Accumath

Toolchest · Datasheet

Accumath Toolchest — Soils: Unified Soil Classification System (USCS)

Dataset file: accumath-soils-uscs.csv Compiled: 2026-10-20 Rows: 24 group symbols — 15 primary groups, 8 borderline dual symbols for coarse-grained soils, 1 fine-grained dual symbol (CL-ML), plus Pt.

What this dataset is

The Unified Soil Classification System (USCS) group symbols, their published group names, the published laboratory classification criteria that assign a soil to each group, and the typical engineering-property ratings the reference tables publish for each group (permeability class and coefficient range, qualitative shear strength, qualitative compressibility, workability as a construction material, and typical embankment use).

The USCS originates with the U.S. Army Corps of Engineers (Waterways Experiment Station, Casagrande/Wagner work of the 1940s–1950s) and is standardized today as ASTM D 2487. This dataset was not taken from the ASTM document. It was taken from two open U.S. government publications that republish the system:

Both are U.S. government publications (public domain). Group names, criteria values, qualitative ratings, and permeability ranges are published facts, recorded here with attribution; no standard's expressive text or table layout is reproduced beyond the factual entries themselves.

How the system divides soils (the published rules)

Soils plotting on or above the A-line behave as clays (C); below it, as silts (M). LL = 50 is the second dividing line: below it, low plasticity (L — ML, CL, OL); at or above it, high plasticity (H — MH, CH, OH). Working rules (FHWA Table 4-9): PI > 7 and on/above the A-line → clay; PI < 4 or below the A-line → silt; PI between 4 and 7 plotting on or above the A-line is the hatched borderline zone → dual symbol CL-ML. - Organic soils (O). Identified by the liquid-limit ratio test: a soil is classified organic when LL(oven-dried) / LL(not dried) < 0.75. (The NRCS Figure 3-17 prints this test in an oven-dry/air-dry form with a 0.7 limit; the 0.75 oven-dried/not-dried form in FHWA Table 4-9, which follows ASTM D 2487, is the one recorded in the CSV.) - Highly organic soils (Pt). Peat and muck, identified in the field by color, odor, spongy feel, and fibrous texture — not by chart position. - Fines of coarse-grained soils are classified with the same chart rules: fines classifying as ML or MH give GM/SM; fines classifying as CL or CH give GC/SC; fines classifying as CL-ML give the additional dual symbols GC-GM and SC-SM (FHWA Table 4-9, note d). Those two symbols are real members of the system but are not given rows in this dataset, which carries the 24 symbols it was scoped to; they are recorded here so the set is not mistaken for the whole story. - Group-name modifiers (not separate symbols): ≥15% sand or gravel adds "with sand"/"with gravel"; 15–29% material retained on the No. 200 adds "with sand"/"with gravel" (whichever predominates); ≥30% retained adds "sandy"/"gravelly"; cobbles or boulders in the field sample are named in the group name. Organic fines in a dual-symbol soil add "with organic fines."

How to read the property columns

Caveats — classification is not a design value

The USCS is an index and communication system: it groups soils that behave alike so engineers can talk about them and anticipate behavior. The ratings in this dataset are qualitative generalizations for compacted material, published for guidance. They are not design properties. Actual values of strength, compressibility, and permeability for a specific soil depend on its density, moisture, stress history, and fabric, and must come from laboratory and field tests on that soil (gradation and Atterberg limits for classification; permeability, consolidation, and shear tests for design). A group symbol on a boring log tells you what questions to ask; it never answers them. This is the same framing rule as the Toolchest concrete dataset: published typicals, honestly labeled, never dressed up as guarantees.

Gaps and source discrepancies (disclosed, not smoothed over)

  1. Dual-symbol rows carry no property ratings. NRCS Figure 3-22 rates the 15 primary groups only. The CSV states this in the cells rather than averaging or interpolating between the component groups' ratings.
  2. Pt is unrated in the source table except for its construction verdict ("NOT SUITABLE FOR CONSTRUCTION," Fig. 3-22; "DO NOT USE FOR EMBANKMENT CONSTRUCTION," Fig. 3-23). Its rating cells say so.
  3. GC and SC permeability. The cm/sec column of Figure 3-22 in the NRCS PDF does not extract cleanly for these two rows; the recorded range (10⁻⁶ to 10⁻⁸ cm/s, 3×10⁻³ to 3×10⁻⁵ ft/day) follows the same chart as originally published by the Corps of Engineers (USCS chart after Wagner, 1957) and is consistent with Figure 3-22's own ft/day column for those rows. All other rows' ranges read directly and unambiguously from Figure 3-22.
  4. Inequality signs. NRCS Figure 3-17 prints "Cu > 4" / "Cu > 6"; FHWA Table 4-10 and the D 2487 criteria use "≥". The CSV records "≥", following FHWA/D 2487.
  5. Organic ratio variant (0.7 oven-dry/air-dry in NRCS Fig. 3-17 vs. 0.75 oven-dried/not-dried in FHWA/D 2487) — see the rules section above; the CSV uses 0.75.

Sources