Rethinking 300 years of Engineering
Research at the University of Sheffield has overturned 300 years of engineering theory and has led to the development of a pioneering computer software tool. The tool enables structural engineers and architects to experiment with building designs and layouts that were previously mathematically impossible to calculate, saving on materials, weight and potentially construction costs.
In the 17th century Dutch polymath Christiaan Huygens stated that a structure following a parabolic curve provides the most efficient form to carry a uniformly distributed load. This belief persisted for the next 300 years, with the 'parabolic curve' now a ubiquitous architectural feature, most notably in bridges such as the Golden Gate Bridge in San Francisco.
However, this fundamental belief has recently been overturned by researchers based in the Department of Civil and Structural Engineering at the University of Sheffield, who have developed state-of-the-art mathematical optimization techniques to identify a structure which is both more optimal and is of very different form.
Dr Mathew Gilbert and Dr Andy Tyas found that when a standard material is involved (i.e. one which can resist both tension and compression forces), a different structure is slightly more efficient, requiring less material, but which is considerably more complex in form than a simple parabolic curve.
The researchers had expected a simple parabolic cable to be the optimal shape even if compression were allowed, and they were trying to reproduce that answer as a test of their software. When their computer program kept producing the more complex structure, the team feared the worst, as Dr Mathew Gilbert explains: "We were trying to work out where we went wrong, and it ended up that the only option was that the current wisdom was wrong."
Drs Gilbert and Tyas developed their structural engineering tool by adapting software used by mechanical engineers, in the aerospace and automotive industries, and applying it to civil and structural engineering problems. However, the complexity of the mathematics involved soon posed problems, so the team collaborated with leading structural engineering firm Buro Happold (designers of the Millennium Dome) to refine their software tool further.
"The tool allows structural engineers and architects to calculate the efficiency of different building layouts. They can then make a decision as to whether the efficiency savings, in terms of weight, building materials and perhaps even cost, are worth the increased complexity of the design" explains Dr Andy Tyas.
The team's software has already found one of its many potential applications in the area of soil slope stability. LimitState, a University spin-out company founded by Drs Gilbert and Tyas, has successfully used the software to solve a soil stability problem which had been dogging the construction of a football stadium for 8 months. The team's software cracked the problem in 2 days.
Dr Gilbert again: "Though the structure we found to be more efficient than the standard parabolic curve is arguably primarily of academic interest, it does suggest that mathematical optimization approaches might also be useful in helping engineers find very significant weight savings in more complex structures".
George Rozvany, a structural engineer at the Budapest University of Technology and Economics, agrees: "The new work is a good example of so-called structural topology optimization, which has already enabled engineers to reduce the weight of new airliners by up to 20%".
For further information, please contact Dr Matthew Gilbert at:
email : m.gilbert@sheffield.ac.uk
Suggested links
Dr Andy Tyas' University web page