FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1929
1929 - 0421.PDF
FEBRUARY 28, 1929 THE AIRCRAFT ENGINEER SUPPLEMENT'TO FLIGHT 1 give in Table II the particulars I have collected. The partially externally braced monoplane, so popular in the U.S.A. for small and medium-sized machines, must in the nature of the case be intermediate in characteristics between tfce internally braced monoplane and the biplane. We may expect to arrive at a clearer conclusion if we confine our attention to a comparison between the latter. Moreover, we have no information on the former type for large sizes except that contained in Dr. Dornier's paper. I have included this in Table II, which serves to bring out one point of interest— there is little difference, between the biplane and the externally braced monoplane, in the proportion of the total wing weight which is due to the external bracing. In regard to the low wing weights quoted by Dr. Dornier, these are all associated with so high a wing loading that it is difficult to make a fair comparison. Of the internally braced monoplanes probably the most interesting are the Junkers and the Fokker. They present a remarkable contrast in material and in general technique. I doubt whether any lighter or more efficient schemes, using the material each prefers, could be devised. The Junkers G.31 and G.24 set us one peculiar problem. Their respective total weights are 17,800 lbs. and 14,300 lbs., and their wings are identical, weighing 2,800 lbs. or 15-6 and 19-5 per cent, of their respective total weight. In such a design it is, I think, logical to add the weight of the " wing structure "' passing through the body, which I estimate to bring the percentages up to 18 per cent, and 22 per cent, approximately. I can only presume that the G.24 was built to higher load factors than are now required, and that the strength of the wings was sufficient for an increased total weight. I believe I am correct in saying that for the G.31 a load factor of just under 4 is required, equivalent on the above assumptions to just under 5 for the G.24. The J.13, an older though essentially similar type, was tested at R.A.E. both structurally and aerodynamicany. The wings weighed 12-7 per cent., which I propose to increase to 15 per cent, for the reason mentioned above. Their load factor at failure was tf-9, which probably indicates a design factor of 6. The Fokker F.VII (three engines) has a wing in one com- plete unit, so that the wing weight (16-7 per cent.) is com- parable with the Junkers figures as modified. The Foeke-Wulf " Moewe," a machine of which the Luft Hansa have a high opinion, is very similar in construction and weight to the Fokker. The weights and factors for the Moewe and the Habicht were given me by the makers, the former being marked " approximate." The B.F.W. machines are as yet little known. They are distinguished by a relatively large span and the weights given are remarkably low. For the Rohrbach machines and the Beardmore " Inflex- ible,"' I was unable to obtain weights. For the "* Inverness," the data were obtained from the R.A.E. test, for which I have to thank the Air Ministry. The weight includes an estimate for the central part inside the fuselage. It will probably be sufficient if we confine our attention to the Junkers and Fokker, representing two widely used types. To appreciate the significance of the figures given some discussion of the strength specification to which they are built must be attempted. The German airworthiness regulations are in my opinion framed in a remarkably logical manner. Their basis is a factor of safety of 1-8 on load factors (for the "normal" class) of n — 2 -f 2/(W - 2), where W is the weight in metric tons, i.e., for the G.31 a total of J -8 x 2-20 = 3• 9t>. For the high speed condition the factor is approximately two- thirds of this. The terminal velocity dive is represented by the same attitude of flight (approximately zero lift) at a speed not greater than 2 s/« X the stalling speed, i.e., 3 X the stalling speed, with the same factor of safety (1-8). Stalling speed is taken to correspond to ordinary wind tunnel data, giving probably about 0-7 for the maximum lift for the Jun- kers and hence a stalling speed for the G.31 of about 70 m.p.h. Taking the moment coefficient of the wings at zero lift to be 0 05, the corresponding tail load (including factor) is about one-third W. To comply with our airworthiness regulations more than twice this load would be required. I will take it that approximately the same conditions apply to the Fokker. This considerable difference in tail load is reflected in a corresponding torque on the wing structure. Taking the weight of the aeroplane x the chord of the wing as a unit, my estimate of the torque (including factor) on the G.31 is 1 • 05. On our requirements it would be 2 • 25. It is difficult to assess its effect, but I think it would probably involve an increase of weight of about 10 per cent. In the Junkers this would take the form partly of a thickening of the covering, since this is the part of the structure mainly concerned with resisting torsional stresses. Probably some increase in the diagonal bracings would also be required, since these are responsible for distributing the air forces (which consist of fairly well localised upward and downward pressures over the rear and forward parts of the chord respectively). The comparatively poor torsional strength and rigidity of any essentially laminar structure is a crucial difficulty in monoplane design, and the greater the torque in com- parison with the maximum transverse loads, the more serious becomes the problem of coping with it. Even with aerofoils of a comparatively high thickness-chord ratio (0-15 and up- wards) it is doubtful whether adequate torsional stiffness can be obtained except by use of a stiff covering for either the whole (Junkers and Fokker) or a part (Dornier and B.F.W.) of the wing. But for wing loadings of the order of 10 lbs./ sq. ft. the large chord, in conjunction with a comparatively heavy covering,* would make the structure uneconomical, and we come to a high whig loading and a high maximum lift. The latter involves a large moment coefficient at zero lift, and hence a large travel of the centre of pressure in normal flight and a large torque on the wing in the terminal velocity dive. The Junkers construction is a brilliant achievement, but it is economical only in conjunction with higher loadings, even in relation to the maximum lift that the model shows (which as I have mentioned does not appear to be attained on full scale), than we consider desirable from the point of view of safety, and a standard of torsional strength less than half that required by British airworthiness regulations. The normal biplane is precisely the form of structure which has a natural torsional rigidity. By the conventional method of calculating the strength of a biplane structure, in which certain bracing (incidence wires) is considered to be inoperative, we deliberately shut our eyes to this essential feature—torsional stiffness. I estimate that such a structure is some 2-i times as stiff in torsion as is iniulied in our methods and generally' appreciably stronger (though not in the same ratio). This high torsional stiffness has an influence which may be important and will in all practical cases be helpful on the distribution of loads in the members of the structure, whatever the nature of the load may be. For the monoplane such considerations are vital, and in so far as we recognise them in monoplane and not in biplane design we are unreasonably handicapping the biplane. Even with this handicap the biplane appears to be lighter than the equivalent monoplane. In estimating the weight of the latter from the information we have accumulated wo have to take into account the following factors :— (1) For the wings, the large chord, involving thicker covering or more internal bracing, or both ; (2) For the rest, the extra length of body (involving also extra depth) and the change in tail surfaces ; and (3) For both, the increase in strength required to cope with the terminal velocity dive. In Table III I give my estimates for the three monoplanes of Figs. 1, 2, and 3. We start from the known weight distribution of the biplane. I think I am justified in treating the one selected as typical of biplane design of its date. There is evidence that the structure weight of modern biplanes, whether of composite or all-metal construction, can be reduced appreciably below the figures taken. We shall therefore run no risk of taking an excessively optimistic attitude towards the biplane. For the monoplanes, the * On the Junkers J.13 the covering, including longitudinal stiffeningstrips, weighs between 07 and 0'4 lbs, per sq. ft. of double surface.
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
