Making Sense Out of DOD's F-35 Test Report
By: Winslow Wheeler | March 5, 2013
When DOD's Director of Operational Test and Evaluation released its annual report to Congress on the performance of US weapons in operational (battlefield) testing, an object of some attention was the section on the F-35 Joint Strike Fighter. The report itself made difficult and cryptic reading, and the news reporting covered whatever highlights harried journalists could readily use in short articles. Little of it probed, in technical terms or otherwise, the meaning and implications.
Lee Gaillard has been writing about aviation, weapons and national security for decades and took it upon himself to analize the F-35 using the DOT&E report as the starting point. His analysis makes important and informative reading for anyone wishing to have more than a superficial understanding of the F-35 and its problems--rooted deeply in the DNA of the aircraft's insanely complex design and its disengenuous (bait and switch) acquisition plan.
This highly informative essay-length analysis is available at Counterpunch and below.
March 04, 2013
'Concurrent Production' Exacerbates Multiple Problems When Money is No Object: the Strange Saga of the F-35
by LEE GAILLARD
On 14 January, very shortly after the Director of Operational Test and Evaluation (DOT&E) released its 2012 annual report on progress in various Pentagon programs (including a 16-page section on the F-35), Turkey announced a one-year delay in the purchase of its first two Lockheed Martin F-35 Joint Strike Fighters. Why? "High cost yield" and flight and combat capabilities that "are not at the desired level yet". In short, the F-35 doesn't work and it's too expensive. (See GlobalFlight.)
That's just the tip of the iceberg for what is the most expensive military procurement program in history. While some will argue that the key word in the Turkish statement is "yet", one must ask whether Turkey or the United States and all other partner F-35 nations will ever get what they were initially promised.
Several sources (Aviation Week & Space Technology, FlightGlobal, et al.) have provided briefer summaries of the DOT&E's F-35 annual report. But few examine the implications of what the DoD has published, or ask questions that should have been asked years ago.
* * *
For its competition against Boeing's X-32, Lockheed Martin built two X-35 prototypes, the first of which flew on 24 October 2000; the first Low Rate Initial Production (LRIP) version flew about six years later, on 15 December 2006. Now, over 12 years since that first flight, roughly 65 F-35 airframes have been delivered-43 of them produced during 2011 and 2012; the 100th aircraft is now on the assembly line.
Not one is combat capable. Even in training flights they face restrictions.
We are dealing with an aircraft that has been produced and tested in fits and starts, hobbled by a massively expensive and ineffective program of what is euphemistically called "concurrent production" where you build, fly, test, repair, redesign, retrofit, re-test-all at the same time, a process patented by R. Goldberg; money is no object.
Part of the problem is, of course, that Lockheed Martin presented us with two versions of what Detroit would call a 'concept car': a one-off only superficially representative design smaller and lighter than the actual fighter of which it was supposed to be a working prototype. The X-35A flew only 27 test flights in the one-month period before its test regimen ended on November 22; the X-35B (converted from the -A) flew 48.9 hours of tests in 66 flights during the roughly six weeks from June 23 to August 6, 2001. And the -C variant's test regime lasted less than a month-from February 12 to March 10, 2001: 73 test flights totaling 58 hours (including 250 carrier-type landings on the runway at Patuxent River; no mention of how successful the arresting hook turned out to be). For the most part, then, test sequences of roughly one month with flights averaging less than an hour each.
Under those conditions, what kind of 'wring-out' testing could these two aircraft do that would reveal future problems with transonic buffet, wing roll off, and the other significant issues that appeared from the start during testing of LRIP aircraft? Thus, when the Pentagon signed on the dotted line for the first lot of LRIP F-35s, it was buying an untested, larger, heavier paper design that hugely increased risks in any 'concurrent production' program. We are now facing the consequences.
For all F-35 versions, according to the DOT&E report, the pilot's helmet-mounted display system doesn't work; the F-35C is not yet carrier-qualified because the tail hook didn't work, had to be redesigned, and only now is being re-tested; the ejection seat in all models would put pilots at serious risk in any non-level flight mode above 500 knots (i.e., most dogfight scenarios); since flight control software is itself still under development, the computerized flight control system lacks crucial intended capabilities; key structural components have cracked and require redesign. The list goes on. Yet Lockheed Martin's Fort Worth plant keeps churning out F-35s in all their defective glory. And those aircraft already produced now need retrofits of software and flight critical hardware.
Let's take a closer look.
In the recently released DOT&E report on 2012 F-35 testing and development, we observe that:
* High-speed high-altitude flight results in delamination and heat damage to the horizontal stabilizers and their stealth coatings (pages 30, 32, and 33 in the DOT&E report; all further numbers in parentheses refer to this report);
* A cracked wing carry-through bulkhead (36) halted durability testing for over a year until it could be analyzed and repaired;
* Weakness in the auxiliary air inlet doors on the -B version led to redesign and retesting and time lost (32);
* A crack was found in a forward rib of the F-35A's right wing root-in addition to the similar crack reported on in the FY11 DOT&E Annual Report (36);
* A crack was found in the right engine thrust mount shear web (37);
* Multiple cracks appeared in the lower fuselage bulkhead flange (37), effectively halting F-35B testing;
* All this in addition to earlier cracks discovered in the -B's right side fuselage support frame as well as under a wing where a pylon and its weapon get attached (37)-and yet another in an internal support structure.
All may require redesigning of parts and subsequent added weight (since strengthening weak parts often involves adding mass to the component as part of the redesign) when for two of the F-35 versions there is less than a one-percent weight gain margin left for the entire remaining development process, and only a one percent margin available to the F-35C. "Managing weight growth with such small margins will continue to be a significant program challenge" (32); that's an understatement. Then there's the issue of retrofit to aircraft already delivered and others on the production line. (There are, of course, other structural issues not listed here-such as the drive shaft for the lift fan (31), now undergoing its second redesign, plus damaged door attachments (31), etc., etc.) Trenchant DOT&E observation: "Results of findings from structural testing highlight the risks and costs of concurrent production with development" (37).
Some obvious questions:
*Why yet another 'spiral development/concurrent production' program when the same kinds of major problems and expenses had appeared years earlier with the V-22 Osprey during whose development 30 Marines were killed? (Not to mention our similar 'concurrent development' fiasco involving the Littoral Combat Ship (LCS): as Rear Adm. Tom Rowden wrote recently in the U.S. Naval Institute Proceedings, "In the interest of quick delivery to the fleet, ship design began before requirements were finalized, and building started before designs were stable." No wonder the Navy has conceded that "LCS vessels are only rated for Combat 1+ levels-lower than a tanker" [as quoted by Mike Fabey in Aviation Week's January 28, 2013 Defense Technology Edition]. Pathetic. Reminiscent of the current barely Block 1 training capabilities of the F-35?
* What was missing from wind tunnel tests and 3D computer modeling studies of flow, weight, and stress that permitted the cracking found in that wing carry-through bulkhead and other basic structural weaknesses to get through?
* Why weren't two representative pre-production aircraft put through the wringer with several months of test flights to find these areas of stress and their causes before completion of final design and authorization of Low Rate Initial Production (LRIP)?
Performance-where the chickens come home to roost. The intended performance envelope for the F-35 is, roughly speaking: altitude capability of 50,000 feet; 700 kts./Mach 1.6 airspeed; maximum g rating of 9.0 (-A), 7.0 (-B), 7.5 (-C) ; turn performance of 5.3 sustained g's (-A), 5.0 sustained g's (-B), and 5.1 sustained g's (-C); acceleration from Mach O.8 to Mach 1.2 intended to be within 65 seconds (See Aviation Week.); angle of attack (AoA) capability to 50 degrees.
At the moment, however, this all seems wishful thinking. Undeveloped software, combined with disappointing results in real-world flight tests ("results of air vehicle performance and flying qualities evaluations" (30) ) have triggered flight restrictions and rolled back overly optimistic Key Performance Parameters (KPPs). For these and a variety of conditions that should not be occurring, flights are limited to top speeds of 550 (not 700) kts. (38) and altitudes of 39,000 feet (38) rather than 50,000 feet; AoA to be no greater than 18 degrees (vs. 50 degrees).as well as the imposition of other "aircraft operating limitations that are not suitable for combat" (38). KPPs for sustained g's in a turn have been weakened-by 20 percent for the -A (5.3 down to 4.6)(30), by 10 percent for the -B (5.0 down to 4.5) (32), and by 2 percent for the -C (5.1 down to 5.0) (33). Transonic acceleration from Mach 0.8 to M. 1.2 suffers significantly: with the -A version, it takes 8 seconds longer; 16 seconds longer with the -B; and a worrisome 43 seconds longer with the -C.an increase of about two thirds. Although the F-35 is essentially a strike aircraft, acceleration capability could be critical in combat.
Transonic roll-off (where one wing loses lift sooner than the other when a shock wave forms at the top of the wing as the airflow reaches the local speed of sound) and buffet (or shaking of the entire aircraft) as more surfaces form shock waves and boundary layer flow becomes turbulent-both were more serious than expected in the -B and -C versions, especially with the latter, whose wingspan is greater than that of the other variants: another possible problem in a combat situation.
Some fighter pilots offered their comments on FlightGlobal: " 'What an embarrassment, and there will be obvious tactical implications,' another highly experienced fighter pilot says. '[It's] certainly not anywhere near the performance of most fourth and fifth-generation aircraft.
'At higher altitudes, the reduced performance will directly impact survivability against advanced Russian-designed "double-digit" surface-to-air missile (SAM) systems such as the Almaz-Antey S-300PMU2 (also called the SA-20 Gargoyle by the North Atlantic Treaty Organization), the pilot says. At lower altitudes, where fighters might operate in the close air support or forward air control role, the reduced airframe performance will place pilots at increased risk against shorter-range SAMs and anti-aircraft artillery" ( See GlobalFlight).
A few questions:
Why didn't earlier wind tunnel tests and computational fluid dynamic modeling predict problems involved in maintaining intended sustained g's in a turn?
Why was not poor F-35 transonic acceleration also predicted-especially for the F-35C, whose eight feet greater wingspan contributes to the significantly larger Mach Cone (the zone of disturbed air behind the shock wave system generated by an aircraft at supersonic speed) that must be dragged during the transonic regime?
Why was there not greater fuselage application of area rule (that pinched waist so visible on the ubiquitous T-38 supersonic trainer), that brilliant 1950s design breakthrough by aerodynamicist Richard Whitcomb specifically to minimize transonic drag?
For the -B model, the lift fan may have prevented such a waist pinch. But why have this tail wag the dog, mandating that commonality be based on the least aerodynamic of the three variants when fuselage area rule could well have been applied to the -A and -C versions, establishing a common baseline design of improved transonic efficiency and performance across the 2243 aircraft intended (in current projections) for the Air Force and the Navy-plus all international customers not intending to order the specialized STOVL version that will be produced in the smallest numbers? Pinched-waist commonality would seem to make sense for the vast portion of the fleet numbering more than four times the 540 -B variants tentatively listed for the Marine Corps and the Royal Navy. As it is, given unique differences in wingspan and arresting gear requirements and STOVL mechanical provisions, each version already differs from the other two versions. Commonality? But applying area rule to 75 per cent of F-35s produced would have added commonality where it is most needed, cutting transonic acceleration time while improving combat efficiency, range, and speed.
Weapons and Guidance Glitches
Most weapons tested for compatibility and safe release have worked so far, but under 1-g conditions in level flight. Have possible wind tunnel-based concerns about post-release unstable airflow around wing and fuselage attachment locations prevented more combat-realistic testing under higher g's and in banking or diving modes?
Then there's the high-tech computer-linked helmet-mounted display system that will control these weapons (already in use with other aircraft and in other air forces)-classified as "deficient". Doesn't work. Why? "Expected capabilities that were not delivered" (35) include latency problems with the distributed aperture system (DAS) in the helmet-mounted video display. Latency-some call it 'transport time'-is the time between aircraft sensors' signal acquisition and its transmission and projection in readable format on the pilot's helmet video display. Currently at .133 seconds, that time delay of over an eighth of a second then has to be added to the pilot's additional physical response time of about .15 seconds if he or she is to react to the data displayed and launch a weapon. In dogfights with closing speeds of over 1000 knots, this cumulative delay of more than a quarter of a second can be potentially fatal, and the latency-derived .133 second margin of error in initial aim point stands as an unacceptable contributor to this dangerous combat deficiency. Then add in deficient "night vision acuity," excessive jitter that degrades data and images, inconsistent bore sight alignment, distracting "green glow" seepage from other avionics, imagery and data unable to be recorded (35). So-those high-tech air-to-air missiles and guided bombs cannot even be launched.
And the 25mm four-barrel rotating cannon with its 180 shells? Intended only for the Air Force F-35A version; -B and -C versions have no cannon, but will require external gun pods mounted by ground crews. Why did F-35 designers intentionally ignore the F-4 dogfighting débacle in Vietnam? The F-4-with no internal cannon and radar-guided Sparrow missiles that did not work at short range-could not shoot down the MiG-17s and MiG-21s thrown against them. Gun pods then provided a poor interim solution before the F-4E was redesigned to carry an internal 20mm cannon.
Now we have the F-35-"F" for its "Fighter" role, although it seems primarily an expensive attempt to replace early model F/A-18s and the Marines' subsonic AV-8B Harrier II STOVL aircraft in their ground attack roles. (Ironically, what Hussein's tank crews feared most was the A-10 Warthog with its GAU-8/A Avenger seven-barrel 30mm cannon, which tore them to bits from above, where their armor was thinnest.)
Why, then, in the DOT&E report are there no results listed from airborne firing tests of the F-35A's cannon? If there have not yet been such tests, has a qualifying 25mm shell even been chosen? (We remember what inappropriate propellant selection did to M-16 rifle performance in Vietnam.) Such testing early on will be crucial in determining the effect of recoil shock on the aircraft's structure and engine operation. Not to mention effects of the muzzle blast and combustion gasses on adjacent stealth coatings given that heat from air friction and radiational heating from the afterburner seemed to do such a job on the skin and coatings of the horizontal stabilizers.
No discussion. So-cannon not tested and other external and internally carried weapons for all practical purposes unlaunchable because of "deficient" sighting system available to pilots, thus rendering all F-35s produced so far as little more than expensive aerial targets for their adversaries.
Vulnerability Increased, Combat Survivability Jeopardized
In a nutshell, the software just isn't ready. We're no longer climbing into P-51s. Since at least the F-16, software has been absolutely essential for onboard computer systems that maintain stability of fly-by-wire aircraft whose design intentionally places them on the thin edge of instability to permit almost instantaneous change in flight path-crucial in a high-speed dogfight or in avoiding a SAM. Without such computers and software, pilots cannot control the aircraft.
Now take the F-35 and all its automated functions-from helmet-cued weapon sighting to datalink sensor transmissions to other aircraft and.the list goes on. It is said that the F-22 Raptor, the F-35's older brother, has 2.2 million lines of computer code; a recent estimate for the F-35's Block 3 (combat capable) mission systems software postulates that the aircraft's own computers will harbor approximately 8.6 million lines of software code-not counting even higher requirements in related ground systems. Yet "Block 3i software, required for delivery of Lot 6 aircraft and hosted on an upgraded processor, has lagged in integration and laboratory testing" (34). Block 2B software is what is required for only the most basic "initial, limited combat capability for selected internal weapons (AIM-120C, GBU-32/31, and GBU-12)" (34), yet DOT&E admits that "the program made virtually no progress in the development, integration, and laboratory testing of any software beyond 2B" (author emphasis; 34)-i.e., no tangible progress toward anything resembling real combat capability. In the wishful thinking department, full combat System Design and Development capability is tentatively scheduled for Block 3F software to be installed starting with production Lot 9 (34), which means on airframe number 214 at the earliest.possibly sometime in 2017. As for that Block 3i software, those Lot 6 aircraft are already on the assembly line (starting with airframe number 96); while delivery may begin in 2014, don't hold your breath: given program history to date, this mission software may well not be ready and Lot 6 aircraft will be in danger of being undeployable-not much better than 'hangar queens' so often grounded for other glitches. (Will Turkey have waited long enough?)
How bad is it? It's all summarized in that Pentagon report: "Flight restrictions blocked accomplishment of a portion of the planned baseline test points until a new version of vehicle systems software became available" (33). And when it comes to internal weapons release and guidance, "basic mission systems capabilities, such as communications, navigation, and basic radar functions" (34), and more-fully coded software is essential. Yet aircraft are being delivered with major variances that defer testing and add "a bow wave of test points that will have to be completed in the future" (34), while such regressive testing of systems that should have been tested earlier but were forced to be deferred massively complicates any results-based linear (not concurrent) testing and development program. At the time of the report's release, even the minimal capability of Block 1 software included in delivered aircraft was deficient by 20 percent (34). Block 2A software was delivered to flight test four months late and 50 percent deficient (34). Let the report speak for itself: "Testing needed for completion of the remaining 20 percent of Block 1 capabilities and 50 percent of Block 2A capabilities will have to be conducted while the program is introducing Block 2B software to flight test. Software integration tasks supporting Block 2B (and later increments) were delayed in 2012 as contractor software integration staff were needed to support Block 2A development, test, and anomaly resolution" (35). So much for any attempt to install mission and flight control software in any logical sequence where later and more complex versions can build on a foundation of previously installed systems. And that's just a small sampling. Sounds like absolute chaos.
Who is the supervisor for software development? For software integration? Why haven't they been replaced?
Better yet, why hasn't software development and integration, at this point, been transferred to a different vendor?
These seem to be some basic questions that no one is asking.
Equally depressing news has appeared in previous Pentagon annual reports on the F-35, and surely these reports have been distributed to members of the House and Senate Armed Services Committees.
Why have they taken no action regarding the mismanagement of the most massive and expensive military procurement program in our history?
More important, when will they start to do so?
Lee Gaillard holds degrees from Yale University and Middlebury College. He served in the Marine Corps Reserve, worked in publishing for Time-Life International in New York, in industry as a senior product marketing specialist for the world's largest manufacturer of semiconductor assembly equipment, and in secondary education as teacher, department head, and school administrator. In 2002, Gaillard attended the Royal Institute of International Affairs defense conference in London, U.K.: "Europe and America: A New Strategic Partnership," subsequently writing two related articles that appeared in Defense News. After Airways Magazine (July 2005) published his examination of the National Transportation Safety Board's flawed investigation of the American Airlines Flight 587 disaster, he served as a consultant to "Airline Cracks," a documentary on load-bearing composite structures in commercial jetliners, telecast by ITV-West (Bristol, U.K.) on Oct. 4, 2005. In 2006, the Center for Defense Information published his monograph on the V-22 Osprey.
Gaillard has been writing about aviation and defense issues for over 25 years. His more than 100 articles and book reviews have appeared in newspapers, professional journals, and magazines around the country-on topics ranging from the role of luck in the Battle of Midway (Naval Institute PROCEEDINGS) to "Submarine Design: Aeroengineering Dimensions" (Submarine Review) and the V-22 Osprey's readiness for combat (Jane's Defence Weekly). He is listed in recent editions of Who's Who in America and is a contributor to the Straus Military Reform Project.