Verification & Validation

This page brings together worked examples in which GEMSai results are checked against independent references — physical model tests and established closed-form and chart-based solutions — so that you can see for yourself how the software performs. Each example sets out the full input data, the reference it is compared against, and the results side by side, including the places where the two differ and why.

We will keep adding examples across the GEMSai foundation analysis suite.

Offshore Pile Foundation Laterally loaded pile in sand vs centrifuge model test API p-y analysis compared with a 50 g centrifuge test — pile head deflection and bending moment

Comprehensive Pile Foundation Laterally loaded long pile vs Reese charts Finite element analysis compared with Reese charts for the n_h soil model — deflection and bending moment with depth

Laterally Loaded Pile in Sand — GEMSai vs Centrifuge Model Test

Module: Offshore Pile Foundation — laterally loaded pile analysis Version: GEMSai Offshore Pile Foundation v1.10 Method: API p-y curves for sand (static) Reference: centrifuge model test at 50 g [1]

This example compares the response of a single pile in dry sand, measured in a geotechnical centrifuge, with the response computed by the GEMSai Offshore Pile Foundation software using the API p-y method. Centrifuge testing reproduces full-scale stress levels in a small model, so a 25 mm model pile spun at 50 g behaves like a 1.25 m diameter prototype pile. That makes it one of the most demanding checks available for a lateral pile analysis: the comparison is against measured physical behaviour, not against another calculation.

The results of this comparison were presented by Prof. V.S. Chandrasekaran in the T.S. Nagaraj Memorial Lecture, organised by the Indian Geotechnical Society Bengaluru chapter in November 2020.

Test set-up

Input data

Pile
	Centrifuge model	Equivalent real life pile dimensions
Acceleration level	50 g	–
Pile diameter	25 mm	1.25 m
Embedded length	180 mm	9.0 m

Soil
Property	Value
Soil type	Dry sand
Effective size	150 µm
Uniformity coefficient	1.5
Grading	80% between 150–250 µm
Angle of internal friction, φ	32°
Dry unit weight	16 kN/m³

The prototype pile and soil were modelled in GEMSai, and the generated static API p-y curves are shown below.

Static API p-y curves for sand generated by GEMSai at depths from 0 to 17.7 m — Static p-y curves generated by GEMSai for the prototype pile.

Results

Chart of lateral load against pile head deflection: GEMSai API p-y prediction and centrifuge test measurements from 0 to 1750 kN — Pile head deflection: centrifuge data (blue) vs API computed with the GEMSai foundation analysis suite (orange), prototype scale.

Bending moment against depth for lateral loads of 287 kN and 515 kN: GEMSai and centrifuge test — Bending moment with depth for lateral loads of 287 kN and 515 kN: centrifuge data (blue) vs API computed with the GEMSai foundation analysis suite (orange).

Note: Based on the two graphs above, the GEMSai pile foundation software results are in line with the experimental centrifuge test results.

Measured horizontal load against deflection and rotation at the load point for centrifuge Test 21, including unloading — Centrifuge model – single pile deflection and rotation at the load point, Test 21 (prototype scale). From Kulkarni, Chandrasekaran & King (1985) [1].

References

Kulkarni, K.R., Chandrasekaran, V.S. and King, G.J.W. (1985). “Centrifugal model studies on laterally loaded pile groups in sand.” 11th International Conference on Soil Mechanics and Foundation Engineering, San Francisco (1985).
American Petroleum Institute. ANSI/API RP 2GEO, Geotechnical and Foundation Design Considerations, 2011.

Laterally Loaded Long Pile — GEMSai vs Reese Charts

Module: Comprehensive Pile Foundation — laterally loaded pile analysis Version: GEMSai Comprehensive Pile Foundation v1.50 Reference: Reese charts (n_h model)

This example checks the GEMSai laterally loaded pile analysis against calculations based on the Reese charts for a pile in soil whose modulus of subgrade reaction increases linearly with depth (the n_h model). The Reese charts and tables are based on the finite difference method, whereas GEMSai uses the finite element method. Both are highly accurate numerical methods, so for the same idealised soil the two should agree closely — and they do.

Input data

Parameter	Value
Pile diameter, D	0.4 m
Pile length, L	24.19 m
Elastic modulus, E	3.3 × 10⁷ kN/m²
Linear variable subgrade modulus, n_h	500 kN/m³
Lateral load at pile head, P_t	25.0 kN
Head condition	Free head

Hand calculation using the Reese chart coefficients

EI = E πD⁴/64 = 41,469 kNm²

T = (EI / n_h)^1/5 = 2.420 m

L/T = 10 → long pile coefficients apply

A_y = 2.435, A_m = 0.772

Pile head deflection:

y_t = A_y P_t T³ / EI = 20.8 mm

Maximum bending moment:

BM_max = A_m P_t T = 46.70 kNm

at depth 1.4T = 3.387 m

Results

Pile head deflection

Reese charts: 20.8 mm

GEMSai: 20.6 mm

Maximum bending moment (at 3.387 m)

Reese charts: 46.70 kNm

GEMSai: 46.19 kNm

Lateral deflection against depth: GEMSai finite element result and Reese chart calculation, practically coincident — Lateral deflection with depth.

Bending moment against depth: GEMSai finite element result and Reese chart calculation, practically coincident — Bending moment with depth (free-head pile).

Comparison at matching depths
Depth (m)	Lateral deflection		Bending moment
Depth (m)	y Reese (mm)	y GEMSai (mm)	BM Reese (kNm)	BM GEMSai (kNm)
0.00	20.80	20.58	0.00	0.00
0.48	18.04	17.85	11.98	11.82
0.97	15.34	15.18	22.93	22.65
1.45	12.78	12.64	32.18	31.79
1.94	10.38	10.28	39.26	38.80
2.42	8.22	8.14	43.98	43.49
2.90	6.30	6.24	46.40	45.90
3.39	4.65	4.60	46.70	46.19
3.87	3.25	3.22	45.13	44.66
4.36	2.11	2.09	42.10	41.67
4.84	1.21	1.20	37.99	37.62
7.26	−0.64	−0.64	13.61	13.46
9.68	−0.43	−0.43	0.00	−0.02
12.10	−0.08	−0.08	−2.00	−2.01
24.20	0.00	0.00	0.00	0.00

Full Reese chart calculation table

Depth z (m)	Lateral deflection		Bending moment
Depth z (m)	A_y	y (mm)	A_m	BM (kNm)
0.00	2.435	20.80	0	0.00
0.24	2.273	19.41	0.100	6.05
0.48	2.112	18.04	0.198	11.98
0.73	1.952	16.67	0.291	17.60
0.97	1.796	15.34	0.379	22.93
1.21	1.644	14.04	0.459	27.77
1.45	1.496	12.78	0.532	32.18
1.69	1.353	11.55	0.595	35.99
1.94	1.216	10.38	0.649	39.26
2.18	1.086	9.27	0.693	41.92
2.42	0.962	8.22	0.727	43.98
2.90	0.738	6.30	0.767	46.40
3.39	0.544	4.65	0.772	46.70
3.87	0.381	3.25	0.746	45.13
4.36	0.247	2.11	0.696	42.10
4.84	0.142	1.21	0.628	37.99
6.05	−0.020	−0.17	0.422	25.53
7.26	−0.075	−0.64	0.225	13.61
8.47	−0.074	−0.63	0.081	4.90
9.68	−0.050	−0.43	0	0.00
10.89	−0.026	−0.22	−0.032	−1.94
12.10	−0.009	−0.08	−0.033	−2.00
24.20	0	0.00	0	0.00

Full GEMSai output table

Depth (m)	Deflection (mm)	Bending moment (kNm)
0	20.579	0
0.484	17.847	11.82
0.968	15.18	22.653
1.452	12.641	31.79
1.936	10.281	38.798
2.42	8.138	43.494
2.904	6.24	45.897
3.387	4.6	46.188
3.87	3.22	44.66
4.36	2.092	41.674
4.84	1.198	37.622
5.32	0.517	32.888
5.81	0.021	27.828
6.29	−0.318	22.748
6.78	−0.528	17.898
7.26	−0.637	13.461
7.74	−0.669	9.56
8.23	−0.648	6.263
8.71	−0.59	3.586
9.20	−0.512	1.509
9.68	−0.426	−0.018
10.16	−0.339	−1.064
10.65	−0.258	−1.706
11.13	−0.187	−2.026
11.61	−0.126	−2.103
12.10	−0.078	−2.007
12.58	−0.041	−1.799
13.07	−0.014	−1.531
13.55	0.004	−1.24
14.03	0.016	−0.955
14.52	0.022	−0.696
15.002	0.024	−0.472
15.486	0.023	−0.29
15.97	0.02	−0.149
16.454	0.017	−0.046
16.937	0.014	0.024
17.421	0.01	0.067
17.905	0.007	0.088
18.389	0.005	0.095
18.873	0.003	0.09
19.357	0.001	0.08
19.841	0	0.066
20.325	0	0.051
20.809	−0.001	0.038
21.293	−0.001	0.026
21.777	−0.001	0.017
22.261	−0.001	0.01
22.745	−0.001	0.005
23.229	−0.001	0.002
23.712	0	0
24.196	0	0

Discussion

The results from GEMSai compare extremely well with the calculations based on the Reese charts. Pile head deflection and maximum bending moment agree to within about 1%, and the deflection and bending moment curves down the pile are practically indistinguishable from one another. The maximum moment occurs at the same depth, 1.4T = 3.387 m, in both. This confirms the reliability of the GEMSai lateral pile analysis — and unlike a chart solution, GEMSai is not limited to an idealised linearly increasing soil modulus, so the same analysis extends directly to real, layered soil profiles.

Try It on Your Own Project

Run these examples yourself, or your own pile designs, with a free GEMSai trial.

Request a Free Trial

Geotechnical Engineering Modelling Software ai

Verification & Validation

Laterally Loaded Pile in Sand — GEMSai vs Centrifuge Model Test

Test set-up

Input data

Results

References

Laterally Loaded Long Pile — GEMSai vs Reese Charts

Input data

Results

Discussion

Try It on Your Own Project