In the world of high-stakes defense contracts, its not unusual to see the military say one thing and do another. But when the Air Force selected the British-made Slingsby Firefly to replace its aging fleet of Cessna T-41s six years ago, even some industry insiders were baffled.
Slingsbys initial entry was so wide of the contract specifications-it was 50 knots slower than specified and had fabric control surfaces, a throwback to the Air Forces biplane days-that it wasnt even flight tested against other aircraft from Mooney, Stoddard-Hamilton, Saab and SIAI Marchetti.
Yet in one of the greatest come-from-behind stories in light aircraft history, Slingsby feverishly re-worked the airplane, installed a larger engine, teamed up with Northrop and snagged an Air Force order for 110 new trainers, plus spares, right out from under the noses of the other companies.
Within a year of the first delivery in 1994, the T3A had posted a sorry operational record. It was plagued by frequent, inexplicable engine stoppages on the ground and in the air and its propensity to depart controlled flight with little provocation prompted the Air Force to restrict the T3A to little more than a touch-and-go trainer, not the full-blown aerobatic workhorse it was intended to be. Students who soloed the old Cessna T-41 in 11 flights, required 18 to safely master the Firefly.
Worse, the engine stoppage problem defied solution and three fatal accidents within 18 months led to grounding of the entire T3A fleet, pending a broad Air Force review of the program. As we go to press, the fleet remains parked, with no solution expected until May of 1998.
What follows is the story of the Slingsby Firefly, a program that has gone seriously awry and whose problems came to light when the staff of our sister magazine, Light Plane Maintenance, was asked by officials at Edwards Air Force Base to consult on the T3As engine problems.
Interestingly, the civil GA connection here is that unless the Air Force finds an acceptable fix, Mooney, Stoddard-Hamilton and others may yet get another stab at the trainer plum. In Mooneys case, that would spin off a dandy two-place sport cruiser. (See page 22.)
A New Trainer
The T3As were bought as part of the Air Forces Enhanced Flight Screener program and not, technically, as trainers. By giving them a taste of how the Air Force flies, the flight screening program was supposed to cull candidates with no aptitude for piloting and thus would, theoretically, save the Air Force money by reducing the washout rate.
The new trainer would be flown from the Air Forces contract site at Hondo, Texas and from the U.S. Air Force Academy at Colorado Springs.
At the outset, some Academy instructors objected to the plan, reasoning that only 30 percent of pilot candidates go into fighter aircraft, making acquisition of an over-the-top aerobat unnecessary.
A common question at the time, says former Air Force Captain Richard Hartlaub, was Why are we spinning students during flight screening? The airplane was simply a screener to determine who qualified to enter undergraduate pilot training (UPT) after graduation from the Academy.
Hartlaub instructed in the T3A at the Air Force Academy, as did his wife Kim. He recently left the Air Force, in part because of his disgust with the Firefly program.
Despite objections from the ranks, the Air Force issued a request for proposals to major aircraft manufacturers through standard Air Force channels, listing the requirements the new airplane should have in order to compete for the contract.
Six companies proposed fly-off entries, including Schweizer, Saab, Stoddard-Hamilton with its Glasair III, SIAI Marchetti, Mooney and Slingsby, a relative unknown in the U.S. but a familiar name in Europe, mainly for gliders.
As they usually do, the Air Force contract listed specific performance requirements, having to do with speed, roll rate, climb capability and, one unusual requirement, resistance to potential hail damage from the storms common in Texas and Colorado.
Further, the entries would have to meet FAR Part 23 certification requirements which, as Mooney engineer Tom Bowen says, was something else entirely and somewhat unusual for a military spec.
And, as is often the case, none of the aircraft entered in fly-off met Air Force specs right out of the chute. The Saab Safari was shy on cruise speed and the Mooneys roll rate was a 1/2 second slower than the six-second maximum specified by the Air Force.
But the real shocker was that the Slingsby entry, its T67M with an 200 HP IO-360 Lycoming purpose-built as a trainer, didnt meet the specs closely enough to even be test flown against the others.
Despite the fact that Mooney and Schweizer were willing and able to make modifications and try again, they were kept in the dark about the results of the fly-off.
Yet at one point, during testing at Wright-Patterson Air Force Base, Mooneys Bill Craig was told by an Air Force colonel that this project is yours to lose. Craig figured the real competition would be from Marchetti, not Slingsby.
While Mooney, Schweizer and the others cooled their heels, Slingsby feverishly reworked the T67M, installing a 260 HP AEIO-540 Lycoming and performing other tweaks to meet the Air Forces wants.
Quite to everyones surprise, the T3A emerged from Britain, with a fresh CAA type certificate that would, theoretically, at least, meet FAR Part 23 requirements.
Yet with its fixed gear, the new T67M260 (later T3A) was still 50 knots slower than the 180-knot Air Force requirement and Mooneys Craig questioned how well the Slingsbys fabric control surfaces would stand up to hailstorms. (According to documents obtained by LPM, this contract requirement was mitigated to allow the T3A to pass the Air Forces muster.)
Firefly Wins
In the end, Slingsby won the contract, despite the fact that Schweizers bid was $3 million lower and Mooneys was more than $6 million less than British companys offer. In all, 110 airplanes were delivered, beginning with the Air Force Academy in February of 1994.
Then the trouble started. Since its introduction, there have been more than 50 ground and air engine stoppages, mostly when the power was reduced for landing or for maneuvers. The Air Force reported engine stoppages just after start-up and before taxi, regardless of whether the engine was hot or cold and regardless of what actions the pilots took to keep the engines running.
Figuring the problem was related to vapor lock, the Air Force cut louvers in both sides of the lower engine cowl. Because the engine was equipped with an automatic mixture control, some in the Air Force believed the problem was with the servo and ordered replacements for the components that were presumed defective.
This action began almost immediately after the airplane was placed in service, along with multiple mixture and idle adjustments and replacement of the flow divider cut-off spring.
Regardless of how the operational anomalies were addressed, the T3A engine continued to run rough and exhibit erratic behavior. No amount of pilot input seemed to help. Electric boost pump activation, mixture-control inputs and jockeying of the throttle had little effect on the engine. Occasionally, the engine just quit without warning; no pilot response would bring it back to life.
In response, the Air Force awarded a $10 million contract to Science Applications International Corporation (SAIC) to find and fix the Slingsby problem. Slingsby performed flow tests on the servo while Baylor University and SAIC personnel in Texas, under contract, began testing a fully instrumented and newly refurbished T3A.
No expense was spared; an all new engine and new accessories were used in the flight test. The goal was to measure inflight parameters for fuel system pressures, temperatures and flow rates. Many in and out of the program pointed fingers at the engine and Lycoming was grilled on the performance of its aerobatic IO- 540.
Bubbles in the Gas
When Slingsby installed the six-cylinder powerplant in the newly designated T3A, a new engine mount had to be configured to accommodate the larger size and weight. The LPM staff noticed this while visiting the Air Force Academy. Of prime interest was the firewall-mounted electric fuel boost pump, which is high on the firewall, requiring sharp bends in the fuel lines. Slingsby officials say the pump wasnt moved from the earlier versions but they cant explain why its so high in the firewall. (Most pumps are mounted as low as possible.)
Adding two cylinders pushed the rear exhaust stacks and muffler assembly close to the fuel filter/strainer assembly on the lower right side of the firewall, about a third of the way up from the bottom of the cowling. It was a tight fit that Captain Hartlaub says caused much of the heating problem experienced with the trainer.
Weight and balance also played a role. The larger engine weighed 80 pounds more than the smaller motor found in the T67M. In fact, all of the other variants of the T67 have much smaller engines.
The heavier powerplant required careful review of the T3As center-of-gravity limits and, while the engine was physically made to fit the airplane, the CG was pushed outside its forward limits. But instead of adding ballast to the tail, says Gary Vogt, a maintenance officer at Edwards Air Force Base, Slingsby opted to remove the fuel system header tank located just aft of the firewall.
This did two things: The CG was re-positioned just inside the forward limit (at the firewall station) and the fuel system lost its ability to relieve vapor pressure and air back to the wing-mounted fuel tanks. Both of these conditions can easily cause fuel flow problems.
Other areas of the engine retrofit were not altered to suit the bigger engine. The air inlet box and air filter assembly were unchanged from the original IO-360 installation and the aircraft fuel line routing and line diameters remained the same.
This meant at full power, the alternate air door opened because the airbox and air filter were too small. It also meant that the engine couldnt develop full power. The impact of the smaller fuel lines was more subtle.
Since the header tank had been removed from the system, the fuel lines were now required to carry a larger fuel demand as well as all the vapor and air produced in the system, a physical impossibility. The larger engine simply placed too much demand on the fuel and air systems.
According to Vogt, testing at Edwards showed that fuel pressure to the engine dropped below 3 PSI during low-power operations, a specification which the Lycoming engine type certificate clearly indicates is too low. The engine simply needs more fuel than that to keep running.
Because the boost pump was located at the highest point in the fuel system, it served as a natural collection point for vapor. In addition, the fuel strainer, located just aft of the muffler, was overheating, causing the fuel to boil and giving off even more vapor. With no return provided to the system, the vapor pushed its way through the boost pump, the engine-driven pump and the fuel servo.
Once the vapor reached the servo, fuel distribution came to an abrupt halt and, in many cases, so did the engine. Subsequent testing of the T3A at Edwards confirmed that this is exactly what can happen during simulated forced landings.
Quick Departures
Although trainers are supposed to be docile, the T3A could be anything but, according to Captain Hartlaub. The T3A is, in my opinion, an experimental airplane, says Hartlaub, In no way is the T3A even close to the old T67M. Every one of those trainers should have had their airworthiness certificates yanked because the airplane did not meet FAR Part 23.
Although Hartlaub admits that the Firefly had good performance for a pilot used to its quirks, he also says the T3A is a lot of airplane for a student pilot. It will recover from an unusual attitude or spin everytime, but it also departs controlled flight a little quicker than youd want for a solo student with only 15 hours flying time.
And perhaps an instructor, for that matter. The T3As bad habits were well known among the instructional community at the Air Force Academy and when a Firefly impacted the ground from an unrecovered stall/spin in February 1995, killing both crew, few were surprised.
It seemed like we were all waiting for it to happen, recalls Hartlaub.
Following that accident, the Firefly was restricted from performing the maneuvers it was originally bought to do. The syllabus was stripped down to essentially a touch-and-go trainer and because the T3A required more flight hours to solo but the goal had to be accomplished in the same calendar time, instructors became frazzled and fatigued because of the heated pace. Morale and proficiency plummeted, according to Captain Hartlaub.
Following more engine stoppages and two more fatal stall-spin crashes-one in May of 1996, a second in June of 1997, the aircraft was grounded at both Colorado Springs and Hondo, pending modifications to at least correct the engine problems, if not the Fireflys handling quirks.
The instructors were further irritated by the fact that Air Force investigators labeled all three crashes as pilot error, despite the fact that the instructors involved were experienced Air Force pilots.
Interestingly, all the stall/spin crashes occurred at Colorado Springs with Air Force instructors who had come up through the transport not fighter pipelines. In Hondo, training-actually screening-was conducted under contract by civil instructors, who presumably have more experience with light aircraft stall characteristics. Further, Hondo isnt subject to the high density altitude days that Colorado Springs suffers.
No Comment
Predictably, the Air Force, Slingsby and Northrop have had little of substance to say on the technical questions we have posed about the T3A, pending the outcome of the Air Force review and proposed modifications to the T3A.
The aircraft themselves languish unused on the Colorado and Texas flightlines and, presumably, Academy training and Air Force flight screening have ground to a halt.
As we go to press with this issue, the T3A story was making national headlines in the popular press, with one Air Force general quoted as saying the Firefly has bumped into some rough spots but isnt necessarily done for.
Come May, well see if hes right.
Also With This Article
Click here to view “FAA Response: A Blind Eye.”
-by Paul Brevard
Paul Brevard is editor of Light Plane Maintenance. For a more detailed report on the T3A Firefly, see LPMs November 1997 and February 1998 issues.
