LRFD calibration of simple limit state functions in geotechnical soil-structure design - Risk and Reliability in Geotechnical Engineering - page 343

(a) 3 (b)

3

2

2

1

1

0 0

n = 71
μ R = 1.00
COV R = 0.356

-1 -1

Lognormal fit
to all data
n = 71
μ R = 1.00
COV R = 0.356

Lognormal fit
to all data

-2 -2

Normal fit to all data

-3

-3

0.0 0.5 1.0

Resistance bias, X R

1.5 2.0 2.5 0.1 1 10

Resistance bias, X R

Figure 8.5 Resistance bias data: (a) CDF plot with normal X R axis and (b) CDF plot with lognormal X R axis.

Figure 8.6 shows resistance bias data plotted against nominal (predicted) resistance val-
ues. In this particular example, only 50 of 71 tests in the source documents reported the
actual pullout test capacity. Nevertheless, quantitative tests of the type previously described
for the load data confirmed that there are no hidden dependencies between X R and R n,
which is also consistent with the visual impression in this figure.

3.0

n = 50

2.5

X R = 1.10 - 0.02 P max

2.0

μ R (1 + COV R ) = 1.36

1.5

1.0

μ R = 1.00

μ R (1 - COV Q ) = 0.644

0.5

0.0

0 10 20 30 40 50 60

R n (kN/m)

Figure 8.6 Resistance bias versus (nominal) predicted resistance (pullout capacity), R n .