Helical (screw) piles

Helical (screw) piles carry load on discrete steel plates screwed into the ground. The helical pile tool computes axial capacity by the two standard limit-equilibrium methods and takes the lesser as governing — the approach of Perko's design guide and the Hubbell/Chance Technical Design Manual — with the AC358 torque correlation for field verification. The released model is explicitly limited to vertical, round-shaft, deep installations; shallow failure is rejected.

The two capacity methods

Q_IBIndividual bearing

Each helix bears independently on its full plate area: Q = Σ q_ult·(π/4)·D² + shaft adhesion. Cohesive bearing uses Nc(z/D)·c ≤ 9c; cohesionless uses σ′v·Nq with the non-displacement Nq (a helix is screwed in with little soil displacement). Governs when helices are widely spaced (≥ ~3 diameters).

Q_CSCylindrical shear

The soil between the top and bottom helices fails as a cylinder of the average helix diameter — soil-on-soil shear (δ = φ) on the cylinder wall, plus bearing of the leading helix, plus shaft adhesion. Governs when helices are closely spaced.

Q_ult = min(Q_IB, Q_CS)

Governing capacity

The released ultimate is then capped by the product's explicit structural axial limit. For an installed pile, a lower capacity implied by the measured torque also caps the released result rather than being hidden behind a pass/fail label.

Shaft & helices

Enter the shaft diameter and embedded length, then the helix plates (depth and diameter, up to 8). Shaft adhesion is counted from grade down to one helix diameter above the top helix — the plate's disturbed zone is excluded (Chance TDM / Perko 2009). All helix depths must lie on the embedded shaft. The public contract requires top-helix embedment 5.1 ≤ H/D ≤ 134 and adjacent spacing 1.55 ≤ s/D ≤ 4.5, matching the published Hoyt–Clemence test population. Every helix must be larger than the round shaft.

The soil profile and groundwater inputs are the same as the axial tool's — helical and straight piles in the same profile are computed with the same NAVFAC DM-7.02 resistance laws.

Compression vs tension

The cylindrical-shear leading helix changes with direction: the bottom plate leads in compression and the top plate in tension. Cohesive-soil uplift can use the released deep-helix expression. Cohesionless uplift requires measured installation torque so the result is capped by field evidence; an analytical sand-bearing result alone is rejected.

Field evidence is direction-specific

Hoyt and Clemence's 91-test population is short-term uplift evidence. The response flags whether a request matches that field population, so a compression result is clearly distinguished from the uplift cases the correlation was derived on. An executable PileCalc comparison against independent centrifuge measurements published by Tsuha and Aoki shows that PileCalc's former full-area NAVFAC sand-uplift expression overpredicts all three medium-sand cases inside the public friction-angle domain. The paper's own torque-derived equation is a different model and reported good agreement. This model distinction is why PileCalc now fails closed without measured torque.

Installation torque (AC358)

T_req = Q_ult / K_t

Torque correlation (Hoyt & Clemence 1989)

Installation torque is an independent field check: capacity correlates with final torque through K_t. PileCalc does not choose K_t from shaft diameter. The request must cite a product evaluation report or project load test, and installed torque must be averaged over the final penetration equal to at least three largest-helix diameters. Pre-install mode returns a target; installed mode compares the measured average and conservatively caps capacity if it does not support the analytical result.

Kt is project or product specific

Published field tests show that the observed torque factor changes with pile geometry, installation, soil, and the selected failure criterion. A literature-average K_t is not a substitute for the applicable evaluation report or project load test.

Factor of safety

The factor of safety is explicit and project-controlled. A value of 2 is common when the applicable evaluation report and installation verification support it, but PileCalc does not infer that suitability from geometry or torque alone.

Reading the results

Governing method — which mechanism limits the pile; the comparison bars show how close the other is.
Per-helix table — the individual-bearing contribution of each plate; undersized upper plates show up immediately.
Required torque — hand this to the installer as the acceptance criterion.
Applicability and convergence — H/D, adjacent spacing, field-population match, and fine/coarse integration differences accompany every result.

Published empirical uncertainty

In the 91-test Hoyt–Clemence uplift study, published actual/calculated capacity ratios had mean / standard deviation 1.50 / 1.18 for cylindrical shear, 1.56 / 1.28 for individual bearing, and 1.49 / 0.88 for installation torque; the reported ranges were 0.07–7.29, 0.03–7.04, and 0.30–4.67. Those aggregates document large empirical scatter but are not a PileCalc holdout score because the individual test inputs are not published in a machine-reproducible table. See the public 1989 paper.

No rock bearing

Helix plates bearing in rock are outside the method — the engine rejects rock layers at a helix depth. Use the drilled shaft or axial tools for rock-socketed elements.

No shallow extrapolation

A top helix shallower than H/D = 5.1 is outside the released method. PileCalc rejects it; it does not reuse the deep bearing factors or clamp the embedment ratio.