FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1955
1955 - 0347.PDF
FLIGHT, 18 March 1955 347 AIRFRAME FATIGUE A.R.B's Chief Technical Officer Gives the Second Barnwell Lecture THE Royal Aeronautical Society's second BarnwellMemorial Lecture, which took place before the BristolBranch of the Society on March 8th, consisted of a paper The Outlook on Airframe Fatigue, by Mr. Walter Tye, O.B.E., B.Sc, F.R.Ae.S. (chief technical officer of the Air Registration Board). In the chair at the meeting was Dr. S. G. Hooker, of the engine division of the Bristol Aero- plane Co., and also present were the R.Ae.S. president, Sir Sydney Camm, and members of the Society's Yeovil and Cheltenham branches. By a coincidence, the consideration of fatigue also formed an important part of Prof. Pugsley's lecture on Structural Safety, given at Preston two days later and reported on pages 363-365 of this issue. Mr. Tye began his lecture by referring to Captain FrankBarnwell and pointed out that it was difficult in 1955 to realize the problems Barnwell and his colleagues had faced in 1905,the year of the first Barnwell designs, as the changes in aeronautical engineering had been so great.Turning to the subject of fatigue, this had been recognized as a distinct form of failure about one hundred years ago, the lecturercontinued, and was first discussed in a paper in 1854. Since then much empirical data had been collected, but the bulk of thishad little relevance to the structure of aircraft. The absence of data had not seriously embarrassed the aircraftdesigner before the last war. When failures occurred they were rarely catastrophic and often associated with some obvious sourceof vibration. Often they were attributable to poor detail design and hence were not difficult to cure. Meanwhile in the years preceding 1945 a number of changeswere occurring. These had been examined by R. V. Rhode, who had concluded that the chief factor which had tended to reducefatigue life was the improvement in static structural efficiency, which had allowed structures to be designed to function at higherstresses. Rhode's calculations showed a staggering reduction in fatigue life of wing structures between 1934 and 1946. Owingto the concentration of effort in improving the structural efficiency with respect to static loading, a demand arose for light alloyswith higher ultimate and proof stresses. On a comparative stress basis, the new alloys such as DTD.363, 364 and 683 had nobetter fatigue properties than the earlier alloys. Thus, for struc- tures of equal static strength, a reduction in fatigue life occurred,and one example given by Rhode showed about a fivefold reduction in fatigue life in transferring from 24 S-T to 75 S-T material. Another more obvious change was the demand in civil aircraftfor longer lives, 30,000 hours being considered reasonable whereas earlier 10,000 hours would have been considered remarkably high.Probably it was this trend rather than the more insidious and far reaching effects previously described that caused a few peopleto think about the fatigue problem. The designers of early post-war aircraft were by no meansinactive in regard to fatigue. Tests were done on typical joints and this work indicated the considerable improvements of fatiguelife which were possible by attention to detailed design. In addi- tion to this ad hoc testing preparations were made (in particularat the R.A.E.) to install large testing machines, and designs of suitable gust-counting recorders were put in hand. One of theobjects was to enable the lives of many components, such as !20r §100 S 2 60 Fig. 1. Frequency of gusts at different heights (from measurements by Taylor). A striking variation is seen between low and high altitudes. 10 20 30 HEIGHT ABOVE GROUND (ttx ipOO) 40 spars, to be established with some certainty. Research work intofatigue, however, had not been proceeding very quickly. It was in fact remarkably slow in the light of subsequent events, but itmust be remembered that there was then no widespread agreement on the importance and urgency of the problem.Important contributions to knowledge of fatigue were made by Pugsley in 1947 and Walker in 1949. Expert opinion as to theseriousness of the problem, however, was by no means unanimous until late 1951 when two events occurred which swept away thedoubts. During an inspection of a fatigue-suspect area of a Viking one of the two lower spar booms was found to be completelysevered by fatigue. At about the same time a Dove in Western Australia suffered a fatal accident due to fatigue failure of thesingle lower boom of the centre-section spar. A sister aircraft which had flown a similar number of hours was found to beseriously cracked in the same region. Following these events an all-out drive was started to establish safe lives for the sparsof wings of existing aircraft; and the operators, the industry, the R.A.E. and the A.R.B. had since worked together towards this end. At first it seemed that tests on the apparently critical portionof spar would suffice. However, an exploratory fatigue test on a complete wing by the R.A.E. showed the value of this typeof test, as the principal failure occurred at a point other than those sections previously considered critical. With the general introduction of pressure cabins after the warthe need for repeated-loading tests on pressure-cabin structure became evident. Most of these tests were made on full-scalerepresentative structures and not on a full-length cabin. The object was primarily to assist in making a reliable structure andit was considered sufficient to include representative samples of typical windows, doors, etc. No specific attempt was made torelate such test-results to a safe fatigue life of the cabin. This attitude towards pressure-cabin tests, the lecturer con-tinued, persisted until 1954 when, following the two Comet accidents, the R.A.E. made a special tank to accommodate afull-length cabin with its wings attached. In this a Comet 1 was tested and a catastrophic type of failure occurred after theequivalent of 9,000 hours' flying, the origin being a small fatigue area which developed into a crack several feet in length. Analarming feature of this test result was the demonstration that skin cracking was not in this case an innocuous matter capableof inspection before reaching dangerous proportions. "Thus, coming to the present day", Mr. Tye said, "we are finding ourselvesfaced with the prospect of extremely costly fatigue-testing embrac- ing tests on components such as spar sections and window cut-outsand culminating in a full-scale check on a co-nn'ete a;rfrrnie." Mr. Tye turned next to discuss problems connected with thelife of spars and in particular determining a saie va.ue of the life. The basic ingredients of the estimate were the losd snec-trum to which the spar was subjected and the S-n curve of the spar. The main fatigue loads in the lower flange of a spar were usuallycaused by gusts so that if the pattern of gusts in the atmosphere were known together with the operational characteristics of theaircraft, it was possible to assess the load spectrum applied to the spar. Information on the magnitude and frequency of gusts hadcome principally from the use of the R.A.E. counting accelerometer, and an analysis of this information had shown that (1) the relativefrequency of gusts of various severities was reasonably constant; for instance, 100 gusts over 10 ft/sec occurred for each gustover 25 ft/sec, and (2) the absolute frequency of gusts of a given severity was primarily dependent on the altitude above the localterrain. Although the time spent in climb and descent at low altitudeswas normally a small proportion of the total flight time, the actual increased frequency of gusts near the ground (see Fig. 1)resulted in this part of the flight being the most damaging. For instance, in a flight of two and a half hours (cruising at 15,000ft),of which only 12 minutes—8 per cent of the total—was spent below 4,000ft, two-thirds of the total fatigue damage occurredduring this short period. Because gusts were so infrequent at 10,000ft or above, the direct benefits to be gained by high-altitudecruising were very small. The high-altitude aircraft gained chiefly from the fact that it climbed more quickly through the loweraltitudes and, in association with longer flight times, this minimized the proportion of time spent at low altitude. Three factors had become prominent due to the importance ofthe time spent at these low altitudes. Firstly, it was important to obtain far more data on gust frequency at these heights; unfor-tunately, but inevitably, only a small fraction of the total records provided relevant data. Secondly, one suspected that such mattersas the nature of the ground overflown would have more bearing
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
