Generative design of soft-armour

Computational models exist to quantify the performance of personal ballistic armour, where ‘performance’ refers to a variety of aspects of personal armour such as ballistic performance or human performance. Balancing the various aspects of performance is (and has always been) the quintessential element in body armour design due to limited load carrying capacity of a person. The Dutch organisation for Toegepast Natuurwetenschappelijk Onderzoek (TNO) has combined these analytical models into a generative design algorithm to aid body armour design optimization by generating an array of potential designs within a given set of constraints.
TNO foresees the use of this design tool to explore how body armour design changes with different performance requirements. Alternatively this design tool can be used to explore how a reduction in one aspect of performance requirements affects the other aspects of performance. Generative design of body armour is achieved through a Differential Evolution (DE) algorithm that combines a discretized approach to body armour parametrization with existing computational models of body armour performance. As a proof of concept, TNO implemented a detailed analysis to quantify protective performance, combined with simplified analyses to thermal, mobility and ergonomic constraints. The body armour is parametrized into squares of material that have varying areal density and can be positioned anywhere on the body. This allows (almost) freeform design of body armour. Protective performance is quantified using TNO’s in house developed ICARUS-suite that evaluates four load cases of a fragmenting threat. The simplified approach to thermal, mobility and ergonomic constraints are constructed by assigning penalties to locations on the body where armour is placed. The algorithm evolves a group of designs over the course of a fixed number of generations. The result is a design that specifies body coverage and areal densities and shows the potential of the tool as a means to integrate and balance various aspects of body armour performance. This paper documents the development of the design tool and the various considerations that come into play when combining existing analytical tools in a generative algorithm.