When Cambridge University's large centrifuge came up for refurbishment, TWI was called in to examine the structure. After an inspection and repair programme the centrifuge was ready for many more years' service.
The Cambridge University centrifuge was designed and built in the early 1970s and came into full operation in 1974. It is a fabricated steel structure 10m diameter which rotates at speeds of up to 186 rev/min, giving peripheral speeds approaching 100m/s. It is used to simulate self-weight stresses in geotechnical structures for a wide range of research and development projects for various organisations. Such projects have included studies related to retaining walls, excavations, tunnels, earth dams, offshore foundations, cratering, pollution migration, earthquake engineering and ice sheet formation.
After extensive use, the University wished to refurbish the centrifuge to guarantee its continued operation. TWI was asked to comment on the integrity of the structure, particularly the welded joints. A preliminary review of the design of the centrifuge, its mode of operation and the environment, suggested that fatigue and corrosion would be the most likely sources of degradation in service, and an inspection programme was devised to cover the areas considered to be most susceptible to problems. Although the outside of the structure was protected from corrosion by paint, there was some concern over the possibility of internal corrosion in box sections.
During a scheduled shut down, staff from TWI visited the centrifuge and carried out an inspection programme. This included thickness measurements using ultrasonics to confirm that internal corrosion had not significantly affected the structure, and magnetic particle and ultrasonic inspection of the welds identified as the most likely to suffer fatigue damage. These included the ends of stiffeners and longitudinal seam welds.
The inspection revealed that the areas examined were substantially free from corrosion and fatigue cracks, except for some fatigue cracking at the ends of stiffeners. A repair procedure was developed and the repairs were carried out in situ. First the cracked regions were ground out and inspected by the magnetic particle technique to ensure that the cracks were fully removed. Repair by MIG welding was then carried out, and the required areas blended into the original structure using a grinding disc. The welded area was hammer peened to minimise the risk of further fatigue problems. Hammer peening induces compressive residual stresses in the weld area and is a very effective way of improving fatigue strength.
The inspection and repair programme was carried out with a minimum of disruption to the operation of the centrifuge. After the repair it was possible to give the university client confidence that the centrifuge would provide many more years of service. Further confidence could be obtained by annual inspections during maintenance.