Isolated Footing Verification Studies
This article summarises how the StatiCAD isolated footing calculation and its connections with the other parts of the program were tested. For the formulas see Theoretical Basis of Isolated Footings, for use see Isolated Footing Design, and for an end-to-end application with screenshots see Worked Example: Reinforced Concrete Building with Isolated Footings.
1. Verification approach
- Calculation core: soil pressure, bearing capacity, section forces, reinforcement and automatic sizing calculations were tested in test programs by comparison with reference calculations independent of the program.
- Textbook examples: solved footing examples were compared line by line.
- In-program checks: saving, interface, report and drawing; soil springs (Y) and shell + Winkler (S) solutions; combined structure–foundation analysis and SAP2000 export were checked on the running program.
- Example with screenshots: all checks of one footing in the worked example were recalculated by hand.
2. Calculation core
| Test | Scope | Result |
|---|---|---|
| Isolated footing calculation | 49 soil pressure cases (40 random, most with partial contact), 14 full calculations, 4 bearing capacity, 2 automatic sizing and textbook examples; edge footing, circular column, steel column base plate, effective area option, top reinforcement and inadequate section behaviours | 2159 / 2159 checks, relative tolerance 10⁻⁹ |
| Partial contact solution | With an independent calculation application (Newton solution); each partial contact case also with a grid integral | Satisfied |
| Compression-only plate solution (S) | Element stiffness, symmetry and rigid body modes; node tributary areas; uniform pressure under uniform load; contact without uplift under central load; uplift at large moment (contact 69%); oblique moment in a rectangular footing with an eccentric column (contact 43%); corner uplift under central load in a wide thin plate (contact 39%); reflection of the tension field in moment, shear, punching and reinforcement. The reference calculation is made by a different route from the program (element stiffness by analytical integral) and each result is also tested by direct minimisation of the energy. | 438 / 438 checks, difference ≤ 1.3·10⁻⁹ |
| Spring matrix of the soil springs (Y) solution | Node at the footing centre, eccentric in x and in both directions; rotated rectangle; rotated and eccentric footing; 90°. The reference matrix is computed by Gauss integration over the base | Satisfied; the diagonal staying at the centre and the transfer rule to the node were also checked |
3. Textbook examples
The solved examples of textbook Chapter 12 "Foundations" are in the test set. Because the book uses gross soil pressure in design, these examples are calculated with gross pressure; where the book has arithmetic errors, the corrected value is taken as the basis.
| Example | Quantities checked |
|---|---|
| SU 12.1 | Bearing capacity qk (Nq 23.18, Nc 35.49, Nγ 27.72; shape and depth factors) |
| SU 12.3 | Axially loaded square footing: punching (Ap 2.4025, up 6.2), shear, moment, minimum reinforcement |
| SU 12.4 | Uniaxial eccentricity: corner pressures, γ, Vpr, shear, moment, reinforcement |
| SU 12.5 | Rectangular column, rectangular footing: pressure, γ 0.9432, shear, moment in two directions |
| SU 12.6 | Biaxial pressure: corners 218.40 / 68.40 / 198.26 / 348.26 kN/m² |
| SU 12.7 | Biaxial eccentricity: corner pressures, γ, Vpr, shear and moment in two directions |
4. In-program checks
| Topic | Result |
|---|---|
| Saving | When the project is saved and reopened all footing fields are preserved; footings are cleared in a new project; Undo works; in a project without isolated footings no isolated footing information is written to the file. |
| Transfer of column forces to the footing axes | The sign mapping was verified with a single-column model. |
| Interface | Drawing, selection, copy / move / mirror / rotate / shift, Find Object, selection filter, selection group and project solution work as expected; rotation is undone with Undo. |
| Report, drawing, quantity take-off, 3D | The report PDF, the foundation plan and detail sheet (DXF), the 3D solid view and selection were tried. |
| Tie beams | 4 tie beams in a 2×2 grid; exact match with the hand calculation (Ln 300, Nb 12.47 kN, minimum reinforcement 4Ø12, Vcr 105.3 kN). |
| Soil springs (Y) solution | In Q and earthquake load cases Uz·kz = −N; Rx·kθX and Ry·kθY exactly equal to the base moments; with a very stiff spring the displacements are 10⁻⁸. |
| Y solution, eccentric node | Columns drawn from the edge (node 25–40 cm off the footing centre), rotated square and rectangular footings: the 3×3 spring matrix going to the analysis is within 10⁻¹⁵ of the independent calculation. In a thick (H = 300 cm, rigid) footing, taking the shell solution (S) as the reference, the difference in combination base moment is 1–3% median, 7–20% at most; the difference in peak soil pressure is 2–5%. The small remaining difference comes from the definition of axial force (in S half of the column self-weight is lumped at the base node). |
| Shell + Winkler (S) solution | In Q and earthquake Σ spring force = N (10⁻¹⁰); in G the difference is half of the column self-weight lumped at the base node (9.2 kN); moment equilibrium 10⁻⁶–10⁻¹⁰, also in a footing rotated 30° and eccentric. In an H = 300 cm footing the pressure is fully planar (rigid behaviour), at H = 60 cm the deviation from the plane is 0.2%. Between mesh sizes of 25 cm and 10 cm the design moment difference is less than 0.2%. |
| Compression-only soil in the S solution | The linear solution of the local plate model differs from the global finite element solution by 6·10⁻⁵ at most. In the example model the base partially lifted in most earthquake combinations and was solved again compression-only (at most 9 iterations, equilibrium 10⁻⁸). In an H = 300 cm footing the result coincides with the rigid partial contact solution with the same forces (qmax difference ≤ 0.4%). Compared with the linear (tension-allowing) field, the peak pressure is 3–4 times and the punching ratio 2–6 times larger. In the combination at the overturning threshold (contact 0.8%), the rigid partial contact solution is used instead of the local solution. |
| Column drawn eccentric | Combination moments exactly equal to M + N·(xnode − xsection); the S result does not change. |
| Vertical load transfer check | In Y and S the spring reactions enter the support reaction sum; no “partial error” warning appears. |
| Combined structure–foundation analysis (isolated footings only) | No “foundation modelling” warning appears; in K, Y and S the period, base shear force and all footing design values are exactly the same as the classical way. When the footing of a column is deleted the warning comes back. |
| Mixed foundation (strip + isolated), force distribution | A strip footing 1 m wide and 8 m long on one axis; two columns on isolated footings. In the classical calculation the strip columns are taken fixed; in the combined analysis the strip relies on the rotational resistance of the soil about its own axis (ks·b³/12 per unit length; 6540 kN·m/rad in total for this strip). Share of the earthquake shear of the columns on isolated footings (along the strip / perpendicular to the strip): in the classical calculation K 50% / 50%, Y 33% / 25%; in the combined analysis K 62.5% / 83.1%, Y 44.7% / 61.8%. Total base shear and period are the same in both ways. An independent 3D frame solver, with the nodes, bars, supports, springs and offsets of the program, gave K 81.5%, Y 61.0% in the direction perpendicular to the strip (difference ≤ 1.6 points); when the strip columns were made fixed it gave the shares of the classical calculation. |
| Mixed foundation warning | Appears in the classical calculation in K and Y solutions; tie beams are not counted as strip footings. It does not appear in the combined analysis or when only tie beams remain; the English text is correct. |
| Mixed foundation with tie beams, combined analysis | Columns drawn from the edge, tie beams pass through the column centre (between two isolated footings and from an isolated footing to the strip). The nodes in the footing outline are rigidly connected to the footing (7 links in each footing); the earthquake shear share of the two symmetric footings is 26.5% / 26.5%. Share of the columns on isolated footings (x / y): classical K 56.0% / 61.3%, soil springs 53.0% / 53.3%, shell 53.0% / 53.2%. The independent 3D frame solver with an exact rigid constraint gave K 57.1% / 61.3%, Y 53.8% / 54.1% (effect of the finite stiffness of the link bars ≤ 1.1 points) and reproduced the program's connected model within 1.5 points. In the connected footing the reaction of all load cases is taken (including scaled vertical earthquake). |
| SAP2000 export | Read back from SAP2000 V27: spring bases with Ux, Uy, Rz restrained, kz 80,000 kN/m, kθ 26,667 kN·m/rad; classical bases fixed. In an eccentric or rotated footing the six terms read back of the coupled point spring are within 5·10⁻¹⁶ of the independent calculation (since the SAP2000 model mirrors the plan y axis, the sign of two terms is reversed). |
| SAP2000 export, mixed foundation with tie beams | Support type “spring”, columns drawn from the edge, tie beams from the column centre: in each footing the points in the outline are in a rigid body constraint, the line springs of the pieces inside the outline are deleted; at the footing points Ux, Uy, Rz restrained, a soil spring in the vertical, the two rotations free. The strip rotational spring is written as point springs: the sum of the 8 points on the strip is 6540 kN·m/rad (same as the program), 22.07 kN·m/rad per unit length in the tie beams. Share of the columns on isolated footings (x / y): in soil springs SAP2000 51.7% / 54.1%, program 53.0% / 53.2%; in the classical solution SAP2000 57.1% / 63.0%, program 56.0% / 61.2%. |
| Automatic sizing | In the example with screenshots (12 footings, 66 combinations per footing) the two sizing rounds take 16.6 and 30.6 seconds. The sizes found by the fast search (angle scan, exponential step and halving) and all numbers of the final design are the same as in the detailed scan (relative difference ≤ 3·10⁻¹⁵). |
5. Hand verification of the worked example
In the three-storey reinforced concrete building in Worked Example: Reinforced Concrete Building with Isolated Footings, 12 isolated footings (soil springs, Y) and 17 tie beams were entered through the interface. The interior footing TT05 (180 × 180 × 60 cm) was recalculated by hand:
| Quantity | Hand calculation |
|---|---|
| Soil pressure | qmax = 285.0 kN/m², ratio 0.95 |
| Punching | Vpd = 377.4 kN, Vpr = 2772 kN |
| Shear strength | Vcr = 801.8 kN |
| Flexural reinforcement | As required 399 mm², minimum 2160 mm² |
| Column starter bars | lb = 484 mm (available 514 mm) |
| Soil springs | kz = 64,800 kN/m, kθ = 17,496 kN·m/rad |
All the values are the same as in the report. In the Y solution two “size → analyse” rounds were needed: as the footing grows, the rotational spring and the base moment of the corner column increase. With the same dimensions, in the classical (K) solution the periods are 26–29% shorter, the base shear is 26–32% larger and the footings overturn; the shell + Winkler (S) solution is within 1–4% of Y (punching ratio of the corner footing 0.25 instead of 0.72).