API: uplift plates & anchors

Runs either the strict Meyerhof–Adams/Das plate-breakout branch or the FHWA GEC 4 straight-shaft ground-anchor branch. The complete scientific scope and applicability limits are documented at Uplift plates & ground anchors.

Endpoint

The request is a union on kind. Both branches requireunitSystem, an exact method ID, and an explicitdesign basis.

• plate requires an explicit circular or rectangular shape, ordered uniform-strength layers, explicit groundwater state, effective plate weight (net of buoyancy where applicable), and uplift coefficient.
• anchor requires complete unbonded/bond geometry, traceable per-length bond transfer, efficiency and reference, tendon SMTS/EA, independent tendon-grout capacity, design and lock-off loads, and a production-test record or explicitnot-tested state.

Authentication, rate limits, and the error envelope are shared by every endpoint — see the API overview. Requests are unit-agnostic: any self-consistent unit system works, and results come back in the same units. The examples below use SI (m, kN, kPa, kN/m³).

Request fields

This table is generated from the same schema that validates the request, so it cannot drift. Where a field appears once per variant row, it belongs to that variant only.

Response fields

Code examples

A complete, runnable request. Replace YOUR_API_KEY (or set PILECALC_API_KEY in your environment).

curl https://api.pilecalc.com/v1/uplift \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "unitSystem": "si",
  "kind": "plate",
  "method": "method.uplift.meyerhof-adams-plate",
  "shape": {
    "kind": "rectangular",
    "width": 1.2,
    "length": 2.4
  },
  "embedment": 3,
  "layers": [
    {
      "type": "cohesionless",
      "top": 0,
      "bottom": 1.5,
      "gamma": 18,
      "phi": 31
    },
    {
      "type": "cohesionless",
      "top": 1.5,
      "bottom": 3,
      "gamma": 18,
      "phi": 31
    }
  ],
  "groundwater": {
    "mode": "dry"
  },
  "plateWeight": 4.5,
  "upliftCoefficient": 0.9,
  "design": {
    "basis": "asd",
    "factorOfSafety": 2
  }
}'

The GET variant returns nominalResistance,designResistance,method.id as a plain-text CSV line — made for Excel's WEBSERVICE() and Sheets' IMPORTDATA() (see Use PileCalc in Excel). Omit fields to get the full JSON response.

For extended-creep, supply testing.extendedCreepCycles: six ordered objects containing load, targetPressure, andreadings of timeMinutes, movement and pressure. Use coherent request units: metres and kPa, or feet and ksf; time is always minutes. GEC 4 Table 23 observation times must be complete. The existing scalar one/ten-minute fields remain required for compatibility but do not qualify an extended-creep record. The declared maximum must equal the highest recorded load. The maximum test load must be at least 1.33 times the design load for permanent anchors (1.20 for temporary ones) and at most 1.50 times (FHWA GEC 4 specification 6.01 and §7.4.2.2); the sixth hold is checked at the applied test load. In creep-susceptible ground GEC 4 says 150 percent may be used, so a lower test load returns a warnings entry recommending 1.50. These are supplied-record checks, not verification of instrument calibration, the unloading sequence or project sampling.

Safety and migration

• The former width + area + soilType plate payload andgroutDiameter + bondedLength + adhesion anchor payload are rejected. They cannot identify plan geometry, units, groundwater, structural failure modes, or test status.
• Plate strength must be uniform. Sand additionally requires uniform unit weight and dry, fully submerged, or water-table-at/below-plate conditions. Clay permits variable unit weight and an internal water table in the integrated soil-weight term. Below the water table the clay soil weight uses the buoyant unit weight: Das (1980) Eq. 1 does not say whether γ is total or buoyant, so the lower-resistance reading is used, no base suction is credited, and warnings states the difference. There are no hidden water-unit-weight, plate-weight, bond-efficiency, or design-factor defaults.
• nominalResistance and designResistance are distinct. Do not apply a second factor to the design result.
• An anchor with testing.kind = "not-tested" returnsdesign-only-not-tested; it is not accepted for service.
• GET/Excel requests carry the identical JSON payload in req; response projection always retains mandatory scientific provenance fields.