Hello:

Develop an exercise in ADONIS and RS2. Find similar results of FOS (1.12 ADONIS and 1.15 RS2) and similar color maps of "total displacement" , But, the results "total displacement" are different, more highs in ADONIS (750 m) y lows in RS2 (5 m). Why?

I share save script of ADONIS.

The exercise of RS2 is: https://www.rocscience.com/help/phase2/webhelp9/pdf_files/tutorials/Tutorial_08_Shear_Strength_Reduction.pdf

Thanks.

thanks for the comment.

750 m is a very high I need to check the code. my guess is that after each FOS calculation iteration the total displacement has not been zeroed out which is the bug in this case. please let me check the subroutine and get back to you.

but I still am not expecting to get similar displacement calculated by RS2 since two programs are using different subroutine to calculate the FOS and also keep in mind that this displacement value is the final displacement at the stage of failure which might be different for each case.

jpulecio

thank to your comment I could find out an existing bug in fos calculation. total displacement/velocity was not set to zero after each fos iteration which caused the disp to accumulate and as a result very large displacement was reported. this bug is fixed now at version 2.2.3. please go ahead and download the latest version and give it a try and let me know of result.

-Roozbeh

Install the new ADONIS.

Excuse me, but, now the contour of total displacement is different.

I share save script of ADONIS. Also, images of each software.

Thanks.

correct, as I mentioned in previous post, there is no guarantee to get similar displacement since this displacement is at stage of failure and different programs are using different criteria and methods to estimate that.

Hello:

1.) ADONIS: total displacement is at stage of failure.

2.) RS2: total displacement is after the stage of failure?
Because, the total displacement is more highs in RS2 (5 m) y lows in ADONIS  (0.5 m).

3.) From your expertise, ¿What other results of are important?

Thanks.

what I mean by "state of failure" is that estimated fos value is right at the border at which model could be stable or unstable. theoretically speaking, epsilon greater than fos value could cause model to fail and cause large displacement. different programs are using different methods to estimate this reduction factor. in order to understand how fos calculation works you need to study reduction shear strength method. I recommend you to read RS2 manual and also you need to study fos calculation in FLAC (which ADONIS fos calculation is based on). you will understand the different between two methods.

Thanks.

... the total displacement is more highs in RS2 (5 m) y lows in ADONIS  (0.5 m).

i think that something still goes wrong.
criteria can be different, but results have to be of the same order.
such difference (5 m is 10 times greater than 0,5 m) is not acceptable.
maybe a bug in the code, very hidden, is still there.

what about flac results for the same ssr test?

it would be a good comparison if you end up getting the exact same FS value which is not the case right now.

I believe the best approach to check the result would be manually modifying the cohesion and friction using following equation and run the model without fos subroutine ("solve()") and check the displacement with the one given by "solve("fos")" if they are different there is bug. also repeat the same procedure with RS2.

cohesion_new = cohesion_original/FS;
friction_new = ATAN(Tan(friction_original)/FS);

1-FS is the factor of safety calculation by each program which is not equal in this case.
2- also one more sensitivity run can be done using exact same FS value for both ADONIS and RS2.

again I really recommend you all to read the Shear Strength Reduction (SSR) method for better understanding the procedure rather than just looking at end result plot from different programs.

Hello:

Applied:
cohesión_new = cohesión_original / FS;
friction_new = ATAN (Tan (friction_original) / FS)

In ADONIS and RS2: FOS = 0.98 --> OK.

Results with the model without fos subroutine ("solve()"): 10 times greater in ADONIS.

i did the test by myself.
here the files used and results obtained.

error in jpulecio procedure was that he used starting values of cohesion and friction angle (that are 5 and 30) with solve() and reduced values with solve("fos").

(part 2 follow, due to dimensions of images)

see part 1 above.

thanks reversi