The production T-37 “Tweet” serves up an enduring mystery : Air Facts Journal (2024)

In very early 1952, I was an undergraduate working part time in Cessna’s Flight Test, Aerodynamics and Preliminary Design Group when a request for proposal for the TX came in from the Air Force. The TX was to be the first, that is the primary, trainer in a series of three new trainers which would finish with the TZ, a supersonic one.

While there were many options included for the TX configuration, the Air Force favored a side-by-side co*ckpit, for more personal communication between student and instructor, and specified particular turbojet and turboprop engines that the government would supply if the bidder chose one of them for their submittal.

At that time, Cessna had no divisions so such an undertaking would be handled by our own aerodynamics and preliminary design sections and would be quite a departure from airplanes like the 170, L-19, 180, 195 and 310 which we had or were developing. And wide, diverse competition from companies like Beechcraft, North American and Temco was expected.

Management decided to enter the competition anyway, and I was immediately put on full time in order to work on the proposal. That still fit with my education goals since at Wichita University, in the Air Capital of the World, Aero Engineering classes were offered only at night so we could work in the industry during days.

We started the proposal work with an internal competition of configurations with single and twin engine versions using the powerplants the Air Force said they would supply. We settled on a twin with jet engines, placed inboard so as to minimize engine out problems for the novice students, and fleshed out the design and prepared a five-volume technical description for submittal in the formal competition. We were so short-handed that I was chosen to completely author two of the volumes, concerned with stability and control, and did a major part of the analysis for a third, but did not write any of that related text.

If our submittal were selected, it would be the first on-purpose jet trainer design ever for the U.S., since existing ones were adaptation of operational jets, like the T-33 derived from the F-80. About six months after we submitted our proposal, we were awarded the contract to build and test the prototype XT-37s. Readers will likely know that the supersonic TZ became the T-38, but for some reason there never was a TY, and I’m not sure what its role was supposed to be.

In 1952, Cessna’s general aviation airplane design and production activity was at the Pawnee Plant on the east side of Wichita. It was to become the Commercial Airplane Division and in 1953 a new division, the Military Airplane Division, for the T-37 contracts, was initially placed in another facility on the west side of town, the Prospect Plant (later expanded into the Wallace Plant constructed on Wichita’s adjacent Mid Continent Airport).

Prospect had been devoted to and continued to do subcontract work for Boeing. (XT-37 flight operations were thus out of a large, leased hangar at Mid Continent till the Wallace plant was first opened.) Our own Chief of Flight Test, Aerodynamics and Preliminary Design was chosen to be the Chief Engineer for the new independent Military division, and the engineering staff was made up of selected transfers from Pawnee plus a lot of new hires.

I wasn’t selected to be transferred, and my old boss, the new Chief Engineer, told me that in the bargaining for internal talent he traded me for two structures engineers. That seemed to me like about the right balance – one aerodynamicist for two structural analysts.

So I missed out on almost the first two years of work on the T-37, which was OK because during that time we did some challenging and successful things at the Commercial Division in bringing the 180 and 310 into production, plus one off-the-shelf military airplane, the OE-2 (for the Marines), the initial design and wind tunnel testing of the M620, the initial design and testing for the 172 (preceded by the 170C) and a little later the 182.

In addition we did some boundary layer control flight research for the Navy with each a substantially modified 170 and L-19, and tested turboprop versions of the L-19 for the Army. I was involved in each of these projects in one discipline and extent or another, except the 182. By the way, my contributions to the T-37 proposal were acknowledged and my wife was invited to the ceremonial first flight of the airplane (our commercial first flights were more discreet, if not downright secretive), but I worked all that day on the other side of town.

But then my chance came to join the Military Division when its Chief of Aerodynamics was promoted to Chief Technical Engineer and I took his place. There was still much development work to do on the T-37, which finally was successfully Qualified, the Air Force term for Certificated, and demonstrated emphatically that our efforts had resulted in a uniquely good all-around airplane.

But I want to focus on a period when that development work was finished and production units of the T-37 were being delivered in quantity. We, of course, did production flight tests on completed units, and the Air Force did a random flight check on some, too.

Let me pause, though, and explain that there were several difficulties with the government-furnished jet engines, and it was easy to call them government-furnished engine problems. One was a very loud, high-pitched whine that came out of the inlets of the engines, which Air Force ground and flight crews identified with a dog whistle, and gave the airplane the endearing, and enduring, nickname of the Tweet. (We actually did a project with vanes with sound-absorbing material surfaces in the inlets that reduced the noise to a tolerable level, but lowered the pressure of the air going to the engines and compromised their performance – and the Air Force, thank goodness for them facilitating unusual nicknames, didn’t think the cost increment and performance degradation with the vanes installed was worth the sound reduction achieved.)

Continuing the historical storyline I just interrupted, at this time of volume delivery of production airplanes, we in engineering were more engaged in finding new missions for the T-37, a very economical airplane to buy and fly, and we wanted to use as much as possible of it in these new configurations. It seems incredible now, and remembering that the engines were GFE, that we sold an airplane that could maneuver like a century series fighter and fly at three-fourths the speed of sound at 35,000 feet for only $80,000. There has been a lot of inflation since then.

We emphasized two quite different uses, both for the military – one in which we plucked out the T-37’s co*ckpit and inboard engines and replaced them with a pressurized four place cabin and more powerful engines to provide a small, cost effective VIP transport. Everything else in the model was original T-37 structure, to which we added wing tip tanks to hold the extra fuel needed for cross-country cruising. This idea was pursued with the Air Force for several years without success.

Then, during my time, we did the initial design, and modest testing, of a light attack version of the airplane with external stores of armaments under the wings, higher thrust engines, and more fuel, again contained in tip tanks. This one ultimately became the A-37 (with variations).

For those who know from another article of my concerns about tip tanks on the 310, understand that I was an advocate for them on these two airplanes, because the extra fuel had to go somewhere and we wanted to save as much as possible of the T-37 unchanged. The tanks for these missions differed in that the ones for the VIP transport had a fuel dumping capability, and for the light attack plane they were instead wholly droppable for quick escape capability in combat.

In addition to trying to expand T-37 missions, we entered the competition for an Army target missile, for which funding was then delayed for a couple of years, by which time the Division had found other interests. And I probably should mention a very important non-airplane project: the design and manufacture of the Minuteman Intermediate Range Ballistic Missile’s huge over-the-road Transporter-erector for the assembled, ready-to-fire weapon, which was concurrent with these others undertakings, and definitely was another major draw on our engineering resources and time.

But we were producing Qualified T-37s and some former experimental test pilots were now production test pilots. One day, one thoughtful one of these came to me and said he believed that some of the airplanes became slightly unstable longitudinally at a certain high speed corner of the flight envelope. I said that was impossible because that was precisely one of the conditions tested in Qualification, and the stability was excellent.

He reluctantly accepted that, but came back a couple of days later and said he was even more certain of this anomaly – which would mean that at least some production airplanes acted differently than the prototypes did, as well as different than the production airplanes the Air Force used to confirm the Qualification.

I told him we would provide him with a handheld force gauge with which he could obtain data to show if this deviation really occurred. It did, on about one out of three airplanes! This was not a safety of flight issue, so we gave hand utilized force gauges to all production test pilots to get further statistics and identify the “bad” airplanes. We then did exhaustive measuring of both these bad ones and good ones for dimensions, contours, and even control cable tensions with the intention of finding and fixing the problem.

But we, perhaps because our measuring skills were inadequate, couldn’t find any difference between the “good” and “bad” airplanes. So we began exchanging ailerons and elevators between good and bad ones, and retesting and still no definitive explanation was forthcoming, as incidences of instability were more or less random.

At this point everyone got kind of paranoid and someone suggested that the ratio of bad-to-good airplanes was about the same as the ratio of units completed on second shift to those completed on first shift, and those lunkheads on second shift must be doing something wrong. That theory was also tested and found wrong.

I was amazed that with all the activity on production airplanes, and the many inspectors, plant reps, and test pilots from the Air Force present, they weren’t curious about what was going on. In our favor, the routine flight checks by Air Force pilots had resulted in no reporting of this problem. And by now we had a large number of T-37s being used daily in the field, and there were no complaints about this flight characteristic.

So, although we did some further collection of data, and with our efforts on developing other missions and products in full force, this mysterious and seemingly unsolvable random light instability problem just faded from our concerns. Yes, I, and we were guilty of not ever reporting the problem to the Air Force customer, but the airplane was getting good reviews in the field, and apparently no harm resulted. So not reporting the problem was not intentional, but just sort of extended over time till the matter was completely forgotten. Still I was bothered that we had no explanation for the condition, and that empty box to check off was in the back of my mind for a long time, even after I left Cessna.

Now skip forward several decades, when the company decided to have a congratulatory event for those of us, from all company departments, who had contributed to that very successful T-37, which ultimately served in its original mission for a record breaking 52 years. But they decided to hold the event on the less celebratory 45^th anniversary of the first flight of the airplane, because they felt if they waited for the 50^th many of us wouldn’t be around anymore.

One of those attending was that test pilot who first brought to my attention the light longitudinal instability problem, who had stayed with Cessna for a few more years. So I asked him if a cause of the problem was ever discovered, fully expecting he would say no. But he said yes, that he was flying in formation with another production T-37 and at that flight condition he noticed a “crinkling” in the elevator of the other plane just when the instability occurred. (To those who read my article here at Air Facts on the 180, you will see a similarity to that light instability at high speed that happened only on the prototype airplane, which had a flimsier elevator than the production models.) He was satisfied that his observation solved the mystery for the production T-37, but I was unfulfilled.

Because why would airplanes with structures built in the same manner on the same tools exhibit this difference? And more importantly why didn’t it show up way back when we switched elevators between supposedly good and bad units? To me, even with his valuable observation of the proximate cause, it was a mystery still unsolved. And I guess it will have to remain so. You’re probably just as dissatisfied with this ending of the story as I am.