Milford Haven Bridge Collapse June 2, 1970 Group 9 The Milford Haven bridge Collapse Introduction This report analyzes and discusses the construction and possible failures of a box-girder supported bridge. In particular the report focuses on the Milford Haven Bridge in Wales which collapsed during construction resulting in four deaths in June 1970. Box-Sections The typical box girder has two webs and two flanges (illustration #1.) However, in some cases there are more than two webs, creating a multiple chamber box girder. (Illustration #1) The added second web in a box girder adds stability and increases resistance to twisting forces. As all the material in a BoxSection is at the extremes of its cross-section the second moment of area is much greater than that of other section types. Extra stiffening along the length to prevent buckling can be achieved by the addition of flanges and diaphragms. Bamboo stems use this same principle as can be seen in illustration #2. (Illustration#2) The forces experienced by box girder bridges are inherently complex to calculate using traditional methods as it is a three dimensional problem comprised of narrow members joined by hinges. Finite element analysis methods using modern computers can now compute the forces with relative ease. Bridge construction using Box-Sections By making a beam out of short sections with wide flanges or closed walls at the ends, a bridge can be assembled by bolting or welding such sections end to end. Such a bridge can be cantilevered out from the piers. When the ends meet, and are joined, with suitable adjustments of the stresses, the result can be a very strong and rigid beam. The problem with this Idea is in construction when the beams are cantilevered out from the piers of the bridge. The stresses on the bridge here are very different and often much larger than experienced after completion. This has led to several collapses in the construction stage of such bridges including the collapse of the Milford Haven Bridge. The Milford Haven Bridge The superstructure of the bridge consists of a deep continuous trapezoidal welded steel girder. The original design consisted of a single continuous box girder of welded steel with no substantial stiffening of the section other than the six pier supports and abutments. This design does not take advantage of the stiffing effects of flanges and diaphragms and so is subject to buckling due to bending moment especially in a cantilevered state. The construction of the bridge was carried out using the technique described in the previous section. However there where no wide flanges or closed walls at the ends of the individual sections. The spans measured from the south 77m, 77m, 77m, 149m, 213m, 149m and 77m. The Collapse The collapse occurred during launching a pre-assembled deck section towards the second pier from one end. The span that collapsed was the second 77m span on the south side, the first having been erected with the aid of a temporary support. The collapse occurred when the last section of box for the second span was being moved out along the cantilever. When the collapse occurred this section slid forward down the cantilever killing four men. Failure was initiated by buckling of the support diaphragm at the root of the cantilever being erected (Figure 3). The diaphragm was torn away from the sloping webs near the bottom of the box, allowing buckling of the lower web and bottom flange to take place. As the diaphragm buckled, it shortened, reducing the overall depth of the box girder; the tendency of the bottom flange to buckle was inevitably increased by this reduction of the distance between flanges which increased the force needed in each flange to carry the moment with a reduced lever arm. The support diaphragm was in fact a transverse plate girder, which carried heavy loads from the webs at its extreme ends, and was supported by the bearings as shown in Figure 3, some distance from its ends. It was therefore subjected to: 1) A hogging bending moment 2) A large vertical shear force The diaphragm plate near the outer bottom corners was subject to a complex combination of actions: 1) The shear of the transverse girder and diffusion of the point load from the bearings was compounded with the effects of inclination of the webs of the main bridge girder which produced an additional horizontal compression action. 2) And an out-of-plane bending effects caused by bearing eccentricity. The load sustained by the diaphragm just before failure was reported to be nearly 9700kN, which agrees tolerably well with independent calculations of strength made after the accident. The calculated design resistance, using design rules that were drafted subsequently and making allowance for likely values of distortion and residual stresses would be considerably less, possibly as low as 5000kN. Result of the Collapse A Committee of Inquiry into the Design and Erection of Steel Box Girder Bridges (1) concluded that the cause of the collapse of Milford Haven Bridge was the inadequacy of the design of a pier support diaphragm. It was of the opinion that BS153 (the only British steel bridge design and construction Code of Practice current in the 1960's) was inadequate for such application and that there was need to consider wider implications for the design of such bridges. The Merrison Committee prepared Interim Design and Workmanship Rules (IDWR). These contained comprehensive rules for the stress analysis for box girders and for the design of steel plated components in complex stress fields, with and without stiffening, and for their connections. The bridge was eventually rebuilt as a cantilever and suspended span in the main span. Further failures of similar bridges around the world occurred. These bridges are, in chronological order: The Fourth Danube Bridge in Vienna (Austria, 6 November 1969) The Milford Haven Bridge (United Kingdom, 2 June 1970) The West Gate Bridge in Melbourne (Australia, 15 October 1970) The Rhine Bridge in Koblenz (West Germany, 10 November 1971). And although no two of the failures where alike a new British Standard was introduced, based largely on the Merrison Rules. The Merrison Rules laid the foundations for the production of a new British Standard, BS5400, Parts 3, 6 and 10 of which, with the additional benefit of further research, now contain comprehensive rules for steel box girders. The implementation of the recommendations led to wideranging changes in contractual procedures and in checking designs and construction procedures. References 1. Inquiry into the Basis of Design and Method of Erection of Steel-Box Girder Bridges. Report of the Committee. HMSO.1973. 2. Report of the Royal Commission into the Failure of West Gate Bridge. Government of the State of Victoria. 1971.