I don't understand your question. If you are asking what design factor of safety should be used, then the answer depends on many factors including the likelihood of the loading combination, how conservative your strengths are, how much you know about the slope and subsurface conditions, assumed water levels, etc.
Agreed. I haven't done slope stability since my research days, but my notes show FS between 1.5 (academic) or 2-3 (practical) depending on the specific problem (mostly level of risk and quality of information).
Thanks for your response.
I got the information that the acceptable FS of a slope is 1.5 for normal condition.
This will be reduced to 1.3 for the extreme condition where the ground water increased to the ground surface.
While in the earthquake condition the FS is further reduced to 1.1 (? - not sure)
I want to know whether the above values are based on practical or the Standards (British, Australian, NZ, ASTM, etc)
I do measure the actual slope and using the assumed effective parameter based on the tested undrained strength. Thanks for your help.
if this is a dam or levee, don't forget about rapid drawdown or liquifaction. You really should be doing a risk assessment to determine what risk tolerance you have against embankment or slope failure.
LOK, the European standards have an overall safety factor of 1.375 for frictional soils and 1.54 for clays. Safety factors can be adjusted by the individual states though.
This if there are no variable loads, otherwise it gets higher.
Also, overall FS has another intrinsic safety margin because it's computed using the characteristic values, cautious estimates of the geotechnical parameters of interest, how cautios depending on many factors.
Some states like Italy require a subjective judgment on the FS, according to the uncertaintes as already mentioned by Geopavetraffic. There is no fixed value. Interestingly enough, this goes against the principles of the Eurocodes, which are supposed to govern in Europe.
The US AASHTO code typically uses minimums of 1.3 and 1.5 for slopes and walls. The higher 1.5 value is used when supporting structures behind a slope or wall.
The LRFD approach to Limit State just takes the reciprocal of these numbers as a resistance factor and is not handled very well at the current time.
The tendency is to use conservative soil assumptions with this criteria so it may be more conservative that it appears.
I must correct my previous post, I realized I've not been exaustive enough.
In Italy the building codes contemplate 2 different situations:
-Artificial or excavated slopes, embankments, engineered works: OFS is the same or similar to that provided by the European standards
-Natural slopes: after the application of the characteristic values procedure, the FS is assigned by the professional by the evaluation of site conditions and consequences of risk.
So, actually European codes in Italy are bypassed only in the case of natural slopes.
It is interesting to notice though, that the FS intervals tend to overlap in the range of 1.35 - 1.5 in various regulations.
As to earthquake conditions, the OFS remains the same but in pseudo-static methods an horizontal component of the load is applied, which tends to lower the FS with respect to the static case. This is compensated by the fact that actions are not factored, but it also depends on the design approach (there are 3 of them and the second has a 4th one which is a subset). Yes, European regulations can be confusing...
If I were in the OP's position and barring the existance of already cited specific instances of liquefaction, rapid drawdowns and so on, I may proceed in two ways:
-After an examination of the various international laws, adopt an FS with a lower bound of 1.4 for frictional soils and an upper bound of 1.6 for analysis using undrained cohesion. These boundaries may be adjusted according to the specific site conditions.
-Adopt a recognized international code and follow it, if there are no local specifications.
Also a tricky and fundamental choice may be the representative values of phi-c and Su. It may suffer of substantial reductions with respect to the lab values.
A friend whose case is presently in court, botched his analysis because he adopted peak values in an already failed slope. His stabilization plan was followed but another slide ensued. He should have adopted residual values even though the sliding surface could not be located exactly. In some cases, constant volume values are advised.
safety factors of slopes are often taken as some prescriptive value. One that can't be crossed!
So, folks take their N-values, do some correlation to friction angle, cohesion, undrained shear strength and simplify geology to rationalize their design assumptions. Often there are no wells, and some short-sighted geologist/field engineer will decree, "looks like perched ground water," so nobody actually takes bouyant unit weights into consideration.
What then truely represents the safety factor? Some calculations from equations, based on YOUR simplifications!
For our DOT, we publish values like 1.3 (non critical slopes) and we publish 1.5 for critical slopes. We define what's critical and what's non-critical. For design of critical slopes (consequence of failure too great, heigh of wall over some value, etc.) we'd require actual laboratory confirmation of soil strength.
Here's what we often get. . . Drained direct shear testing on a fat clay with a time to failure of 8 minutes! I mean, no determination of Mv (T50) or any of the consolidation characteristics that the DDS would require to set the shearing rate. So, sure, base your design on 1.3 or 1.5. What's the input data look like? That's where the real issue resides though.
If we had a real robust lab program, a geologist that actually applied geology, boundary conditons set on real ground water measurements, I'd temper any of these prescriptive safety factors.
I also require folks dismiss long term cohesion and use fully-softened shear strength for shale fills or stiff-fissured marine clays. It's a rare case that I'll use cohesion for rotational shear failures. We're looking for a 75 year design life, which is in sharp contrast to many developers more typical 20 year design life.
@Mccoy, thanks for your detail explanation.
I understand the analyse either in frictional soils and undrained cohesion for engineered slopes or natural slopes.
I agree that is a tricky and fundamental choice to adopt the right phi-c and Su parameters
Some papers mentioned the failed slope should be treated using the parameters from Direct Simple Shear Test
Please elaborate, what the meaning of the OFS?
@fattdad, thanks for your valuable explanation.
I understand either for critical or non-critical slopes.
Actually it is a bit odd if we have to get the actual laboratory strengths, which ones? there are 3 conditions in slope failure such as: compression, shear and extension.
Most of the clients won't pay for the laboratory cost. At the end we only take the correlation strength parameter.
What the meaning of DOT? One more question is regarding design life, it is for earthquake design?
LOk, the OFS or overall safety factor is a term which refers to the total effect of all the various partial safety factors used in the analysis. In the European codes, there is a safety factor for each loading condition (permanent and variable), there are safety factors for each geotechnical parameter and there is a global safety factor for sliding.
The actual minimum FS in the Eurocodes is actually = 1 after you have applied all the previous factors.
The OFS does not consider the reduction of the soil strenght due to the adoption of the characteristic values, which gives a further margin of safety.
In Italy, if the landslide is not a big one and there are budget concerns, we usually have a few lab tests performed on a single parameter like direct shear (peak values) or residual shear (residual or fully softened values). Determination of Su is trickier as Fattdad told.
I can add further, confirming what Fattdad says, dismissing the cohesion value in slope stability analyses is pretty frequent in designs performed in Italy as well. This is an intuitive adoption of safety margin used by the old school of geotechs.
I don't do much earthquake design work - I've studied it, but in Virginia (USA) we rarely see earthquakes that damage earthworks. So, I was talking about design life from the, "loss of cohesion" or "fully-softened shear strength" perspective.
If you, "actually have to get shear strengths" you'd want to consider the stress path for your failure mode and decide whether to use triaxial compression, triaxial extension, drained direct shear, UU or other such approach (i.e., CPTu, DMT, etc.). No way I can address what's correct for some generic project. Also, please recognize in the United States (i.e., via ASTM) the selection of C and Phi is not returned from the laboratory. Such engineering interpretation of strength data is not an obligation of the testing laboratory. It's considered a professional service.
I will not encourage N-value correlations for the assignation of shear strength! I know it's done, but. . .
@Mccoy & fattdad, I appreciate to your detail explanations.
fattdad, I am still eager to know how to analyse using a design life. Please inform me the papers or example for this. Thanks
As mentioned above by the others FOS is based on much on the situations but FHWA report for slope stability by nail suggests 1.35
and for walls with soldiers and anchors between 1.2 - 1.3, but it should be considered FHWA is normally for highways so if your going to have a deep excavation near an existing structure such as a building, it is better to use a higher FOS. normally 1.5 would be logical
