The objective of this session is to review best practices in checking, reviewing, calibrating, and verifying the results of a finite element model for concrete structures. This session will focus on proper interpretation of the FE analysis results and presentation of those results in terms commonly used by the engineering community. Attendees will learn techniques used to build their confidence in finite element models. The session is intended to provide an overview of some of the current strategies in the practice for sanity checks and reliability reviews used in structural assessment of concrete structures and foundations from conceptual engineering mechanics, benchmark studies, experiments, and theoretical values.
Sanity Checks and Self-Checking of Finite Element Models
Presented By: Paul Schuman
Affiliation: Simpson Gumpertz & Heger Inc
Description: A sanity check is a basic test used to quickly evaluate a finite element model to confirm whether the results represent realistic behavior, to ensure that the model is working as expected, and to confirm whether the model is providing results that make sense. A sanity check is not a detailed test or check of inputs or other modeling decisions, nor does it take the place of comprehensive quality control and quality assurance. Rather, a sanity check involves engineers evaluating results at important features and critical model locations and comparing them to their understanding of expected structural behavior to identify potential errors. It is good practice for all engineers to incorporate sanity checks throughout the modeling process to both save time and money, as well as to confirm the engineers’ understanding of the model. Performing a sanity check requires engineering judgment and a practical understanding of what they expect from the model in advance of the analyses. Not performing regular sanity checks on finite element models can result in significant inefficiencies in both the analysis and design process due to unnecessary re-work. More importantly, not performing sanity checks can lead to errors in the design and detailing of concrete structures. When to perform sanity checks, the goals of sanity checks, the benefits of sanity checks, the importance of sanity checking a model for concrete structures, and some basic types of sanity checks are addressed in this presentation. The presentation will conclude with examples of sanity checks used in real-world design and analysis applications.
Quality Processes for Bridge Analysis Models
Presented By: Serhan Guner
Affiliation: University of Toledo
Description: The design of bridges often involves the use of FEA models of varying degrees of complexity, including:
- 1D line elements
- 2D line elements (with/without shell elements)
- Linear-elastic finite elements
- Nonlinear finite elements
A variety of analysis software are used to analyze concrete structures. This process can be complex with a large number of input parameters and output data. The engineer must understand the limitations of the analysis method and software, possess experience in developing analysis models with proper approaches and assumptions, and correctly interpret the results. A simple check of the program input values is not an adequate way of ensuring the accuracy and validity of these models. Quality assurance (QA) and quality control (QC) are two essential processes for the quality management of analysis models. Verification and validation (V&V) play a critical role in the QA/QC process.
The objective of the current NCHRP Project 20-05, Topic 54-11, Quality Processes for Bridge Analysis Models, is to identify and document state department of transportation (DOT) practices related to the quality processes for bridge structural analysis models. The study investigated processes for:
- Identifying appropriately qualified staff
- Choosing an appropriate analysis method and software
- Verifying and validating the software
- Modeling with proper approaches and assumptions
- Verifying the analysis results, and
- Reconciling discrepancies between independent models.
Fifty state and District of Columbia DOTs participated in the study. In addition, five state DOTs were examined as Case Studies, including California, Colorado, Iowa, Louisiana, and New York DOTs. This presentation will present the findings with an emphasis on V&V processes.
Case Study Examples of Finite Element Model Sanity Checks Used in the Design and Analyses of Concrete Structures
Presented By: Nicholas Cramsey
Affiliation: Nicholas G. Cramsey
Description: A finite element model is only as effective as its modeler. Models can vary significantly in size, complexity, time spent on model building/generating, and the overall purpose for the evaluation. Additionally, modelers are responsible for many things throughout the modeling process: understanding and interpreting the overall and local model behavior, incorporating updates into the model, and tracking effects of those updates on behavior, performing design checks, evaluating loadings to be included, meeting project deadlines, …the list goes on. It may seem there are perils and distractions at every turn. How does a modeler stay sane throughout the modeling process and produce an effective model that can be relied upon? Sanity checks!
Sanity checks are basic tests a modeler performs regularly to understand if the model is producing reasonable results. There is no “one size fits all” verification or sanity check process that covers all types of analysis models, but there are recurring aspects to the checks, and much can be learned from past experiences. This presentation will present specific case study examples of sanity checks that were used on real-world projects. These sanity checks were a necessary part of the analysis process that steered the overall project approach, increased efficiencies, allowed the projects to stay on schedule, amongst many other benefits.
Best Practices for Verifying and Validating Complex Non-Linear Finite Element Models of Concrete Structures
Presented By: Robert MacNeill
Affiliation:
Description: Nonlinear analysis requires enhanced rigor to ensure accurate predictions of physical phenomena. Consistently applying established best practices informed by accumulated engineering analysts’ experience is critical. Best practices can be applied to modeling methodologies like the meshing approach, element choice, and constraint/load application techniques. Another vital consideration is the material modeling approach. Nonlinear concrete response can be complicated to model, especially on an application-specific basis. Choosing an appropriate constitutive model is a fundamental decision for the analyst. Populating the model with proper inputs is equally essential. Concrete properties can be highly variable. For example, two materials with similar unconfined compressive stress can behave differently under confined pressures.
For practical problems, the accuracy of predicted nonlinear concrete response is highly dependent on a wide range of model inputs. Typically, analysts are limited to unconfined compression test and sample weight data, which do not sufficiently characterize the material. We can make rule-of-thumb assumptions about tensile strength to refine the model further. Depending on the loading environment, we may need to better characterize the tensile strength or understand how the material responds under confined pressures, cyclic loading, or at high strain rates. Additional material tests may be necessary, depending on the model fidelity required for the problem.
This presentation will highlight some key considerations for applying best practices for nonlinear concrete modeling and identifying situations that may require extra testing to improve model response accuracy.