Issue 5

We wish to compliment the author for his thorough analytical and experimental investigation.

Professor J. Heyman: There is an enormous amount of new material in this paper and it is going to take us a long time to assimilate it. I would like to make one or two broad comments and not discuss the detailed results. One of the most interesting features is the use of realistic flexible joints rather than the rigid joints that were used in the previous tests. If I can summarize the test results, the beam behaviour using the flexible joints accorded pretty well with the predictions of the Joint Committee’s report, the behaviour of the columns was not so close, but the ultimate collapse loads reached by the columns were certainly on the right side-the Joint Committee’s method was conservative. All of these observations, while differing in degree, were in fact results repeated from the previous test with the rigid joints.

Mr. D. E. Thorp: I would like to give some of the main reasons that led to the final solution of the single raft over the whole area of reactor building. The raft is resting on about 30 m (100 ft) of medium sand with a relative density of about 50 to 60 per cent. I understand that little is yet known in the field of soil mechanics about stresslstrain relations within sand masses, and sophisticated analysis still assumes, I believe, a semi-infinite elastic medium. Hence the results must be treated with some reserve. Conscious of these uncertainties, we were faced with the design of superstructures whose differential settlements had to be strictly limited to avoid damage to finishes and plant. In particular, the goliath crane for the fuelling machine, which operates on a longitudinal rail system supported consecutively on the fuel handling unit, intervening steelwork and concrete pressure vessel, is particularly sensitive to undue differential settlement.

The paper presents a theoretical analysis of the failure probability density function and cumulative failure density function for timber columns of known slenderness ratios and known distributions of the material properties. This analysis then allows comparisons to be made of failure probabilities in columns of various slenderness ratios and other properties. D.A. Newton and R. Ayaru

Internationally conoidal reinforced concrete shell roofs have been used attractively. They can of course also be made, and sometimes are made, from laminated timber, glass reinforced plastics (glass fibre), aluminium, steel, etc. C.B. Wilby and M.M. Naqvi

The graphical procedure presented in this paper is an extension of the linearized deflexion theory for suspension bridge analysis which brings the theory more into the region of preliminary design. By construction of graphical models to represent equations derived by a reformulation of the classical deflexion theory approximate values of deflexion, bending moment and hanger load distribution across the span can be quickly calculated for all typical loading cases.