Found an interesting article from 1962...
Historical note on the 1.5 Factor of Safety for Aircraft Structures
F. R. Shanley, Professor of Engineering, and Head of the Structures Division, University of California, Los Angeles
June 22, 1961
THE ULTIMATE FACTOR of safety of 1.5 was first introduced in the U.S. civil-airworthiness requirements in the early 1930's. Since that time it has remained unchanged, except for special provisions regarding fatigue, pressurization, etc. During the past several years it has become evident that the rather arbitrary use of such a factor may impose a severe penalty on structures such as missiles, space platforms, etc. It seems desirable, therefore, to review the situation that existed when this factor was first introduced.
When I joined the CAA (now FAA) in 1930, the airworthiness requirements (Bulletin 7-A) specified only a "load factor," the value of which was in the order of 6.0 for a typical airplane. This was generally interpreted to mean that the structure should not fail before the airplane developed 6g in flight.
Around 1932 I was assigned the job of "rationalizing" the structural airworthiness requirements. At about the same time the Air Force (then the Army Air Corps) and the Navy (Bureau of Aeronautics) were engaged in similar efforts. The main idea was to relate the air loads more closely to actual flight conditions. This resulted in the introduction of gust-loading conditions, the V-g diagram, the use of moment coefficients, etc.
At that time there were very few statistical data on loading conditions. The best evidence we had was that airplanes that had been designed for the required "load factors" had in general performed satisfactorily. We therefore adopted the principle that, in providing for higher flight speeds, gusts, etc., we should adjust the ultimate load requirements so as to coincide, at least roughly, with those that had been used for airplanes which had good service records.
Everyone concerned realized that the old "load factor" (say 6.0) contained a factor of safety. The big question was, how much of the load factor was actually expected to be developed under the most adverse flight conditions? If we had expected to develop only 3g, for a load factor of 6.0, the corresponding factor of safety would of course be 2.0. Therefore, by specifying a "limit" load factor of 3.0 and a factor of safety of 2.0, we would end up with the same "ultimate" load factor (6.0). (The terms "limit" and "ultimate" were adopted later in a joint meeting of Army, Navy, and Commerce Department representatives, the forerunner of the group that later originated the A-N-C publications.) At that time the most popular structural material was 17ST aluminum alloy. This had a ratio of ultimate tensile stress to yield stress of approximately 1.5. The corresponding ratio for 4130 steel (also widely used) was somewhat lower. This appeared to mean that we had not been developing more than about two-thirds of the total "load factor" in flight, otherwise there would have been many cases of permanent deformation. This is often cited as the basic reason for the choice of 1.5, rather than some other number. However, even at that time I recognized that this was not a very sound argument, for at least two reasons: (a) it did not apply to compression members that failed by buckling, and (b) tension members were almost always critical at joints, for which the efficiency was generally below 80 percent.
The main reason for my favoring the 1.5 factor was very simple. If we had used 2.0, and worked backward from (say) a load factor of 6.0, the "limit" factor would have been 3.0, while if we used 1.5 it would be 4.0. Although we had not yet written any requirements relating the limit load to absence of permanent deformation, I anticipated this as a logical interpretation of limit load and therefore wished to see the limit loads remain relatively high. The only real significance that can be attached to the ratio of yield stress to ultimate stress is that since this ratio was then 2/3 or more, for all materials being used, no penalty was imposed on existing airplanes by working backward from the existing load factors, using a factor of safety of 1.5.
For the requirements on gust load factors, also introduced at that time, the factor of safety had to be interpreted more directly, because the gust load factors were actually "limit" factors. It seemed that by providing for half again as much load as that actually expected in the worst cases, we would be covering all the unknowns, such as statistical variations in material properties, tolerances in manufacturing, etc. Experience has shown that this has been true, for "static" loading. Whether 1.5 is too high is another matter that requires careful and thorough investigation of all factors influencing the final results.
The main purpose of this note has been to clear the atmosphere surrounding the 1.5 factor of safety. It is not "sacred." It was selected in a rather arbitrary manner. Its main virtue is that it has worked rather well for subsonic aircraft except for fatigue. In this connection it should be pointed out that US commercial aircraft have, to some extent, been protected against fatigue failures by at least two supplementary (multiplying) factors of safety, the 1.15 extra factor for fittings and the 1.33 extra factor for pressurized cabins.
It would seem that there is need for an intense study of structural airworthiness requirements, with particular emphasis on factors of safety and related matters. This should include commercial, military, and space vehicles. A truly rational system should apply equally well to all types of structures, because it would permit decisions to be made on the basis of acceptable risk, type of structure, type of loading, material variations, dimensional tolerances, etc. It should include fatigue, aeroelasticity, optimum-design theory, etc. Such a project would be a much bigger job than it was in 1932. It should be a joint effort by all the services, the manufacturers, NASA, the universities, research laboratories, etc. Furthermore, it should he international in scope, at least for commercial aircraft. A start has already been made by various individuals, as indicated by the references following this note (these are by no means complete).
In the meantime some immediate progress can and should be made in specific areas, such as fatigue, missiles, space vehicles, sonic effects, etc. But we must be careful not to throw away anything that we have learned from many years of actual operations of aircraft.
These notes were written from memory and no attempt has been made to document the opinions expressed, or to give credit to the many others who were involved in the early development of the airworthiness requirements. It would be interesting to collect similar notes from all of these people.
Shanley, F. R. (1962, February). Historical note on the 1.5 Factor of Safety for Aircraft Structures. Journal of the Aerospace Sciences, 29(2), 243-244.
