An extremely powerful but very specialised area of earthquake engineering research is carried out using geotechnical centrifuges, writes Paul Wheeler
Only one of the half dozen geotechnical centrifuges in the UK has the capability to model earthquakes.
This phenomenally specialised machine, at Cambridge University, has led development of the centrifuge in the West since the mid-1970s. From there, researchers have moved on to establish world leading geotechnical centrifuge centres, from Australia to the US.
In principle the centrifuge is simple: The behaviour of any structure or material that is dependent upon its selfweight can be simulated in a laboratory-scale test, providing it can be subjected to an acceleration greater than the Earth's gravitational field.
Though mass is unchanged, the effective weight increases in proportion to forces generated by the centrifuge whirling around at high speed - typically creating many hundreds of g.
Stress created by the simulated increase in weight in the model is identical to stresses experienced in its full-scale counterpart.
One of the most powerful geotechnical centrifuges belongs to the US Army Corps of Engineers' Waterways Experiment Station (WES) at Vicksburg, Mississippi. Tackling large earth dams, locks and river control structures, environmental problems and military research, WES needed a centrifuge with a large payload capacity and high G capability. The facility was put together in 1989 with technical knowledge from Cambridge.
'Very often, centrifuge testing provides the only practical means of validating numerical simulations or design methods.
This is especially so when the events of interest are catastrophic, ' says Scott Steedman, one of the facility's designers and now a director at consultant Whitby Bird.
The main driving force in the US came from Dr Bill Marcuson, then head of the Army Corp of Engineers' Geotechnical Laboratory, and Richard Ledbetter, who became the first director of the Centrifuge Research Centre.
'Both were renowned geotechnical earthquake engineers, with particular expertise in liquefaction, dam design and remediation, and this undoubtedly contributed towards the early emphasis on the development of a large centrifuge shaker, ' says Steedman.
'Experience has shown that the centrifuge technique is particularly valuable in earthquake geotechnical engineering, partly because of the difficulty of conducting field experiments, but generally because of the success that has been achieved in replicating field phenomena of pore pressure generation, soil degradation, soil-structure interaction and ground failure.'
WES commissioned a large beam centrifuge from Swiss motion simulation specialist Acutronic. It can accelerate an 8t payload to 143g, and 2t to 350g, and has a platform area of 1.3m 2.This enables problems which in a real physical environment would measure up to 300m wide and 300m deep to be simulated under a wide variety of loading conditions.
'The Vicksburg centrifuge typically models liquefaction effects down to around 100m, compared with most centrifuges that can model to around 20m depth, ' Steedman explains.
'While these are highly appropriate for modelling the performance of harbour walls, embankments and dykes, you need to be able to look deeper for larger structures such as earth dams.'
Circle of development The story behind the development of the geotechnical centrifuge could almost provide the plot of a John Le Carre novel.
Back at the first International Conference on Soil Mechanics & Foundation Engineering (ICSMFE) held at Harvard in 1936, the Russian scientist GI Pokrovsky published a paper on the potential application of centrifuges in geotechnical work. But during the conference the great Professor Karl Terzaghi - the grandfather of modern soil mechanics and at the peak of his influence - made a characteristically impassioned dismissal of all small-scale physical modelling as 'utter futility'.
Probably as a consequence, the geotechnical centrifuge was forgotten about completely in the West for nearly four decades.
One of those who started reinvestigating its application was the brilliant and still youngish researcher Andrew Schofield who, with Professor Peter Wroth, had already developed the theory of critical state soil mechanics.
By a twist of fate, Moscow hosted the 1973 ICSMFE. 'At that time we thought there had been difficulties by which Pokrovsky's techniques proved to be less useful than he hoped in 1936, ' reports Schofield. Nevertheless, he thought he might be able to glean some historical details on Pokrovsky's work, since the published literature on the centrifuge had dried up completely.
According to legend, once in Moscow, Schofield placed a few centrifuge diagrams, a copy of Pokrovsky's 1936 paper and a Union Jack on a table. It did the trick and introductions were made.
'Our Soviet hosts invited all who were interested in centrifuge techniques to a meeting for open discussion with Pokrovsky and other Soviet engineers.'
It was to Schofield's utter surprise and massive excitement that the Soviets had continued to develop the centrifuge through the Cold War. 'I was surprised to see in a book by Pokrovsky that the Soviets had successfully modelled nuclear weapons craters, ' says Schofield. Transmission through soils of movement caused by nuclear explosions is in many ways similar to that of seismic waves.
Schofield also discovered that after his early centrifuge model experience, Pokrovsky worked on weapons effects during and after the Second World War, and as a direct consequence of the Moscow revelations, the US Defence Nuclear Agency later sponsored crater tests using the Boeing centrifuge in Seattle. This led to an order of magnitude reduction in the predictions of the size of craters resulting from nuclear explosions.
In short, says Schofield.
'Geotechnical centrifuge testing showed that Allied weapons would have been dramatically ineffective.'
INFOPLUS www-civ. eng. cam. ac. uk/geotech www. wes. army. mil/centrifuge/ index. html