Sand dykes and sills and vented sand boils on the Fraser River delta and adjacent Serpentine River floodplain in southwestern British Columbia record one or more major liquefaction events. The source of the dykes is a shallow subsurface saturated sand unit that was deposited in foreslope and distributary-channel environments during Holocene progradation of the Fraser delta. The dykes cut steeply through a crust of delta topset muds and locally flatten out as sills in overlying peaty sediments. At two sites, evidence was found for venting of sand onto a subaerial or intertidal surface. Liquefaction and upward movement of sand and water locally deformed the intruded sediments, causing some subsidence and uplift of the delta surface. All observed liquefaction features are younger than ca. 3500 BP, and at least some are younger than 2400 BP. It is not clear, however, whether they formed during one or several separate events. A review of possible causes suggests that the liquefaction features probably are seismically generated; they thus may provide the first direct evidence for moderate to large, prehistoric earthquakes in the Vancouver metropolitan area.

Dynamic soil–structure interaction analyses were carried out for the seismic retrofit design of the immersed George Massey Tunnel, both to predict and study soil liquefaction and related tunnel movements and to design ground improvement. The proposed ground improvement included ground densification using vibroreplacement stone columns along both sides of the tunnel and seismic gravel drains adjacent to the outer edge of the densified zones. The den sification and drainage were proposed to locally mitigate soil liquefaction and reduce displacements of the tunnel to tolerable levels. Centrifuge model tests with base shaking to simulate earthquake effects were conducted to verify and calibrate the numerical models. This included simulating the effects of ground densification and drainage on reme diating tunnel movements. This paper presents the principal results from the dynamic analyses, the centrifuge model design and testing procedure, the class A predictions of the centrifuge tests, and discussions of the centrifuge test results and numerical model calibrations.Key words: immersed tunnel, seismic retrofit, soil liquefaction, design verification, centrifuge testing, numerical calibration.

A centrifuge testing program was conducted to assess the earthquake performance of countermeasure retrofit techniques for an existing concrete tunnel embedded in a liquefiable sandy deposit. In a series of three separate heavily instrumented model tests, a tunnel section-foundation system was studied first without retrofitting, and then with a densified-wall and a gravel-wall liquefaction countermeasure techniques. The test results provided valuable insights into the underlying mechanism and effectiveness of each countermeasure and served as basis for calibration and validation of computational modelling of the George Massey Tunnel seismic retrofit design. Test results showed that the implemented countermeasures significantly reduced the tunnel movements. The gravel-wall technique was found to be particularly effective in reducing the extent of liquefaction as well as reinforcing the tunnel-foundation system against earthquake induced deformations. The centrifuge test results and the calibrated numerical model studies led to a final ground improvement scheme that included both stone column densification and seismic gravel drains.

No abstract

No abstract