GAMA and FAA: in talks to explore changes to Part 23 Certification 4

[WKTaylor]

GAMA & FAA engineers have started discussing changes to the Part 23 aircraft certification process/requirements.

These changes could dramatically reduce the costs of certification by stratifying aircraft more along lines of complexity and/or weight and/or capacity. The lighter/less-complex/lower-capacity aircraft would face less challanging certification hurdles that are costing millions of dollars... which spawned the LSA catagory [low speed, 2-place, day-VFR].

Listen to the GAMA director of engineering and manufacturing describe what has been initiated... and hope that sanity prevailes.

http://www.avweb.com/podcast/podcas...A_Part23Revisions_205273-1.html?kw=AVwebAudio

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[tbuelna]

wktaylor,

I don't know if I would agree with having less stringent part 23 certification requirements simply based on complexity/weight/capacity metrics. I would propose that the level of certification effort for any particular aircraft design should instead be based on a "risk" scale. The risk scale would take into account many variables, such as how/where the aircraft would be operated, how the aircraft is manufactured and maintained, what is the predicted overall failure/reliability rate of the aircraft, etc. If the aircraft design and operational envelope can be shown by analysis to present a very low risk to the occupants, other aircraft, and the general public, then there might be a good case for reducing some of the certification hurdles. However, I don't think it's a good idea to waive certification/safety procedures just because the aircraft is a VFR piston single seater with mechanical controls.

While I'm mostly a proponent of limited federal government regulation, I would accept that having FAA certification of aircraft designs intended to operate in the public airspace is a legitimate function of the federal government. And since the certification process is primarily for public safety, one thing that would be very helpful to GA aircraft manufacturers would be for the FAA to have a fixed cost and time period for the certification effort.

The recent history of GA aircraft is littered with small/start-up GA aircraft manufacturers that have failed during the certification process (Adams, Eclipse, etc.). For these small businesses, a delay of 12 or 24 months in certification causes them to go bankrupt due to lack of cash flow. If the cost and time involved in certification were fixed and predictable, it would be very helpful.

Regards,
Terry

 

[Sparweb]

I suppose it would only apply to new certification project applications, not existing ones, and existing certificated aircraft requirements wouldn't change.

I'm not convinced that there would be a big advantage, because these new aircraft manufacturers seem to insist on complexity: MFD's in the panels, all-composite wings, explosive safety features like parachutes and (soon) airbags. The consequences of failure of such things can quickly become catastrophic, requiring high high reliability, and then you get fault-tree analysis percolating throughout the machine... so much for the streamlined certification process.

The real problem may be that customers expect something in their brand-new aircraft that they just can't afford.

It does make me wonder if anyone has ever done a breakdown of FAR 23 requirements and tabulated what each one of them costs (approximately) the manufacturer in design, and the customer in ongoing maintenance. Take a random requirement, such as 23.997 Fuel strainers. The strainer itself costs a few bucks, the fitting and holder a few more, the drawing and factory assembly instructions take a bit of time, location and support may be debated a bit on the floor, and then somebody has to put it in the IPC and write a paragraph for the maintenance manual. Dollars. Then the customer has to pay his mechanic to clean it from time to time, part of regular maintenance and so on. More dollars. Simple calculation. But-

How the heck are you going to take regulations like that and decide that "it's not as important as we thought" and eliminate it!? If it's not that, is it "climb over a 50-foot obstacle", or "emergency procedures in the flight manual", or is the 9.0g forward ultimate load factor going to get shaved down?

Exactly what is going to go?

If it's about cost, then some bean-counter has to stick his head up and say that the estimated cost of the regulation "X" is much higher than the estimated cost of the failures "Y".

You wouldn't catch me up there making that argument! Ford tried that with the Pinto rear-end accident lawsuit(s) and it was a PR disaster.

Steven Fahey, CET

 

[verymadmac]

The current certification process I have seen results in aircraft flying out the door (largely) matches the drawings, the loads etc are documented, handles as said and is largely safe.

So sure its not perfectly efficient but it works. There are changes that can be made to the system BUT they are not going to reduce costs by more than a couple of percent. Any other changes would likely reduce the level of safety

Its not the reg's that are killing aviation its the reduction in peoples will to fly.

 

[berkshire]

It is the stupid stuff that runs costs up, Some years ago Scwheizer aircraft in Elmira NY were in the process of certifying under part 23 an all metal single seat sailplane. During this process an FAA inspector asked them if they were going to allow smoking in the cockpit. The designer replied that it was an all metal aircraft, there was no fuel on board and nothing to burn, so why not. Big mistake, they had to prove that statement( burn tests), then install an ashtray and a placard stating when smoking could and could not take place.
Needless to say this did not help their bottom line.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[thruthefence]

"It is the stupid stuff that runs costs up"

If you get the chance, look at the leading edge of a Lear 60. About every forth fastener, spanwise, you have a button head machine screw. The rest of the screws, scores of them, are flush. I was told the following by a very knowledgeable Bombardier tech rep:

"At some point in the process the aircraft was "tufted" and flown to observe airflow on the leading edge, with the "yarn" attached to the aforementioned button head screws. These tests were completed, and the aircraft proceeded through the certification process, but the button heads were never replaced by flush screws. At some point in time, this was noticed, & plans were made to install the proper countersunk screws, but the FAA would have none of it, without re starting the entire process."

This may be BS, but having had some contact with SOME FAA guys in the past, I would not be surprised if there was some truth in it.

 

[tbuelna]

thruthefence,

Sometimes, the relationship between FAA DERs and the OEMs can get confrontational. On one hand, since the DER must sign-off on the design and thus take some responsibility for it, the DER tends to be very conservative and cautious. On the other hand, since the DER must take some responsibility for giving their approval, they are also given significant authority over the process.

The FARs themselves are also sometimes a bit subjective. The FAA has a difficult task when writing the FARs. They don't want to make them too specific, which would lead to incomprehensible FARs that resemble ObamaCare or IRS regulations. But at the same time, the FARs need to be specific enough to minimize any creative interpretations that might present a safety issue.

As I noted, anything that can be done to improve the working relationship between the FAA, DER and OEM would be of great benefit. Controlling certification costs and schedules, standardizing the certification process, and getting the DER intimately and continuously involved from the very beginning of the design process would all be very helpful.

With small OEMs, having an experienced DER point out potential certification issues early on would reduce costs and delays in the actual certification process. The FAA budget could surely afford to assign a dedicated full-time DER to any GA OEM that has applied to certify an aircraft. Such an approach would promote a proactive, cooperative relationship between the FAA and OEMs, instead of the reactive, adversarial relationship that seems to sometimes exist. Rather than having a situation where the OEM wants to get through certification with the minimum amount of effort, and the DER mostly wants to make sure he has his rear-end covered, we need a process that would encourage both parties to do the best job.

 

[dgapilot]

The certification of the aircraft (or engine) is only part of the problem. One someone gets a TC, they then have to get a PC and establish a quality system. The manufacturing and quality systems can be worse than the actual certification.

An interesting point on this is to look at history. Take a look at the number of NEW TCs issued since Part 23 went into effect in the early 1960s. I'd bet that there aren't more than about 100 or so. Thats 50 years of aviation history and very few new designs. Now look at the period from 1927 to 1939, just of=ver a 10 year period, and most of it during the depresion. During that period there were over 700 TCs issued. Something to be said for Aero Bulletin 7a and CAR 4a. Also take a look at the number of ADs issued on Part 23 airplanes and compare to airplanes certified under the CARs (with exception of the PA-28 and the Beech 18). Looks more like the newer designs are released before they are ready for prime time. There are lots of old airframes that have NO ADs yet stuff that comes off the line today have lots of ADs before the paint is dry.

 

[blakmax]

I sure hope that they consider design and process issues for adhesive bonded joints. Many people do not realise that it is possible to design an adhesive bonded joint (especially in metals) where the bond can be stronger than the surrounding structure. In other words, the metal fails outside the joint, and the adhesive does not fail.

Now consider this; if the metal always fails, then all of the certification tests for details (as per AC 20-107B) will always measure only the strength of the metal and we can look that value up in MMPDS-3. So why do all those tests? Just think of the reduction in certification costs by eliminating so many tests.

To take this approach requires a number of changes to design and process methodology. Firstly, the rediculous and invalid average shear stress design method must be scrapped and replaced with Hart-Smith's load capacity method where the variablity in strength of a joint caused by changes in thickness, elastic modulus, thermal expansion coefficients for adherends is accounted for by calculation, not meaningless tests. Design such that the bond is stronger than the structure and failure will always occur outside the joint. With the average shear stress method this is not possible so testing is always essential.

The next issue is to provide an adequate overlap to ensure that the joint can achieve the required overlap.

Next, there needs to be assurance of bond durability and that depends totally on the method of surface preparation, and in particluar how valid testing to demonstrate bond durability is undertaken.

These issues are explained in DOT/FAA/AR – TN06/57, May 2007.

http://www.tc.faa.gov/its/worldpac/techrpt/artn0657.pdf

As a separate issue, the application of damage tolerance to management of bonded joints needs to be carefully addressed, because current methods are only valid immediately after production. Application of damage tolerance to service defects is meaningless and unsafe because the current methods assume that the adhesive surrounding hte defect maintain virgin strength, when in reality if a joint is disbonding, then the surrounding bond is degraded so any assumption of pristine bond strength is invalid. See the reference attached.

Now if the issue of bond durability is addressed, there is a potential to significantly reduce the cost of NDI for the aircraft owner because bond degradation should not occur. The only need for inspection then is to evaluate physical external damage.

Regards

Blakmax

 

[thruthefence]

Murphy strikes again.

Unbreakable composite bond breaks.

http://www.aopa.org/aircraft/articles/2011/110922faa-proposes-fine-for-cessna-corvalis-defects.html

 

[blakmax]

thruthefence

Which is exactly why I said the processes must be validated. In that particular case, I am told that the bonds were performewd in a hot environment and the workers had used fans fitted with water jets to coll them down. Water and adhesive bonds just do not mix.

Get the design right, get the processes right and the bonds should not break.

 

[Sparweb]

Get the design right, get the processes right and the bonds should not break.

... and the factory climate control, and the training of the technician, and the quality control of the adhesive supplier, and the NDT inspector's tool calibration... It's endless.


How many of us have assembled two sheets of aluminum with a series of MS rivets, pulled until the rivets broke, and checked the failure load? It would give you a lot of confidence in the Mil-Hdbk-5 numbers.

Try the same thing with composites, and you find the results scatter all over the place.



Steven Fahey, CET

 

[blakmax]

Steven Fahey said"How many of us have assembled two sheets of aluminum with a series of MS rivets, pulled until the rivets broke, and checked the failure load?"

If you did the same with an adhesive bond which had been designed and processes correctly the failure load should approach Fult as per MIL HDBK 5 (now published as MMPDS). You can never achieve anything approaching that value with fasteners.

I agree with your comments about processing issues with composites. It is very easy to produce a bad composite and also a bad adhesive bond if you do not know what you are doing. However, this is not the fault of the technology; it is a problem with education and implementation.

Regrads

Max Davis

 

[thruthefence]

"it is a problem with education and implementation"

Spot on, I agree completely.

But unfortunately life outside the lab is driven by bang-for-the buck, and what's the first thing that falls by the wayside in bad times?

Training, IMHO

 

[blakmax]

I have done this outside the lab. When adequate procedures were validated and implemented at a major defence repair facility, the repeat repair rate fell from 43% to almost zero for almost twenty years, and in those cases where failures did occur, we could determine that it was technician malfunction, not the processes.

The economics don;t stack up. A quick and dirty process that saves a few hours but which is repeated frequently is not more economical that a validated effective process that is done well....Once.

Regards

Max

 

[SWComposites]

"A quick and dirty process that saves a few hours but which is repeated frequently is not more economical that a validated effective process that is done well....Once." > try to convince the MBA types of that; all they care about is this quarter's budget/profits. Training - that's just a cost to be cut. Process documentation - another cost to be cut. Follow the process - why when it is cheaper to take shortcuts. Third world back alley repair shop offers to do repairs for 1/10 the cost - woo hoo, sign em up. Sadly the commercial aircraft repair world is very different from a military repair depot.

tbuelna - "The FAA budget could surely afford to assign a dedicated full-time DER to any GA OEM that has applied to certify an aircraft. " > ah no; the FAA does not employ DER's; all DER's are paid by companies, either as direct employees or as consultants. In theory the FAA has engineers in the ACO's to help applicants, but those engineers are overwhelmed with work as it is now. Increasing the FAA budget - good luck getting that thru Congress.

 

[thruthefence]

"but those engineers are overwhelmed with work as it is now."

Anyone attempted an RVSM approval lately?

 

[Sparweb]

Hi Blakmax,

(I should point out that the following are personal opinions, I guess).

I don't run down composite structures, I just debate how appropriate they are when the full life-cycle of an aircraft structure is considered. There are some direct operating costs that can be saved by producing aircraft structures with composites, and the first owner will probably realize these benefits.

My world revolves around the airframes that are still flying after 30 years (40, 50, 60, 70, even) and they've changed hands dozens of times. It's a very different perspective.

The argument that a well controlled factory can produce a superior aircraft is not in question. I'd agree with that. It's the continued airworthiness costs that are my concern. A de-lamination zone in a CF spar - scrap the wing. Find corrosion on the web of an aluminum spar - clean/replace/reinforce and it's on its way.

Once the composite airframe design becomes the norm, then aircraft will not live as long as they did with metallic airframes. The utility, economy, and accessibility of general aviation is something that we take for granted today. Making airframes cost 5x more and last 1/2 as long will be detrimental to the industry as a whole.


Steven Fahey, CET

 

[berkshire]

Sparweb

""A de-lamination zone in a CF spar - scrap the wing. ""

Then why would you have a repair schedule from some manufacturers for a 100 to 1 scarf joint to take out the damaged area.

Your statement does not fly.

B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[blakmax]

I am not so sure that composites are so susceptible to being scrapped for minor damage as suggested by Steven. Many repair methods are available for composite structures including adhesive bonding and mechanical fastening. My main concern is that many OEMs would use resin injection to "repair' a delamination. I have yet to see realistic tests which demonstrate a restoration of strength and restoration of fatigue properties after resin injection. I HAVE seen(old)data which showed that injection repair of impact damage reduced the fatigue life of a 24 ply composite by 2/3rds compared to leaving the damage alone.

However I am also not a fan of scarf repairs as suggested by BE, especially at a scarf ratio of 1/100. Suppose you have a structure which is 0.25 inches thick. If a scarf of 1/100 is applied then the scarf will be 25 inches long, on both sides of the defect, giving a 50 inch repair. I can't see any DER approving a process to remove 50 inches of good material.

Now my expertise is mainly in metal bonding, and if Steven's metal aircraft were bonded using appropriate design methods and valid processes, they would last even longer than the alternate mechanically fastened method.

Regards

Max