By Bob Koerner
In formulating a factor-of-safety value for a specific geosynthetic application, a design engineer is essentially calculating a ratio of resisting forces-to-driving forces. This contrast is even further recognized when using load and resistance factor design (LRFD) which is apparently in our not-to-distant future.
Regarding the resisting forces, they are primarily the soil’s shear strength and the geosynthetic’s tensile strength. In the writer’s opinion, both of these strengths are capable of being determined reasonably well. Even if one designs on a probability-of-failure hypothesis (Duncan, 2000) the statistical variations within the required performance tests (e.g., wide-width tensile, direct shear, and transmissivity) are well established. [George Koerner’s proficiency test program has a wealth of such data which is available in published form (2002)].
It is the required forces that are the focus of this particular column and they can be subdivided into dead and live loads. The dead loads are typically gravitational forces created by the mass of involved soil and they are quite deterministic. The live loads, however, are an altogether different set of circumstances and subjective estimates are often necessary.
Three situations will be commented upon further: surcharge loads, hydrostatic loads, and seismic loads.
Most mechanically stabilized earth walls and slopes have a surcharge load at, or near, their crests. Building loads are usually tractable but temporary live loads from oversized stationary vehicles, unanticipated storage of products or materials, unannounced swimming pools, and stockpiled snow or ice loads have been troublesome in the past. A conservative selection of such surcharge loads anticipating all aspects of long-term use of the site is necessary.
Hydrostatic loads require careful deliberation in their selection since intense storms have caused wall, slope, veneer stability, and even waste stability failures. Hurricanes, cyclones, and other intense weather situations seem to be occurring with greater frequency than in the past. Clearly coastal communities and infrastructure have seen their share of problems, most recently with Hurricane Sandy causing several veneer slope failures in its wake. In this regard, one wonders if a designer should select the 50-year, 100-year, or most-probable precipitation and, furthermore, just how good are these established values in an age of global-warming?
The selection of a seismic coefficient in earthquake prone applications is extremely important in calculating FS values. As an illustration, the following factor-of-safety variation of a 30m-long veneer cover soil using different seismic coefficient (Cs) values is readily seen.
In all three of these situations, illustrating appropriate “live load” selection, the significance to the design community is heightened by the recent decision in Italy as reported in The Economist magazine. No additional commentary is necessary!
“In Italy, sloppy seismology can lead to prison
The earthquake that destroyed L’Aquila, a city in central Italy, on April 6th 2009 killed 309 people at the time. But it took until Oct. 22nd of this year for it to claim its latest casualties. Those casualties were seven men-three seismologists, two engineers, a volcanologist and a public official-convicted on that day of manslaughter, for misleading L’Aquila’s inhabitants about the risks they faced. Each was sentenced to six years in prison, though that may be reduced on appeal.”
Robert M. Koerner, Ph.D., P.E., NAE is the director emeritus of the Geosynthetic Institute in Folsom, Pa.
Duncan, J.M. (2000). “Factors of Safety and Reliability in Geotechnical Engineering,” Jour. Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 126, No. 4, pp. 307-316.
Koerner, R.M. and Koerner, G.R. (2002). “Beyond Factor of Safety: The Probability of Failure,” Proc. GRI-16 Conference, GII Publ., Folsom, Pa., pp. 1-18.
The Economist, October 27, 2012, pg. 80.