Aeronautical Engineering - Bonding with Redux

by: Charles J. Moss, B.Sc.

Reproduced from : THE AEROPLANE

No: 329
SEPTEMBER 20, 1946

Note: The tables referred to in this article have been scanned from originals and may be viewed by clicking on the relevant hyperlink. As graphic files they may take a while to load.
We recommend you close the new window each table will generate after viewing.

IN RECENT YEARS the advances in aircraft performance have been, very striking. In 1939 speeds of the order of 350 m.p.h. were exceptional, but now they are almost commonplace. Such advances have been made possible by improvements in power unit output, and by aerodynamic refinements, and as a result the designers of aircraft structures have been faced with increasingly difficult problems. They must design their aircraft to withstand very heavy loads, and at the same time they have to bear in, mind the stringent aerodynamic requirements. For example, the thickness of wings must be a minimum and their surfaces should be as smooth as possible. Undoubtedly these requirements will become even more important, and consequently the difficulties of the structural designer will be more acute.



(Above) MODERN WING CONSTRUCTION - A wing panel with "Reduxed" stringers, immediately after removal from the press.

They must design their aircraft to withstand very heavy loads, and at the same time they have to bear in, mind the stringent aerodynamic requirements. For example, the thickness of wings must be a minimum and their surfaces should be as smooth as possible. Undoubtedly these requirements will become even more important, and consequently the difficulties of the structural designer will be more acute.

The advance from the Mosquito to the Hornet is a good example of how structural designs have developed to meet more exacting conditions. The Mosquito wing spars have wooden tension and compression booms, but this would have been impossible for the Hornet, because of the large cross-section of wood necessary for the tension booms. . As a result the tension booms were made of aluminium-alloy extrusions, while the remainder of the spars were of wood, i.e., the compression booms and the web. The, metal and wood were then welded together so as to form a final spar of remarkably low weight and high strength, and of small depth. One result of the success of this design, and of the testing carried out by the Royal Aircraft Establishment, Farnborough, was the approval given for Redux bonding by M.A.P A D.T.D Specification, No. 775, will shortly be issued to cover the process.

Application to All-metal Aircraft

However, it seems, likely that in the, next few years high performance aircraft will, in a majority of cases, be of all metal construction. The need for smooth, thin wings is greater than ever, and the hitherto normal riveted construction must necessarily come under very severe criticism. Most aircraft skins are literally covered with rivet heads, and consequently determined efforts are being made to diminish their aerodynamic effect, either by polishing operations or by finishing the surface with special paints. Other methods are being tried, including spot welding, but even this leaves a small mark where the electrode makes contact with the skin.

This represents one general line of approach to the problem, but there is an alternative which is already, in use on one production aircraft. The de Havilland Dove has fuselage and wing stringers " Reduxed" to the skins. Unlike riveting and spot welding, this does not consist of a large number of small local attachments, but is a bond covering the whole area of contact between stiffener and skin. Hence it achieves two important objectives. It avoids numerous small stress concentrations, and leaves the skin perfectly "clean." Theoretically it seems quite wrong to make a hole in a skin as a first step to attaching a stiffener, and Redux now makes it possible to avoid this.



(Above) A MORE EFFICIENT STRUCTURE - The photograph shows the rear fuselage in a dove in which "Reduxed" stringers are fitted.

A comparison between the relative strengths of riveted, spot Welded and Redux-bonded structures, is interesting. Table I compares the results obtained by I. G. Bowen on a large number of tests on riveted joints in Alclad sheet with those obtained with Redux joints. All the joints were 1 in. wide. The strength of the Redux joints is considerably greater than that of the riveted joints, especially when countersunk rivets are used. In Table II will be found a similar comparison between the results of R. Della-Vedowa and M. M. Rockwell, of Lockheeds, on spot-welded joints and Redux joints in 24 S.T. aluminium-alloy sheet.



(Above) STATIC LOAD TESTS - This test is being done on the D.H. Dove with the bonded stringers. When a similar wing of riveted construction is tested, the buckles run through the rivet- holes.

Once again the Redux joints are considerably stronger than the others. Nevertheless, these are only small-scale tests and the following results, published by the courtesy of the Aircraft Division of the English Electric Co., Ltd., Preston, are of more direct interest to aircraft designers. Tests were made on flat-ended 16 s.w.g. and 18 s.w.g. panels, each 22 in. long, and of three different aluminium alloys. On the bonded panels rivets were fitted to the end of each stringer. The results are given in Table III. With regard to the results it will be seen that in every case the Redux-bonded panel failed at a higher load than the riveted panel. Further tests on Redux-bonded panels at + 60° C. and -40° C. and under repeated loading at normal temperature all behaved satisfactorily.

With the high strength aluminium alloy materials in use at present, wing spars will undergo considerable strains before failure occurs and as a result the panels must be capable of carrying a load not only after the buckling of the skin, but after the initial buckling of the stringers themselves, i.e., after the skin and stringer combination has passed its maximum load. In consequence it is not sufficient to substitute bonding for riveting without first giving a careful consideration to the area of the bond between the stringer and the skin. However tests have shown that, given an adequate area of bond, the Redux will take a considerable load even after buckling.

Full-scale Tests



(Above) TEST TO DESTRUCTION - It can be seen that although the stringers on this Dove wing have fractured (between ribs 6 & 7) the bond is still intact.

Finally, we are able to give, by courtesy of the de Havilland Aircraft Co., Ltd., photographs of the full-scale tests on the wing of the Dove. It will be seen from these photographs that "top hat" stringers are used, and it is de Havilland's experience, based upon a large number of unpublished panel tests, that panels with Reduxed "top hat" stringers will develop shear stresses up to 125 per cent. of those developed by exactly similar panels with riveted and spot-welded joints.

The results of the full-scale test fully support this conclusion. The appearance of the wing under load is quite different from that of one of riveted construction. In the latter case the buckles all run through the areas of stress concentration, i.e., the rivet holes. In the case of the bonded wing it will be seen that the skin is held firmly along the whole length and the whole width of the stringers. Such a construction is, of course, much stronger than one in which the skin is perforated like a sieve and in which the load is taken on a large number of small areas surrounding the perforations.

The wing was subjected to the following series of loadings:-
No. of cycles Load cycle
(1) 1,200 0 to 2 g.
(2) 500 0 to 2.45 g
(3) 5 0 to 3.4 g
(4) 10,000 0 to 1.5 g
(5) 5 0 to 3.86 g
(6) 5 0 to 4.37 g
(7) Destruction test by static loading to 5.5 g,
which is 108% of the design load.

After more than 11,500 stress cycles the wing was in perfect condition, and failure did not occur until after the fully factored load had been passed by an appreciable margin. The failure was an impressive demonstration of the strength and reliability of Redux bonding, because even where the wing skin fractured, only one stringer parted from the skin for a very short length. There can be no doubt whatever that had a conventional riveted structure been tested in a similar way, either the rivet heads would have come off or they might have pulled through the skin.

Other Advantages of Redux

An improvement in strength is not the only result of using Redux. The skin of the Dove, for example, is in striking contrast to that of most other aircraft because it is so clean. On high-speed aircraft this will become of greater importance in the future. The other advantage is saving in cost. Aircraft riveting is quite an expensive business and Redux is already effecting appreciable economies in aircraft production.

In the case of the Dove, Messrs. de Havillands use a large press and sets of stringers are bonded to fuselage and wing skins in sizes up to about 4 ft. by 12 ft. It must be made clear that all the attachments required are made in one operation. When riveting is used it is necessary to jig drill each hole in the stringer and skin as a separate operation, and to countersink each hole in the skin before the actual riveting is carried out.

This discussion has mainly dealt with the attaching of stiffeners to skins. There are, however, many other uses for Redux. It is used for constructing the floor of the Vickers-Armstrongs "Viking." In this case a comparatively thin plywood floor is reinforced by top hat stringers. The rolled light alloy sections are first of all bonded to 1 mm. Veneers. Even in a small 6-ft. by 3-ft. press it is possible to bond about 180 ft. in one hour. The veneered sections are then cold glued to the plywood. The floor is then free from any rivet or boltheads and is also strong for its weight.

Bonding also makes it possible to attach local reinforcements either to metal or to wood. Good examples of this are to be found in the folding wing attachments of the "Hornet" and the "Mosquito."



SIMPLIFIED SPAR CONSTRUCTION - On the left is shown the large box spar assembly on the Mosquito, which makes an interesting comparison with the smaller unit of the Hornet, on the right, with Redux bonding.



DETAIL COMPONENTS - Built up from sheets of Alclad, this provides an example of "Reduxed " local reinforcement.

Future Developments

There is no doubt whatever that the main difficulty in the development of Redux-bonded structures is its application to doubly curved surfaces. Hitherto flat and singly curved surfaces have caused little difficulty, but at present it is necessary to make up form tools to bond doubly curved panels. At present no other method has been evolved for applying the heat and pressure necessary for bonding. Such tools can, of course, be castings, but even so they tend to be expensive in relation to the numbers of aircraft produced.

Conclusion

Redux bonding is only at an early stage in its development, and we are convinced that it will have further uses both in the aircraft and in other industries. Briefly its advantages are that it gives stronger joints than do riveting or spot welding, it gives an aerodynamically clean external surface, it is cheaper than riveting and spot welding and, finally, it makes possible structures (such as those referred to above) which have hitherto been impossible.



(Above) UNDER TRIAL - The prototype de Havilland Sea Hornet XX, which completed its trials on H.M.S. Ocean, the Royal Navy trials-carrier.

References:

C. Bowes, Aircraft Engineering, 1945, 17, 83.
R. Dell-Vedowa and M. M. Rockwell, The Welding Journal, 1942, 21,514.
"Aeroplane," January 25, 1946.
"Aeroplane," March 1, 1946.

Bibliography

"Redux Process"- Light Metals, May. 1943. pages 219-221.
"Strength of glued joints" --Aircraft Engineering, Volume XVI (1944) pages 115-118 and page 140.
"Superglue "- Life (U.S.A.), March 5, 1945.
"Some recent advances in synthetic adhesives"--Plastics, Volume IX,May. 1945, pages 228-234.
"Metal Bonding --Automobile Engineer, September, 1945, pages 354-6.
"Adhesives for Metals" --Metal Industry, September 14, 1945. Pages 162-4.
Northern Aluminium Co., Ltd., Research Department investigations published as M.A.P. Scientific and Technical Memoranda Nos. C. 9/44 and 13/45.