Forum: General F-35 Forum

JW, I think you are making the assumption that static test loads have never been verified in, and adjusted by, flight test loads subsequent to the static testing to 150 DLL. While easy to infer from the article, static testing to 150 DLL did not really complete static testing, it merely verified the accuracy of the analytical load predictions prior to the beginning of loads flight testing (where an incremental approach similar to what you suggested near the top was followed). If the static testing was really "complete" at the date you suggested, what were they doing for the other 2+ years the article was in the test rig? Similarly the STOVL static article completed 150 DLL testing in late 2009, but has remained in the rig since; same for CV.

F-16.net Sponsor

Perhaps this article helps with getting to know the 'new way' of doing things. Don't ask me.

F-35’s performance benefits from BAE Systems integrated structural testing regime

http://www.hbm.com/en/menu/applications ... e-systems/

"...A key aspect of BAE Systems’ approach to the F-35’s structural testing regime was the need to develop improved methods of managing test information. The company wanted to streamline its approach by adopting a new mind-set that challenged previous methods....
&
Static and dynamic tests
"Both fatigue tests and static tests are undertaken by S&DT. Fatigue tests take much longer to complete than static tests and are based on data trending which looks for changes in the structural response over time. These can easily signify the possibility of cracks or other in-service related conditions.

The tests are undertaken by applying various loads to different structural components in an accelerated timeframe to reproduce potential operational damage. Complex PC-based test control and data acquisition systems are used to control how the forces are applied to the different test structures and the responses carefully monitored. Test specimens are usually instrumented with strain gauges enabling real-time recording of strains generated...."
&
Sharing data
"Another challenge was to successfully share test data in a timely manner with multiple sites. These included the BAE Systems’ sites at Brough and Samlesbury along with three Lockheed Martin plants and Northrop Grumman’s El Segundo factory in the USA...."
&
FTI Tests
"Another area in the development of the F-35 where BAE Systems’ expertise is important is in de-risking the Flight Test Instrumentation (FTI) system on the aircraft. This proved especially challenging and required a unique approach. Normally the crucial task of testing the data acquisition system on modern military aircraft is completed using the aircraft’s own FTI system.

In the case of the F-35, however, testing the on-board systems needed to be carefully co-ordinated as the system is fitted at Fort Worth, Texas, USA but the tail sections are built in the UK. This means that BAE Systems at Samlesbury has to power up and function check every fitted transducer and the associated aircraft wiring during building. The approach gives a high level of confidence that the whole system is functioning...."

ONLY parts of the long article excerpted here so best to read entirety at URL. Tah.
_____________

CAPTION: "BAE Systems’ structural tests on the F-35 utilise HBM’s equipment to gather data from some 3,800 channels in one of the world’s largest structural DAQ systems."
http://www.hbm.com/typo3temp/pics/d297880d5a.jpg

Structural Test Labs

http://www.lockheedmartin.com/us/aerona ... tlabs.html

"The Structural Test Labs support a wide range of activities that relate to structural testing of full-scale aircraft and large component structures. Our facilities equipment and experience allow us to accommodate large-scale tests with minimum transition time. A number of test frames are available that can be easily modified for many component tests. The lab designs and fabricates unique test frames for large component and full-scale tests. Marietta and Fort Worth facilities have structural test lab capabilities. The Palmdale site also performs testing to support early program development.

A variety of full-scale and component tests are conducted in the structural test labs:

• Full-Scale Component and Airframe Static and Durability Testing

• Aircraft and Component Calibration

• Mass Properties/Moment of Inertia"

quicksilver wrote:

JW, I think you are making the assumption that static test loads have never been verified in, and adjusted by, flight test loads subsequent to the static testing to 150 DLL. While easy to infer from the article, static testing to 150 DLL did not really complete static testing, it merely verified the accuracy of the analytical load predictions prior to the beginning of loads flight testing (where an incremental approach similar to what you suggested near the top was followed). If the static testing was really "complete" at the date you suggested, what were they doing for the other 2+ years the article was in the test rig? Similarly the STOVL static article completed 150 DLL testing in late 2009, but has remained in the rig since; same for CV.

To begin with, the static test article is now in Fort Worth. I don't think it was magically teleported there from the UK in an eyeblink. Obviously the 170 load cylinders had to be removed, the whiffle tree system (thousands of components) had to be removed, the load application pads (thousands) had to be removed, the instrumentation system had to be removed, the instrumentation wiring removed, and the airplane removed from the rig. Then the airplane had to be inspected for damage and prepared for shipping, then trucked to a UK port and loaded onboard. Then shipped to some US port (I assume by ship), then unloaded and trucked to Fort Worth. Alternately it may have been disassembled and loaded in a rather large airplane and flown to FW.

How long did that take? Beats me. I simply believed the test completion date as trumpeted by BAE, June 9, 2010. If, as you say, additional static test loads were applied after flight test data was available, then BAE and LM were less than truthful about ST completion, weren't they?

Now can you explain to me how applying un-verified test loads somehow verifies the accuracy of analytical design loads? That test verifies the strength as at least 150% of the un-verified loads and checks the accuracy of stress analysis, that's all.

johnwill wrote:

To begin with, the static test article is now in Fort Worth.

Is it? The article referenced for this thread both supports and contradicts that assertion.

johnwill wrote:

How long did that take? Beats me.

Then we can safely say you don't know what they did with the test article for the intervening period between Jun 2010 and Aug 2012. Given your experience, I'm sure you don't believe they just let it sit there, nor that it took two years to dismantle and ship the thing back the the US.

johnwill wrote:

I simply believed the test completion date as trumpeted by BAE, June 9, 2010. If, as you say, additional static test loads were applied after flight test data was available, then BAE and LM were less than truthful about ST completion, weren't they?

I'm not inclined to believe someone is lying to me because my understanding is different than theirs. We are also talking about a press article, which it's safe to say, is not exactly the horse's mouth.

johnwill wrote:

Now can you explain to me how applying un-verified test loads somehow verifies the accuracy of analytical design loads? That test verifies the strength as at least 150% of the un-verified loads and checks the accuracy of stress analysis, that's all.

I didn't say 'design loads' -- I said load predictions, the accuracy of which 1) provided earlier confidence that they weren't going to have major structural mod requirements, and 2) allowed more rapid envelope expansion with a higher degree of confidence that they wouldn't have catastrophic failure. Even so, loads flight testing has been conducted at a very deliberate pace. Where the flight test results conflicted with the static tests they conducted additional testing.

quicksilver,

From your 8/13/2012 11:51 pm post "static testing to 150 DLL did not really complete static testing, it merely verified the accuracy of the analytical load predictions prior to the beginning of loads flight testing", you still haven't explained how that process works. "design loads" are in fact "load predictions" at that stage of the program and remain so until verified (or not) by flight test. The point is, static test does not provide any confidence in design loads, analytical loads, flight test loads, or any other kind of loads. Static test provides confidence in strength and in stress analysis.

"Even so, loads flight testing has been conducted at a very deliberate pace. Where the flight test results conflicted with the static tests they conducted additional testing."

Do you have any documentation to verify this claim? "Deliberate pace", is being kind. The F-16 flight loads test airplane went to 9g on flight 7. How about the F-35?

Another point - external store loads have not yet been flight tested, as I understand it. Therefore, static test external store conditions could not possibly have been correlated with flight test. How will they static test those conditions if the need arises?

Your defense of the static test scheduling is commendable, but you have offered nothing to verify your claims.

Are any of the following possible?
a) As suggested an arbitrary margin was added to the analysis loads
b) Loads measured from AA-1 were used to verify the analysis loads
c) Flight test got behind schedule and someone (or a lot of someones) didn't think to delay/re-schedule static testing
d) The test article (AG-1) won't be disassembled until after flight test is completed in 2016

Sufa Viper

johnwill wrote:

Do you have any documentation to verify this claim? "Deliberate pace", is being kind. The F-16 flight loads test airplane went to 9g on flight 7. How about the F-35?

And the F-16 also almost crashed on its first flight. Things have changed considerably over the past 30 years, and you have understand the context of 21st century aircraft development to make an accurate comparisons regarding any subject. Both programs have achieved commendable successes (in engineering, production, and test) in very different development environments.

johnwill wrote:

Your defense of the static test scheduling is commendable, but you have offered nothing to verify your claims.

Noone needs to bring unreleased data (which may be F-35 PI, LMPI, or FOUO) into the public domain for this discussion to make a point.

sufaviper wrote:

Are any of the following possible?
a) As suggested an arbitrary margin was added to the analysis loads
b) Loads measured from AA-1 were used to verify the analysis loads
c) Flight test got behind schedule and someone (or a lot of someones) didn't think to delay/re-schedule static testing
d) The test article (AG-1) won't be disassembled until after flight test is completed in 2016

Sufa Viper

Sure, all of those are possible. I'm simply trying to understand how they got around the potential situation being discussed. Not sure, but I don't think AA-1 had loads instrumentation. I think (c) is unlikely, as those are smart people in charge.

Another possibility is the customer approved the plan without realizing the risks involved.

LMAggie wrote:

johnwill wrote:

Do you have any documentation to verify this claim? "Deliberate pace", is being kind. The F-16 flight loads test airplane went to 9g on flight 7. How about the F-35?

And the F-16 also almost crashed on its first flight. Things have changed considerably over the past 30 years, and you have understand the context of 21st century aircraft development to make an accurate comparisons regarding any subject. Both programs have achieved commendable successes (in engineering, production, and test) in very different development environments.

johnwill wrote:

Your defense of the static test scheduling is commendable, but you have offered nothing to verify your claims.

Noone needs to bring unreleased data (which may be F-35 PI, LMPI, or FOUO) into the public domain for this discussion to make a point.

No, it was the YF-16 that almost crashed on its first flight. I was there.

Believe it or not I was leading flight loads test programs well into the twenty first century, so have a reasonably good grasp of how things are done now. Both programs have done well, as you say, but I am not at all impressed with the "deliberate pace" of the flight loads test of the F-35. Some would call it incredibly slow.

I'm not asking for the release of any sensitive data, just some kind of updated PPT chart to show any retest activity.

LMAggie wrote:

And the F-16 also almost crashed on its first flight. Things have changed considerably over the past 30 years, and you have understand the context of 21st century aircraft development to make an accurate comparisons regarding any subject. Both programs have achieved commendable successes (in engineering, production, and test) in very different development environments.

John, I am learning tons just reading this thread, you should be charging for lessons.

One question occurs to me based on LMAggie's quote, IF things have changed considerably, would it not be reasonable to assume that the changes were positive and that they would have resulted in shortening the amount of time needed for testing? I am sure you remember the AUDIO testing set up for the F-111? I was briefly involved with that when the Aussies bought the cold test chamber for their fleet.
I don't believe that project management techniques have advanced nearly as well as the technical side of the house. That's my guess for the pedestrian pace of the program.
fisk

Fisk,

Changes have been made, of course. That's why the F-35A flight loads test will take three years and the F-16 (with a larger flight envelope) was essentially complete in one year. Evidently the static test for F-35 was done in record time - well done!

You're right, the program management capability has not kept pace with technical side of things. I certainly recognize the complexity of the F-35 management task compared to the F-16. Just having the Navy aboard is reason enough for that.

Thanks for the kind words. I am also learning much .

fiskerwad wrote:

LMAggie wrote:

And the F-16 also almost crashed on its first flight. Things have changed considerably over the past 30 years, and you have understand the context of 21st century aircraft development to make an accurate comparisons regarding any subject. Both programs have achieved commendable successes (in engineering, production, and test) in very different development environments.

John, I am learning tons just reading this thread, you should be charging for lessons.

One question occurs to me based on LMAggie's quote, IF things have changed considerably, would it not be reasonable to assume that the changes were positive and that they would have resulted in shortening the amount of time needed for testing? I am sure you remember the AUDIO testing set up for the F-111? I was briefly involved with that when the Aussies bought the cold test chamber for their fleet.
I don't believe that project management techniques have advanced nearly as well as the technical side of the house. That's my guess for the pedestrian pace of the program.
fisk

My theory on increased development span in the aviation industry as a whole is the fundamental change from federated subsystems to integrated subsystems. In legacy aircraft, each subsystem was developed largely on its own and integrated at the tail-end of the development cycle. Because they were developed without the need (or very minimal) to function with the other systems, there weren't as many surprises during build and test due to this. (And of course there were suprises but for other reasons)

What you have now are systems that are completely interwoven, and that includes both software, hardware, and structure. And because every aircraft (now and then) is initially designed by predictions, changes are inevitable. With modern development, a necessary change in one subsystem triggers a waterfall effect to many, many other systems. And the change to one of those of systems may waterfall into other systems.

A perfect example is the STOVL flight control system. STOVL flight requires highly integrated actions between propulsion, control surfaces, and structures (doors). So if any one system has an issue, the waterfall can affect them all. And each subsystem has to mature at the exact same rate or else you lose efficiency (schedule). The least mature system is what keeps you from moving forward. And there are thousands of other depedencies just like this example on a modern integrated system like the F-35, 787, A380, etc. Not to mention integration between aircraft to aircraft, aircraft to ground station, aircraft to maintainer.

This was exacerbated on F-35, in my opinion, due to SWAT. SWAT trimmed the fat on every subsystem, many times replacing 'comfortable' margins and redundancy used from legacy practices with 'adequate' margins and redundancy. The system is still safe and effective, but makes follow-on design corrections and accomodations more difficult IMO.

Everyone, from contractor to customer, is much more risk averse now too. Noone wants to take the next step without being sure that the system is ready. So the increments of maturity (i.e. how long before I think Im good to take the next step in design/build/test) has increased.

And to cap it all off, project management is still trying to figure out how to best manage all of this integration. The DoD has been scrambling in the past 10 years to stand up a robust systems engineering model to account for the new integrated systems that are needed now. I think anyone would agree that the ability to manage large scale integration is a big challenge for this industry.

Anyways, thats my Two Cents

A valuable statement, thanks. But none of which helps to explain why there may be unverified loads used in F-35A static test or why the pace of flight loads test seems so slow.

It's human nature for each generation to think they invented the wheel, so to speak, and that previous generations had it so easy. None of that messy integration stuff. Since you have no idea how airplanes were developed thirty, forty, or fifty years ago (I doubt you were even around then), let me be the first to break the news to you - considering the primitive tools available, the engineers of those days were giants who developed the methods and tools you use today. And the generation before them developing WWII airplanes with slide rules only. Try doing F-16 load analysis on a HP-9820 desk top calculator, try writing down thousands of items of test data on paper spread sheets during structural lab tests. You think "legacy" systems had comfortable margins? If the F-111 SWIP program (already mentioned) had not been successful, that Plant 4 you work in would have been used for grain storage for the past forty years, with no F-16 or F-35 program.

You are right about the current generation being risk averse now. Sometimes I think it is more CYA than anything else. You might be shocked at some of the risks taken in the F-16 flight test program, yet no pilots or airplanes were lost in development.

I wasn't trying to imply that the previous generation had it easier. I was simply contemplating why it was taking longer to develop modern aircraft and that it's because there is more workload (the culprit being integration). Just like it took more workload to develop the F-16 than WW2 aircraft. Thats not to say it was any less and more hard to achieve. Each generation had it 'hard'; breaking different barriers and pushing the envelope in different ways that paved the way for future generations to take on new challenges. Which is why I think it's difficult to make an apples to apples comparison.

I'm not sure what prompted you to be so defensive and condescending, but thanks for reminding me why I stopped coming to this forum.

Forum: General F-35 Forum

F-35 testing turned full circle UK StructuralWise