Developing the realistic blast loading associated with an internal detonation occurring within a pressure vessel or heat exchanger is challenging. Unlike evaluation of external blast loading on structures due to far-field explosions, where typical overpressure-time histories can be reasonably defined based on empirical data, investigating confined detonations presents additional complications. The subsequent impulsive peak reflected overpressure from confined detonations acting on a structure can be extremely high due to the close proximity of the blast source to the vessel wall or pressure boundary. This establishes the possibility of significant structural damage for process equipment subjected to an internal detonation, even for relatively modest amounts of concentrated explosive products.
This paper discusses the underlying theory of blast analysis and examines the practical application of non-linear, finite element based, explicit computational techniques for simulating the load acting on a structure due to internal and external blasts. The investigation of a recent, real-life industry failure of a heat exchanger due to a suspected internal detonation is discussed. Explicit, three-dimensional blast analysis is performed on the heat exchanger in question, and an internal detonation is simulated to reasonably replicate the considerable damage actually observed in the field. This analysis permits the determination of an approximate amount of concentrated product that caused the accidental explosion; that is, the plausible equivalent amount of explosives is back-calculated based on the predicted damage to the finite element model of the equipment in question. Computational iterations of varying charge amounts are performed and the predicted amount of permanent damage is documented so sensitivity to the hypothesized charge amount can be quantified. Furthermore, explicit blast analysis of nearby equipment is performed.
In this investigation, computational results for both the heat exchanger (subjected to internal blast loading) and surrounding equipment (subjected to external blast loading) are in good agreement with the measured plastic deformations and failure modes that were actually observed in the field. Commentary on the likely detonation event that caused the significant damage observed is provided. Additionally, an advanced finite element failure criterion that is driven by plastic yielding is employed where portions of the computational model are removed from the simulation once a user-defined strain threshold is reached. This approach facilitates simulation of the gross heat exchanger pressure boundary failure actually observed in this case. The explicit finite element based analyses discussed in this study reasonably predict the structural response and damage characteristics corresponding to a recent, real-life industry failure.