Failure of a Test Embankment on Sensitive Champlain Clay
- 1 day ago
- 8 min read
Revisited with DeepEX software
1- Introduction
The failure of a test embankment constructed on sensitive Champlain clay at Saint-Alban, Québec, remains one of the most instructive case histories in geotechnical engineering. The experiment was conducted to investigate the behaviour of embankments on soft clay foundations and was extensively documented by La Rochelle et al. (1974). The study included detailed monitoring through vane tests, settlement measurements, pore pressure instrumentation, and field observations of the failure mechanism. A decade later, Talesnick and Baker (1984) revisited the case and compared calculated slip surfaces with the observed failure path using optimization-based slope stability methods. Their work demonstrated that advanced search procedures can reproduce both the failure surface and a safety factor close to unity, as expected at failure.
In this article, the Saint-Alban embankment failure is revisited using DeepEX, considering only the stability of the embankment. The case is analysed using both Limit Equilibrium Method (LEM) and Finite Element Method (FEM) approaches within the same modelling environment. The objective is to reproduce the failure mechanism, compare computed safety factors with those reported in the literature, and evaluate how well modern numerical tools capture this classic geotechnical case history.
1.1- Description of the Failure
The Saint-Alban test embankment was constructed in stages on a deposit of sensitive Champlain clay until failure occurred. The embankment reached a height of approximately 4 m when instability developed. The failure occurred rapidly once the critical height was exceeded. Observations indicate that the main movement took place within about one minute, followed by slower deformation and settlement. The crest of the embankment experienced significant settlement, while heaving occurred near the toe. Field instrumentation and observations allowed the researchers to determine the position of the failure surface. The slip surface developed primarily within the soft clay foundation and followed a rotational shape close to a circular arc.
Figure 1 – Observed failure mechanism and slip surface (Adapted from Talesnick and Baker, 1984).
The availability of detailed field measurements makes this case particularly valuable for evaluating numerical stability analyses. Both the location of the failure surface and the soil strength profile were well constrained by site investigation data.
2- Site Conditions and Geology
2.1- Geological Setting
The test site is located within the St. Lawrence lowlands of Québec, where thick deposits of marine clay were formed following the retreat of the Champlain Sea. These deposits are known for their high sensitivity, high water content, and relatively low undrained shear strength, making them susceptible to instability and progressive failure mechanisms. The soil profile at the site consists primarily of soft marine clay layers overlying deeper granular deposits.
Figure 2 – Soil stratigraphy and foundation conditions (Adapted from Talesnick and Baker, 1984).
2.2- Foundation Soil Properties
The foundation soil profile consists of multiple clay layers with varying undrained shear strength. The main characteristics include:
· A thin surface crust
· Several layers of soft to medium sensitive clay
· Increasing undrained shear strength with depth
· A relatively stiff layer at greater depth
Undrained shear strengths measured through field testing ranged approximately between 11 kPa and 30 kPa in the clay layers.
For the present analysis, the soil parameters were implemented in DeepEX using undrained conditions for the clay layers. Table 1 summarises the soil properties used in the numerical model.
Table 1 – Soil parameters used in DeepEX modelling
Soil | γ (kN/m³) | c′ (kPa) | Su (kPa) | φ′ (deg) | OCR | Elastic parameter |
Fill | 18 | 0 | – | 44 | 1 | E = 25000, ν = 0.45 |
L1 (undrained clay) | 18 | – | 30 | 0 | 1 | E = 20000 |
L2 (undrained clay) | 16.5 | – | 11.5 | 0 | 1 | E = 2874 |
L3 (undrained clay) | 17 | – | 18 | 0 | 1 | E = 2874 |
L4 (undrained clay) | 17 | – | 24 | 0 | 1 | E = 2874 |
L5 (undrained clay) | 18 | – | 28 | 0 | 1 | E = 20000 |
The clay layers were modelled using undrained shear strength (Su) parameters, consistent with short-term stability conditions during embankment construction.
3- Embankment Geometry and Properties
The embankment was constructed using granular fill material with relatively high shear strength.
Typical properties of the fill include:
· Unit weight: approximately 18 kN/m³
· Friction angle: approximately 44°
· Negligible cohesion
The geometry of the embankment reproduced in the DeepEX model corresponds to the dimensions reported in the original studies.
Figure 3 – DeepEX numerical model geometry and soil layering
The model includes the embankment, the layered clay foundation, and the groundwater conditions consistent with the field observations.
4- Numerical Modelling with DeepEX
To revisit the case history, both LEM and FEM analyses were performed in DeepEX.
The main objective of the modelling was to evaluate whether the computed failure mechanism and safety factors are consistent with those observed in the field and reported in the literature.
4.1- Limit Equilibrium Analysis (LEM)
The LEM analysis was performed using the Bishop simplified method, which assumes circular slip surfaces and satisfies moment equilibrium.
The analysis procedure included:
· Definition of the soil stratigraphy
· Assignment of undrained shear strength parameters
· Generation of potential slip surfaces
· Search for the critical failure surface
The analysis identified a critical circular slip surface passing through the soft clay layers beneath the embankment.
Figure 4 – LEM slip surface, slices and Global FS contours (Bishop method), failure mechanism.
The calculated factor of safety was approximately 1.13, indicating instability and consistent with the observed failure condition.
4.2- Finite Element Analysis (FEM)
In addition to LEM analysis, the problem was analysed using FEM within DeepEX.
The FEM analysis employed the Shear Strength Reduction (SSR) technique to determine the factor of safety.
The modelling process involved:
· Generation of the finite element mesh
· Definition of elastic-plastic soil behaviour
· Gradual reduction of shear strength parameters
· Identification of the failure mechanism
The FEM results provide displacement contours that highlight the development of the failure mechanism within the clay foundation.
Figure 5 – FEM displacement contours, LEM+FEM total displacements and failure mechanism.
The FEM displacement pattern clearly indicates a rotational failure mechanism similar to the observed field behaviour.
5- Results and Discussion
5.1- Comparison of Safety Factors
The safety factors obtained from the DeepEX analyses are consistent with those reported in the literature.
For the Saint-Alban embankment:
· LEM (DeepEX): FS ≈ 1.13
· FEM (DeepEX): FS ≈ 1.14 (strength reduction factor)
· La Rochelle et al. (1974): FS ≈ 1.20
· Talesnick and Baker (1984): FS ≈ 1.03
The results show that the computed factors of safety fall within the same range as those obtained by previous investigators. On the other hand, these values are all close to unity, as expected for a case at failure.
Differences between the values are expected due to variations in modelling assumptions and search procedures used to identify the critical slip surface.
5.2- Comparison of Failure Mechanisms
A comparison between the numerical predictions and field observations shows strong agreement.
Key observations include:
· The LEM critical slip surface closely matches the circular failure arc observed in the field.The FEM displacement contours identify a zone of large deformation that aligns with the same failure mechanism.
· The predicted slip surface passes through the weaker clay layers, consistent with the measured shear strength profile.
The FEM results also highlight the progressive development of deformation within the clay mass prior to failure.
5.3- Comparison of LEM and FEM Results
To better compare the two analysis approaches, Table 2 summarises the main outcomes from the DeepEX modelling and relates them to the observations reported in the original studies. The comparison shows that both methods capture the same overall instability mechanism, even though they express failure in different ways. The LEM analysis identifies a critical slip surface and factor of safety directly, while the FEM analysis highlights the deformation pattern and the zone of progressive failure through displacement contours and strength reduction. The overall agreement with the field behaviour is good, and the calculated safety factors remain within the same order as those reported by La Rochelle et al. (1974) and Talesnick and Baker (1984).
Table 2 – Comparison of LEM and FEM results for the Saint-Alban embankment case
Aspect | DeepEX LEM (Bishop) | DeepEX FEM (SSR) |
Analysis type | Limit equilibrium | Finite element strength reduction |
Soil strength approach | Undrained shear strength profile | Undrained layered soil model with elastic parameters |
Failure indicator | Critical slip surface and FS | Displacement localization and strength reduction factor |
Computed safety factor | 1.13 | 1.14 |
Failure mechanism | Rotational, circular slip surface | Rotational shear zone shown by displacement contours |
Agreement with field mechanism | Good | Good |
Comparison with La Rochelle et al. (1974) | Close to observed failure condition; same mechanism | Same order as reported factor of safety |
Comparison with Talesnick and Baker (1984) | Similar magnitude and similar failure geometry | Similar order of safety factor; confirms instability trend |
Both methods capture the same overall instability mechanism, although they express it differently: LEM through a discrete slip surface and FEM through a continuous deformation field.
5.4- Circular versus Non-Circular Slip Surfaces
One of the interesting points raised by Talesnick and Baker (1984) is the distinction between analyses that restrict failure to a circular arc and procedures that allow more general slip surface shapes. They noted that in many slope stability problems, limiting the search to circular surfaces may overlook the actual critical mechanism, especially when the failure path is influenced by layering or strength contrasts. In the Saint-Alban case, however, their optimization-based analysis still produced a slip surface that was approximately circular and located close to the observed one.
This is consistent with the DeepEX results. The Bishop LEM analysis reproduces the observed mechanism well, while the FEM results confirm a rotational failure pattern. This indicates that, for this case, a circular failure assumption is appropriate and sufficient.
5.6- Graphical Overlay of Observed and Computed Failure Surfaces
To make the comparison clearer, the article can include a single overlay figure showing the observed field failure surface together with the DeepEX LEM slip surface and the FEM deformation pattern. This type of figure is especially useful because it shows, at a glance, that the numerical analyses are not only producing comparable safety factors, but also reproducing the same overall failure geometry described in the original case history.
Figure 6 – Overlay of observed and computed failure mechanisms
Observed failure surface from field interpretation (after La Rochelle et al., 1974, and Talesnick and Baker, 1984), compared with the DeepEX LEM critical slip surface and the FEM zone of maximum total displacement. The three results show good agreement in both depth and overall rotational shape, confirming that the numerical model reproduces the embankment failure mechanism with reasonable accuracy.
6- Conclusions
The Saint-Alban embankment failure remains a benchmark case for evaluating slope stability methods in soft clay.
The DeepEX analysis leads to the following conclusions:
· Both LEM and FEM reproduce the observed failure mechanism with good accuracy
· The computed safety factors are consistent with published results
· The failure mechanism is well represented by a circular slip surface
· FEM analysis provides additional insight into deformation and failure development
· Combining LEM and FEM offers a comprehensive understanding of the problem
Overall, the study demonstrates that modern numerical tools can successfully reproduce classical geotechnical failures and provide valuable insight into the behaviour of embankments on soft soils.
References
La Rochelle, P., Trak, B., Tavenas, F., & Roy, M. (1974). Failure of a test embankment on a sensitive Champlain clay deposit. Canadian Geotechnical Journal.
Talesnick, M., & Baker, R. (1984). Comparison of observed and calculated slip surface in slope stability calculations. Canadian Geotechnical Journal.
Trusted By
Still not persuaded? Let us show you how to reduce your design time by up to 90%!
