Tuesday, April 11, 2006

Eclipse - Four Topics for Discussion

This is a four part critical analysis of the Eclipse Aviation program. The First Part will discuss aircraft weight, believed to be the Achilles heel for this program. The Second Part, the fallacy of friction stir welding. Third, the business plan, and finally, Vern Raburn's penchant for publicity and his on going promotion and self-promotion.

One can not fault Vern for trying to breathe new life into business aviation, but I do fault his attitude. He claims the industry is fossilized and he is going to show the world how it should be done.

In 1963, we brought to life the Lear Jet Model 23. At certification, $13 million was invested, the aircraft sold for less than a half-million and cruised 100 mph faster than the Eclipse. climbed to 41,000 feet in 13 minutes (19 minutes to 35,000 feet for the Eclipse) and out performed the commercial airliners of the day! The airplane was certified 10 months after first flight. We had no CAD systems, no Blackberrys, no e-mails, no FAX machines, no cell phones, the company's first computer did not arrive on the scene until just before certification. The handful of engineers (30) was managed by a chief engineer...the company managed by one President (Bill Lear) and one VicePresident.

So Vern, where is the progress? The Eclipse bloated organization consisting of a CEO, a COO, a CFO, a Senior Fellow, and 12 Vice-Presidents???

Part 1.

Airplanes are designed around their engines. Each pound of thrust will support between 2.5 to 3.0 pounds of aircraft weight. If the number is down in the 2.5 region, the airplane will be a real hot-rod. Three and above, the performance might better be described as limp for a jet.

The Eclipse was designed around two Williams 770 lb thrust EJ-22 engines. Eclipse designers went to the upper limit (3.05) and established a takeoff weight of 4,700 lbs.

Empty weights for a twin engine executive jet should be about 60% of the takeoff weight. If the design team is really good, they might achieve a lower number. The Eclipse engineers (or Vern) thought they were really good and set the target at 57% and 2,700 lbs became the weight bogie. It was a tough assignment to stay within 2,700 lbs ...in my opinion, impossible.

Never-the-less, they built a Williams powered prototype. It flew once, barely. We have no idea what it weighed, it was not a complete airplane. The shortcomings of the Williams engine have been well reported and no need to re-plow that ground. It was not viable for the program and a search for a suitable engine began. Pratt & Whitney agreed to develop a new engine that would produce 900 lbs of thrust.

Eclipse had to revise the design. I suspect the engineers looked at how much both the empty and gross weight would have to grow to accomodate the bigger, more fuel thirsty engines and added these weights to the original numbers. The new takeoff weight became 5,640 lbs. Weight to thrust went to 3.13 (a bit worse than the original 3.05). For this round, empty weight was established at a more realistic level of 60% (3,390 lbs) but still a huge reach for a 6-place pressurized business jet.

Cessna has delivered over 170,000 airplanes. They have a reputation for building light weight designs. When they decided to compete in the Very Light Jet market with a 6-place entry, the lightest jet they could conceive will gross 8,000 lbs. That is 40% heavier than the Eclipse. There is not that much difference in cabin size...not that much difference in performance. Is Vern Raburn 40% smarter than Cessna?

What Vern does not realize is that it is tougher to build a small airplane than a larger one. Some things just do not scale down. The engine Pratt is developing for the Eclipse is one inch smaller in diameter than the version being developed for the Cessna Mustang, The Mustang engine produces 50% more thrust and will have better fuel specifics. The accessory gear boxes will be nearly the same size. People and baggage do not scale down. Seats, systems, interiors will all be nearly the same size and weight. How do you build a comparable airplane 40% lighter?

If it were a horse race, I would bet on the Mustang.

Here is why empty weight is important. Vern claims the Eclipse will have the ability to fly Phoenix to Chicago with 4 occupants and a tailwind. In the first place, his 4 occupants are based on a 170 pilot with 30 lbs of baggage and three 170 lb passengers (no baggage).

In 1964 when I started Lear Jet's Sales Engineering Department, all of our flight profiles were based on 170 lb occupants each with 30 lbs of baggage. Forty years later, we are now in the Big Mac era. The FAA says the average occupant weighs 184 lbs plus their lap tops and baggage.

Even if Raburn can hit his 3,390 lb bogie, an honest range, with four 184 lb occupants each with 30 lbs of baggage, will be closer to Phoenx-Peoria with a tailwind which is fine if they want to visit the Peoria based Caterpillar Tractor Company.

For every 100 lb of weight growth over 3,390 lbs, the range will drop by about 120 miles. My guess is the airplane will weigh empty more than 3,700 lbs resulting in range closer to 800 nm with four occupants and reserves.

One would think this would be a real issue at Eclipse. But in spite of the fact they have five prototypes flying, none are fully equipped with upholstered fully certified seats, sound proofing, de-ice and anti-ice equipment, so no one can be sure what the final weight will be.

Nearly all new airplanes miss their empty weights. The F-35 Joint Strike Fighter, the Airbus 380 and even Boeing's 787 Dreamliner are all reported to have weight problems (in the case of the 787, some report a 3% or 7,176 lb overweight problem).

In the early 80's, Learjet was developing the Model 60. Thanks to a strong order book, the next available delivery position was serial 186. By the time the aircraft was certified and a fully equipped aircraft put on the scales, the empty weight had ballooned causing a loss of range/payload . The result, over half the prospects canceled their purchase orders.

One final amusing (at least to me) note. The Eclipse website profiles three prospective owners...all skinny guys. Could all of Vern's 2,000+ prospects be skinny guys and gals?

Part 2.

At every opportunity, Vern Raburn likes to highlight his use of Friction Stir Welding to speed assembly and cut costs. In an article in Business & Commercial Aviation, Dr. Oliver Masefield, a Senior Fellow at Eclipse, stated FSW eliminated 7,400 rivets and 1,000 hours of assembly time and that an Eclipse airframe could be built in 600 hours.

Talk about your smoke and mirrors! Looking at the green airplane on display at Oshkosh 2004, I would guess that 80% of the assembly is still riveted together. And Senior Fellow Oliver failed to mention that the equipment to do FSW is very expensive as is the tooling. As far as his assembly in 600 hours, that does not include assembly hours incurred at suppliers!

Even that is not a new idea. In the late 70's, Learjet's Board of Directors ordered the Tucson Production Manager Don Howard to reduce assembly hours for the Model 55. Don promptly farmed out all the small assemblies reducing direct man-hours but nearly doubling manufacturing costs.

I had not heard of FSW until Vern started talking about it. I put in a call to Cessna's head of manufacturing research and development and said, "tell me about FSW". He said Cessna had been looking at it for a number of years and continued to evaluate the process. Their issue was the probability of long term corrosion associated with the weld. This concern was validated in an article in the 2003 issue of Forming and Fabrication which consisted of a Q&A with John J. Tracy, PhD, VP Structural Technologies and Raj Talwas, Manager-Unitized Metallic Processes; both with Boeing St. Louis (formerly McDonnell Douglas).

Here are a couple of excerpts:

F&F: Is this effort (reducing costs) progressing smoothly?

Raj: There are still many technical concerns. We are interested in friction stir welding commercial fuselages, wing structures, and wing torque boxes. We want to do that in the next two to three years. We are moving slowly, one thing at a time. There are still a lot of technical hurdles to overcome.

John: One of the two main technical concerns is the tooling required. Because this method applies a downward force, unless the geometry is just right, providing proper backup, clamping to the tooling can be difficult. Backup tooling resists the force of the upper tool. For complex shapes, which is really where we want to go, backup tooling is difficult.

F&F: What is the other issue?

John: The other significant issue in commercial aircraft is corrosion. There is concern about corrosion in certain joints that needs to be walked through.

Raj: When we friction stir join two plates, the tool comes in from above. Sometimes it leaves a defect on the under-side, called a root defect. This defect is so tight, we can't see it with ultrasonic, x-ray, or even dye penetrant inspections. It is not a big concern on a rocket that has a very short lifetime. But it is a big concern in commercial aircraft.

The Eclipse position on the process is that FSW is FAA certified. While I have not seen the documentation, I would guess the certification only applies to structural integrity. I don't believe there is any FAA requirement to address corrosion until it shows up in the field.

In September 2005, the Society of Automotive Engineers held a forum on friction stir welding in Albuquerque. The fee was $600 and I was not that interested in the technology so I did not attend figuring there would be reports of technical breakthroughs or what ever. But I have not heard even one peep out of Vern Raburn after the forum.

Had there been good new out of the event, me thinks Vern would have been on his half-billion dollar soap box crowing about his leadership in FSW technology.

Part 3.

The genius of Vern Raburn can be found in some parts of his business plan. Use sombody else's money and shift as much risk and development costs as possible to the supplier base. The investor is reported to be Bill Gates. He is a big boy and can take care of himself.

The suppliers? Well there has been a huge turnover rate. In my opinion, the lucky ones are out of the program. I know one in particular, Chuck Taylor, the owner of Metal Craft in Cedar City, Utah. Metal Craft is a respected and capable supplier of aircraft structural components. They built a lot of parts that went into the Eclipse prototype. Not anymore, won't go into details, but let's just say it was a messy divorce! Vern calls companies like Metal Craft non-performers, but there are two sides to every issue.

Several years ago, we built tools for the Eclipse as a second tier supplier. Enough to get an up close and personal look at a lot of unrealistic engineering. In the past year, I have no-bid over $300,000 in Eclipse tooling and hope that whoever took on the tooling realize what they were getting into.

I would like to see the purchase agreement between Eclipse and Pratt & Whitney. If the contract reflects a mid-2006 certification date that Vern claims and a fast ramp rate to 1,000 units (2,000 engines plus spares), P&W would need to get cracking...bricks and mortar, production machinery, trained people, scarce and exotic metals. There is not much doubt that P&W can produce at high rates provided they are given sufficient lead time. But if Vern's program stalls or even goes belly up, who is going to be there to pay for those engines? And Pratt will be in a pickle because nobody else is using that engine.

Many have suggested Vern can not build the Eclipse for the $1.2 million price tag. In fact, Cessna President Jack Pelton made that same statement to me two years ago in a private conversation. His comment was based on an internal Cessna study.

Another question is, how does Vern hope to recover his development costs?

With his drunken sailor spending habits, and what has been reported in the press, development costs may approach a half-billion by the time the first production unit will deliver.

Let's run some amortization numbers for these costs not including cost of capital:

Deliver 10,000 units add $ 50,000 per unit to cover development costs
Deliver 5,000 units add $100,000 per unit to cover development costs
Deliver 2,500 units add $200,000 per unit to cover development costs
Deliver 1,667 units add $300,000 per unit to cover development costs
Deliver 1,125 units add $400,000 per unit to cover development costs
Deliver 1,000 units add $500,000 per unit to cover development costs

The obvious conclusion is that this program can only pay off with delivery rates unprecedented in the industry.

I have one other question for God's gift to aviation. What in the hell were you doing during the two years when Pratt was developing a new engine? Probably 40-50% of the certification items could have been cleared during this time. As of the posting of this blog, your seats are not even certified. You did not need a engine hanging on the airplane to qualify the seats or do fuel slosh tests on the wing or dozens of other tests required for certification.

I suspect the certification process was not understood at Eclipse. That was evident when the FAA Type Inspection Aurthorization was scheduled at the front end of the certification program. The FAA will not issue a TIA until the company completely tests the airplane and proves it safe for FAA personnel to board and begin flight testing. The FAA is not in the business of going where no one has gone before. The airplane has to be pretty well wrung out by flight and ground tests before the the TIA is issued.

But I like the weasel words on the Eclipse web site justifying rescheduling the TIA to the end of the program:

"After discussions with the FAA, it was determined that a more appropriate means to accomplish this would be to allow development flight tests to be used for compliance testing as long as we have an appropriately conformed test article and approved Test Plans."

What you should have said is, "now we understand what we have to do to get a TIA!"

But other things have annoyed me as well.

In the fall of 2004, Vern addressed a CEO forum at the Forbes Magazine's New York headquarters. Bloin and goin, he tells them he is using high-end technology like Ungraphics, a full 3-D CAD and CAM system that had been previously limited to use by only the largest of the aerospace companies. I had to burst his bubble and send him pictures of our high-end Unigraphics full 3-D CAD and CAM system Tom Prescott and myself purchased in 1983 (with the help of second mortgages on our respective houses). The system was used to design the Prescott Pusher, which for the record, was the first ever airplane fully designed in a computer. I am still running Unigraphics today in my tooling business.

Then somebody wound his spring again. Off he goes spouting off about his FARO Laser Tracker, being on the leading edge to utilize this new technology that can ensure component accuracy to .0001. There you go again Vern, you have exaggerated the accuracy by a factor of 10, it's .001. The device while very expensive ($120,000), is common in the industry. We have had ours for several years.

So last summer, I'm flying United Air Lines to attend the Oshkosh show. I thumb through United's in-flight magazine and there is an article by Patrick Heck, Manager of Flight Training for United. He announces the United tie in with Eclipse and proclaims this is the first time a major airline will provide training for a business jet manufacturer.

Partick, you have been hanging out too long with Vern, and his BS is infectious. So I had to send you a letter with a copy of an article from the 60's showing a picture of United's Learjet and their plan to train Learjet pilots. For the record, the United training program did not work out so LJ went to Flight Safety and asked them to manage an in-house training program which became the standard for nearly every major general aviation manufacturer today.

Part 4.

Vern claims to have over 2,000 orders which means his production line is sold out for the next 3-4 years. Why then, all the promotion and self-promotion?

Air shows, city marketing tours, advertising and other promotions are all a drain on company resources. If I were running the program ( and I am glad I am not) I would husband all of my resources and concentrate 120% on certification and getting a firm grip on empty weight.

It is almost certain the airplane will come out heavy. If the weight issue is as bad as I think, then you don't have much of an airplane and deliveries will never reach the levels required to be profitable and enable Eclipse to recover development costs. Better you cut your losses and fold your tent.

Williams did not do the company any favors when it failed to deliver as promised. So Raburn played the hand he was dealt and tried to recover by using a different engine. But airplanes are point designs and through the re-design it is compromised. For instance, the tip tanks. No self-respecting aircraft engineer would design an airplane in the 21st century with tip tanks. It takes the fuel in one tank just to offset the weight and drag penalty for both tanks.

Raburn has several strikes against him. He is trying to build a cabin-class business jet lighter than what may be physically possible plus he has some real cost recovery issues to deal with. Any added expenditures to try and fix his technical problems will only add to the costs and the difficulty of getting out of his sea of red ink.

It is premature for Vern to be taking bows, but perhaps this is as good as it is going to get.

As for the Collier Trophy, let's just say I am embarrassed for the National Aeronautic Association which selects the winner.