Degree Name

Doctor of Philosophy


Department of Materials Engineering


The objective of this research work was to develop new Fe-Mn-Si based shape memory alloys with property and cost advantages over existing commercial Fe-Mn-Si and Fe- Mn-Si-Cr-Ni alloys. The work involved the investigation of the effect of alloying elements, heat treatment and precipitation on the shape memory effect and corrosion resistance of candidate alloys. The mechanism of stress induced 𝑦→Ρ martensitic transformation and its reverse transformation, and the effect of thermomechanical training on the shape memory effect were systematically studied.

TEM observations supported the concept that the regular overlapping of stacking faults can result in the formation of bulk e martensite plates. Stacking faults were also found to exist in e martensite plates, and it is inferred that these faults can act as embryos for Ρ→𝑦 reverse transformation.

It was found that the shape memory capacity of Fe-Mn-Si based shape memory alloys varies with annealing temperature, and this effect can be explained in terms of the effect of annealing on 𝑦↔Ρ transformation. Annealing at about 873K was found to be optimal to form the dislocation structures which are favourable for stress induced martensitic transformation, thus resulting in the best shape memory behaviour.

Strengthening of the austenite matrix is generally considered to be an effective method for improving the shape memory effect of ferrous shape memory alloys, and although ausaging has been found to improve transformation reversibility and shape memory effect in Fe-Ni based alloys, precipitation of coherent Ni3Ti particles in austenite was found in the present work to degrade the shape memory effect of Fe-Mn-Si based shape memory alloys. TEM results showed that precipitates can pin Shockley partial dislocations, suppressing stress induced 𝑦→Ρ transformation. Optical microscopy and X-ray diffraction results confirmed that the amount of stress induced e martensite decreased when precipitation occurred.

It was found that alloying Fe-Mn-Si ternary alloys with Al and Cu significantly enhanced their resistance to hydrochloric acid attack. For example, Fe-28Mn-6Si-lAl-1Cu and Fe-20Mn-6Si-7Cr-lCu (wt%) alloys developed in the work exhibited both good shape memory effect and corrosion resistance to hydrochloric acid. Furthermore, Fe-Mn-Si-Cr-Ni-Cu corrosion resistant alloys developed on the basis of conventional Fe-Mn-Si-Cr-Ni stainless steel alloys exhibited improved and increasing resistance to hydrochloric acid corrosion with increasing Cu content. The new Fe-Mn-Si-Al-Cu and Fe-Mn-Si-Cr-Ni-Cu alloys also showed better corrosion resistance than their conventional counterparts in sulphuric acid. Immersion tests in 3.5% NaCl solution showed that Cu did not significantly improve the corrosion resistance. Among the alloys examined, the reference Fe-Mn-Si-Cr-Ni alloy gave the best corrosion resistance in 3.5% NaCl solution. However, the corrosion resistance of Fe-Mn-Si-Cr- Cu alloys in 3.5% NaCl was relatively high and increased with increasing Cr concentration. Potentiostatic tests on Fe-Mn-Si-Cr-Ni and Fe-Mn-Si-Cr-Ni-Cu alloys indicated that the addition of Cu is beneficial in facilitating passivation.

The shape memory effect of the alloys investigated was significantly improved by thermomechanical training, with the efficiency of training being a function of the number of training cycles, the training strain, and the recovery annealing temperature. Generally, the shape memory effect increased with increasing training cycle number, and reached a saturation value after 4 to 6 training cycles. The optimal training strain was found to depend on the alloy composition. An excessive training strain induces slip deformation which does not contribute to the recovery strain. The recovery annealing temperature was found to be crucial to the thermomechanical training, and the current study indicated that annealing at about 873K was most effective for improving the shape memory effect.

Thermomechanical training makes stress induced martensitic transformation easier, because the training process creates dislocation structures which promote the nucleation of martensite. Training resulted in a finer martensite plate size, reducing the local transformation shear and volume strains which need to be accommodated in the austenite. Thermomechanical training also influenced Ρ→𝑦 reverse transformation, with the Af temperature decreasing with number of training cycles. In contrast, the Ms and As temperatures remained nearly constant. Therefore, the improvement of shape memory effect on training is due to the facilitation of both stress induced martensitic transformation and its reverse transformation.

Several of the new alloys, and particularly those based on Fe-Mn-Si-Cr-Cu, showed excellent strain recovery after training. Recovery strains of up to 5.4% were obtained, significantly higher than those obtained for the Fe-Mn-Si and Fe-Mn-Si-Cr-Ni reference alloys and higher than strains previously reported in the literature for Fe-Mn- Si based alloys. The corrosion resistance of the Fe-Mn-Si-Cr-Cu alloys was also similar to that of the commercial stainless alloy (Fe-Mn-Si-Cr-Ni) and because of the replacement of up to 10wt% Ni with lwt%Cu, the new alloys are much less expensive.