Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON THE APPLICATION OF ARTIFICIAL INTELLIGENCE TO CIVIL AND STRUCTURAL ENGINEERING
Edited by: B.H.V. Topping
Collapse Load Factor for Rigid-Plastic Analysis of Frames using a Genetic Algorithm
A. Kaveh and K. Khanlari
Building and Housing Research Center, Tehran, Iran
A. Kaveh, K. Khanlari, "Collapse Load Factor for Rigid-Plastic Analysis of Frames using a Genetic Algorithm", in B.H.V. Topping, (Editor), "Proceedings of the Seventh International Conference on the Application of Artificial Intelligence to Civil and Structural Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 33, 2003. doi:10.4203/ccp.78.33
Keywords: plastic analysis, frames, collapse mechanisms, collapse load factor, genetic algorithm.
The minimum and maximum principles are the basis of all the general analytical methods for plastic analysis and design, Baker, Horne and Heyman . The most widely applicable method of analysis based on the minimum principle is that of the combination of elementary mechanisms, developed by Neal and Symonds [2,3].
Plastic analysis and design of rigid-jointed frames has been cast in the form of linear programming by Charnes and Greenberg , as early as 1951. Further progress in the field is due to Heyman , Baker and Heyman , Jennings , Theirauf , and Kaveh , among many others. Considerable progress has been made in the past decade, a complete report of which may be found in Munro  and Livesley .
In the method of combination of mechanisms, for a given frame and loading, every possible collapse mechanism can be regarded as the combination of certain number of independent mechanisms. The number of independent mechanisms, IM, is given by IM = NCS - DSI, where NCS and DSI are the number of potential plastic hinge positions (critical sections) and DSI is the degree of static indeterminacy of the frame.
For each possible collapse mechanism, a work equation can be written down from which the corresponding value of the load factor is found. The actual mechanism is distinguished from among all the possible mechanisms by the fact that it has the lowest corresponding value of , by the kinematic theorem, Neal . The independent mechanisms with low values of are therefore examined to form a mechanism, which gives an even lower value for .
In this article, a direct method is developed rather than using a mathematical programming approach which has its own difficulties. In this approach an additional degree of freedom (DOF) is added to each member of the main structure. Therefore each element contains five degrees of freedom, consisting of four independent translation DOFs and two dependent rotation DOFs. The external work and internal work are then calculated and using the principle of virtual work, the load factor for each mechanism is obtained.
Once the basic collapse mechanisms and their load factors are calculated, the genetic algorithm is used to combine the mechanisms and identify the one corresponding to the least possible load factor. Examples are included to illustrate the efficiency of the present method compared to the use of simple genetic algorithm.
Genetic algorithm is shown to be a suitable tool for calculating the collapse load factor of frames. The hybrid crossover-mutation of this article reduces the number of generations needed for optimization and increases the speed of convergence to the load factor. This also enables to perform analysis for those frames for which the simple genetic algorithm fails to converge.
purchase the full-text of this paper (price £20)