Identifying the right numerical modeling approach

By Georges Kassab | May 10, 2023

Numerical modeling is the process of representing a complex physical system as a mathematical model using numerical tools in order to study system performance under various real-world scenarios.

This process is routinely used in geotechnical analysis to predict and understand the behavior of dams, foundations, tunnels, landslides and more. This typically involves simulating the behavior of soil and rock in response to different loads and boundary conditions. As the number of available modeling methods and technologies continue to grow, so does the importance of identifying the appropriate approach to ensure project success and sustainability.

Numerical models

A variety of numerical approaches can be used to analyze the behavior of soil types and structures, such as retaining walls, embankments, and bridges. Key considerations when creating numerical models of geosystems include identifying behaviors of interest, building a robust data model, selecting and calibrating a constitutive model, and determining a procedure for interpretating results.

Applications of numerical modeling in geotechnical engineering continue to demonstrate benefits by improving project designs and reducing risks. While numerical simulation can be extremely helpful in predicting the behavior of systems and structures, selecting and examining the wrong parameters can lead to unsuitable system representation and interpretation. Adoption of applicable assumptions regarding soil properties and/or appropriate modeling of groundwater flow can ensure results adequately represent stability and settlement of structures.

One famous example taught in engineering academia is the Tacoma Bridge collapse in 1940 which was due in part to the inadequate modeling of wind loads. With today’s advancements in modeling approaches and technologies, threats of such collapses can be significantly mitigated or eliminated.

Different approaches

With the variety of tools currently available for geotechnical engineers to assess ground and structure behaviors, it has become increasingly necessary to consider the strengths and weaknesses of the various approaches that can be used to develop models and analyze their results.

A variety of approaches can be taken for system representation, each with their own benefits and caveats:

  • Finite element analysis (FEA) is a useful approach for complex geometries and non-linear materials but can be computationally intensive and require a high level of expertise.
  • The finite difference method (FDM) is recognized as being simpler and more expedient than FEA but is limited to regular geometries and linear materials.
  • The boundary element method (BEM) is efficient in addressing problems with large surface areas but can be less accurate than FEA or FDM.
  • The discrete element method (DEM) is a useful approach for modeling the behavior of granular materials. However, it can also be computationally intensive and requires specialized software.

Common mistakes in geotechnical engineering numerical modeling include neglecting to account for soil variability, underestimating the importance of boundary conditions, and using inadequate or incorrect constitutive models to analyze soil conditions.

Results can be qualified and checked by carefully considering the assumptions and limitations of a chosen model, validating the model with field data, and utilizing the appropriate software and modeling techniques.

Why numerical modeling is vital

Several geotechnical engineering projects may have seen more positive outcomes had effective numerical models been considered. In the aftermath of the collapse of the Mount Polley Mine tailings dam in British Columbia in 2014, an independent panel concluded that the dam’s breach resulted from a failure in the embankment foundation and that its design did not adequately take the complexity of the foundation into account. Following its forensic investigation, the panel recommended the adoption of best applicable practices (BAP) such as advanced geotechnical analysis tools, including numerical modeling, to evaluate tailings dams and their foundations so that similar incidents can be prevented in the future.

Numerical modeling has also been attributed to the success of many well-known projects. The Confederation Bridge, a 12.9-km-long structure connecting Prince Edward Island to New Brunswick, Canada, is revered as an engineering marvel. Numerical methods were used extensively during the bridge's design and construction stages to simulate soil behavior and to measure the structure’s stability under a vast variety of loading and environmental conditions. Modeling allowed the team to optimize the design of bridge piers and abutments and to evaluate the structure’s stability under both common and extreme traffic scenarios. Numerical modeling was highly instrumental in ensuring the bridge’s ongoing safety and reliability. Confederation Bridge has been in continuous operation since opening to traffic in 1997.

Over the next five years, advances in geotechnical engineering modeling are likely to include improvements in software and hardware capabilities, the integration of data from remote sensing and other sources, and the incorporation of artificial intelligence and machine learning algorithms to improve model accuracy and efficiency. Additionally, a growing emphasis on sustainability and resiliency in geotechnical engineering projects will undoubtedly require new modeling approaches and techniques.

Hatch’s Dynamic Earth Solutions (DES) practice continuously refines its numerical modeling expertise and acumen by acquiring knowledge of the most current modeling approaches and applications. Contact us to find out how we can help you.