FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1958
1958 - 0048.PDF
50 FLIGHT MAKING THE P.I WING ... the corner and retained on two outer faces by a wall extending slightlyabove the level of the block. A bolt passes through the rubber block and metal pad into a retained nut capable of free vertical movement. Tensionon the bolt applies pressure to the metal pad and consequently to the rubber, which is thus forced into the desired intimate contact with thesurfaces of the corner. The actual hole in the extreme corner is not itself sealed, but the leak-paths to it arc isolated by blanking off the approaches. This corner joint includes certain of the seals developed to preventleaks through gaps between certain structural members. The design again relies on a rubber block being forced into intimate contact with the sur-faces surrounding and across the leak-path. Through both adjacent members where the gap exists, and through a top packing plate, a hole isdrilled to suit the rubber plug, stopping at the inside skin level. Again, the same technique of applying pressure to the rubber by means of a boltand pad is used as in the corner joint. An additional item is the top packing plate, to provide a complete circumferential seal above the levelof the gap. No attempt is made to fill the cavity between the abutting parts, but simply to seal all the faces which are in contact with the rubber.Sixty seals of this type are fitted in each wing. After over four years of continuous use, these are still functioning satisfactorily and no replace-ment has yet been necessary. Coupled with the design of the corner sealing was the developmentof a satisfactory rubber material, to withstand total immersion in kerosene and drying out over a temperature range of — 50 to + 100 deg C, and alsothe manner in which the various shapes of sealing blocks were to be manufactured. The material ultimately chosen was Hycatrol H.E.4 (Fireproof Tanks, Ltd.).Since the arrangement of the design for scaling depended on a pressure application, and this is also a condition of rubber moulding, experimentswere carried out with the intention of simplifying, for small quantity pro- duction, the moulding die manufacture. This lesulted in building upfrom representative structural sections, replicas of the corner joints iden- tical to those on the aircraft. In this way all the peculiar shapes, radii,tapers, etc., were produced naturally by the structural sections much more rapidly and more cheaply than by the usual die-sinking techniques. Two types of panels form the closure of the tank unit and, since theyform the webs of the spars, they are required to carry shear loads. To ensure maximum shear efficiency f iom the joints of the webs to the boomflanges, no gasket sealing is employed. Instead, for panels requiring infrequent removal, the sealing is by a nylon monofilament O.OOSin thicklaid in two parallel lines close to, and on either side of, the bolting lines A cross-section through one of the wing rivets and skin reproduced larger than lite size. The method of sealing, which is described in the text below, is to employ a sharp-edged projecting ring around the counter- sinking in the upper skin which bites into the mating part of the rivet when the /otter is closed. and retained in position by a thin adhesive coating. Panels providingaccess for servicing of equipment, and for tank inspection, have an integral seal permanently fitted, which does not require refitting on subsequentreplacement of the panel. This type of sealing was really derived from the development of thedies for its manufacture. Requirements were for a bead of rubber pro- jecting slightly above the surface to present an interference with thesurface to which it should contact, and adjacent cavities either side of the bead, providing space for the deformation of the rubber. The die ismade by passing a roller over the surface of a steel plate, the effect of indenting by the roller throwing up a burr on either side. These burrsmould into the rubber the <3>vitics described previously. The access panel is grooved coincident with those in the die, equal to the width over theburrs. The rubber section is moulded directly on to the access panel by using the latter as one-half of the die, and applying to this a bondingagent. The seal is in this way permanently formed and attached to the panel. Due to the limited access to the interior it was essential that a reallysatisfactory sealing of all bolts and rivets was obtained. If leaks were found during the routine pressure testing on production wings, the break-down of the structure to permit repairs would be a serious interruption. A special type of bolt design was ultimately chosen. A small ridgeis produced on a standard countersunk head which bites into the material in which the head seats and becomes self-sealing. As an additional safe-guard sealant is applied to the bolt shank before insertion; though all tests were carried out without any sealing medium other than the ridge. Rivet sealing is very similar but "in reverse." Since the rivet materialis softer than the skin, the ridge would flatten during riveting, so instead this is formed in the countersinking of the hole. The first type of cuttermade to produce this type of hole was purely a standard countersinking bit, with small right-angle nicks in each cutting face. Though this madea ridge, the corner was radiused. and experience had shown that a really sharp edge was essential for sealing. The final type of tool develop:dcomprised two cutters, the first flat-sided to produce the bore between the lidge faces and the second a hollow truncated countersinker located over the shank of the first, and forming the top surface of the ridge and theremainder of the countersinking. Fatigue testing of both rivets and bolts showed no adverse effect; it was thought the notching effect of the boltsin the skin material, and similarly in the rivet head, would have proved serious, but this was not the case. To ensure that the fuel compartment retains its pressure tightnessunder all circumstances and conditions in service, the sealing wherever possible is at least duplicated. The surfaces of adjoining parts are coatedwith an interfaying sealant and a fillet of slightly heavier sealant applied along each edge. All rivets, though self-sealing, are dipped before closingin the same material used for the structure sealing, and a fillet of sealant applied to the panned head. The ridged-head bolts are similarly coatedwith sealant before insertion and, in addition, special ring washeis which seal both the shank and the inside face of the structure are fitted underthe nuts. Bolts thiough the outer spar web and access panels have sealing washers both under the heads and under the nuts. Mr. Taylor's Paper • -- ' w-\. -~. It is my task, as one of the production people, to talk about the manu-facturing considerations. The main problems arising from the design proposals were concerned with utilizing the facilities already availablefor the rapid production of the prototypes. It was decided that, despite the strength of the skins and their allied structure, jigging would have tobe substantial and in conformity with recognized production standards to achieve the degrees of accuracy required. The length of skin ultimately exceeded the length of the availableFainhatn rolls. It was possible, however, suitably to modify the machine by extending the length of the rollers, except for the last 4in, when theframe of the machine interfered; making the last supporting upper bear- ing profile identical with the top roller diameter served to "iron" theremaining portion. A special router was schemed and built for machining the profile,rebates and recesses in the skin panels. This provided the bare essentials, a table and longitudinal slides on which ran a cross-beam carrying therouter head. A vacuum table was added later. The curve required was only slight, and this fortunately was able tobe sprung flat during the routing and returned to shape after machining. Manufacturing lugs were provided at intervals around the perimeterof the skin which served for holding-down and the location of local templates. The laige recessed areas outboard in the top skin were routed,a tracer for contact with the template being also used as a skid on the top surface of the skin. The operation was carried out by sweeping thecutter in increments from side to side, always keeping metal forward of the cutter for contact with the skid, and finally to track around the peri-meter of the rebate to finish and correctly ladius the boundary edges. The template, of course, had to be proportionately oversize to cater forthe increased diameter of the tracer. No drilling was carried out on the skin at this stage except in the manufacturing lugs. In order to follow the skin curvature the top face of each spar boomwas twisted. As first designed they were to be made from T-extrusions, the top flange of each being extruded 90, 87.5 and 85.5 deg. This was aconvenient compromise to the twist required on each spar boom (spar No. 3, 90 deg; Nos. 2 and 4, 87.5 deg: Nos. 1 and 5, 85.5 deg.). Theextrusions were later redesigned, which practically eliminated at least one of the complications of the bending and twisting requirement,with particular emphasis on spars 1 and 5. These were bent to a 56in iadius to the shape of a hockey stick. This bend—in the least helpfuldirection relative to the T-section—was further complicated by a rise in a vertical direction, in order to fit the camber of the inside of the skin asit followed around from the leading edge to approximately the wing centreline. This was carried out on a Hufford A.46. Here again further complications were apparent, for the flanges had tobe profiled, which would create considerable variation in cross-sectional area; if profiled after bending the release of metal would most likelydisturb the "set," and the spar was much too long to be gripped at its extremities. The decision was made to profile the flanges to size fromthe commencement of the bend to the outboard end. To leave uniform thickness from the commencement of the bend to the inboard end, threescallops should be cut in each of the three flanges, to engage with pegs of a removable clamp. The former was mounted on the die bolster andthe free end was connected by a substantial link to the free jaw of the Hufford, in order to restrain the tool and component as the bent'end waswrapped and stretched. A snake was used during the forming to support the unsupported flangeof the section protruding from the tool, and the springback of the component after the first wrap was such that a second solution treatmenthad to be given, with a further set-up in the machine and a final wrap and stretch. This was followed by adjusting the long outboard end tocorrect the distortion caused by the heat treatment, within six hours after being taken from the bath. In assembly, the spar booms are located in their individual assemblyjigs together with their various attachments, the inside skin line being carefully taken care of, and all holes, in the way of the diagonals andbracing members drilled and reamed and, with fitting bolts, bolted together. The bolts arc not locked, as the whole assembly has to bedismantled later. The spars are located in a vertical assembly jig and the ribs, undercarriage brackets, and other structure assembled, all holesbeing reamed, but assembly being only with fitted bo'.ts and without sealant. Each skin is then offered up in turn for drilling and reamingof all rivet and bolt holes. Next, the wing structure is dismantled. The wing components arekept as a set, and any finishing (e.g., countersinking and deburring) is carried out. The skins and the members to be riveted to the skin are thenthoroughly cleaned, the mating surfaces coated with sealant and the com- ponents joined in a special riveting machine. It was decided that anyattempt at positioning either the skin in suspension, or a strong enough squeeze riveter in suspension, over a rivet head, would be quite inefficient,and the machine illustrated provided vertically balanced moving plat- forms with 6ft of vertical travel and 18in of horizontal travel for thedolly and the h6ld-up tool, following the curvature of the skin.
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
