Bishop's method is a fine.  OMS may give you more 'stable' results with some geometries and soil conditions; and Janbu will theoretically give you "better" results.  But I very seldom have a reason to stray from Bishop -

Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.  See FAQ158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

As for how to do the anayalis, you should run an analysis of the existing conditions, i.e. with no additional fill.  Then starting adding the fill, not as a surcharge but as soil layers in the analysis.  Just keep adding the soil and changing the geometry until you are happy with the results.

Geopavetraffic,
what about running a prior analysis on the existing slope, as you suggest, and then adding a generic distributed surcharge, of the same lenght of the design fill and simulate progressive fill elevation by increasing the value of such distributed surcharge ?
This sounds less awkward then re-punching the data for the slope profile + layers profile, unless there is some specific reason to follow your method...

         

That will work, unless you expect that the fill will have some strength that you wish to model as well. (Possible.)  It also won't work if the presence of the fill alters the phreatic surface such that it is within the fill as well. (Unlikely.)

Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.  See FAQ158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

Using a surcharge may work just fine, assuming as Focht3 pointed out that the material has no strength, and causes no changes to the existing soil.  But many software packages treat surcharge loads funny and cause problems with the analysis.

Thanks for your suggestions everyone.  I am doing the calculations in an Excel spreadsheet that I am creating.  If I model the fill as a surcharge, would I simply add the surcharge stress to the weight of the underlying slices?  Or is there also a horizontal stress acting on "downslope" slices as there would be if this were a retaining wall design?

Thanks again.

How tall is this slope?

The existing slope is about 16 feet tall, vertically, with a 1:1 slope.  It is a river bank and the river is about 6 feet deep over the toe.  The slope consists of about 18 feet of fill material above about 26 feet of silt, underlain by stiff silt (44 feet from the top of the stiff silt to the top of the slope).  The final construction will consists of about 6 feet of fill set back about 50 feet from the top of the slope.  During construction, the soil will be preconsolidated with (TBD)feet of fill set back about 20 feet from the top of the slope.

                                                               ________
                                             <- 20'->    / New fill
                            _________________/________
                           /            
______________/___ WT (10')__________  ______
__River______/               Exist Fill (18' thick)

                 ______________________________
                   Compressible Silt (26' thick)
____________________________________________________________
             Stiff Silt

JasonG:

I agree with all of the advice given so far.  One additional point that I would make is that you should rule out the possibility of a thin weak horizontal layer or preexisting failure plane and that you may have a noncircular failure surface.  We just looked at a slide that was triggered on a riverbank where about 20% of the proposed fill was placed and a very long planar failure surface was activated.  It turns out that this was probably a preexisting slide plane and its geometry in relationship to the location of the new load triggered the slide.

In my opinion, prior to running the analysis, you should develop, from fundamental principles, an understanding of the likely failure modes and then use the stability analyses (and parametric studies) to confirm or disprove the postulated failure modes.

Glen

I would suggest that you purchase a good quality commerical slope stability program instead of trying to use a spreadsheet.  The analysis requires that the factor of safety be iterated and the forces properly accounted for.  It can be done by hand, however, it takes a long time and prior to putting it into a spreadsheet you will have to solve the problem by hand to check the answer.

Unless you have done a lot of stability analysis by hand, don't start now.

As far as the iteration goes, spreadsheets like Excel do have iteration capabilities, and I applied them in this specific case of figuring out the stability safety factor.  
But, with the exception of relatively simple models, slope stability spreadsheets can really drive you nuts, expecially so when applying the Bishop method, locating the failure circle(s) and assigning the strenght parameters along its lenght. And you are supposed to do it for tens of circles!
I too suggest a commercial software, today they're pretty much inexpensive. One advantage of spreadsheets over most commercial software though is that, by means of macros or commercial add-ins (or by approximation methods such as the one described by Mike Duncan) you may carry out a probabilistic stability analysis.
One commercial product which advertises probabilistic and reliability capabilities is Rocscience's Slide5. It comes at US $ 1495, not too bad. You come across pretty good stuff at very reasonable prices, in the region of US$ 300.
      

Unfortunately, purchasing software is not an option for me.  I'm making headway on an Excel spreadsheet.  With some intense logical funtions, geometry equations, and macros, I should be able to duplicate the basic funcionality of the software package.  It's labor intensive but a great learning experiance.  I'm not a big fan of using canned software without paying my dues by hand first anyway.

You will find the need to recalculate the slice weight, inter-slice forces, lines of thrust, etc. for each small change in the geometry is very tedious.  If you are a practicing engineer, the small cost of commercial software ($300-$500 U.S.) makes hand calculations look exorbitantly expensive (considering the loss in billable time.)

If you're a student, this isn't a "homework help line"...

Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.  See FAQ158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

Focht3:

Yes, I am an engineer and yes this is tedious.  Once the spreadsheet is complete, the calculations will be revised automatically as the circular slip surface is varied with a macro. Getting there is quite a task though.

Can you suggest a software in the price range you mentioned?  The couple I looked at were around $1,000.  I may be able to justify to the "higher ups" the purchase of a software that is only a few hundred dollars.

Get Snail from Caltrans' web site. It's basically a soil nail wall design program that is also used for simpler slope stability problems.  It's free and downloadable.

Try this link:
www.dot.ca.gov/hq/esc/geotech/request.htm

Just a short note - if you really want calculate the slope stability the old fashioned way - use hand drawn sections and a calculator and the table from Terzaghi and Peck.  This is far more tedious than any spreadsheet program - but, after having done it a number of times, it sure helped to give insight to how to pick the critical circles quicker!
Secondly, perhaps a commercial software company (example G-Slope) will sell you an old DOS-based version for peanuts especially if they know you WILL NOT spring for the $1000 version.  The DOS version is quite a good program - not as fancy, but still it does the job.  You can do a "set" of analyses for a particular configuration for a simple case in less than 15 to 30 minutes.

I am not sure what you intend to build on the site interms of structure but noting that you will have six feet of fill finally and that your river bank exists at a 1:1 slope etc I wonder if I would not spend my time examining why the bank is standing at a 1:1 slope first before undertaking slope stability analysis of the surcharge. I presume that tis is a proposed a building site as it seems that you are taking pains to consolidate the existing ground.

Personally, I think you will be wasting money applying a surcharge to the ground as presented. Existing Fill and compressible silt are terms that need to have some test data for proper engineering interpretation. Fill can be so described from visual observations of the stratigraphy at the site but has good engineering characteristics. The bank standing at 1:1 says something.

 

JasonG
you may have a look at this site:

www.geostru.com

this italian firm sells a slope stability software at
US$ 370.21 (today's rate-of-exchange euro-dollar)

I have used it in my latest 2 projects and it is pretty good, sure good quality/price ratio. Lots of capabilities.

JasonG

Waht is the existing fill, and what will the new fill be?

regards

The existing fill is sand and gravel.  The new fill will probably be similar.

I think you are to praised for building a spreadsheet.  Programs are only as smart as the people that use them and building your own program is a time consuming, but highly beneficial way of understanding the method of slices.

One useful resource with the spreadsheet is "Slope Reliability Analysis Using Generalized Method of Slices."  It is in the Journal of Geotech & Geoenviron. Engineering Vol 124, No 4 - April, 1998.  It shows how to set up Morgenstern and Price and Spencer methods (which satisfy all conditions of equilibrium) and takes advantage of Excel's solver to do some searching for you.

TerraBoy,

Thanks for the kudos.  It was a daunting task, but I was able to develop an excel spreadsheet that completely automates the simplified bishop method.  The output is practically identical to that of commercial software.

a disadvantage to writing your own software is getting the agencies / regulators to accept it.  Maybe the only way for them to approve your work is for them to run another program and try to duplicate your results.

Has it come to that?  What difference?  Do they not accept hand versions of Bishops analyses (aka vide Terzaghi and Peck's book) like we used to do before personal computers??  Why do they want to "know" the programme?  Just a bunch of questions but . . .

It's been my experiance that regulators must accept the calculation method (i.e. bishop), not the method used to perform the calculations (i.e. software).  They may not accept the results from a software package that uses it's own method.

Hmmm,

I always use the software to "zoom in" on the critical circle, then confirm the answer with hand calculations.  It's quite efficient from a time use standpoint -

Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.  See FAQ158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"