This invention relates to structural joints for piles, columns and like structural members comprising first and second end-fittings for the respective structural members to be joined, the first end-fitting having a central socket and the second end-fitting having a central spigot for entry into the said socket, the socket and the spigot having mutually engaging cantilevered lugs, those on the socket projecting radially outwards and those on the socket radially inwards and being so spaced around the periphery of the joint that in one relative rotational position of the end-fittings the lugs are disengaged and the end-fittings are freely separable in an axial direction and in another relative radial position of the end-fittings the lugs are engaged in pairs to prevent any substantial relative axial movement.

Such joints are regularly used to connect the sections of precast reinforced concrete "segmental" piles, and have been considered for other structural applications, for instance for mounting columns on floor slabs in reinforced concrete buildings, connecting steel columns, or fixing lighting colummns, heavy-duty fence-posts or gantries for railway electrification equipment.

Economic considerations dictate that such joints are fabricated from steel sheet, and the current practice is for the lugs to have a thickness nominally equal to half the thickness of the sheet, allowing the sheets forming the ends of the members to be joined to be coplanar. The result of this is that the entire tensile load is transmitted through lugs which are half the thickness of the steel sheet as well as at most half its circumference in aggregate, and it follows that the loading in the full-thickness part of the

sheet must be less than a quarter of the load it is capable of bearing (or the lugs would fail) . It is not practicable to overcome this weakness simply by locating the two sheets out- of plane to allow the lugs to overlap without reducing their thickness, since this would create gaps that could not be reliably filled with concrete and would thus be a potential source of weakness and/or corrosion risk. This invention seeks to increase the efficiency of loading of the steel and thus allow a reduction in its thickness while allowing the two sheets to remain coplanar.

The invention is characterised in that at least some pairs (and preferably all of them) consist of lugs each of which is tapered towards its unsupported end to form mutually engaging surfaces having a dimension in the direction of the axis of the joint that is at least a major part (and preferably substantially the whole) of the dimension of the lug in that direction.

For simplicity of fabrication and assembly, the mutually engaging surfaces preferably conform substantially to surfaces of revolution about the joint axis. An ideal curved surface that would load the lugs equally along their radial length could be computed, but as practical engineers we prefer a simple surface, in particular a portion of a cone; a cone with a semi-angle of in the range 25-55° is considered suitable and a semiangle of about 45° is particularly preferred. Curved surfaces or composite surfaces, made up for example of two conical surfaces of different taper, could be used if desired.

Normally all the lobes will be equal in circumferential length and equal (apart from a small allowance for clearance) to the spaces between them. Four lobes per joint member is

appropriate for typical applications. Smaller or large numbers (down to two and up to at least about 8) could be used. Unequal lobes could be used if it were desired to ensure that unsymmetrical members should be correctly assembled, but are inconvenient otherwise. It will be apparent that mutually engaging lobes should normally be the same (or at least nearly the same) length and that, when assembled, each lobe should normally be adjacent to a space between lobes of the other member big enough for it to pass through.

Preferably one of the end-fittings also comprises a projecting pin and the other a recess to receive it, as an aid to alignment during assembly.

When it is to be used for connecting segmental reinforced concrete piles or other reinforced concrete members, each end-fitting will normally include a plain relatively thin backing sheet adjacent to the structural member from which the lugs project to form part of an enclosure into which the concrete is to be cast and either reinforcing bars projecting to be embedded in the concrete or means for securing to the end-fitting reinforcing bars separately assembled (such as the main reinforcing bars of the structural member) . Each may also include a skirt member to extend peripherally around the end of the concrete body. Reinforcing bars, whether pre-fixed to the end-fitting or not, may be attached to it in a number of ways. We prefer to pass the end of the bar into a tapered aperture in the structural member of the end-fitting and secure it by welding on both sides of the structural member, including filling the wide end of the aperture (the end from which the rod does not project) with weld metal. Alternatives include the forming of

a head on the bar by a forging or swaging process, the use of screw-threaded connections, and other welding techniques.

The invention will be further described, by way of example, with reference to the accompanyinig drawing in which:

Figures 1 and 2 are partly-sectioned side views respectively of the male (upper) and female (lower) member of a joint in accordance with the invention for a segmental precast concrete pile; Figure 3 is an underneath plan view of the joint member of Figure 1;

Figure 4 is a plan view of the member of Figure 2; Figure 5 is an enlarged fragmentary section showing the members assembled together and sectioned on the lines V-V in Figures 3 and 4; and

Figures 6 to 9 are further enlarged fragmentary cross- sections showing alternative designs, that of Figure 6 being a current conventional design for comparison.

The male joint member of Figures 1 and 3 has three sheet-metal parts: a structural plate 1 which is basically a circular disc apart from four radially outwardly projecting cantilever lobes 2 (to be described shortly) ; a plain square backing plate 3 in face-to face engagement with the structural plate; and a skirt 4 around the edge of the backing plate. It also includes an axially-projecting aligning pin 5.

Correspondingly, the female joint member of figures 2 and 4 has three main sheet-metal parts, namely a structural plate 6 which is square with a central aperture which is circular apart from four radially inwardly projecting

cantilever lugs 7, a square backing plate 8 and a skirt 9; a socket 10 is provided to receive the aligning pin 5. In each case, reinforcing bars 11 are welded in countersunk holes 12 in the respective structural plates 1, 6 which are nominally filled with deposited weld metal 13

(Figure 5) . These reinforcing bars may either form parts of the end fittings (in which case they will be relatively short and will overlap with the main reinforcing cage of the reinforced concrete shaft 14 of the pile) or they could be the ends of longitudinal members of the main reinforcement (in which case, obviously, the welding step will be a step in the assembly of the end-fitting with the shaft rather than a step in the fabrication of the end-fitting) .

As so far described, the end-fittings might be entirely conventional (except that the welding of the reinforcing bars is unusual) . In this case, the lobes 2 and 7 (figure 6) would each be half the thickness of the structural plates (less an allowance for clearance) and would engage each other on flat surfaces 14 that are normal to the axis of the joint. In accordance with the invention, the lobes 2 and 7 in the form of joint shown in the detail of Figure 5 engage each other on surfaces 15 that are inclined to the axis. In this particular design, these mutally engaging surfaces 15 are portions of a cone with a semi-angle of about 45° and their vertical dimension is substantially the same as the whole thickness of the structural plates 1, 6 from which they are formed. The lobes are thinnest at the free ends of the respective cantilever where they are least loaded and approach the full thickness of the plates at the roots of the cantilevers, where the load is greatest: compare the conventional design of Figure 6, where the lobes are just as

thick at their tips, where the load as practically zero, as at their roots, where they must bear the full tensile load.

Figures 7-9 illustrate minor modifications; in these figures, the clearances have been exaggerated for clarity. In Figure 7, the ends 17 of the lobes have been blunted, so that the vertical dimension of the mutually engaging surfaces 16 is less than the whole thickness of the plates, but still a major part of it - say at least about 75% of the total thickness and preferably more. Figure 8 illustrates the use of composite mutually engaging surfaces, shown for this purpose as two conical surfaces 18 and 19 which are both coaxial with the joint but differ in semi-angle. Figure 9 illustrates the use of curved mutually engaging surfaces 20. These variations could be combined if thought desirable in any particular application.

As with conventional joints, the joint members described are first assembled with the reinforcement of the pile or other structural member with which they are to be used, and concrete is appropriately cast. After curing (and in the case of a pile joint after one of the pile segments has been driven into the ground) the mating end-fittings are presented to each other and aligned until the pin 5 enters the socket 10. The respective segments are then relatively positioned at 45° from their intended engaged position, so that the lobes 2, 7 may pass each other and after they have done so are rotated back (or onwards) by 45° so that the lobes become fully engaged. Preferably the joint parts are then locked to prevent rotation and eliminate any real risk of accidental disengagement, for example by the use of a (or more than one) grub screw inserted in a bore (not shown) extending radially across the boundary between plates 1