Uplift plates & anchors

The uplift tool contains two separate analyses: explicit plate breakout by the declared Meyerhof–Adams/Das formulation, and straight-shaft grouted ground-anchor design and test-record screening under FHWA GEC 4. They do not share a generic “uplift factor,” and neither path silently supplies units, groundwater, bond strength, or design factors.

Status and scope

• Plate: horizontal circular or rectangular plates with uniform strength in one drained or undrained failure mechanism.
• Ground anchor: one straight-shaft grouted anchor with traceable ground-grout and tendon-grout capacities.
• Not included: helical anchors, belled shafts, anchor groups, mixed drained/undrained failure mechanisms, or arbitrary strip anchors.

Use the dedicated helical-pile method for helices and thedrilled-shaft method for shaft or bell uplift.

Plate breakout

A rectangle uses its physical shorter side B and longer side L. For a circle, B = L = D. Area, perimeter and aspect ratio are reported explicitly; aspect ratios above 10 are rejected.

The square-plate critical embedment ratio is multiplied by min(0.133L/B + 0.867, 1.4). In deep mode, the breakout factor freezes at the critical depth; effective overburden remains evaluated at the actual plate depth.

Clay critical depth is B · min(0.107c_u + 2.5, 7) · min(0.73 + 0.27L/B, 1.55), with c_u converted to kPa for this empirical equation. The shallow transitionβ uses linear interpolation of the digitized average curve in Das (1980), Figure 4; the source-reading ordinate uncertainty is approximately ±0.02. The response reports nominal and ASD/LRFD resistance, shallow/deep mode, critical depth and component forces.

References: Das and Jones (1982), sand and Das (1980), clay.

Layers and groundwater

Layer records must begin at zero, remain contiguous, and cover the complete embedment. Every record must have the same friction angle or undrained strength. Heterogeneous strength is rejected: selecting the weakest layer does not establish a conservative breakout mechanism.

• Sand: uniform unit weight and 20° ≤ φ ≤ 45°. Groundwater may be absent, at the ground surface (fully submerged), or at/below the plate. A water table inside the embedment is unsupported.
• Clay: uniform positive undrained strength. Variable unit weight and an internal water table are supported through the exact integrated effective soil-weight term.
• Hydrostatic inputs: explicit water-table depth and water unit weight; submerged unit weight must remain positive.

Mixed drained and undrained profiles are rejected because the model has no validated composite failure mechanism.

FHWA GEC 4 anchors

The anchor path follows the preliminary load-transfer and structural screens in FHWA-IF-99-015, Ground Anchors and Anchored Systems. Ultimate transfer is supplied per unit bond length from a named GEC 4 table, qualified project test, or other explicit source. The engine does not infer it from grout strength.

• Soil and intermediate-geomaterial bonds are limited to 4.5–12 m; competent rock to 3–10 m.
• Bar unbonded length is at least 3 m; strand unbonded length is at least 4.5 m.
• Inclination is 10–45 degrees below horizontal.
• Bond-zone center cover is at least 4.5 m; shallower ground-mass breakout is unsupported.
• Preliminary ASD factors are at least 2 for soil/intermediate geomaterial and 3 for competent rock. LRFD requires a project factor and reference.

Ground-grout bond, tendon rupture, and tendon-grout bond are shown separately. Grout compressive strength is retained only for material traceability. Tendon-grout resistance requires qualified product or test evidence such as ASTM A981 data.

Production testing

GEC 4 requires every production anchor to be load tested. The tool records proof and performance scalar observations, checks maximum test load and apparent free length, and applies the published short/extended hold screens. For permanent anchors, extended-creep testing accepts six complete load-hold records through 10, 30, 30, 45, 60 and 300 minutes. Each record includes the hold load, target pressure and measured time, movement and pressure observations. Every final logarithmic cycle must satisfy the 2 mm criterion; no passing last hold can hide an earlier failed hold. The 4.5-minute observation is interpolated on log(time). Missing records remain rejected. Temporary extended-creep schedules remain unsupported.

• Proof/performance short hold: movement from 1 to 10 minutes must not exceed 1 mm.
• If the hold is extended: movement from 6 to 60 minutes must not exceed 2 mm.
• Apparent free length must be at least jack length plus 80% of the design unbonded length.
• An untested anchor remains visibly design-only-not-tested and is not accepted for service.

Fail-closed limits

The request is rejected, rather than extrapolated, for:

• missing unit system, groundwater data, design basis, method ID, or evidence reference;
• gapped, overlapping, reversed, or too-short plate and bond profiles;
• heterogeneous plate strength, mixed plate failure mechanisms, or mixed soil/rock anchor bond zones;
• unsupported aspect ratio, friction angle, anchor inclination, cover, or bond length;
• inconsistent SI/US water unit weight or nonpositive submerged unit weight;
• test records whose loads, movements, or observation chronology are impossible.

Validation evidence

Deterministic gates

• Component and ASD/LRFD algebraic identities.
• SI/US physical twins for plate and anchor results.
• Published sand and clay aspect-ratio equations, digitized clay transition, supported groundwater cases, and orientation invariants.
• FHWA GEC 4 worked example: 100 kN/m over 8.9 m gives 890 kN nominal and 445 kN at FS 2.
• 60/70/80% tendon limits, creep screens, apparent free length, and negative applicability tests.
• Identical strict request/result semantics through core, contracts, Convex, REST/Excel serialization, saved rerun, and reports.