FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1921
1921 - 0196.PDF
MARCH 17, 1921 machines while being ttsted. Vibration often leads to un-expected results, and is beyond the scope of calculation. In one machine brought to the writer's notice the drag wirescontinually broke although strengthened by three to four times the calculated strength. After putting rubber blocksunder the engine no further trouble was experienced, and the original wires were returned. (b) Parts under combined stresses in practice should betested under similar combination of stresses. This point is of the greatest importance. For instance, the principalwood used in the construction of aeroplanes is about twice as strong in tension as in compression, i.e., 5,000 lbs. persq. in. and 12,000 lbs. per sq. in. respectively. When a spruce specimen is tested in pure bending itbegins to fail on the compression side, and as the load in- creases the compression failure travels inward, causing adisplacement of the neutral axis, until the failure is nearly two-thirds across the spar. Typical failures are shown inFigs. 4 and 5. This results in a stress diagram similar to that shown in Fig. 6. Since all systems of engineering calculation are based onthe theory that the material is elastic and bends in the arc of a circle, the neutral axis remaining unchanged, it neces-sarily follows that the apparent maximum skin stress is considerably greater than the actual skin stress at failure. • 8,500 IBs/sHIt^ h fW w beam. When wooden beams are tested by the usual laboratorymachines they invariably fail on the tension side, as shown in Fig. 8. Hence they give entirely misleading results. Distribution of Pressure, Travel of Centre of Pressure, andEnd Effect.—Before proceeding with any method of calcula- Fig. 6. The apparent skin stress in spruce in bending is about 8,500,whereas the actual stress is only 5,000 lbs. per sq. in. No serious error is likely to arise from this misapplication oftheory if the figure 8,500 is used in spars under pure bending. When, however, the spars are under combined bendingand compression as in all wing spars the discrepancy is serious. Then if the spars begin to fail on the compressionside it deflects, and the end load exerts a bending moment which increases the primary bending moment. The resultis that the spars fail at an apparent stress much below the apparent failing stress in pure bending. A safe figure forgrade A spruce in bending and compression is 5,500 lbs. per sq. in.The importance of these points may be gathered from the fact that all machines calculated by the R.N.A.S. for thefirst two and a half years of the War were based on the assumption that the strength of spruce in combined bendingand compression is 8,500 lbs. per sq. in., whereas during the same period all machines calculated by the Military Aero-nautics Directorate and the Royal Aircraft Factory used the figure 5,500 lbs. per sq. in. In other words, all naval machineswere 35 per cent, below calculated strength. As no system of mathematical calculation gives results varying more than3 per cent, the futility of striving for mathematical perfection is apparent.No blame should be attached to the R.N.A.S. calculators for this and other similar practical points which only college•engineers of wide experience and intuition could be aware •of. As however certain of the R.N.A.S. calculators reportedthat the Military aad Factory methods were wrong, this fact should be men-[_ fj Jfc—L tioned in fairness to these departments. Method of Testing Materials. — One ofthe several suitable methods for testingwood is shown in Fig. 7. In this casethe bending moment is applied by twoloads at a distance from the supports,and the end load isFig. 7. applied in a known relation to the bending moment. Thebending moment diagram, which is shown sectioned, approxi- mates to the bending moment diagram of a universally loaded FftACTim£.v YELLOW PINE BEAMS. '*M. Loweo »T CCMTKI. •TE.NSIOIJ SIDE SAP YftOP ArtO KMOT OH TOUBK S10& T FAIL STREV» • 3540 lav •. Slot, TENSION SIPE. CowmtssioH Sine. Fig. 8. tion, the distribution of pressure over the whole wing, theend effect, and the travel of the centre of pressure on wings of the section to be used should be ascertained from the veryvaluable data provided by the N.P.L. For the purpose of sand test and calculation, it is usual to assume that theforward position of the centre of pressure is at 0.28 width from the front edge and that the rear position is 0.5 width. Sandtests under these conditions of loading were also carried out at the writer's request on the ribs. The distribution of thepressure over the wing surface from back to front is met by the above tests, but special consideration should be givento the end effect produced by the leakage of air from the sides Curves showing the variation of pressure over the extensionsmay be considered, but according to fair average practice the pressure over the extension may be assumed to be parabolicup to a distance not exceeding 4/5 chord. According to good practice the overhang of the extension should not exceed 4/5chord. Preferably, it should not exceed 2/3 chord, In calculating for down load the distribution cannotbe assumed parabolic as above. Research is still required to determine the shape of the curve of end loading underdown load. Methods of Calculation.—Any sound method of calculationwill give equally satisfactory results if based on the correct foundation. A reliable and simple method used in cal-culating all British machines on active service at the Armistice is set forth in outline in Chapter IX of the writer's book onElementary Aeronautics. A more highly mathematical method evolved by Mr. Berry can be recommended. Thedifference in the results obtained by the two methods does not exceed 3 per cent., which is negligible fh comparisonwith the variation of the factors upon which they are based, an engineer's insight and judgment being of more importancethan any mere method of calculation provided the method is sound, and in this opinion I am supported by the Americans. Sources of Errors in Methods of Calculation.—All systems ofcalculation are based on the assumption that wood, steel, and other materials of construction are perfectly elastic up t0the point of failure, and that the neutral line does not change its position relative to the section. They are also based onthe further assumption that the points of support of the spars at the junction with the interplane struts remain in the samestraight line independent of the joints in the spars, incidence, bracing, etc. These assumptions differ so far from fact that any systemof calculation based upon them is vitiated unless supported by factors derived from practical experience. • I96
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
