Safety factors and limit states analysis in geotechnical engineering: Discussion

Abstract

The paper is a further valuable contribution by Dr. Meyerhof on the subject of safety factors and limit states design in geotechnical engineering. It is understood that the paper also represents the basic approach that will be adopted with respect to limit states design in the second edition of the Canadiaiz foundation engineering manual. The load and resistance factors given in Table 2 of the paper are indicated to agree with those specified in the National Building Code of Canada (1985) and in the Ontario Highway Bridge Design Code (1983), respectively. They also agree very closely with those given in the Danish Code of Practice for Foundation Engineering (1978). The author indicates that the total and partial safety factors for geotechnical analyses 'are interrelated, and that the partial factors suggested for use in limit states design have been obtained by calibration with conventional geotechnical analyses. It is apparent from the paper that a constant value for the overall factor of safety is obtained only if the partial factor, f+ for example, varies with the value of the friction angle, 4'. ~ l t e r n a t i v e l ~ , if fm is held constant, then the overall factor of safety will not remain constant. The writer has examined the variation of the overall factor of safety for simple slopes, cantilevered rigid retaining walls, flexible sheet pile walls, and shallow foundations, designed to satisfy the ultimate limit states criterion using the author's recommended partial factors and the results are bgeflv outlined herein. The design of simple slopes using shear strength parameters in terms of total stress may be carried out using Taylor's stability charts (d, = 0 analysis). For this particular case the author recommends a partial factor, f, = 0.65, which is equivalent to an overall factor of safety, F, of 1.54 on cohesion or height, which compares closely to the overall factor of safety of 1.5 employed for the design of such slopes in conventional practice. The design of simple slopes with uniform soil and ground water conditions using an effective stress approach may be carried out using the Bishop and Morgenstern (1960) stability coefficients. The slope inclinations obtained using the limit states analyses show increasing overall factors of safety with increasing pore water presswes, from 1.25 with r, = 0 to about 1.8 for ru = 1.5. It is customary in conventional practice to select an overall factor of safety, F, which is dependent upon the loading conditions and the consequences of having a slope failure-occur. Highway, railway, cut, and embankment slopes tend to be designed using a lower overall factor of safety than would be employed for a slope where a structure is to be constructed near the top or base. The recommended limit states