SECTIONAL TRANSACTIONS.— G. 363 



(3) To enhance durability by increasing the stiffness, thickness or riveting of 

 parts or members above that required by considerations of strength only. 



(4) To codify methods of calculation of loads and stresses so that engineers working 

 independently ma_y get strictly comparable results. 



(5) To give structures a strength capable of withstanding a standardised system 

 of loading while the stresses in their members do not exceed specified maximum 

 intensities. 



Examples of each type of rule are given. 



The adoption of well-considered rules of the first four types aid good design and 

 results in what is sometimes called good practice. Selections of rules based on 

 extensive experience have been put together by various engineers in the form of 

 Designing Specifications. Examples of rules and specifications are discussed. 



The idea that a multiplicity of rules and regulations can take the place of experience 

 or prevent bad or weak design on the part of the inexperienced is deprecated. 



In relation to loads and the corresponding limiting stresses, designing specifications 

 are of two fundamentally different types. In one type the maximum intensity of 

 stress allowed in a member is adjusted to suit the position of the member and the 

 severity of the loading conditions in that member. In the other the stress is limited 

 to a nominal maximum for all members, but the loads carried by different members 

 are adjusted according to an empirical formula which increases the loading where 

 the conditions are more severe. 



Regulations relating to stresses have an important influence on weights and costs 

 of structures, and for structural work sent abroad the country adopting conservative 

 practice and low limiting stresses, though producing comparatively robust and durable 

 structures, maj^ be seriously handicapped in international competition. 



Prof. C. Batho. — Theory and Experiment in Structural Design. 



A field of investigation of importance to the structural engineer is the development, 

 by analysis and experiment, of exact methods for the calculation of the stresses and 

 deformations in a structure, and the determination of the conditions which govern 

 its breakdown under excessive loading. The results must often be complicated, and 

 efforts should be made to reduce them to forms applicable in practice. 



The problems involved fall into two groups : (1) consideration of the structure as 

 an elastic framework, (2) the study of the distribution of stresses and deformations 

 in the members and of the manner in which the connections transmit the forces and 

 affect the strength and stability of the structure. 



Statically indeterminate structures are often necessary or desirable. The 

 difficulty of calculation of redundant structures is often exaggerated owing to neglect 

 of the systematic treatment, by means of deformation diagrams and influence lines, 

 developed mainly in Germany. The calculation of frameworks without diagonals, 

 such as building frames, is possible by various methods, but often very lengthy. 

 Recent investigations indicate that reliable approximate methods are possible. Begg"s 

 experimental method is of value in these investigations. 



Problems of the second group are largel}^ experimental and include (a) intensive 

 laboratory explorations of the strain and deformations in structural elements, connec- 

 tions and built-up members, (6) laboratory tests to destruction, (c) experiments on 

 actual structures. Examples of these are discussed in the paper. It appears to be 

 desirable to investigate the simpler elements of construction before attempting to 

 analyse the results of tests on complete structures. Tests to destruction should be 

 directed towards the observation of the conditions of initial breakdown. Experiments 

 on the conditions of elastic breakdown of materials under non-uniformly distributed 

 loads are desirable. 



Investigations such as are outlined have led, and should still lead, to increased 

 economy by making possible more rational methods in design. 



Discussion. 



Report of Committee on Electrical Terms and Definiliovs (Prof. G. W. 0. 

 Howe). 



