High Fidelity Modeling of Building Collapse with Realistic Visualization of Results
ASI was accepted for Phase I of the DTRA solicitation number 082-005, titled “High Fidelity Modeling of Building Collapse with Realistic Visualization of Resulting Damage and Debris.” The purpose of Phase I was to validate the Applied Element Method (AEM) as it’s implemented in ASI’s Extreme Loading® for Structures Software (ELS) for blast analysis and progressive collapse. The text below summarizes the project or you can download the full report here.
The project started with collecting available information needed to show the theoretical background and developments of AEM from the time it originated in 1995 at Tokyo University through the report date in 2009. This was followed by gathering internal numerical tests that were performed with AEM for different materials including concrete, steel, bricks, pre-stressed concrete, etc., under both static and dynamic loading conditions.Since AEM was relatively new compared to the Finite Element Method (FEM), it was important for ASI to run the sample cases for this study as blind numerical tests in order to gain further acceptance of those who have been using FEM for decades. To accomplish this, a DTRA contractor was tasked to release only the information needed in order for ASI to perform the numerical analysis while retaining all output results until after the AEM analysis was completed. By following this approach, ASI avoided speculation that modeling parameters were adjusted to better match the FEM or test results. During the project ASI performed 22 blind numerical tests including 9 scenarios of walls under blast loading, 2 scenarios of columns under blast loading, and 11 case studies of a 5-story structure under progressive collapse.
Walls under Blast Loading:
The first 9 blind test cases included the evaluation of walls under blast loading. Each scenario had different boundary conditions, thicknesses and distances to the blast source. No physical test data was available and AEM was compared to FEM (LS-Dyna® software). All the 9 cases showed excellent agreement with FEM in terms of load-displacement and failure pattern.
Columns under Blast Loading:
Next, ASI conducted two blind tests with reinforced concrete (RC) columns referred in the report as Sample 1 and Sample 2. In these cases, the column in Sample 1 was severely damaged, but did not fail due to a blast, while the column in Sample 2 was damaged from a structural point of view. The blind numerical tests showed that AEM obtained reliable results. The general conclusion was that AEM’s failure shape, displacement mode, and failure modes all agreed with the physical test findings.
5-Story Structure Evaluated for Progressive Collapse:
Finally, ASI was tasked to perform progressive collapse analysis for 11 case studies that was conducted on a 5-story reinforced concrete building composed of columns and flat slabs with varying reinforcement details. Three columns were removed, one at a time, from the building perimeter. The results of the simulations were compared to a proprietary FEM software program.
In cases where no collapse occurred, AEM obtained similar results to FEM, confirming the conclusion that both methods have the same accuracy for highly nonlinear behavior when no collapse occurs.
The study also showed that the AEM results were closer to engineering judgment than the
FEM results since all failures were local and occurred around the removed columns.
It further indicated that the problems of suspected hour glassing of elements that occurs with the FEM for large deformations range did not occur with AEM simulations.
It was found throughout the study that AEM can be used as an effective engineering tool to provide high fidelity modeling of building collapse with a realistic visualization of the resulting damage and debris. Modeling and analysis time took approximately 10-12 hours per scenario using AEM making it more time efficient than FEM, which can take much longer depending on modeling complexity. AEM provides a practical engineering tool for modeling and analyzing such complicated phenomena found in extreme loading events.
Download an read the full report here.