No abstract

The behaviour of granular solid–liquid mixtures is key when deforming a wide range of materials from cornstarch slurries to soils, rock and magma flows. Here we demonstrate that treating semi-solid alloys as a granular fluid is critical to understanding flow behaviour and defect formation during casting. Using synchrotron X-ray tomography, we directly measure the discrete grain response during uniaxial compression. We show that the stress–strain response at 64–93% solid is due to the shear-induced dilation of discrete rearranging grains. This leads to the counter-intuitive result that, in unfed samples, compression can open internal pores and draw the free surface into the liquid, resulting in cracking. A soil mechanics approach shows that, irrespective of initial solid fraction, the solid packing density moves towards a constant value during deformation, consistent with the existence of a critical state in mushy alloys analogous to soils.

Grain refinement via semi-solid deformation is desired to obtain superior mechanical properties of cast components. Using quantitative in situ synchrotron X-ray tomographic microscopy, we show an additional mechanism for the reduction of grain size, via liquation assisted transgranular cracking of semi-solid globular microstructures. Here we perform localized indentation of Al-15wt.%Cu globular microstructures, with an average grain size of ∼480 μm, at 555 °C (74% solid fraction). Although transgranular fracture has been observed in brittle materials, our results show transgranular fracture can also occur in metallic alloys in semi-solid state. This transgranular liquation cracking (TLC) occurs at very low contact stresses (between 1.1 and 38 MPa). With increasing strain, TLC continues to refine the size of the microstructure until the grain distribution reaches log-normal packing. The results demonstrate that this refinement, previously attributed to fragmentation of secondary arms by melt-shearing, is also controlled by an additional TLC mechanism.

A first attempt at the three-dimensional evaluation of naturally initiated surface connected and internal fatigue cracks is presented. Fatigue crack initiation and growth in air and vacuum environments have been investigated through Xray microtomography in air and vacuum environments at elevated temperatures (350• C), accompanied by post-mortem electron microscopy of the fracture surfaces. In vacuum (< 10 −5 mbar), multiple internal and surface-connected crack initiation was observed, but only the surface-connected cracks grew. In contrast, fewer cracks formed in air, these were mostly surface-connected and all were observed to grow. In all instances the initiation features were associated with globular primary α. An improved fatigue life was found in vacuum, which was mostly a consequence of delayed initiation, but was also due to slower fatigue crack propagation. The non-propagation of internal cracks was taken to imply that even the good laboratory vacuum obtained here was insufficient to simulate the conditions obtained for an internal crack in a component. The crack shape evolved towards a semi-circular shape a/c = 1 in air during fatigue crack growth, whilst the vacuum cracks remained semi-elliptical (a/c 1.4). This was taken to imply that oxide-induced crack closure played a role in fatigue crack growth in air.

Lithium-ion cells can unintentionally be exposed to temperatures outside manufacturers recommended limits without triggering a full thermal runaway event. The question addressed in this paper is: Are these cells still safe to use? In this study, externally applied compression has been employed to prevent lithium ion battery failure during such events. Commercially available cells with Nickel Cobalt Manganese (NCM) cathodes were exposed to temperatures at 80 °C, 90 °C and 100 °C for 10 h, and electrochemically characterised before and after heating. The electrode stack structures were also examined using x-ray computed tomography (CT), and post-mortems were conducted to examine the electrode stack structure and surface changes. The results show that compression reduces capacity loss by −0.07%, 4.95% and 13.10% respectively, measured immediately after the thermal testing. The uncompressed cells at 80 °C showed no swelling, whilst 90 °C and 100 °C showed significant swelling. The X-ray CT showed that the uncompressed cell at 100 °C suffered de-lamination at multiple locations after test, and precipitations were found on the electrode surface. The post-mortem results indicates the compressed cell at 100 °C was kept tightly packed, and the electrode surface was uniform. The conclusion is that externally applied compression reduces delamination due to gas generation during high temperature excursions