Blast survivability of a fatigued naval surface platform

An assessment of the blast survivability is a requirement of modern naval vessels during both the initial design stage, and when life-of-type extension analyses are undertaken. Changes in the operational profile of naval vessels throughout their service life combined with the use of high strength maritime steels have resulted in structural fatigue becoming an increasingly important consideration as vessels age. Literature suggests that fatigue ageing may impact the high strain-rate performance of a structure. A lack of understanding of the effects of fatigue aging may therefore result in overestimations of the blast survivability of a naval platform, possibly putting sailors‚ÄövÑv¥ lives at a greater risk than if the effects could be accounted for. This PhD thesis aims to investigate and quantify the effect that fatigue has on blast survivability of naval structures made of DH36 steel, a common naval structural steel, by means of experimental and numerical investigations. The experimental methodology was developed and implemented to determine any effect of fatigue ageing on the high strain rate material properties of DH36. The numerical investigation determined the effect of fatigue cracking on structural response to blast loading. Following this, two machine learning tools were developed to enable the consideration of the results of this research in the design and analysis of a vessel. The novel experimental methodology included the development of a fatigue ageing process for DH-36 specimens from which smaller specimens for high strain rate testing were obtained. These smaller specimens were tested using a bespoke Split Hopkinson Tension Bar setup. The results show that there is no significant change in the high strain rate material properties of fatigued DH36, however a reduction in quasi-static yield stress and flow stress curves is evident. The numerical investigation was aimed at determining the effects that fatigue cracks may have on the structural response of a representative ship stiffened panel. The novel methodology includes extensive validation of an application of crack propagation elements in LS-DYNA using experimentally obtained explosive bulge tests data. The inbuilt LS-DYNA Friedlander-equation-based load models are utilised to model the blast loading over a wide range of charge sizes, stand-offs, and locations for a variety of crack lengths determined by crack propagation modelling. It was found that, for the investigated stiffened panel, a reduction in the highest survivable charge size of up to 37.5% exists, when compared to the survivable charge size for a crack that is not detectable by eye. This finding suggests that cracks can play a significant role in the structural response to blast loading for ship stiffened panels. The finding displays a need for designers and analysts to account for the effect of cracks on blast survivability. To assist the design and analysis of the fatigue effect on blast survivability a machine learning tool was developed based on a Neural Network algorithm. The tool allows for the rapid assessment of the charge required to initiate plate failure of a representative stiffened ship panel and is based on inputs of the charge size and location, and the length of fatigue crack that is present in the stiffener. The tool can readily integrate new or extended datasets of blast event results, and once an appropriate dataset is used for training, allows designers to determine the effect of different fatigue crack scenarios on the survivability of a given geometry when a fatigue analysis is incorporated. Finally, a method to determine the blast survivability of a representative ship stiffened panel is presented. The methodology utilises the results of the numerical investigation into blast loading and fatigue cracks, alongside crack initiation and growth models to determine change in blast survivability over time. The blast survivability of a representative ship stiffened panel is suggested to have been reduced by an average of 9% over a 20-year period due to fatigue crack growth. This is a significant reduction which, if unaccounted for in design and life-of-type extension analyses, will lead to the overestimation of the ability of a vessel to withstand blast loading.