Solved Morgenstern Price Problem

[SLL78]

I need to do a "hand" calculation of the MOrgenstern Price Method for a simple two material slope with a known failure surface. I cannot use slope stability software as what I need to do is a check of the output of the software.
Is there an example I can be pointed towards using hand calcuations and/or excel based calcs? (I can use excel spreadsheets for this).
I am getting stuck on how to solve the problems and cannot find any examples in the materials I currently have and am not sure where to look for additional information and solved problems in this method.
Thank You!

 

[dgillette]

M-P by hand? No thanks.

Any reason you can't use a simpler method of slices like simplified Bishop, which CAN reasonably be done by hand? (Or a chart solution, or 2-3 sliding blocks?) Haven't done it in 25 years and would have to go look it up, but you could probably do SB as a spreadsheet much easier than with a calculator. Oldestguy and some others around here have probably done it with a sliderule and a beam compass to draw the circle on "D-size" graph paper.

My recollection is that simplified Bishop gives slightly lower FS than more "complete" methods like Spencer or MP. In many situations, the difference could be within the "noise" of material properties, subsurface geometry, piez conditions, etc.

 

[SLL78]

Unfortunately no. We have to do a software verification as part of our QA program (we are highly regulated by our clients) which involves either solving a single failure surface by "hand" or inputting a solved M-P method problem from an established source (i.e. journal or text book) into the program to see if they both spit out similar problems.
The engineer who did this calculation also used Janbu method (which I think I may be able to do but would also appreciate assistance on). We had instructed the engineer to use Bishops but unfortunately she decided that our advice was not to her liking, even though this was her first time doing such a calculation. I'm basically cleaning up a mess at this point and its going to be a difficult journey!

Too much information I know, but I understand why Bishops would be suggested, it lends itself to being solved by hand/spreadsheet far more. In fact, I have a pretty nifty fill in teh blank spreadsheet for that and for Ordinary Method.

 

[dgillette]

Ouch. Pain. Might be just as quick to recrank several representative cases with SB, then check them by hand with SB, to show that all are consistent with the MP results. Can't recall right off, but doesn't MP vary the interslice-force angle for each slice, and maybe the height as well? I'm home today, and don't have my ref library.

 

[dgillette]

"Can't recall right off, but doesn't MP vary the interslice-force angle for each slice, and maybe the height as well? "

My point here was that there is an awful lot of iterating to get an answer, once the MP spreadsheet is set up.

DRG

 

[Dirtguy4587]

I've seen a spreadsheet set up to crank out M-P Analyses, but to simplify it enough to make the spreadsheet useable, it did make a number of assumptions. I would suggest that it will be quite challenging. Certainly it is do-able, and I would suggest digging out the original papers by Morgenstern et al. to work through the theory. In grad school, we did ordinary and bishops by hand, and were told that we'd always have software to do the more advance methods...

I've never seen a client request a "hand calculation" to back-up software. If I were the client (which I often am), I would tend to challenge the basis for the analysis input, rather than the analysis itself. If you've got the inputs right, shouldn't be a problem with the analysis (providing you are using a "commonly accepted" piece of software).

 

[ntschwanz]

Checking results is always a good idea. Unfortunately, the more rigorous the analysis the more difficult it is to check. We used to do hand checks of simpler analyses but since we have migrated to using Spencer procedure we check results of one program by using a completely separate program (i.e. Spencer result from software A is checked/verified against the Spencer result of software B). Perhaps this would be acceptable to your clients.

 

[cvg]

why not get the software manufacturer to provide the QA for his software? Certainly they should have done this when they wrote the program...