Case Study

2D or 3D LE? That is the Question - Yeager Airport Case Study


When is 2D slope stability analysis sufficient and when should 3D analysis be used in conjunction with 2D analysis?

In February’s RocNews, we highlighted the case study of a 72-meter reinforced soil structure (RSS) which failed at the end of runway 5 of the Yeager Airport near Charleston, West Virginia in 2015. That article focused on a paper by Vandenberge et al, where 2D limit equilibrium (LE) analysis in Slide2 was used to help identify the cause of the slope failure.

Figure 1: Yeager Airport RSS during construction with black primary geogrid and green geosynthetic wrap (Left). The same RSS after failure (right).
Figure 1: Yeager Airport RSS during construction (left) and after failure (right).

Combining 2D with 3D LE analysis allows for a more detailed look into your model and can account for a variety of additional factors that could potentially generate different factors of safety (FS). This paper by James Collin, Timothy Stark, Augusto Lucarelli, Thomas Taylor, and Ryan Berg looked at the same Yeager Airport failure and conducted 3D LE analysis in Slide3 to compare with results from 2D analysis in Slide2.

Runway 5 Reinforced Soil Slope

To comply with airport regulations, Runway 5 of the Yeager Airport needed to be extended in length by 150 m. Due to the mountainous terrain, a RSS needed to be constructed. Two different uniaxial geogrids were used for the primary reinforcement zone and a lightweight geogrid face wrap was used to support the surface. Both types of geogrid can be seen in Figure 1 (left).

Figure 2 shows the cross section of the original design for the RSS (left) and the cross section as it was built (right). Field changes were made to the design and the RSS was placed on approximately 3m of unreinforced fill as opposed to rock. Also, the length of the geogrid at the bottom of the RSS was shortened from 53.4m to a minimum length of 20.7m.

Figure 2: 2D Cross Section of RSS. The original design (left) and the as-built design (right).
Figure 2: Cross Section of RSS. Original Design (left). As-built (right).

2D & 3D Limit Equilibrium Analyses

With unreinforced slope scenarios, the difference between 2D and 3D FS results are due to differences in slope geometry, side forces and material properties. However, in cases which use reinforcement, there can be even greater differences between the 2D and 3D FS results.

What makes the Yeager Airport failure so important to study is that it was one of the largest mechanically stabilized earth (MSE) slopes and was extremely complex. Because of its size, combination of soil and rock materials, nonlinear material behaviour and the numerous layers of geogrids, accurately modeling this RSS was a huge task achieved using Slide3.

In this case study, the 3D analysis took into account:

  • 3D driving forces
  • 3D slope and groundwater conditions
  • Overlapping geogrids
  • Anisotropic geogrid tensile resistance
  • Different types of uniaxial geogrids
  • Construction damage to the geogrids

2D LE and 3D LE analysis was conducted for 5 slope scenarios (Table 1)

  1. Initial Design
  2. Revised design
  3. End of construction peak soil strength
  4. End of construction fully softened strength
  5. Failure
Table 1: Five scenarios analyzed in 2D and 3D LE Analyses
Table 1: Five scenarios analyzed in 2D and 3D LE Analyses

L1 corresponds to the original design in the contract plans and L2 represents the as-built case. These scenarios are represented in Table 1.

The post-failure investigation obtained samples from the primary geogrids and bearing soil at the toe of the slope. Samples from the soil-rock interface failure surface were tested and the fully softened and residual strengths were used in the stability analysis. S1, S2 and S3 represent the soil-rock interface properties and their values are shown in Table 2.

Table 2: Engineering Parameters used in Stability Analyses. Moist Unit weight 21.2. Effective stress Friction 36 degrees. Effective Stress Cohesion 0.
Table 2: Engineering Parameters used in Stability Analyses

Measured and estimated geogrid material properties were used for stability analysis represented as G1, G2, and G3.

  • G1- measured initial design minimum long-term design strength (LTDS)
  • G2- designed geogrids accounting for installation damage
  • G3- designed geogrids accounting for installation damage and creep

Figure 3 (left) shows the complexity of the geogrid layout at an elevation of 281.5 m of the RSS. 3D analysis was able to model the actual layout of each geogrid strip based on information from the construction sketches. A MATLAB version 9.7 script was written to compute the coordinates of every geogrid strip to build out the 3D model in Slide3, shown in Figure 3 (right).

Figure 3: Geogrid layout diagram created during construction at elevation 281.5m (left). Slide3 output for slope at time of failure with the high groundwater condition and anisotropic geogrid tensile strength (right).
Figure 3: Geogrid layout diagram created during construction at elevation 281.5m (left). Slide3 output for slope at time of failure with the high groundwater condition and anisotropic geogrid tensile strength (right).

The results of the 2D and 3D LE analyses are listed in the Table 3. 2D analysis used isotropic conditions whereas each 3D slope scenario was analyzed with the tensile resistance of the geogrids modeled as isotropic and anisotropic to assess the difference in tensile resistance between the MD and XMD.

Table 3: 2D & 3D factors of safety for analysis cases 1-5
Table 3: 2D & 3D factors of safety for analysis cases 1-5

This paper investigates the Yeager Airport runway failure in more depth by using Flac3D to look at the stresses and deformations in the RSS prior to failure to better understand the kinematic forces that were at play.

Conclusion

This paper highlights the combined application of 2D and 3D slope stability analysis which verified the failure mechanism identified in the post event forensic subsurface investigation. It also highlights the importance of accurately modelling the factors that contributed to the failure. In this case study, the 3D stability analysis conducted was in excellent agreement with the field observations and conditions at the time of the reinforced slope failure.

This brings us back to the question: When is 2D analysis sufficient and when should 2D and 3D analysis be used together?

We encourage you to read the full paper to see the full details.

If you use 2D programs for slope stability analysis, start a free trial of Slide3 today and see how 3D LE analysis can benefit you.

Back to top

More from Rocscience