Integrating 2D and 3D finite element analysis allows geotechnical engineers to better predict ground behaviour and optimize support systems while saving considerable time. RS2 and RS3 provide powerful, complementary tools for tackling civil and mining geotechnical engineering projects at all difficulty scales – from simple, everyday problems to the most challenging projects. They seamlessly bridge between the dimensions.

Traditionally, geotechnical engineers analyze problems in 2D to obtain starting insights and transition to more complex 3D modelling when necessary. This article argues for a shift in this workflow paradigm, more often starting with 3D and using 2D for some focused or critical areas.

Why 3D Analysis Should Precede 2D in Modern Geotechnical Engineering

Similar to other disciplines, modelling in geotechnical engineering has evolved from simplified two-dimensional analyses to sophisticated three-dimensional simulations. This path follows the historical development of computational power and methodological advancements. It emerged not necessarily because it was optimal but because of historical limitations in computational power and software capabilities.

Now, the geotechnical profession is equipped with advanced, user-friendly and computationally efficient 3D modelling software like RS3, which allows engineers to evaluate the mechanical and hydrogeological response (stress distributions, deformation patterns, structural stability, and groundwater flow) of both surface and underground excavations more realistically. These tools can simulate complex excavation geometries, ground conditions and pore pressure regimes.

This evolution in capability calls for us to reconsider our modelling approach despite history.

The Case for Starting with 3D Analysis

The fundamental argument for starting with 3D analysis is simple: real-world geotechnical problems are inherently three-dimensional. Excavations, slopes, tunnels, and foundations all exist in 3D space and interact with geological conditions that vary in all directions. A 3D analysis captures these spatial relationships more accurately than any 2D approximation can.

3D analysis excels at capturing geometric nuances that 2D models inevitably simplify. Current 3D finite element programs can better represent complex soil and rock anisotropic behaviour. These differences are not merely academic—they can significantly impact design decisions and safety assessments. 2D methods can both overestimate or underestimate the stability of excavations, support, or reinforcement requirements. Starting with a 3D analysis allows engineers to identify such discrepancies early in the design process and helps avoid costly remediation or overly conservative designs.

RS3 offers an RS2 Section Creator feature that allows engineers to readily capture 2D cross-sections from RS3 models and generate RS2 files, which enables engineers to refine or optimize critical sections identified in their 3D models at lower computational costs.

The Role of 2D Analysis

We are not suggesting that geotechnical engineers abandon 2D analysis by advocating for a 3D-first modelling paradigm. Instead, we are calling for 2D modelling to be repositioned as a powerful tool for sensitivity analysis and rapid iteration after establishing a reliable 3D baseline.

One of the most valuable aspects of performing 3D analysis first is that it provides a reference point (baseline) against which to evaluate 2D results. By comparing factors of safety or predicted deformations between 3D and 2D models, engineers can quantify the "dimensional error" introduced by 2D simplifications for their specific project.

Design optimization typically requires numerous iterations to explore different configurations, support systems, or construction sequences. Once the relationship between 3D and 2D results is understood, 2D analysis becomes an efficient tool for this design optimization and sensitivity studies. After establishing the 3D-2D relationship, engineers can use calibrated 2D models for these iterations.

Practical Implementation of the RS3-RS2 Workflow

We will next outline steps that you can apply to implement the proposed 3D-to-2D workflow using RS3 and RS2.

Step 1: Develop 3D RS3 Model

Begin your analysis by developing a detailed 3D model in RS3. This model must incorporate the problem's actual 3D geometry, material properties, groundwater conditions, and boundary conditions. As part of this 3D model building, you can define excavation and support sequences.

Step 2: Identify Critical Cross-sections

Next, compute your 3D model and analyze its results to identify critical sections, i.e., areas that merit their detailed investigation. Typically, these are regions of high stress, maximum displacement, lowest factor of safety, or sections with complex geological conditions.

Step 3: Extract Cross-sections, Compare 2D and 3D Predictions, and Adjust

Then, use RS3's "RS2 Section Creator" to extract 2D cross-sections through the critical locations. The RS2 Section Creator transfers geometries, material and support properties, water tables, and loads from the 3D model directly to RS2.

After running the 2D models, you must compare their results with the corresponding sections in the 3D model to establish how well the 2D results match or depart from the 3D outcomes. This exercise will help you to account for 3D effects that may not have been captured in the 2D models.

Step 4: Perform Iterative Changes and Sensitivity Analysis in RS2

Once you develop 'calibrated' 2D models, you can conduct quick sensitivity analyses and design optimizations in RS2. You can more easily explore different support configurations, material properties, or excavation sequences that work for your design without the computational burdens of 3D analyses.

Step 5: Verify Final Designs in RS3

You can verify the final solutions in the 3D model to ensure that your new designs perform as expected after being optimized or refined through 2D analysis in RS2.

Challenges in Implementing the 3D-First Paradigm

3D geometry building difficulties and increased computational demands can be challenging at the initial stages of implementing the 3D-first approach. However, today's reasonably priced computing resources and software advances have significantly reduced this barrier. RS3 provides both direct 3D geometry modelling tools and simplified modelling with block or voxel models to facilitate the 3D-first approach.

Direct 3D Geometry Modelling

RS3 explicitly represents the boundaries of geological domains and excavations, which allows for precise modelling of complex geological features and smooth interfaces. This precision typically leads to faster numerical convergence and higher fidelity in the analysis results. RS3 also provides adaptive meshing techniques which you can employ to focus computational resources on areas of interest, such as stress concentration regions, discontinuities, or locations with high displacement (or shear) gradients. However, for complex problems, direct 3D geometry modelling comes with longer model setup times, mainly due to the need to clean up input geometries.

Block/Voxel Modelling

Block/voxel approaches offer modelling simplicity and rapid setup. However, they introduce limitations that can compromise geotechnical analysis quality. These methods create "jagged boundary defects" ("staircase artefacts") at geological interfaces, which can require substantially higher block density to achieve reasonable accuracy. This discretization effect can undermine the fidelity of models to actual geological conditions, particularly in scenarios where the exact geometry of interfaces is critical to stability assessment. One example is the presence of thin geological features and discontinuities that may be critical to stability but cannot be adequately captured at the resolution of the block model. Block/voxel models can also experience numerical instabilities, which may require specialized stabilization techniques.

Tradeoffs in Practical Implementation

This brief discussion underscores the importance of understanding the tradeoffs between model setup efficiency and model accuracy. Careful consideration of the technical limitations of the direct geometry and block/voxel approaches will help engineers ensure that their modelling decisions do not inadvertently compromise the quality and reliability of results.

Facilitating the 3D-First Modelling

Building 3D models and interpreting their results requires more expertise than 2D analysis. However, modern software offers increasingly user-friendly interfaces, which, in combination with training, can address this challenge. RS3, for example, provides you with tools to apply either direct geometry or block/voxel modelling methods and helps you weigh the respective technical and workflow advantages and limitations. These features help you carefully consider project requirements, available data, and the specific technical questions being addressed against workflow efficiency and results accuracy.

Although a 3D-first approach may initially increase project timelines and costs, it ultimately offers design optimization and risk reduction benefits that can outweigh the initial modelling investments.

Concluding Remarks

The traditional approach of starting geotechnical finite element analysis from 2D and progressing to 3D emerged from historical developments rather than from best practices. The relatively cheap computational capabilities available today and integrations between 2D and 3D software like RS3 and RS2 invite us to reverse this paradigm — starting with comprehensive 3D analysis to establish an accurate baseline and then leveraging the advantages of 2D tools to develop efficient designs.

The 3D-first modelling practice acknowledges that real-world geotechnical problems are inherently three-dimensional while recognizing the practical benefits of 2D analysis. By combining 2D and 3D analysis in the suggested workflow, engineers can achieve a good understanding of complex ground behaviour and develop safer, more efficient designs.

Starting with 3D and moving to 2D establishes a new paradigm that better serves the challenges of modern geotechnical engineering projects.

More from Rocscience