## Difference Moderately Conservative And Characteristic Values

values that are 'very unlikely to be exceeded' (but not worst physically possible)

Safety factors applied to soil strength

apply ...

generous values

less conservative values

total stress analysis

1.5 (temporary works)

not recommended

(Fs)

1.1-1.2 (temporary works)

1.2-1.5 (permanent works)

1.0 (temporary works) 1.2 (permanent works)

Equivalent safety level

5% fractile?

0.1% fractile

According to CIRIA 104's successor, CIRIA C580,15 soil parameters selected by this method are equivalent to representative values defined in BS 8002 (see Section 5.3.3) and to characteristic values defined in Eurocode 7 (Section 5.3.2). The difference between CIRIA 104's moderately conservative value and Eurocode 7's cautious estimate is again merely one of semantics.

5.3.5 How much ground is involved?

An important factor in selecting the characteristic value of a ground parameter X is an assessment of how much ground is relevant to the occurrence of the limit state.

Failure of part of the ground may not lead to an ultimate limit state being exceeded by the structure - for example, forces are often redistributed from highly-stressed regions to adjacent lower-stressed regions. Because of this, it is the average value of the material strength (or other relevant material property) that governs the occurrence of the limit state. The characteristic value Xk should be selected as a cautious estimate of the spatially averaged value of X representing the relevant volume of soil or rock. In statistical terms, Eurocode 7 requires a 95% confident assessment of the mean value of X.

When a small volume of ground is involved, the characteristic value Xk should be selected as a cautious estimate of the spatially averaged value of X within that small volume (otherwise known as the 'local' value of X). This value may be significantly lower than the one selected for a larger volume of ground, since there is less averaging being done. Statistically, the characteristic value Xk in this instance will be closer to the 5% fractile discussed in Chapter 2.

For example, consider the settlement of a footing resting on a layered soil profile, in which one particular layer (e.g. peat) is much more compressible than the others. In this situation, the coefficient of compressibility (mv) of the peat dominates the calculation of total settlement and hence we are particularly interested in the local value of mv in this layer.

Figure 5.8 shows our evaluation of the characteristic undrained strengths of the London and Lambeth Clays first shown in Figure 5.3 and chosen for pile design.

The black symbols give values of undrained strength cu derived from 100 standard penetration tests (SPTs), assuming the following ratio between cu and SPT blow count N:

which was chosen from Stroud and Butler's correlation with plasticity index shown in Figure 5.6.

The white symbols give values of cu measured in 91 triaxial compression tests (TXs) on U100 samples. We have ignored the circled data points on Figure 5.8 since they give unrealistically low strengths, believed to be the result of sample disturbance.

In the London Clay (above - 20m OD), we have chosen a cautious estimate of the spatially averaged strength, suitable for calculating the average skin friction over the pile shaft with 95% confidence. Because the points are not particularly scattered, the selected line is close to the middle of the data. If we had been worried about locally low values controlling shaft friction, we would have chosen the dashed line, which is a lower bound to the SPT results. We have attached greater importance to the SPT results since they provide - albeit indirectly - measurements of the clay strength in situ, whereas direct measurement of strength in laboratory tests on small samples may not be representative of field conditions.

In the Lambeth Clay (below - 20m OD), we have chosen a cautious estimate of the spatially-averaged strength, for a smaller volume of soil (down to -33m OD), suitable for calculating the pile's end-bearing resistance. Because the data points are highly scattered in the Lambeth Clay and end-bearing involves a relatively small volume of soil, we also gave serious consideration to choosing the cautious estimate of the local strength given by the lower dashed line. There is some disparity between the strengths obtained from the SPTs and those from the triaxial tests, the dashed line being a better fit to the latter.

Cautious estimate Cautious estimate of focal value of spatial mean

Figure 5.8. Difference between local and spatially-averaged characteristic values

### 5.3.6 Well-established experience

Eurocode 7 requires the selection of characteristic values to be 'complemented by well-established experience', i.e. practical contact with and observation of facts or events and knowledge or skill acquired over time.16

Well-established experience also includes simple rules-of-thumb for estimating soil parameters, such as the following17 for determining the peak (9) and constant-volume (^cv) angles of shearing resistance of siliceous sands and gravels:

10, rounded If 0, uniform 2, sub-angular l + i 2, moderate grading 4, angular J | 4, well-graded

### 5.3.7 Standard tables of characteristic values

Geotechnical engineers are sometimes required to estimate ground parameters for a particular stratum without the benefit of test results in that stratum. This is often the case for made ground or other near-surface soils, for occasional layers of coarse soil (i.e. gravel or sand), and for thin layers of fine soil (i.e. silt or clay). In the absence of test results, ground parameters may be selected from standard tables of values, such as those that appear in many geotechnical text books.18

The 'standard table' below shows how to determine a clay's constant-volume angle of shearing resistance (^cv) from its plasticity index (Ip):19

0, rounded ' f 0, uniform q> = 30° + -j 2, sub-angular I + -j 2, moderate grading !- + -4, angular J I 4, well-graded

Angle of shearing resistance, <pcv 30° 25° 20° 15°

When choosing a characteristic value from tables such as these, Eurocode 7 requires a 'very cautious value' to be selected. It is important, therefore, to understand the engineering basis of the table, so that the degree of caution embodied in the published values can be taken into account when selecting the characteristic value.

### 5.3.8 Summary of geotechnical characterization

Figure 5.9 summarizes the ways in which a characteristic value Xk may be chosen according to Eurocode 7. In most situations, a cautious estimate of Xk is made from derived values of the geotechnical parameter X; when there is sufficient data, statistical methods may be used to select a 95% confident value (upper or lower, as appropriate); in the absence of data, reference made to standard tables of characteristic values may be used, with increased caution. In all cases, the resultant value should be checked against well-established experience.

## Post a comment