This work addresses the behaviour of replicated microcellular pure aluminium under multiaxial stress states and in the presence of stress and strain localization sites. Processing of the foam was conducted in-house, using the replication process. The main processing steps are: (i) infiltrating a bed of sieved sodium chloride (NaCl) particles with pure molten aluminium under Argon pressure, (ii) solidifying the casting directionally, (iii) machining samples in the solidified composite obtained, and (iv) dissolving the salt in water containing chromate corrosion inhibitors. The main advantage of this process is its great versatility; notably, it allows tailoring independently both the size of the final pores in the replicated foam (by sieving the initial particles), and the relative density of the obtained foam material (by compacting the salt particles before infiltration). The first part of this work is dedicated to the study of the multiaxial yield behaviour of replicated foams. For this purpose a new triaxial loading frame was built to test cubical specimens. In Chapter 3 preliminary tests on pipe specimens are presented together with the results obtained on the triaxial device for 75 µm pore size foam cubes under axisymmetric and biaxial stresses. These show that the yield surface depends on all three stress tensor invariants, a dependence that has been suggested recently by theory and finite element simulation, but hitherto not shown in microcellular metal. A simple empirical expression taking the influence of the third stress tensor invariant into account is proposed based on the data. Experiments on 400 µm pore size foams under axisymmetric and biaxial stress and in three different deviatoric II-planes (planes in stress space perpendicular to the hydrostatic axis) confirm the dependence of the yield surface on the third stress tensor invariant, the yield surface becoming trilobal at elevated hydrostatic stress (tensile or compressive) in the II-planes, while being circular at zero hydrostatic stress, a feature that is captured by the empirical expression proposed on the basis of the present experiments. The second part of this thesis addresses foam deformation and fracture in the presence of cracks, notches or holes. The fracture toughness of replicated 400 µm pore microcellular aluminium is explored using disc-shaped compact tension testing broadly along the lines of the ASTM E1820-08a standard. The material shows marked R-curve behaviour, and the presence of bridging ligaments in the crack wake. Fractography reveals that the crack propagates via the rupture of struts normal to the crack plane, this being accompanied by a degree of plastic deformation of the struts near the crack plane that is made visible by slip markings along the strut surface and that increases in extent with Vm. Measured crack initiation and steady state propagation J values vary roughly as Vm3; this is a stronger dependence than has been observed in commercial (closed-cell) metal foams. A simple model is proposed to describe the initiation toughness of ductile open-cell foams, based on an estimate of the crack tip opening displacement at the moment when a strut aligned in the loading direction fails by ductile rupture just ahead of the crack tip. The model agrees well with the data up to a relative density of 20%, accounting in particular for the observed scaling of the toughness of replicated foams on the third power of the relative density. Finally the tensile flow and failure of flat dog-bone specimens containing a cylindrical hole, and of cylindrical notched samples of microcellular aluminium is studied. Both 400 µm and 75 µm pore size foams exhibit a notch strengthening effect, i.e., the peak failure stress increases as the depth of notches in cylindrical samples increases. In dogbone samples, the presence of a hole ranging from 0 to 4mm in diameter (in a sample 9mm wide) does not affect the net section peak failure stress of the 75 µm foam while the 400 µm pore size foam exhibits a slight increase in net section failure stress as the hole diameter is increased to 2 mm. Plastic flow curves for notched and hole-containing samples are well predicted by a finite element simulation based on the Deshpand – Fleck flow law, showing that the observed trends in the data are predominantly mechanical in nature, and strongly linked with the presence of triaxiality at the center of the notched samples.