No abstract

No abstract

Additive manufacturing (AM) enables the design of new cellular materials for blast and impact mitigation by allowing novel material-geometry combinations to be realised and examined at a laboratory scale. However, design of these materials requires an understanding of the relationship between the AM process and material properties at different length scales: from the microstructure to geometric feature rendition to overall dynamic performance. To date, there remain significant uncertainties about both the potential benefits and pitfalls of using AM to design and optimise cellular materials for dynamic energy absorbing applications. This experimental investigation focuses on the out-of-plane compression of stainless steel cellular materials fabricated using selective laser melting (SLM), and makes two specific contributions. First, we demonstrate how the AM process itself influences the characteristics of these cellular materials across a range of length scales, and, crucially, how this influences the dynamic deformation. Secondly, we demonstrate how an AM route can be used to add geometric complexity to the cell structure, creating a versatile basis for future geometry optimisation. Starting with an AM square honeycomb (the reference case), we add porosity to the walls by replacing them with a lattice truss, while maintaining the same relative density. This geometry hybridisation is an approach uniquely suited to this manufacturing route. It is found that the hybrid lattice-walled honeycomb geometry significantly outperforms previously reported AM lattices in terms of specific strength, specific energy absorption, and energy absorption efficiency. It is also found that the hybrid geometry outperforms the benchmark metallic square honeycomb in terms of energy absorption efficiency in the intermediate impact velocity regime (i.e. between quasi-static loading and loading rates at which wave propagation effects begin to become pronounced), a regime in which the collapse is dominated by dynamic buckling effects.

No abstract

In principle, stress gauges mounted to measure lateral stresses in a shocked matrix allow the shear strength of the material to be determined. However, interpreting the resistance profiles from lateral stress gauges is hindered by the fact that the stress field in the vicinity of the insulating layer in which the gauges are embedded can differ significantly from the stress field that would be generated in the sample if no gauge were present. A series of high resolution Eulerian hydrocode simulations have been run which suggest that the stresses in the insulating layer vary with distance and time in a way that depends on the thickness of the layer, the shock strength and the elastic and plastic properties of both the layer and the matrix. In particular, if the shock velocity in the matrix material is high the stress at a typical gauge position initially rises to a sharp peak then falls with time, but when the shock velocity in the matrix is low the stress rises relatively gradually throughout the time of interest. The shapes of the stress versus time profiles predicted by the hydrocode compare well with the results of lateral gauge experiments on several different materials. It is concluded that lateral gauges can be used to measure the dynamic strength of materials provided high resolution computer simulation is used to take account of the perturbation of the stress field in the shocked sample caused by the gauges.

No abstract

Manganin piezo-resistive gauges mounted in an insulating layer of thickness typically 100 µm, oriented perpendicular to the wave front, may in principle be used to measure lateral stresses in a shocked metal sample. However, high resolution simulations strongly suggest that the layer can perturb the stress field in the sample, thereby introducing significant errors into the observed stress measurement. In order to compensate for the perturbations we need to be confident that we are computing them correctly. Experiments are described in which velocity interferometry is used to obtain information about the perturbation, which is independent of that provided by the lateral gauges. The experimental results support the simulations and, therefore, provide strong evidence that lateral gauges, especially when mounted in metals, can significantly perturb the stress field in the sample.

The mechanism of shock-induced dynamic friction has been explored through an integrated programme of experiments and numerical simulations. A novel experimental technique has been developed for observing the sub-surface deformation in aluminium when sliding against a steel anvil at high velocity and pressure. The experimental observations suggest that slight differences in conditions at the interface between the metals affect frictional behaviour even at the very high-velocity, high-pressure regime studied here. However, a clear finding from the experimental work is the presence of two distinct modes of deformation termed deep and shallow. The deep deformation is observed in a region of the aluminium specimen where the interfacial velocity is relatively low and the shallow deformation is observed in a region where the interfacial velocity is higher. A 1D numerical treatment is presented which predicts the existence of two mechanisms for dynamic friction termed ‘asymptotic melting’ and ‘slide-then-lock’. In both modes there is a warm-up phase in which the interface temperature is increased by frictional heating. For high initial sliding velocity, this is followed by the onset of the asymptotic melting state, in which the temperature is almost constant and melting is approached asymptotically. This mechanism produces low late-time frictional stress and shallow deformation. For lower initial sliding velocity, the warm-up terminates in a violent work hardening event that locks the interface and launches a strong plastic shear wave into the weaker material. This slide-then-lock mechanism is characterized by sustained high frictional stress and deep plastic deformation. These predicted mechanisms offer a plausible and consistent explanation for the abrupt transitions in the depth of sub-surface deformation observed in the experiments. A key conclusion arising from the current work is that the frictional stress does not vary smoothly with pressure or sliding velocity. Instead the pressure and sliding velocity determine whether the impulse will be very high or very low. The next generation of friction models for hydrocodes will need to account for these factors.