Air blast load generation for simulating structural response

Emily L. Guzas, Christopher EarlsSteel 2010 · DOI

TL;DR

This paper introduces a novel, transparent methodology for generating air blast loading for structural simulations by consolidating open-literature data into a user-friendly computer code. This approach aims to demystify blast load modeling, offering a more accessible and auditable alternative to proprietary methods, and demonstrating its utility through simulations of steel plate and girder responses.

What Problem Does This Paper Address?

In recent years, the threat of explosions to structural integrity has become a critical concern. However, accurately simulating the complex, transient loads imposed by air blasts is challenging. Existing methods for generating these blast loads often rely on proprietary software or data that isn't publicly accessible, hindering transparency and independent verification for engineers and researchers. This lack of transparency makes it difficult to understand the underlying assumptions and limitations of blast load models, which is crucial for reliable structural analysis, especially in scenarios involving conventional high explosives. The paper aims to bridge this gap by providing a detailed, open-source approach to air blast load generation.

Methods

The core of this research is the development of an air blast load generation computer code. This code synthesizes air blast parameters (peak overpressure, arrival time, duration, and decay constant) from various open-literature sources, primarily Kinney and Graham (1985) and Brode (1977). Unlike some existing methods that use graphical data, this code employs piecewise polynomial fits to derive these parameters as functions of scaled distance.

Key methodological contributions include:

  • Consolidation of Open-Literature Data: The paper compiles and presents a comprehensive set of equations for blast parameters, making them readily available for use.
  • Development of a Load Generation Code: This code translates the blast parameters into time-history nodal forces for finite element analysis.
  • Inclusion of Shielding Effects: A ray-tracing technique is implemented to account for how structural geometry can shield certain areas from direct blast impact, applying reflected loads to unshielded regions and side-on loads to partially shielded areas.
  • Angle of Incidence Consideration: The code incorporates the effect of the angle at which the blast wave strikes a surface, modifying the reflected overpressure and impulse accordingly.
  • Equivalent Triangular Pulse Approximation: The code also supports the generation of simplified triangular blast pulses, which are often used for design purposes, by preserving peak overpressure and impulse while modifying duration.

The developed code is designed as a preprocessor for commercial finite element software like LS-DYNA, allowing for detailed nonlinear dynamic analysis.

Datasets

The paper does not introduce new experimental datasets. Instead, it relies on and synthesizes data from existing, publicly available sources, including:

  • Kinney and Graham (1985): Used for developing piecewise polynomial fits for arrival time and decay constant, and for peak overpressure in free air.
  • Brode (1977): Used for peak reflected overpressure calculations.
  • Baker (1973, 1983): Referenced for general blast phenomenology and data on reflected and side-on air blast cases.
  • Kingery and Bulmash (1984): Used as a comparative benchmark, as their equations form the basis for widely used (but proprietary) blast load generation programs.
  • TM 5-1300 (US Army 1990): Referenced for design considerations and the concept of equivalent triangular pulses.

Key Insights and Ideas

  • Transparency is Paramount: The paper strongly advocates for the use of open-literature data to ensure that engineers and designers can understand and critically evaluate the assumptions behind blast load models, moving away from "black box" approaches.
  • Small Parameter Changes, Big Response Differences: The case studies highlight that even seemingly minor variations in blast loading parameters (e.g., using different sources or approximations like triangular pulses) can lead to significant differences in structural response, underscoring the sensitivity of structures to blast loads.
  • Shielding and Angle of Incidence Matter: The inclusion of shielding and angle of incidence effects in the load generation code is shown to be crucial for more realistic simulations, especially for complex geometries.
  • Direction of Deformation is Significant: The study observed that the direction of final plate deformation (towards or away from the blast) can be influenced by the loading model and has implications for structural stability.
  • Kingery-Bulmash vs. Open-Literature: The comparison between the Kingery-Bulmash (KB) parameters and the proposed open-literature parameters (Current) reveals notable differences in load duration and peak overpressure, leading to distinct structural responses.

Key Results

The research successfully developed and demonstrated an air blast load generation code that utilizes open-literature data.

  • Plate Example: Simulations of a steel plate under spherical explosion showed that different loading parameter sets (Current, Current-Tri, KB, KB-Tri) resulted in varying transverse displacements. While initial responses were similar, plastic deformation led to divergences. Notably, the KB-loaded plate deformed away from the blast, unlike others. Error metrics indicated significant differences in displacement profiles between the loading methods.
  • Girder Example: For a wide-flange girder, simulations also showed distinct responses based on the loading parameters. The proposed equations (Current) and their triangular equivalents (Current-Tri) produced similar results, as did the KB and KB-Tri variants. However, larger differences were observed between the Current and KB-based loadings. The study highlighted that peak displacements and flange deformation profiles varied considerably across the different loading scenarios.
  • Frequency Content: Analysis of the frequency content of the air blast loading revealed that while the energy is spread across a wide range, the decaying exponential and triangular pulses concentrate energy at zero frequency. Differences in frequency spectrums were observed between the proposed and KB approaches, suggesting potential impacts on structural response time histories.

Limitations and Future Work

  • Pressure Relief and Clearing Effects: The current code does not account for pressure relief or clearing effects, which occur after a reflected blast wave reaches a free edge. The authors note that there is no universally accepted model for pressure relief, and existing methods have shown inaccuracies.
  • Far Face Loading: The model neglects loading effects on the "far faces" of structural surfaces (i.e., surfaces not directly facing the explosion). This is acknowledged as a simplification that introduces conservatism.
  • Mach Stem Effect: The calculation of peak overpressure does not account for the Mach stem effect, which can significantly increase overpressure in certain conditions.
  • Experimental Validation: While the paper uses open-literature data, direct experimental validation of the developed code's predictions against new test data is not presented.
  • Future Work: The authors suggest future work could focus on improving the calculation of peak overpressure to include the Mach stem effect. Further research is also needed to better assess the inherent uncertainty in air blast loading and its impact on structural response.

Summary updated

Abstract

The current research presents a detailed methodology for generating air blast loading for use within a finite element context. Parameters describing blast overpressure loading on a structure are drawn from open literature sources and incorporated within a blast load generation computer code developed for this research. This open literature approach lends transparency to the details of the blast load modeling, as compared with many commonly used approaches to blast load generation, for which the details are not publicly available. As a demonstration, the load generation code is used with the finite element software LS-DYNA to simulate the response of a steel plate and girder subjected to explosions modeled using these parameters as well as blast parameters from other sources.
BibTeX
@inproceedings{guzas2010air,
  author = {Emily L. Guzas and Christopher Earls},
  title = {{Air blast load generation for simulating structural response}},
  booktitle = {Steel},
  year = {2010},
  doi = {10.12989/scs.2010.10.5.429},
}
Figure