Handicapped A4D-1



Home; Aviation; Cameras; Fiction; Health&Safety; Military; MS-Apps; Non-Fiction; Submarine; Technical; Trains; Watches; Transportation


Handicapped A4D-1

Written by Jack E. Barth, June 15, 1992, a Douglas Aircraft Structural Engineer during the 1940s, 1950s, and 1960s.

The structural design of an aircraft follows a necessary sequence. After a contract is signed with the U. S. Navy, the next step is to agree on a design criteria. This criteria sets weights and speeds as well as many other important requirements. After this agreement, the designing flight conditions are determined. The resulting report is titled "BASIC FLIGHT LOADS". This report determines vertical and longitudinal load factors and, more importantly, the balancing tail loads. The magnitude of the tail loads are critical in determining the wing and fuselage loads as well as the horizontal tail load. Using these loads, the structure can be designed and strength checked.

At the start of the A4D 1 program, I was responsible for the strength of the fuselage. We were seriously undermanned, and I had only three engineers working for me. I assigned them the job of determining the longeron and cover skin loads and the subsequent analysis. I personally analyzed all of the frames and bulkheads. Part way through this program, the F4D 2 project was started and engineers were diverted to that project. As for the A4D 1, the basic flight loads report had not been "formalized" or, of course, submitted to the Navy, even though the drawings had been approved and the shop was building parts. I was then pulled off the fuselage and was directed to complete the flight loads report.

I had never worked on flight loads, and needed some guidance. The engineer, Louis R., who had done the rough work, showed me his calculations and told me it was simple. He had a computer program that would grind out all the answers as fast as the parameters were supplied. I was new to both the aerodynamic coefficients and the tail loads program as well as the large computer. I was not willing to blindly use the program and decided to hand calculate the tail loads for several points on the flight envelope.

The 20,000 foot altitude flight envelope is shown on page 25 of report number 17906. The velocities are equivalent rather than "true." The curved line on the left is determined by Cl max., and the vertical line at 560 knots is M (mach) = 1.25. All velocities between these limits up to 7.0 G and down to 3.0 G are considered. Page 26 from the same report shows the derivation of the balance equation, the equation for the tail load.

I began to get a feel for the affect of each coefficient. After calculating several points on the flight envelope, I began to search for higher tail loads. The aircraft was designed for a maximum tail load of 17,700 pounds and I soon suspected that I was going to find a more critical case and, in fact, I found a flight point with a tail load in excess of 24,000 pounds. Page 29 of this same report shows the A4D 2 condition number 129, and that is the one I found. I, of course, reported my finding to my boss, Pete Shaw. Louis R. was called in to justify his 17,000 pound results that differed from my 24,000 pound load. He soon agreed that he had missed the critical case. His program was O.K., he just had no feel for what coefficients the tail loads were sensitive to.

Now, what effect did this higher tail load have? The design load on the horizontal tail and aft fuselage was increased by 50 percent, and these parts were already being built! We were too far along to accept these higher loads. As a solution, the first 54 aircraft had a restricted flight envelope. The restriction on the A4D 1 is shown superimposed on the A4D 2 V/N diagram. This restriction eliminates those areas in the flight envelope where the tail load would exceed 17,000 pounds. I had caused enough trouble, and someone else finished the report. Louis R.'s reputation was forever clouded, and he left Douglas a short time later.

This incident led me to be cautious of people using computer programs without any understanding of what the program was doing. During the DC 10 project, I questioned some of the stresses at the outside corner of the passenger door. The engineer doing the calculations could only assure me that he had plotted the numbers from the computer correctly, but he had no way to evaluate the correctness. This example is not typical, but it does occur.