Soilicon

Verification

Compared with published and exact results

Every case below is a Soilicon project you can download, import and run yourself. The table is generated from the same files that Soilicon's test suite runs on every change, so the values shown are those of the latest development version.

Status: beta. These comparisons are evidence, not a formal sign-off. Every module still reports MODULE_UNVERIFIED. For the first three slope cases the geometry was reconstructed from published slice tables, so agreement within about 1 % is the realistic target.

Slope stability

CaseReferenceReference FSoilicon FDifferenceProject
Loading…

Steady seepage

CaseReferenceReference q (m³/s per m)Soilicon qDifferenceProject
Loading…

Using the files

  1. Download a project (a .json file in the soilicon-project format).
  2. In the app, choose Import in the project list and pick the file.
  3. Open the project and run its analysis. The project description names the source and the reference value.

To load everything at once, ask your AI assistant to import the files: it can call import_project on each one.

Internal checks

Besides the published cases, each release is tested against checks with exact answers:

  • Planar block: Spencer and Morgenstern–Price reproduce the closed-form wedge factor of safety to 10−15, dry, with water and with a strip load.
  • Bishop identity: GLE's moment factor at λ = 0 equals simplified Bishop to 10−9 on every test circle, including loads and ponded water.
  • φ = 0: on a circle, all methods give F = c L R / Σ W d, also with seismic coefficients.
  • Loads: strip, line and inclined loads match an independent Bishop iteration to 10−15.
  • Invariants: a mirrored model gives the same F, the choice of moment point does not matter, and loads outside the slip mass change nothing.
  • Seepage:
    • A manufactured solution converges at the expected order.
    • Sheet-pile flow under a pile is within 2 % of the exact Harr/Zhukovskii solution.
    • The Kozeny parabolic dam is within 2.5 % on discharge and 1–2 % on the phreatic line.
    • The mass balance is better than 10−8.

Next

Still to add: ACADS 1(c) and 2(b) (layered and large models), Fredlund & Krahn (1977) with ru, Arai & Tagyo (1985), published non-circular GLE cases, a pseudo-static benchmark, and a seepage-face height check against Polubarinova-Kochina.

Sources

  1. Rocscience, Slide2 Verification Manual, Hand Calculations 1 and 2 (after Sharma, in Abramson et al. 1996).
  2. Seequent, GeoStudio example Verification – Hand Calculations (SLOPE/W).
  3. Giam, P.S.K. & Donald, I.B. (1989). Example problems for testing soil slope stability programs. Monash University, Civil Engineering Research Report 8/1989.
  4. Harr, M.E. (1991). Groundwater and Seepage. Dover.