Giulia Lombardi1,Gael Sebald1,2,Gildas Coativy2,Jacques Jay2,Atsuki Komiya1
Tohoku University1,Institut National des Sciences Appliquées de Lyon2
Giulia Lombardi1,Gael Sebald1,2,Gildas Coativy2,Jacques Jay2,Atsuki Komiya1
Tohoku University1,Institut National des Sciences Appliquées de Lyon2
The elastocaloric effect of natural rubber (NR) was first studied at the beginning the 19<sup>th</sup> century, by John Gough in 1805, and James Joule in 1850, giving rise to what is today known as the Gough-Joule effect [1]. This describes the property of NR to give out heat when stretched and to contract when heated. Despite these findings now date back to more than 170 years ago, it was only recently that this elastomer gained interest among the scientific community for space cooling applications. Contrary to most caloric materials, NR is non-toxic, abundant and low-cost.<br/>When stretched between elongation of 1 to an elongation to 6, it was proved experimentally an adiabatic temperature change up to 10~15 K with rather low mechanical losses leading to a ratio between potential heat exchange and the mechanical work around 3~5 (also known as COP<sub>mat</sub>) [2]. This attractive elastocaloric activity is however not applicable because of the excessive geometry change during operation. For example, the surface at an elongation of 1 might be too small for fast heat exchange compared to the fully elongated one (change of exchange area combined with the large increase of the thickness). Therefore, up to now, we can mention only two works that explored the development of elastocaloric cooling based on NR: the single-tube system developed by Sebald <i>et al.</i> in 2020 [3] and the membrane-inflation system developed by Greibich <i>et al.</i> in 2021 [4]. This is highly in contrast with the number of published works of caloric cooling systems based on other elastocaloric materials, the most used being NiTi-based shape-memory alloys (SMAs) mainly due to its high adiabatic temperature (between 15 K and 40 K). However, these materials require high values of applied tensile force (300 MPa or more), which are not ideal for system scale-up.<br/>The objective of this work is devoted to the experimental evaluation of the elastocaloric activity of elastocaloric elastomers, that will be presented in detail. As a focus, we demonstrate that NR is indeed a good candidate for elastocaloric cooling applications, despite the limited adiabatic temperature exhibited by the material when subjected to cyclic elongations. Cyclic elastocaloric properties of NR tubes were studied by imposing trapezoidal cycles at 0.1 Hz under various elongations. The maximum temperature variations on the material were recorded for elongations between 400% and 600% of its initial length, giving a Δ<i>T</i> of 4 K. The main advantage of NR lies on the applied tensile stress, two orders of magnitude lower than that required by SMAs. We recorded, in fact, a maximum stress on the NR tubes of 2 MPa for cyclic elongations at 0.1 Hz between 350% and 550% of its initial length. The corresponding mechanical losses were found to be equal to 0.27 MJ m<sup>-3</sup>. In order to estimate the material efficiency for elastocaloric cooling, we calculated the material coefficient of performance COP<sub>mat </sub>, defined as the potential absorbed heat divided by the hysteresis loss. We found a COP<sub>mat</sub> value of 22 when considering the first 50 cycles, which is one of the highest recorded values among elastocaloric materials. As the characterization proceeded until 10<sup>4</sup> cycles, the COP<sub>mat</sub> reached 53, thanks to a drop of the mechanical losses of the material. These values, along with the low tensile force required, demonstrate the high elastocaloric potential of NR, a future key material to be used in cooling applications, given that further improvements both on the material and on the system side are achievable.<br/><br/>[1] G. A. Holzapfel and J. C. Simo, Comput. Methods Appl. Mech. Eng. 132, 17 (1996).<br/>[2] Z. Xie, G. Sebald and D. Guyomar, Appl. Phys. Lett. 108, 041901 (2016).<br/>[3] G. Sebald, A. Komiya, J. Jay, G. Coativy and L. Lebrun, Appl. Phys. 127, 094903 (2020).<br/>[4] F. Greibich, R. Schwödiauer, G. Mao, D. Wirthl, M. Drack, R. Baumgartner, A. Kogler, J. Stadlbauer, S. Bauer, N. Arnold and M. Kaltenbrunner, Nature Energy, 6, pages260–267 (2021)