L-3805’s (Aero Commander) first flight from the Ron Smith Collection

LOOKING BACK Chapter 8, Part 3: The Aero Commander Takes Wing

For his book Stars and Commanders, Dave Duntz conducted years of research, including with Ted Smith’s unpublished memoir. The memoir was published in chapter 8 of Duntz’s book, and will be presented here in three parts. This is the third installment. You can purchase Duntz’s book at www.starsandcommanders.com.

During the first flight of the prototype, Johnny Martin had made a quick check of stall characteristics and at that time it appeared we would be okay. It is necessary, during certification tests, to determine stall speeds, following airspeed calibration as the stall speeds are the basic parameters from which other flight testing is based. These speeds vary a small degree between forward and aft center of gravity ranges and the basic loads. Structural criteria are set up with a specified range of center of gravity travel. We established the basic loads for the primary structure to handle what is termed a rectangular envelope of C.G. travel with a center of gravity range of 20% to 35% of mean aerodynamic chord. Many aircraft have a sloping envelope with a specified gross weight (generally lower) at the forward end of the envelope since in some cases trim power is not sufficient or the forces at the wheel get too high or the airplane becomes unstable generally at aft C.G. locations.

In the stall regime it was found that the aircraft was capable of handling the full range of C.G. travel, power on or off, gear and flaps up or down, and in any attitude. The airplane was controllable through all configurations in the stall regime and no wing drop in any attitude or configuration including very high angle of attack power on stalls.

In static stability we had slight difficulty in two areas: one with the rudder and the other with elevator. Regarding the elevator, the problem was not unusual for a high-performance aircraft and was not a serious problem to overcome. In a high-performance aircraft such as the Aero Commander, with its highly balanced surfaces, one generally finds that when the elevator is placed in a relatively high angle up position, that the aerodynamic balance unports and lets the elevator more or less float and the forces become very low. This tendency of the elevator to float was found during a check for static longitudinal stability at aft C.G. as the speed was reduced with up elevator to a speed of 1.1 times the stall speed. The tolerance for the return to the original trim speed is plus or minus 5 M.P.H. In our case the airplane would nose down when forces at the wheel were relaxed, but it passed through the trim speed and kept nosing over steeper and steeper without any tendency to return to a nose up position.

The fix: Very simple. Provide a down spring in the elevator system of sufficient tension to provide the opposing force in the system to hold down the elevator float and also increase the elevator force gradient at the wheel.

For the prototype we quickly installed a pair of rubber bungee rings of 3/8 diameter with one end attached to the elevator bell crank in the tail of the airplane and the other end attached to the structure. With full up elevator the bungee was putting in a down spring force of about 30 pounds at the control wheel.

For the prototype we quickly installed a pair of rubber bungee rings of 3/8 diameter with one end attached to the elevator bell crank in the tail of the airplane and the other end attached to the structure. With full up elevator the bungee was putting in a down spring force of about 30 pounds at the control wheel.

The installation was made in less than two hours, and we were once again in the air climbing for altitude to recheck the static longitudinal stability at aft C.G. Bert trimmed the airplane for 1.3, gradually reduced speed at 1 M.P.H. per second according to requirements to 1.1, relaxed aft pressure—the nose of the airplane gradually pitched over and returned to the exact trim speed with “0” tolerance.

On the downhill side, it was the same way. The nose slowing pitched up and stabilized again at the 1.3 trim speed, again “0” tolerance. Our first guesstimate of force required was correct. We later tried a heavier force in the bungee, but we did get a slight divergent characteristic, so we kept the force that was first established and both our aft and forward static stability regimes had been satisfied. We ran into a little trouble later on when the first production airplane was certified out of the Fort Worth region where they would not accept the rubber cord bungee. In time the rubber bungee would deteriorate and so would its effectiveness. Our answer was to replace the rubber cord bungee with steel springs.

Static stability with the rudder at first went perfectly. When the prototype was yawed 20 degrees or more to left or right, holding the wings level with aileron it returned to the original heading exactly. The prototype with its large vertical surfaces, 50% of which was in the rudder, gave dead beat static directional stability in this mode. This meant the airplane would return to the original heading prior to a yaw within the 5% tolerance set by the C.A.A.

However, when placed in a side slip configuration to a bank of 30 degrees and aileron pressures released, the wing should return to a wings-level position using top rudder. Here we had a great surprise: with rudder fully deflected to the left, the aerodynamic balance on the rudder would become unported (30-degree rudder travel) and would lock in the full deflected position!

Here we were faced with our first real problem. Theoretically, it would have been cause for discontinuation of the C.A.A. flight tests. However, as stated earlier, the C.A.A., George Haldeman’s group had been extremely cooperative, allowing Bert to act as Company test pilot as well as project pilot for certification by the C.A.A. and the entire project group wanted to provide as much help as possible within the boundaries of the regulation.

On our way down to the landing at Culver City Airport thoughts were being generated in my mind as to how to overcome the rudder lock problem. There appeared in my mind three possible solutions: one to restrict the rudder travel. It was believed that this was not a good solution as it would have an adverse effect in other areas such as stalls, and with one engine out when Vmc {minimum control speed on one engine) had to be determined later in the flight test program. Or, as we had done with the elevator, provide a spring device to provide an opposing force to prevent the rudder from locking in when deflected fully to the left.

Another method that came to mind was to change the shape of the aerodynamic balance on the nose to one of a diamond shape, but this would require a major structural change and would take days to complete. Still another way we used on production aircraft was to provide a dorsal fin. This would also take considerable time to fabricate and install.

It might be mentioned, however, that the rudder locking over could be neutralized by applying a force with foot pressure of about 100 pounds on the opposite rudder.

Upon landing Bert called his boss, Mr. Heimendinger who was Chief of Flight Test for the region and asked that he come over to fly the aircraft with us. Heimi did so and together we decided the condition was not acceptable and we would have to provide some type of fix or redesign.

Quickly putting my mind to work on a simple method of correction, I decided to design a small fitting attached to the floor containing a compression spring that as the rudder was deflected about 60% of its travel, the pedal would contact the spring and as the rudder was fully deflected, the spring would continue to compress and, therefore, provide the opposing force to keep the rudder from locking in when fully deflected. Worth a try for the prototype and later when the airplane was placed in production, a more refined system could be designed. It seemed only logical in this respect that a diamond shaped nose piece would be a more refined and more professional way to correct the problem for production aircraft along with a dorsal fin. As it turned out later this was the way production aircraft were built and with the new nose and dorsal, the rudder pressure gradient at the rudder pedal was increased about 30% and the force was more linear and had a much better feel.

Anyway, that afternoon I built the spring system described and had it installed in the aircraft within about three hours, leaving time in the day to run a flight check to determine its effectiveness. It worked well and another problem was resolved with a very simple device. The only adverse thing about the spring was that the first 60% of rudder travel was normal; the last 40% of the rudder push was against the spring and outside of being a little unusual to have normal build up forces with right rudder, it was a little strange to have an asymmetrical feeling at the rudder penal on the left side. Nevertheless, it did the job and was accepted by the C.A.A.

Takeoff and landings were conducted simulating 50-foot obstacles and the response of the aircraft was excellent. Trim changes for the landing configuration were negligible; placed on a heading and a 10-degree angle of descent, it would hold heading with no deviation, steady and inherently stable, aileron control was extremely responsive, as were rudder and elevator, but the rate of roll with ailerons alone were in the order of 90 degree per second or better depending on speed.

To establish Vmc or minimum control speed, it must be demonstrated that when full power is suddenly lost on one side, the airplane should not yaw more than 20 degrees toward the inoperative engine, then returned by rudder to the original heading and continue to maintain normal flight with the inoperative propeller in its low drag configuration. For our aircraft there was no full feathering propeller so our minimum drag position would have to be established by awindmilling propeller without power which does provide additional drag to the aircraft. We did all of our single engine work with a windmilling propeller which resulted in very conservative results.

We continued our single engine flight test, making 30-degree banked runs into and away from the inoperative engine. The airplane proved to have excellent control in these regimes. Bert was pleased. After each day’s flight Bert would reduce all data collected with the assistance of a flight engineer from the C.A.A. Bert was dedicated to his position in the certification program, and as each day passed, he became more and more interested in seeing the airplane through the many tedious hours of flight test toward ultimate certification. He was a great person to work with, and we all deeply appreciated his efforts on our behalf. Johnny Martin from Douglas, who was an excellent engineering pilot, could not devote the necessary time to the program.

I did not at that time have very much engineering flight test experience nor did I have a multi-engine rating. Pete Leaman had neither a multi-engine rating nor was he engineering oriented to act on our Company’s behalf as an engineering flight test pilot. So, it was all on Bert’s shoulders with the blessing of the Chief of the Region, George Haldeman. We were fortunate to have this kind of cooperation and were grateful for it.

By now both Bert and myself were, I believe, a little over-confident since the majority of the flight tests had been completed and had gone well with the exception of the two problems mentioned. We did have to install a safe flight stall warning system as stalls with and without power gave little or insufficient stall buffet warning prior to the stall. I decided future aircraft that I design and build would be configured such that heavy aerodynamic buffet would be designed in so that no artificial means of stall warning would be necessary and in later years we will see that this commitment was fulfilled.

Bert and I went out one evening to check single engine stalls. We positioned ourselves over the water at about 2000 feet at the west end of the Santa Monica Pier. Bert reduced the power on the left engine to “0” thrust and applied full power to the right engine. Speed was reduced by raising the nose with elevator and as I recall the speed reduction was a little rapid as we were both confident nothing out of the ordinary would occur, since stalls in other regimes had been previously shown to be mild and fully comfortable, but we were now in for a big surprise: The airplane stalled without any warning whatsoever, rolled rapidly to the right; Bert reached to pull the R.H. throttle back, but his hand missed the lever. I then reached over and brought the R.H. throttle full back. The airplane, by this time, had assumed a very steep nose down attitude in a right turning gyration. The time was late evening and it was dusk. Bert made instant recovering following the episode, but to this day I remember the main gear wheels almost touching the little white caps of Santa Monica Bay.

We had come close and neither of us showed much concern at the time, but late that night I awoke in a cold sweat with the final realization. I guess that we had come within fractions of seconds of making headline news in the next morning papers.

Once again, we were blessed by the Old Man Upstairs having his hand on our shoulders. Vd (max dive speed) was satisfactorily completed and immediately following a post Type Board meeting with the C.A.A., it was concluded that we had met all of the flight test requirements of Part 4 flight. All engineering drawings had been approved but one thing remained: service testing, then a requirement of the C.A.A. and still is today in certain instances if the aircraft is of unusual design or has features where historical data is not available to work from. After discussion it was agreed that 100 hours of service testing would be necessary to conclude type testing.

The crew was established – Bert Bantle, Ernie Rice and myself. Ernie was to keep an accurate record of hours flown, the number of takeoffs and landings, the type of runways used, and an accurate record of any malfunctions of any type whether it be airframe, engine, propeller, landing gear, a check on brake wear, propeller operation, and all accessories. By this time we had been able to purchase and install a pair of Aeromatic propellers. As explained the Aeromatic propeller, due to its inherent design containing counterweights acting against aerodynamic forces on the blades, was theoretically intended to maintain constant speed and to automatically feather in the case power was lost from either side of the aircraft. In theory everything was convincing, but in practice during a climb the pitch angle of the blades would become too high to obtain good climb performance and with an engine out, the propeller would not feather or even semi-feather. In straight level cruise condition, the Aeromatic propeller gave excellent performance, but with its other shortcomings it was not a satisfactory propeller.

We started our 1 cross-country service test flights in May 1950 and, since we started late—about 6:00 P.M., we landed at Palm Springs for an overnight stop so that we could get off to an early start the next morning.

Taking off from Palm Springs the next morning we flew to Winslow, Arizona for our first landing fueling since we still had a limited fuel supply, but this was good from another standpoint as it gave us the opportunity to make more takeoffs and landings than if we had more fuel for longer stage lengths.

I was as much interested in potential breakdowns as was the C.A.A. We flew on into Albuquerque, New Mexico and stayed the night. Next day we flew on into Oklahoma City to visit with Amis and others who were interested in the aircraft. Then on east from Oklahoma City arriving in Indianapolis on the evening of May 29. Here Bert met some of his aircraft industry friends: Fish Salmon from Lockheed, Tony Levier from Lockheed, and I, too, met some friends from Douglas.

Before the evening was over, we had all obtained tickets for the Indianapolis 500-mile Memorial Day race to be held the next day. We all slept in the large dormitory on the second floor of Roscoe Turner’s big hangar that was provided with sleeping cots, lavatory facilities and showers.

The service flights were continued and the aircraft performed beautifully with no problems or consequence. By the time we returned home we had flown coast to coast. The important and most pleasant to us all was the completion of the type certificate service tests without incident and without any breakdowns. After the C.A.A. studied the service data accumulated on the trip, George Haldeman issued the type certification on June 30, 1950. A treasured document indeed! It was the result of hard persevering work by the C.A.A. and our own dedicated group of people. The type certificate, dated June 30, 1950, issued to us, was the first type certificate issued in the region after the change had been made for the C.A.A. to become the Federal Aviation Administration (FAA).