P.SATRE J.AIRCRAFT E Table 7 Effeet of cold working on the creep strength of AU2GN 12 AU4G1￥2021 g:130c AU2GN兰2618 Drop in creep strength, Condition % Quenched and tempered 0 Quenched and stretched 1.5,tempered 16 0.4 AU2GN Quenched and stretched 6%,tempered 23 Quenched and stretched,tempered,stretched 1.5 40 0 10.000 20000 30.000 40.000 under 20 cyeles/hr approximately.This is schematized in Fig.15. Fig.1I Creep strain of AU4GI and AU2GN sheet metal I should like to add that,from the fatigue standpoint, specimen (ONERA results). testing of large structural clements bore out the structural solutions we adopted,in addition to the choice of the alloy out of the 30,000 hr of life,the test was designed to guarantee itself. initially.Based on the accumulated in-flight experience,the In short,we have developed satisfactory solutions with actual fatigue life will be much greater. AU2GN for most of the airframe,though it also includes Moreover,these damages occurred only on a limited steel-mainly for the landing gear-and titanium.I should mumber of aireraft (usually fewer than 10 and never more than like to repeat here that to put the question "titanium or no 20)out of more than 200 CARAVELLE in service,and titanium'is yet another misconception.Just what use is generally did so in the course of the first 10,000 flight hours. made of titanium depends primarily on the mission envisaged We consequently place great reliance on this over-all for the aireraft.I have already said that titanium was no test which enables appropriate preventive measures to be doubt necessary at Mach 2.5,but not at Mach 2.I would taken in the great majority of cases.Yet it is no more add that even it is not necessary,its use can be justified all than a check and an ultimate precaution,intended merely the same,though of course it would be introduced gradually. to bear out the struetural options we made.How did we Subsonic aircraft use titanium.So far the CONCORDE go about making these choices? has made seant use of it:about 1.5%of the struetural The fundamental criterion in the choice of a light alloy weight on the production aircraft,whose definition is in the was the resistance to creep.Figure 11 shows the good process of being frozen. strength characteristics of the selected alloy (AU2GN Yet in the prototype defined three years earlier,the figure similar to your 2618)and why a very well-known alloy like was only 0.5%,and there can be no doubt that it will rise AU4GI (similar to 2021)had to be discarded in spite of all further still.For instance,we are studying the possibility its qunhties:good mechanical properties and impervious- of introducing titanium rivets instead of the present monel ness to stress corrosion,and the fact that it is an alloy which rivets.If it is adopted,this modification alone will raise the is now fully mastered in the industrial sense proportion of titanium by 0.5%.If a somewhat higher speed Our studies have since led us to set a limit of 0.1%on were envisaged for a secoud-generation CONCORDE.the creep strain (Fig.12).It should be noted also that cold- leading edges could in turn be made of titanium.However working reduces the creep strength of AU2GN,as shown in I shall not dwell on this any louger as there are a few more Table 7.All conventional sheet-metal forming had therefore misconceptions I must dispel. to be ruled out on CONCORDE,which implied some revamp- ing of our workshops.In fact the CONCORDE was in- Improving Reliability and Safety strumental in bringing about another change as well:it was partly responsible for the advances made in numerically- Although we now think this problem has been fully mas- controlled machine-tools in Europe. tered,we must inelude it among what I have called the true Figure 13 shows a 20-ft panel being machined by such problems.For,I must say,to have applied existing rules methods,and Fig.14 a frame web machined from the solid, purely and simply would not have made any sense. where 335 lb are machined down to 25 lb. As I have already said,we are not building a powered Becuuse it is desigued for creep strength,the aireraft is buggy.Yet some of the things one reads or hears suggest over designed for creep/fatigue interaction at frequencies of that that is precisely the way the SST is thought of in some CREEP AT 150C X 1750c 0.05 CT88图CREEP AILURE车 15 0,01 WITHOUT CREE中CRACKS 90 TES 51 OUT PRIOR CREP 3 104 103 Fig.13 Numerically controlled machining of an upper Fig.12 Post-creep fatigue strength. wing skin panel.P. SATRE J. AIRCRAFT 0.8 30,000 40,000 TIME(HOURS) Fig. 11 Creep strain of AU4G1 and AU2GN sheet metal specimen (ONERA results). out of the 30,000 hr of life, the test was designed to guarantee initially. Based on the accumulated in-flight experience, the actual fatigue life will be much greater. Moreover, these damages occurred only on a limited number of aircraft (usually fewer than 10 and never more than 20) out of more than 200 CARAVELLE in service, and generally did so in the course of the first 10,000 flight hours. We consequently place great reliance on this over-all test which enables appropriate preventive measures to be taken in the great majority of cases. Yet it is no more than a check and an ultimate precaution, intended merely to bear out the structural options we made. How did we go about making these choices? The fundamental criterion in the choice of a light alloy was the resistance to creep. Figure 11 shows the good strength characteristics of the selected alloy (AU2GN, similar to your 2618) and why a very well-known alloy like AU4G1 (similar to 2021) had to be discarded in spite of all its qualities: good mechanical properties and impervious￾ness to stress corrosion, and the fact that it is an alloy which is now fully mastered in the industrial sense. Our studies have since led us to set a limit of 0.1% on creep strain (Fig. 12). It should be noted also that cold￾working reduces the creep strength of AU2GN, as shown in Table 7. All conventional sheet-metal forming had therefore to be ruled out on CONCORDE, which implied some revamp￾ing of our workshops. In fact the CONCORDE was in￾strumental in bringing about another change as well: it was partly responsible for the advances made in numerically￾controlled machine-tools in Europe. Figure 13 shows a 20-ft panel being machined by such methods, and Fig. 14 a frame web machined from the solid, where 335 Ib are machined down to 25 Ib. Because it is designed for creep strength, the aircraft is over designed for creep/fatigue interaction at frequencies of Table 7 Effect of cold working on the creep strength of ____ AU2GN Drop in creep strength, Condition % Quenched and tempered 0 Quenched and stretched 1.5%, tempered 15 Quenched arid stretched 6%, tempered 23 Quenched and stretched, tempered, stretched 1.5% 40 under 20 cycles/hr approximately. This is schematized in Fig. 15. I should like to add that, from the fatigue standpoint, testing of large structural elements bore out the structural solutions we adopted, in addition to the choice of the alloy itself. In short, we have developed satisfactory solutions with AU2GN for most of the airframe, though it also includes steel—mainly for the landing gear—and titanium. I should like to repeat here that to put the question "titanium or no titanium" is yet another misconception. Just what use is made of titanium depends primarily on the mission envisaged for the aircraft. I have already said that titanium was no doubt necessary at Mach 2.5, but not at Mach 2. I would add that even it is not necessary, its use can be justified all the same, though of course it would be introduced gradually. Subsonic aircraft use titanium. So far the CONCORDE has made scant use of it: about 1.5% of the structural weight on the production aircraft, whose definition is in the process of being frozen. Yet in the prototype denned three years earlier, the figure was only 0.5%, and there can be no doubt that it will rise further still. For instance, we are studying the possibility of introducing titanium rivets instead of the present monel rivets. If it is adopted, this modification alone will raise the proportion of titanium by 0.5%. If a somewhat higher speed were envisaged for a second-generation CONCORDE, the leading edges could in turn be made of titanium. However, I shall not dwell on this any longer as there are a few more misconceptions I must dispel. Improving Reliability and Safety Although we now think this problem has been fully mas￾tered, we must include it among what I have called the true problems. For, I must say, to have applied existing rules purely and simply would not have made any sense. As I have already said, we are not building a powered buggy. Yet some of the things one reads or hears suggest that that is precisely the way the SST is thought of in some 0,05 0,02 0,01, 0,05 EQUIVALENT HbURS AT 120°C CREEP AT 150°C X 17S°C « «- —— 4 —— 105 2. 10* ___ 1750C ___ 150 «»C O 4 _ ] CREEP AILURE! -* —— v 4 * ' > > WIT CREE •IOUT > CRACKS 9O TESTS WITHOUT PRIOR CREEP • Fig. 12 Post-creep fatigue strength. 2 OF CYCLES AT20°C Fig. 13 Numerically controlled machining of an upper wing skin panel