Pressure-impulse diagrams for vented hydrogen deflagrations in 20-foot shipping containers- Numerical predictions using the Impetus Afea solver and validation against data from full-scale experiments
MetadataShow full item record
- Master theses 
A pressure-impulse (P-I) diagram is a graphical representation of the level of structural response of a given structure to pressure loads characterised by different combinations of pressure and impulse. Weak structures, such as buildings and shipping containers, can experience considerable structural damage if subjected to relatively modest internal pressure loads, and it is important to dimension both enclosures and venting devices in a way that prevents structural collapse and the formation of projectiles. Shipping containers are often utilized for housing process equipment, such as compressors and electrolysers. Fires and explosions represent a significant hazard for such installations, and knowledge about how a specific structure responds to internal pressure loads is useful for risk assessments and safe design. In the present study, P-I curves were created numerically for several wall displacements using the non-linear explicit finite element (FE) tool Impetus Afea Solver. The characteristics of the curves agree well with the theoretical characteristics of non-ideal explosions with finite rise time. The curves from the finite element method were compared to unique results from full-scale experiments conducted as part of the EU funded HySEA project. Most of the experimental results were located in the dynamic region of the P-I diagram. To make it easier to compare the experimental results with the different areas of damage in the P-I diagram, the experimental results were divided into categories of maximum displacement. Whereas some of the experimental results correspond to the areas in the P-I diagram with the same level of damage, others do not. There are several sources of uncertainty associated with both the numerical approach and the experiments. Sensitivity studies were performed to study the impact of varying the yield strength of the material in the container walls, moderate damage to the corrugated structure, and different pressure-time profiles for the pressure loads. Reducing the assumed yield strength of the material resulted in a significant increase in wall displacement. Damage to one of the walls affected not only the wall displacement of the damaged wall but the entire structure. Despite the deviations between model predictions and experiments, the use of P-I diagrams may still be valuable for safety and design purposes. The primary limitation from an engineering design point of view will most likely be the reliable prediction of the relevant pressure loads for a given structure. The suggestions for further work include the use of a more detailed model for the FE tool, combined with direct comparison of the structural response obtained with the measured pressure-time histories for repeated loading of the same structure. This will require more computationally intensive calculations than could be included in the present study.