Little known subterranean UoM centrifuge is decommissioned

Hidden underneath a grassy bank between the Simon building and Oxford road lies a 32 metres square geotechnical centrifuge designed to expose packages of 3.5 tonnes to a gravitational force of 200g (that's equivalent to spinning two cars so fast they weigh the same as 2/3rds of a Blue Whale, or the same as the Space Shuttle Endeavour). William Craig, retired Reader in Engineering and academic visitor tells us more about the history of the centrifuge.

2018 is the 50th year since the UK’s first purpose-built centrifuge for geotechnical engineering research was built and commissioned (1969) in the Pariser Building, by Prof. Andrew Schofield. He had previously tested very small models on a makeshift centrifugal rotor on the horizontal axis of another piece of rotating machinery in Cambridge and had hired time on a larger aero industry centrifuge in Luton. Coming to a Chair at UMIST he had a pit excavated beneath the floor of the hydraulics laboratory and installed a machine capable of generating 150g at a faceplate radius of 1.68m. The maximum payload mass (geotechnical model + its support container of maximum size 1.0x0.9m in plan and 0.4m deep) was 750kg, balanced by dead weights at the opposite end of an all-welded box rotor, Figure 1.

At the time I was an MSc student working with Prof. Peter Rowe in the Dept. of Engineering, on Brunswick St., in the Faculty of Science. Rowe had seen three papers published by Schofield and by Japanese and Russian authors at an International Conference of Soil Mechanics and Foundation Engineering in Tokyo and had been looking at the possibilities of building an even bigger machine in the Simon Building since 1968. He arranged for me to use the Pariser centrifuge, under Schofield’s guidance, to test models of a failing compacted clay embankment in a container 406x254x157mm internally, using only 20% of the machine capacity.  With virtually no ancillary equipment or instrumentation the preformed slope was brought to failure by increasing the rotational speed of the centrifuge – the so-called ‘gravity turn-on’ approach. By the summer of 1970 I had produced the first Manchester Master’s centrifuge thesis which was soon followed by more substantial PhD theses from UMIST students Alf Lyndon, later to run another centrifuge at Liverpool University, and Charles Hird who moved to Sheffield University. Schofield moved back to Cambridge where a large machine had been designed and installed while he was in Manchester and this was later followed by several other machines and the build-up of a large, world-leading group of centrifuge modellers which continues today. UMIST centrifuge work continued under the guidance of Malcolm Bolton, Prof. Andrew Palmer (both later to return to Cambridge) and David Maddocks. Following Maddocks’ untimely death the machine fell into disuse, was decommissioned and removed some years ago. Evidence of its existence can still be seen in the form of the cover plates to the machine pit just visible beneath the current large wave flume.

Following my MSc, I was appointed to a lecturing post within Rowe’s group and was tasked with supervision of the construction of a new laboratory being built specifically to house an even bigger centrifuge on the Oxford Road side the Simon Building. The machine has a symmetrical fixed-beam rotor with faceplate radius of 3.2m, designed to take a 3.5tonne model package to 200g, Figure 2. It was designed with a view to modelling major soil embankments and with a model container having internal dimensions 2.0x1.0m in plan and 0.6m deep would have been able to simulate structures 400x200m in plan x 120m high. Finance constraints at the time limited the motor installed to a 200HP unit rather than the 300HP the designers suggested would be needed to generate the radial acceleration of 200g. The machine ran, at rotational speeds up to 210rpm generating maximum accelerations around 130g, for almost 40 years, last running in 2007, after completing several thousand experimental simulations of a multitude of geotechnical problems. The eventual fall in usage reflected loss of staff and logistical difficulties following movement of the Simon group to the North Campus, but the terminal problems were the need for a replacement drive control system and fatigue failure of a single bolt on the rotor which undermined the integrity of the whole machine. The centrifuge is now to be removed and the laboratory refurbished for alternative purposes.

In the early years work on slopes, which had determined the machine size and configuration, predominated but by the mid-seventies the development of North Sea oil and gas fields dictated a change of focus. Permanent offshore platforms, both piled and gravity, presented major novel foundation problems when subjected to wave loading. Temporary installations for rigs involved in site investigation used jack-up structures or floating units tethered by spreads of marine anchors. Pipelines on or in the seabed buckle under cycles of thermal loading. All these areas provided opportunities for generic as well as site-specific design-related research over more than thirty years. We spent more than three years working on design options for the final section of the storm surge barrier which protects the Netherlands from the sea and a similar period simulating possible foundation configurations for the Troll field gas production platform which would be built some time later.


When Peter Rowe retired the centrifuge laboratory was named in his honour by Andrew Schofield, who presented him, Figure 3, with a detailed 1:32 scale model of the centrifuge manufactured in-house by Bob Glasper, Figure 4. Academic support for the centrifuge was substantially reinforced by the arrival of Caesar Merrifield, initially appointed to teach management, who was an integral part of the centrifuge geotechnical group for more than twenty years and led a number of non-traditional engineering modelling projects – salt water intrusion into aquifers, blast loading and impact, sub-sea gas blow-outs and the sorption of radioisotopes in ground water flow. On the technical support side Martin Cruickshank ran and maintained the machine and designed our equipment and models for thirty years and Bob Glasper produced many of the structural components for our geotechnical models as well as most of the test rigs for almost as long. They were invaluable - big research requires teamwork.


Lots of exciting research using two large machines was backed up by use of a smaller machine, designed and built in-house and used for teaching geotechnics laboratory classes for over thirty years, but now condemned.

In recent years Drs. Abuel-Naga and Carlos Lam attempted, without success, to raise funding to either refurbish the machine in the Rowe Laboratory or purchase a small replacement for the North Campus. In 2018 a new opportunity unfolds as Domenico Lombardi has raised the necessary finance and commissions a very neat new machine, equipped with modern on-board electronic systems for model loading, data monitoring and visualisation, in the basement of Pariser, Figure 5, to be used for both teaching and research.


I hope he has as much fun and excitement as we had.












▲ Up to the top