Causes of high stresses

a. Welds produce high thermal stresses as the metal is first hot and molten, then freezes and shrinks but is restrained from shrinking by the cold steel either side. This inevitably leaves tension in the weld metal and shear in the interface between the weld and the parent metal. The maximum stresses appear at the ends of straight welds. One particular place where this is evident is at the corners of rectangular box sections, where two straight welds meet, usually at a place where bending and tensile and shear stresses coexist. The thicker, longer and hotter the weld, the greater the stresses. Research is needed to find out if multi pass thin welds are better or worse than single fat welds.

b. Unequal thicknesses. Thin metal heats up more quickly than thick steel in the hot molten zinc. If a thick plate is welded to thin material, the thin material expands quickly compared to the thick. Surprisingly, the forces between the plates are concentrated at the extremities of the connection and not distributed evenly along the welds. As examples, an end plate on a box section, a flange plate or web plate on a beam, a stiffener within a web of a beam can give rise to these high stresses at the extremity of the weld. This does not include regular end plates welded to open ends of .

c. Pattern of heating. As a member is dipped into a tank of zinc, dipped parts expand, bending the member. This bending can provide compressive, tensile and shear stresses in components of the member. The more slowly the member is immersed, the hotter the hot parts, the greater the stresses. This is made worse in the following ways.

 

i) If the member is too deep for the tank then the bottom half of the member can be completely up to bath temperature when the top half is still at ambient temperature. A clever dipping schedule analysed by plane frame can reduce stresses.

ii) The ambient temperature itself can alter thermal differences, particularly when very cold, and the member is wet from pickling, and exposed to cold and wind.

iii) Tubes take much longer to dip than open sections. The air makes them float in the molten zinc. They have to be lowered in slowly while air escapes.

iv) Trusses are subject to a wider range of different stresses as they heat up. They tend to be very stiff so forces are higher. Shear stresses are not evenly distributed along hot/cold interfaces but are point loads at the lacings. Joists at lacing nodes are likely to include different thicknesses; and to experience peak moments; shear; tensions. The more members and redundancies and stiffness in a truss, the more prone to LMAC.

v) Relative stiffness of truss members can mean the more limber members flex but the stiffer members get higher local stresses at connections and the stronger members will be the ones to fail first.

vi) Cold rolled tube or box sections have elements at yield stress in bending and are usually in high tension all along the seam. They should not be galvanised unless they are simply open ended tubes or only have really well vented thin end plates: never in a truss.