Reply to the discussion by Mesri and Wang on “Test fill on soft plastic marine clay at Onsøy, Norway”

Abstract

The author is grateful for the clarification given in the discussion regarding the Cc and C values. The comparison of the Ko-values from the empirical equations referred to in the Discussion with the hydraulic fracturing Ko-value given in the article is interesting, although it is perhaps a bit unusual to take ′-values from DSS tests, where ′ is the friction angle, when triaxial test values are available. The 24 h virgin compression line for the CRS oedometer tests was determined by drawing a line parallel with the CRS virgin compression line (stress in log scale) through the end point where the stress has been kept constant overnight. I agree that the 24 h void ratio – effective vertical stress relationship for 20 mm thick oedometer specimens with a maximum drainage path of 20mmcannot be expected to be exactly the same as the e– v ′ relationship for a 110 mm high triaxial test specimen with a maximum drainage path of 55 mm, but in my opinion the difference should not be very pronounced. Equation (D3) in the Discussion is the same one used at the Norwegian Geotechnical Institute (NGI) to compute the coefficient of permeability, k, fromCRS oedometer tests. In the paper by Sandbaekken et al. (1986) this equation is referred to as originating from Wissa et al. (1971). The k-values calculated for a particular CRS oedometer test are found to group along a straight line (the k-line) when plotting logk versus vertical strain, 1, for the virgin part of the stress–strain data. The calculated k-values for stresses below the apparent preconsolidation stress, pc ′, are usually significantly higher than corresponding to the values from the k-line. A possible reason for this may be that CRS oedometer tests at NGI are carried out without back pressure. The k-values from the k-line usually agree well with constant-head permeability tests when performed at stresses above pc ′, and are therefore normally used to calculate cv-values, also below pc ′. To perform permeability tests below pc ′ in the oedometer cell is considered to be a bit risky due to the possibility of getting by-pass of water between the stiff oedometer ring and the specimen in this stress region. For the Onsoy clay with pores with humus components and black spots (indicating high concentrations of iron–sulfur in the spots) down to a depth of about 8 to 10 m, it was suspected that the k-line values from CRS tests could be too low for stresses below pc ′. Constant-head permeability tests on triaxial test specimens with an initial height mostly about 110 mm (initial diameter mostly about 54mm) with back pressure about 300 kPa were therefore carried out. The permeability tests on the triaxial test specimens could be performed at effective stresses below pc ′ because no fluid by-pass can take place due to the soft lateral boundary in these tests. The constant-head permeability values measured on triaxial test specimens, shown in Fig. 20 and especially in Fig. 21 in the paper under discussion, are seen to be significantly higher than the oedometer k-line values. However, the difference between the two sets of k-values varies considerably; k-values back-calculated from fieldmeasurements were therefore used to correct the k-line values when estimating rate of settlement in the field.