Structural Systems Optimization for the Building Industry
Research Theme: Computational Design
Despite large quantities of research conducted in the field of structural optimization, uptake in structural design practice has been very limited.
Whilst widespread use is made of commercial optimization packages in the aerospace and automotive industries, this has not been mirrored in the building industry.
This project, sponsored by Arup, attempts to develop suitable structural optimization methods for practical use.
Motivation
As well as obvious cost savings, stochastic optimization and search techniques offer the capacity for rapid generation and evaluation of design alternatives. However, design tasks are often highly complex, with many loadcases and constraints and objectives that may be difficult to express mathematically.
Objectives
- To understand the reasons for slow uptake of optimization methods in thebuilding industry.
- To appreciate the required functionality of methods and tools to be usedin practical design
- To develop methods suitable for use in the building industry
Method
Observations have been made on a number of live projects in an attempt to gain a better understanding of the design process, as well as a perception for the potential role of optimization techniques. Further discussions with structural designers have been useful in assessing their attitudes towards such methods.
Case studies have been conducted to assess the potential for application of existing optimization techniques in design practice. These have included a small study in material removal methods for continuum optimization and a form of pattern search applied to the design of diagonal bracing for the steel tubular frame of a tall building.
Findings
Direct involvement and the application of directed search methods on a live project produced a set of highly efficient designs, with substantial variation in appearance. Both the structural design team and architects were very enthusiastic about the methods used and results obtained. However, the general applicability of the approach taken is substantially limited.
Details
This project has included involvement in the design of a steel tubular-framed tall building. The optimization task seeks to find structurally efficient configurations for the diagonal bracing elements that are required to provide adequate resistance to lateral loading. The problem is complicated by multiple loadcases and the difficulty of handling strength-based constraints.
The architectural vision considered bracing elements to be grouped into spirals emanating from the base of vertical columns. The spirals are to be of varying length, terminating at different points up the building. The search procedure used a stochastic form of pattern search to add or remove elements from spirals to find a variety of designs with minimal numbers of bracing elements, whilst still meeting the strength constraints. Figure 2 shows a set of such designs. The variation in appearance allows designers and architects to choose aesthetically pleasing configurations from a selection of efficient designs.
Acknowledgements
- Support is provided by Arup.
Selected Publications
- BALDOCK, R. and SHEA, K. (2004) 'Structural systems optimization techniquesfor the building industry' in Next Generation - Intelligent Systems inEngineering, the 11th International Workshop of the European Group for IntelligentComputing in Engineering (EG-ICE), Weimar, Germany, 199(4),11-15
- BALDOCK, R., SHEA, K. and ELEY, D. (2005) 'Evolving optimized braced steelframeworks for tall buildings using modified pattern search' in InternationalConference on Computing in Civil Engineering, ASCE 2005, Cancun, Mexico
- BALDOCK, R. and SHEA, K. (2005) 'Combining continuum topology optimizationand structural design guidelines for bracing system design in multistory steelbuilding frames' in Soft Computing in Engineering, the 12th InternationalWorkshop of the European Group for Intelligent Computing in Engineering (EG-ICE),Radziejowice, Poland .
