Buy Before You Fly (III)

February 19, 2001

 Comment: #406

Discussion Thread - Comment #s 405 & 401


[1] Andy Dworkin, "Obsolete Parts Plague Defense Industry: Chip technology changing faster than weapons," The Dallas Morning News, April 25, 1999. Excerpts attached.

[2] Tony Capaccio, "F-22 Production Decision In March, Official Says," February 15, 2001. Excerpts attached.

Separate Attachments:

[3] Phillip Coyle, "F-22 Test and Evaluation Status," Memorandum for Under Secretary of Defense (Acquisition, Technology, and Logistics), 20 December 2000.


Last Friday, Steven B. Plummer published the following rebuttal to the Op-Ed Adm Shannahan and I published in the Washington Post Outlook on February 11 [also distributed as Comment #405]. I urge you to read his letter carefully before reading my response below.

Plummer's Rebuttal

Defending the F-22
Letters to the Editor
Washington Post
Friday, February 16, 2001; Page A24

It's hard to tell whether Franklin Spinney's and John Shanahan's opinion [Outlook, Feb. 11] is based on a misinterpretation of the data or on a lack of it. In either case, their arguments about the F-22 are just plain wrong. At this point in its development, the F-22 is the most rigorously tested military aircraft in history.

Mr. Spinney and Adm. Shanahan assume that military aircraft that must last for at least 30 years can be developed and purchased the same way civilian automobiles are. Nothing could be further from the truth or common sense.

They also argue that the F-22's 860 flying hours are too few to justify low rate initial production. In fact, the YF-22 and F-22 together have logged more than 1,000 flight test hours—over four times as many as the F-15 achieved before it began low rate initial production.

The writers also fail to mention that at the same time the F-22 has been flying, we have spent more than 11,000 hours testing its avionics in a lab and on a flying test bed (a Boeing 757 outfitted with actual F-22 hardware and software) and thousands more hours ground-testing its critical systems, airframe and stealth capabilities.

Had we performed these tests sequentially, the F-22's development would have taken many more years and billions more dollars. With our process, the F-22 will save American lives on foreign battlefields much sooner than possible under a sequential development strategy. That will be the most important measure of our success.

The Defense Department does everything possible to ensure our aircraft acquisition programs are managed professionally, competently and with integrity. The Air Force's F-22 program meets these standards.

Principal Deputy
Office of the Assistant Secretary

End Plummer's Rebuttal

Plummer alleges our op-ed describing the pathologies of the "Buy Before You Fly" policy defied truth and common sense with respect to the Air Force's F-22 fighter. He says we either misinterpreted the data or were ignorant of it.

But to prove this point, it is Plummer who defies truth and common sense.

The first set of Plummer's counterpoints relates to his assertion that we made factual errors in describing the F-22's flying hour test program: Plummer claims the F-22 is the most rigorously tested aircraft in history, with "more than 1,000 flight test hours" on the YF-22 and the F-22, which are "four times as many as the F-15 achieved before it began low rate initial production."

Note that Plummer chose to include the YF-22 in his total of more than a thousand flight hours. The YF-22 flew a total of 153.2 hrs in the early 90s, but that airplane was very different from the F-22 now being tested—it had no mission avionics, no stealth features, different structures, was lighter, and had an underpowered engine. Its primary purpose was to demonstrate the capability to fly supersonic without an afterburner (and some limited other tests—like weapons separation for the Sidewinder missile). The YF-22 flight demonstration did not even perform the combat persistence tests needed to determine if it had a meaningful cruise range at supersonic speeds in non-afterburning power (so called "mil-power"), an omission that could come back to haunt the Air Force, given the low fuel fraction of the F-22.

The differences between the YF-22 and F-22 make it illogical and absurd to include the YF-22 flight tests in any evaluation of the F-22's current readiness for production.

But let's give Plummer the benefit of the doubt and include the YF-22's 153.2 hours in the F-22's flying hour total.

As of 31 January 2001, the last day for which we had data prior to writing the op-ed, the five F-22s in flight test had accumulated at total of 853.4 flying hours. By the time of Plummer's letter was published on 16 Feb, one more F-22 had made its first flight, and the six flying aircraft had accumulated a total of 869.1 hours.

Now if we add the most recent numbers (i.e., 869.1 hours) to the YF-22's 153.2 hours, we come to a grand total of 1,022.3 flying hours. To be sure, 1022.3 is 2.2% greater than a 1000, so Plummer is technically correct when he says the F-22 has accumulated more than 1,000 hours—but not much more, and remember, this assumes we are willing to accept his specious inclusion of the YF-22.

But does a total of 1022.3 flight test hours make the F-22 the most rigorously tested aircraft in history?

One piece of supporting factual evidence Plummer uses to discredit our op-ed is that "more than a 1,000 flight test hours" (i.e., 1022.3 hours) is "over four times as many as the F-15 achieved before it began low rate initial production." [emphasis added]

This claim is flat wrong.

According to the 1974 McDonnell Douglas Annual Report as provided by the company's chief historian, the F-15 made its first flight on July 27, 1972 and had accumulated 1,350 flight test hours by the time of the limited rate production decision on February 19, 1974. In other words, the F-15 flew 32% more test hours than the combined total of the YF-22 and F-22 or 55% more than the 869.1 hours for the F-22.

There is more.

If the F-22 "is the most rigorously tested military aircraft in history," then why was its total flight test program cut back by 37% since it was defined in November 1994? At that time, the Air Force said the F-22 (not the F-22 plus the YF-22) would accumulate 1,400 hours of a 5,191 hour flight test program before its initial production award. But the total program was reduced to 4,337 hours in February 1998 and to 3,757 hours in September of 1998. (Coincidentally. the current total of 869.1 hours is 38% less than 1400 hour production threshold—a number that will not change substantially if the F-22 is approved for production in March).

Presumably, less testing is be needed when a program is progressing ahead of expectations and has fewer problems than originally expected.

Such is not the case for the F-22; in fact, the opposite has been occurring. The delays and reductions in testing reflect cost-based adjustments to accumulating problems. These problems are now so serious that Phillip Coyle, the Pentagon's Director of Operational Testing and Evaluation, recommended a delay in the production decision as recently as 20 December 2000. He made this recommendation in a revealing memo to the Under Secretary of Defense for Acquisition, Technology, and Logistics. [Ref 3,]

Coyle's memo explains why the F-22 test program has "fallen considerably behind schedule again this year" [emphasis added]. He cites late deliveries, and a plethora of technical and mechanical problems. As of December 19, 2000, he reported the F-22 had completed only 55% or 322 hours of the 590 hours planned for calendar year 2000 as recently as December 1999. Moreover, of the projected 7,745 "flight test points" (i.e., particular flight conditions to be explored) planned for CY 2000, only 3,075, or 40%, had been completed by the end of November 2000.

Coyle said also that the "continuing late deliveries of the avionics flight test aircraft" (i.e., a/c #s 4004-4009) had led to a loss of another 21 available aircraft months of avionics flight testing during CY 2000, bringing the total to 42 aircraft months that have been lost to avionics testing.

Based on these and other problems, Coyle concluded that the Initial Operational Test and Evaluation could not be started in August 2002 without unacceptable risk. He recommended that the engineering and manufacturing development program be extended at least nine months to a year or more to complete essential development testing. Therefore, he concluded, "from a test point of view, there is no reason to authorize LRIP (low rate initial production) at this time and some justification to delay LRIP." [emphasis added]

If the F-22 is the most rigorously tested program in history, then Plummer ought to explain why the Pentagon's Director of Operational Testing says the F-22 is not ready for production.

Plummer, of course, is silent on the issues posed by the Coyle memo [Ref 3].

A second aspect of Plummer's claim for testing rigor is that the F-22 testing includes over 11,000 hours of avionics testing in the lab and on the flying test bed (a Boeing 757 outfitted with F-22 hardware and software). According to a Boeing press release (Sept 19, 2000), actual testing included 15,000 in the lab but only 562 hours had been on the flying test bed when the Block 3.0 avionics began testing in the flying test bed on September 1, 2000. Moreover, according to the most recent flight test report, F-22 aircraft #4005, the only F-22 now carrying the Block 3.0 avionics, had flown only one sortie for nine-tenths of one hour.

Based on this kind of factual analysis and reasoning, Plummer trashed our "buy before you fly" recommendation by saying, "With our process, the F-22 will save lives on a future battlefield much sooner than the sequential development strategy."

Much sooner?

This is a truly bizarre conclusion. For one thing, the F-22 has been in development in one form or another since Lockheed was selected to design and build the YF-22 demonstrator in October 1986 - or far longer than any other fighter in history.

Ironically, the F-22 may be also the first plane in history to have its major systems based on technology that was grossly obsolete before it went in production. Andy Dworkin, for example, reported almost two years ago in the Dallas Morning News (on April 25, 1999, Ref 1 below) about the growing concern that the five-volt computer chips powering the F-22's avionics are already so obsolete that it may be difficult, if not impossible, to support the F-22 logistically, if it ever enters the fleet.

Five-volt chips were state of the art when the F-22 entered development. But with the advent of the smaller, faster 3.3 volt Pentium chip in 1992 and the even faster 1-volt Pentium III chip now in widespread use (and even faster chips are coming off the drawing boards), the computers powering the as yet untested F-22 Block 3.0 and 3.1 avionics are being left behind in the dust bin of technological history.

Moreover, one can not simply install a 1-volt chip into a system designed for 5-volt chips, because a power surge within the 5-volt limits could fry a 1-volt chip, just as a normal voltage spike can blow an insufficiently sized fuse in your house's electrical system. There is no known way to fix for this problem, so the F-22 program office has been compelled set up a $1 billion program to address material and part shortages over the lifetime of the F-22. The problem is that no one knows how many obsolete parts to buy, because the pattern of parts consumption will not stabilize until the F-22 has been in the fleet for several years.

So why did Stephen Plummer, a Principle Deputy Assistant Secretary of the Air Force, write the letter to the editor to rebut an op-ed that focused primarily on the Marine Corps V-22 Osprey and only briefly introduced the F-22 as an example?

The answer can be found Ref 2 below.

The F-22 is in trouble.

David Oliver, the Pentagon's top acquisition official, told Tony Capaccio that a production decision must be made in March, otherwise the F-22 program will run out of money by March 31.

Put another way, we must put it into production to protect the contractor—now this is a conclusion that is getting dangerously close to the main point of our op-ed.

Oliver said the F-22 was not being held up by President Bush's much ballyhooed strategic review, but it is under review because officials are trying to determine what its production costs should be.

Costs are a problem because Congress passed a law saying the Pentagon must certify the F-22 program can meet the $37.9 billion cost cap established by Congress. If the AF can not meet the cap, it must submit a plan on how it will reduce costs. And in January, the Air Force's top acquisition official, Darleen Druyun, acknowledged the Defense Department can't certify the cap will be met. Furthermore, the breach could be very large, because Capaccio reports that the Pentagon's independent cost analysts now estimate that the program will exceed the production cap by about $9 billion.

So the Air Force has concocted a very unusual plan, to put it charitably. In contrast to the traditional lament that production stretch outs increase costs, the Air Force now claims a stretch out will save money.

Specifically, that Air Forces claims it can narrow the $9 billion breach and reduce costs by deferring the purchase of 13 of the 86 fighters scheduled through 2004. The Air Force says this stretch out can fee up enough money to fund future productivity enhancements that will reduce the future costs enough to meet the cost cap over the long term.

Justifying a pre-mature production decision on this kind of futuristic economic hokus pokus brings us full circle to the main point of our op-ed—a point that Steven Plummer hoped to duck but has reinforced inadvertently—namely that the "buy before you fly policy" is preferred by the acquisition mafia because reduces political risks to the contractors and their allies in Congress and the Pentagon while increasing the risks to soldiers (increased technical risk) and taxpayers (increased economic risk) [See Comment #401].

Oh, by the way, there is one other thing that Stephen B. Plummer left out in his letter to the editor.

Note that he signed his letter as "Principle Deputy, Office of the Secretary of the Assistant Secretary of the Air Force." But he is not a civilian political appointee or career civil servant working in an oversight capacity in the Air Force's civilian secretariat, as is suggested by his title.

Plummer is, in fact, a Lieutenant General—on active duty in the U.S. Air Force.

Chuck Spinney

[Disclaimer: In accordance with 17 U.S.C. 107, this material is distributed without profit or payment to those who have expressed a prior interest in receiving this information for non-profit research and educational purposes only.]

Reference #1

Dallas Morning News
April 25, 1999

Obsolete Parts Plague Defense Industry
Chip technology changing faster than weapons

By Andy Dworkin,
The Dallas Morning News

Lockheed Martin Corp. designed the new F-22 "Raptor" to be the world's most advanced fighter plane.

But parts of it are already obsolete.


Simply put: The chips that let kids fly fighters in computer games are often far more advanced than the chips that run the real fighters.


"The cost to us to sustain our past will begin to exceed the cost to insert new [weapons] systems," said Malcolm Baca, chief operating officer of Yorba Linda, Calif.-based TACTech, which tracks when semiconductors leave production.


"There is no such thing as a 5-volt PC made today. There is no such thing as a 5-volt cellular telephone - those were the old ones that looked like a lunch bucket," Mr. Baca said. "The 5-volt market as we know it will be pretty well collapsed by 2004, 2005.


"Much like a fuse, if you put a 5-volt surge over a 3-volt, 2-volt or 1-volt device, you would blow it," he said. "It's a real nasty problem."


One option is a "life-cycle buy" - to stock up on all the components you think you'll need before a part is no longer made. But that solution takes a lot of cash up front, experts said, and it can be hard to estimate just how much to buy, since production and maintenance can drag on for 50 years or more.


A final, possible solution is engineering around a part - finding or designing a replacement component or system. But this is also pricey - officials said it can cost from $150,000 to $1 million just to work around an old part.


The only real solution, Mr. Borky said, is to fundamentally change the way weapons systems are designed. Designers need to use "open architecture" - dividing the whole system into logical, individual parts that connect in clearly defined ways. That lets engineers swap parts in and out, at a lower cost, when new designs are needed.


While the F-22's decades of life mean many more obsolescence problems lie in store, (Lockheed Martin's) Mr. Hammond said the new team to track and fix obsolete parts should mitigate the problem. It might even save more than its $1 billion cost, he said.

"I don't want to make this problem sound insignificant, because it's not," Mr. Hammond said. "But we're on top of it."

Reference #2
February 15, 2001

F-22 Production Decision In March, Official Says

By Tony Capaccio

Washington—The Pentagon will decide next month whether to award Lockheed Martin Corp. a $2.1 billion contract to produce the first batch of F-22 jet fighters, said David Oliver, the Pentagon's top acquisition official.


Investors have worried this decision might be delayed as the Bush administration studies U.S. tactical aircraft needs as part of a broader review of military forces, strategy and spending.\


``In the context of the broader uncertainty that exists about the review, the Street is worried a lot of things can slip, so the fact you can nail some of these things down, all the better,'' Callan said. ``It's an important program for the company, it's an important program for the earnings and cash flow.''


``The secretary will have to make a decision'' because the program runs out of money on March 31, said Oliver.

A signed production contract would cap Lockheed Martin's efforts since last year to restore investor confidence after a series of problems with its aircraft, missile and space rocket programs. Shares of Bethesda, Maryland-based Lockheed Martin have risen 97 percent in the last 12 months. Lockheed last month raised its profit outlook for 2001 because of corporate restructuring that has reduced debt.


In deciding to approve a contract award, the Pentagon must certify to Congress that the F-22 program can meet a $37.9 billion cost cap required by Congress or else submit a plan on how it will reduce costs. The Pentagon can't certify the $37.9 billion production cap will be met, said Darleen Druyun, the Air Force's top acquisition official last month in an interview.

Independent Pentagon cost analysts estimated the program will exceed the production cap by about $9 billion.


The program calls for building 339 planes with the first entering the service in December 2005. Each plane is estimated to cost $182 million in inflation-adjusted dollars that include research, production, support and air base construction costs.

Lockheed Martin shares fell 13 cents to $35.69 in trading of 1.77 million shares on the New York Stock Exchange.