts end up in the Q store (No doubt the last batch is there, but some from the
earlier batch's might there if they got damaged).
How many of these parts uninstalled are left.
Is there any difference between the current part configuration and the original test article that could be used.
Does it look like there is much margin in the parts design, what is its critical failure mechanisms (you maybe able to
reduced the required material properties).

Taking an actual sample from each aircraft may not be out of the question. It doesn't need to be a standard size as its likely cheaper to calibrate for a odd shaped sample than a cutting out a bigger sample. Does the part need any improvements that could cut out a suitable piece like a new access panel (sometimes one is lucky) or a corner that could be chamfered or has a non-optimized area.

do you think they supplied the correct temper, wrong temper, or some random HT, if you are just comparing tempers conductivity might still work.

Is there anyway a test load could be applied to the item in the field
is it small, fish scales etc
if it is large can a local load be applied to an element of the part to show yielding & ultimate is sufficient. I am not sure how this would look except that maybe one could tap a bolt into the part and do a pull out test using a small portable test rig designed to limit the damaged area (although without thinking about it too much it this would appear to give you just shear strength, unless you damage a larger area) and then repairing the hole.

i hope there a few new ideas there for you.

 

[VEBill]

"...how to prove something is of sufficient strength..."

In my experience, the QA and Airworthiness folks would be interested in compliance to defined requirements and documentation of same.

The "sufficient strength" (or not) only impacts the formal Risk Assessment, and (if not) possibly restricting or grounding the affected fleet.

Disclaimer: I'm not an SME in exactly this, so stand by for others to confirm. Defer to your QA and AW folks.

Good luck.

 

[Lescombes]

Thank you to everyone for their thoughts.

tbuelna, rb1957, VE1BLL and wktaylor - I have been disappointed with the response from my QA/QC department regarding this issue, and I am considering exploring some of the more serious regulatory/legal options that everyone has described. This potential problem is actually keeping me awake at night.
In the meantime Mr Taylor has hit the proverbial rivet on the head (or tail?) and correctly identified the root cause being that our Procurement team have been far from clear on our outbound PO. I believe this to be a longstanding oversight which considerably predates my employment with the company.

verymadmac - I appreciate your providing some new ideas, and I'll throw them into the mix of things to consider. Problem I have is that it may be random heat treatment, it may be incorrect temper - I have no idea as to what sort of material properties I have and what sort of scatter there may be. There some older components lying around that I could test in-house, however I am wondering how much variation there may be in the semi-finishing process and whether tests of a couple of samples from a couple of batches would be enough to show the (literally) hundreds of components we have in service are 'good'? The suspect part is unfortunately a long continuous piece which may preclude some of your very nifty suggestions for surreptitiously sampling in-service components.

I know that our naturally inquisitive minds are frustrated at the lack of detail here, however the situation that Mr Taylor described as his first aerospace ethics lesson may yet play out within the company here and therefore I am reluctant to speak too far 'out of school' on a public forum at this stage.

I do thank everyone again and will update when I can.

 

[verymadmac]

If they overcooked the parts, this would likely show up as higher than expected corrosion events in service particularly SCC and intergranular corrosion. Customer support work in the form of repair schemes and complaints, etc is likely to give an indication if this is occurred.

 

[Sparweb]

Perhaps being devil's advocate here, but have you verified that the loads on the part (nominal, fatigue, limit, ultimate) justify your concern about the temper?
If you assumed the lowest possible strength in the material, would the margin of safety still be positive?

I have heard claims from makers of NDT equipment (ultrasonic, I think) that the temper of aluminum alloys can be determined. I have never challenged them to prove this claim, nor do I know how certain the result would be if I tried.

STF

 

[Lescombes]

Hi SparWeb and everyone,

Unfortunately the part was statically substantiated by test 'back in the day' however has had a safe life fatigue analysis performed. I have been working on characterising the internal stress in response to various load cases, and whilst a gross overload failure is unlikely for the part even in the O condition, the implications upon its fatigue life should it be O are hard to pin down at this stage.

So my company has, after much hand wringing, advised the applicable regulatory body and commenced internal investigations. We have also conducted a tensile test upon one part from the current batch of 25 that are in stock, which thankfully showed that the particular part was good.

The question I'm now facing is - how many tests are sufficient to at least prove the current batch is good....? which is more a rhetorical question that we're still pondering.

 

[WKTaylor]

Lescombes... my perspective...

1. Based on the unknown-unknowns: are You sure the correct aluminum alloy 'raw stock' was used? Are the original MILL-certs available and trustworthy?

2. What aluminum alloy heat-treatment spec: AMS2770, AMS2771, AMS2772, AMS-H-6088, ASTM B917 or ASTM B918 [was ASTM B597]... Other?

3. Situations like this mandates 100% evaluation of EACH PART per AMS2658 Hardness and Conductivity Inspection of Wrought Aluminum Alloy Parts. This includes parts already installed in structure.

Accomplish hardness and conductivity testing on multiple [widely separated] locations on EACH part.

IF the conductivity/hardness is narrowly consistent and are in the mid 1/5-to-4/5ths of the acceptable ranges and are approximately [reasonably] consistent from part-to-part AND location-to-location on each part, then likely these results are adequate to confirm parts are acceptable; and can be recertified.

IF readings vary widely within the min-to-max range, part-to-part and location to location, then worry. Destructive testing of 'worse case reading' [low and high] parts [~10% random selection] is needed. IF any tensile/elongation abnormalities are noted, then condemn the lot.

IF readings vary wildly/erratically within or just beyond min-to-max range, part-to-part and location to location, then worry A lot! Destructive testing of ALL parts with readings at/beyond range limits is needed. Depending on many factors, any inconsistent hardness/conductivity readings combined with tensile-strength/elongation failures should cause the lot of parts to be condemned.

4. WARNING

IF Your HT process varied from the original drawing requirement [old parts especially] then the HT/Forming operations in a specific sequence are critical to minimize adverse grain-growth and warpage.

On one acft I worked-on, many parts [sheet metals and extrusions] were hot formed in the fully heat treated state [1950s and 1960s]. After forming no other operations were required and the part had highly consistent grain and strength properties. Short term heat exposure and forming had minimal/no effect on the base alloy.

However, current vendors do not have hot-forming tooling and are generally ignorant of the process. The current flavor is to form in the 'O' condition then SHT/age to the final temper [straightening if/as needed]. Depending on how much straining occurs, the highly deformed areas can show adverse grain-growth and loss of strength/toughness. I always insist that hardness/conductivity readings be taken at these areas of 'high strain'... along with minimally-strained areas for comparison.

The CORRECT/BEST SEQUENCE is to SHT, quench and form/straighten in the [AQ/W] unstable temper; then HT-age to final temper. Consistent grain size, minimal warpage and consistent allowables are easy to attain with this method.


Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[tbuelna]

Since this situation involves an aircraft structural component, there are only two things to consider. If the vendor supplying this partially finished structural component provided a legitimate C of C confirming that the product delivered conforms to all of the requirements defined by your purchase order, then you have no cause to reject it. If the vendor delivered a non-conforming product, due to insufficient documentation or any other reason, then your receiving inspection should not have accepted it. In any case, you are ultimately responsible for the quality of any product you deliver to a customer, and not your suppliers.

You can always try submitting a disposition request to your customer asking them to accept a non-conforming product, but it is not likely they will agree when it involves an aircraft structural component.