by L. D. Alford
The T-39, Model 40 Sabreliner, is one of the most forgiving and fun jets to fly. It may be the best aircraft in its class ever built. In the Air Force, we modified it extensively for flight test. One of the most creative and useful modifications was when we installed a pod adaptor on the belly for the use of ECM (Electronic Countermeasures) pods for aircraft and pod flight test. The T-39 could carry an ECM pod for about 15 to 20 times less cost per hour than a fighter. It was a real money saving idea, but it was a modification that many in the government predicted could never be done. I was the test pilot and the project manager. I had the dirty job of encouraging the many naysayers and making sure the modification became a reality.
I luckily was assigned a team of young engineers who were mostly convinced the project could be accomplished. The problem was when they returned to their home offices, their old head engineer bosses gave them a difficult time. I knew the project could be completed and flight tested. I knew we could accomplish it on cost and schedule. This is what I drove my team to do.
Unfortunately, you can’t make an omelet without breaking some eggs. Forcing the government engineering community to recognize that this project could be made a reality was ninety percent of the project. We were living in the era of Total Quality Management (TQM), and this project could never have been accomplished through gentle means. Literally, from the first moment the program started, certain groups tried to kill it. I was lucky my boss, the wing king (wing commander), wing vice commander, and most of my young engineers were behind the program. The rest of my problem was convincing the others and preventing the program from being submarined.
The T-39 Pod Mod was an excellent example of a program that required a test pilot engineer at the helm. Because of my experience and expertise, I was able to review all the program documents, the design prints, and all the blue prints. During the program many changes had to be made because the design engineers didn’t fully understand the aircraft systems. A simple review of the preliminary prints showed the problems with the designs and enabled changes to be made before any metal was cut.
Later, design decisions during manufacturing were very easy for a test pilot again because of familiarity with the aircraft systems. The flight test development was handled by an outstanding flight test engineer. She wasn’t enamored with the project at the beginning, but quickly came on board and developed an outstanding test plan and overall management of the engineering side of the project.
Wright Patterson AMX (quick reaction modifications) accomplished the modification to the aircraft. They were led by another outstanding engineer who made the program move forward in that stage. Unfortunately, the one time I went on a flight test mission to Edwards, the program bogged down and halted for almost a month. All that needed to be down was make two decisions: whether to go off base for heat treating new bolts for the adaptor and the need to do a vibration nodal test to support the flutter analysis. I could have made the decisions in a phone call—yes and yes. They waited until I returned to the base. In spite of this, the program still wasn’t delayed.
We were very lucky we completed a nodal analysis because the vice wing commander’s number one question was about flutter. There is very little chance of flutter in a centerline modification, but it was one of the vice commander’s bugaboos, and a great question to ask about such a substantial modification.
What is flutter, you might ask. It’s not the feeling you get when you look at your love. In aviation, flutter is an undamped (self-increasing) vibration where aerodynamic forces couple with a natural vibration mode of the structure. The resulting oscillating motion will normally end in the failure of the structure. In other words, the wing comes off. This is a really bad thing. All aircraft are susceptible to flutter. The limiting top speeds of most aircraft are based on flutter and must be fifteen percent less than the first flutter node of the aircraft.
Flutter is a big deal, but, as I said, it isn’t usually affected by changes in the centerline of the aircraft. Still, flutter isn’t a phenomenon to mess around with and our vice commander was very concerned about the possible effects of the pod and adaptor to the flutter envelope of the aircraft.
Finally, everything was completed and we were almost ready to move into flight test. This is the point in an aircraft program where the paperwork must match the weight of the aircraft prior to first flight. The trick is to get all the sign offs. And this is where the AM (aircraft modification) branch tried to make a final grandstand to submarine of the program.
I had an incredible electrical engineer who was the deputy program manager and in charge of the EMI/EMC (Electromagnetic Interference/Electromagnetic Compatibility) testing and characterization. He produced a comprehensive EMI/EMC report with test data to support the program. When the head of AM made the final modification signoff, he did so without the EMI/EMC report. Then he turned around and called me to his office to officially complain that we had mishandled the modification because we didn’t have the required EMI/EMC data. The AM branch chief was afraid of the wing king because the wing commander supported the program. That’s why he signed off on the modification, but he wanted to submarine the program by finding the slightest problem with the program that he could use to cut it off. This was in spite of the fact that the modified aircraft already sat on the ramp ready for flight test.
When the AM chief called me into his office, I brought my deputy program manager with the EMI/EMC report in hand. My intelligence system was up and running—I had already heard through Wing HQ what was going on.
The AM branch chief launched right into us and accused us of dereliction of duty. I simply handed him the EMI/EMC report and asked why he signed off the modification if he didn’t have all the data in hand. That turned the tables on him. In actuality, by signing off a modification without all the paperwork, he had breached Air Force regulations and engineering policy. Bad on him. We didn’t hear a single peep from AM for the rest of the program.
We did have some other problems. A week before flight test, the test director and I went to check out the aircraft modification and mark the flight controls for testing. During flight test, you must make specific and measured inputs to the controls and you must be able to measure the movement of the flight controls. We installed a protractor to measure the rotation of the yoke and tapes for the forward and aft movement of the yoke and rudder pedals. The problem is when we rotated the yoke to move the ailerons, the yoke moved almost twice as far to the left as to the right. The aircraft had already been signed off by QA (Quality Assurance). This was bad. We had a flight control issue just before a major flight test.
We reported the problem to QA and the aircraft crewchief. After a night of investigation, they determined the flight control pulleys on one side had been installed the wrong direction—whoops. Lots of heads fell. I’m glad. We could have been dead, or worse, the test could have been delayed.
Finally, we were ready for the flight test. Everything had been signed off, and the aircraft was ready to go. A few weeks prior to this, we took out a T-39A model to get a baseline on the aircraft. We accomplished all of the tests we would make on the actual test aircraft for practice and to get a good baseline read on the aircraft’s characteristics. The fact that we used an A model was a mistake we had no idea would affect the flight test results. We also determined there was no reason to fly the pod adaptor in a separate test. We based these decisions on the data we had from our analysis of the aircraft systems and on the good engineering data. These decisions didn’t hurt the program, but resulted in some problems for our test program and for flight of the aircraft later.
The first fight happened with me as the aircraft commanders and my squadron commander as the copilot. There was no way the squadron commander was going to miss an opportunity like this. This was likely the only first flight we had accomplished on such a highly modified aircraft at Wright-Patterson in over ten and maybe more years. It was the chance of a lifetime in the Air Force since not many test pilots have the opportunity to make a first flight at all. Most first flights are accomplished by contractor pilots.
The first flight occurred on 18 March 1993 and was considered a higher risk than normal flights. We started with taxi tests and moved on to a takeoff. We had a T-39 chase the entire flight. I have video. The flight was awesome. I made the first takeoff and the first landing. The flight test engineer and I acted as a continuous crew for the entire flight test program. We changed out the copilot based on the test requirements and crew availability. This is likely one of the few programs since the 1950s where the same test pilot and flight test engineer were able to complete the entire program from start to finish.
This was similar to the kinds of programs my father participated in as a test pilot for the Boeing Company in that decade. I felt like I was a pioneer of flight test like those great aviators in the past. It was one of the finest aviation experiences in my life.
The flight test wasn’t without issues. Of note our flight test engineer was fantastic. She met the main test of all great flight test engineers—she took and analyzed data flawlessly while using a barf bag. Many of the test points we had to hit are what we call “hiacka.” They required uncomfortable entries and uncomfortable recoveries. The yaw and g limit points are especially bad for crew in the back. Our flight test engineer kept us safe. In fact, she likely saved our lives. The seventh sortie was the only flight where we had to “knock-it-off” and return to base (RTB). This was due to the fact that we exceeded the maximum yaw by two degrees during a steady heading sideslip. A steady heading sideslip is where the aircraft is held at a constant altitude and airspeed while cross controls (rudder and aileron) are bled into the aircraft to the limits of the flight control system. When we made the baseline tests using a T-39A, the limit of the T-39A was easily within the structural limits of the pod and adaptor at a maximum of twelve degrees. When we tested the pod on the T-39B, we exceeded the limits on a high speed test point by two degrees to fourteen degrees total. Our flight test engineer noted the problem and calculated the forces on the aircraft. Based on her analysis, we called a “knock-it-off,” flew gingerly back to base, and made a very gentle landing. We didn’t want the tail to fall off—that would have been bad.
Inspection of the aircraft found no damage to the tail or other structures, but structural overload could have been a real problem. We determined that the T-39B has a relaxed yaw stability compared to the T-39A. The inertial characteristics of the B are very different from the A. The pod and adaptor cause the yaw stability to decrease that makes the rudder more effective on the B. We thought the aircraft would be the same and they aren’t.
The reduced yaw stability of the aircraft was not enough to be a real problem for flight, but it became a WARNING in the flight manual. I discovered later that the pod adaptor also reduced the yaw stability. On a flight for the AF Test Pilot School, while a test pilot student was made a low speed steady heading slide slip (restricted with the pod), he released the controls too quickly. The aircraft almost departed controlled flight. I had to take the aircraft and input anti-spin controls. That scared the “you know what” out of me. I came back and wrote a new WARNING for the flight manual concerning flight with the adaptor only on the aircraft. Since we assumed pod adaptor alone wouldn’t have that much effect on the flight characteristics, we never tested it. Oh well. This just is an excellent example of how difficult it is to predict aircraft performance. Flight test is a necessary inconvenience.
So the lessons learned in this program are these: No matter how close they are, don’t assume different models of the same aircraft are really aerodynamically the same—they aren’t. Second, accomplish some degree of flight test for all expected configurations of a new external modification. External modifications to an aircraft make them different and that difference cannot be predicted. It requires flight test.
The T-39 Pod Mod program was an outstanding program. I got a first flight out of it. I was able to manage a complete program from conception to completion. Air Force flight test gained an asset that was worth its weight in gold. To protect fighter flight test resources, they eventually put the aircraft in the boneyard as a result of budget cuts. That resulted in much higher costs for pod flight testing—especially pod target testing. I think it is stupid to get rid of low cost assets that save costs for the purpose of keeping high cost assets, but that’s the way government bureaucracies work. Someday someone will want to make another low cost pod carrying flight test aircraft. The T-39 Pod Mod program shows that it can be done.
The author is a retired Air Force test pilot. His other aviation, technical, and fiction writing can be referenced at www.ldalford.com.
– The End –