Role of thermal expansion anisotropy on the elastocaloric effect of shape memory alloys with slim-hysteresis superelasticity

We reveal the mixed-mode mechanism of elastocaloric effect in the course of slim-hysteresis superelasticity in a severely deformed NiTi shape memory alloy with tailored (negative/zero/positive) thermal expansion anisotropy. It is shown that in addition to the latent heat of phase transformation, the reversible heat from elastic deformation (known as the thermoelastic effect) plays a significant role in the overall elastocaloric response. The magnitude of the adiabatic temperature change associated with the thermoelastic effect $|\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}|$ can reach 3.1 K, which is comparable to that from the phase transformation latent heat ($|\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{PT}}|=2.9$ K). The sign and magnitude of the $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}$ scales with the sign and magnitude of the coefficient of thermal expansion (CTE) along the stressing direction. For the directions with negative CTEs, the $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}<0$ while for those with positive CTEs the $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}>0$ upon rapid release of tensile stresses. As such, the cooling performance is strongest when the material is tensioned along the direction of the strongest negative CTE $\left(–14.3\times {10}^{-6}{\mathrm{K}}^{-1}\right)$ owing to the synergistic interplay of $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{PT}}$ and $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}$. The results reveal the importance of thermal expansion property on the elastocaloric effect.

Figure 1

Microstructure, mechanical, and thermal expansion properties of the 46% cold-rolled NiTi. (a) Bright-field TEM micrograph. (b) Dark-field TEM micrograph captured using the diffraction arc (containing the intensities of both B2 and B19′ phases) marked in (a). (c) High-resolution TEM image. (d) Fast Fourier transformation images of the boxed regions in (c). (e)–(f) Crystallographic textures of the (e) B2 and (f) B19′ phases represented by ${\varphi }_2={45}^{\circ }$ and ${\varphi }_2={84}^{\circ }$ ODF sections. (g)–(h) In-plane tensile stress-strain curves. (i) In-plane anisotropic linear thermal expansion.

Figure 2

(a)–(e) Adiabatic temperature variation $\mathrm{\Delta }{T}_{\mathrm{ad}}$ of the 46% cold-rolled NiTi sheet upon rapid release of different uniaxial tensile stress levels along different in-plane directions at $T_{\mathrm{amb}}=292\mathrm{K}$.

Figure 3

Evolutions of $\mathrm{\Delta }{T}_{\mathrm{ad}}$ of the 46% cold-rolled NiTi sheet with the applied (a) stress and (b) strain levels.

Figure 4

In situ neutron diffraction characterization of the cold-rolled NiTi at room temperature. (a) Diffractograms measured under stress-free state, and 600-MPa and 1000-MPa tensile stress along RD. (b) Variations of the lattice parameter and lattice volume strain of the B2 phase with stress level. The lattice parameters ($a_0$) are obtained from the measured neutron diffractograms by Rietveld refinement and are used to calculate the lattice volume strain according to ${\varepsilon }_v=3\frac{\mathrm{\Delta }{a}_0}{a_0}$.

Figure 5

Contribution of the thermoelastic effect to the measured $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{total}}$ of the cold-rolled NiTi sheet. Stress dependence of $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{total}}$ along (a) $\theta =0^{\circ }$ (RD) and (b) $\theta ={90}^{\circ }$ (TD), and (c) direction dependence of $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{total}}$ for $\sigma =200\mathrm{MPa}$ due to the CTE anisotropy. The differences between $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{total}}$ and $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{TeE}}$ are due to the $\mathrm{\Delta }{T}_{\mathrm{ad}}^{\mathrm{PT}}$ from the stress-induced phase transformation.