Thermag IX, 2021

Stefano Dall’Olio, Urban Tomc, Katja Klinar, Simon Nosan, Andrej Kitanovski

University of Ljubljana, Faculty of Mechanical Engineering, Ljubljana, 1000, Slovenia, stefano.dallolio@fs.uni-lj.si

Abstract

To exploit the potential of magnetic refrigeration, a prototype of wine chiller based on an active magnetic regenerator (AMR) was designed and built in our lab. The device consists of two regenerators working in parallel, containing around 100 g of gadolinium (Gd) in total. In our system the magnetic flux is provided by an electromagnet and the applied flux can reach a magnitude of 1 T in the high field mode, while being close to zero when the coil is not activated. The magnet can work at a frequency up to 20 Hz. To increase the cooling power and the coefficient of performance (COP) of our magnetic prototype we focused on the optimization of the core component of the technology, i.e. the active regenerator. In our prototype the fluid passes through a porous media consisting of a packed bed of Gd spheres having an average diameter between 100 and 300 μm. The regenerator has a cross section of 167 mm2, length of 55 mm, and is 3D printed with the technique of stereolithography (SLA). The optimization of the regenerator has been achieved by focusing on three main areas: structural characteristics of the housing, flow distribution and heat losses of the regenerator to the environment. For each of the previous points, there was an iterative process based on numerical simulations carried out in ANSYS to decrease the amount of dissipations. For the mechanical design of the regenerator, we considered 8 bar as the highest flow working pressure, and we consequently optimized the geometry to have a maximum deformation of the housing lower than 150 μm, threshold which corresponds to the average sphere diameter.

The flow maldistribution inside the regenerator is another important source of dissipation, together with the amount of the dead volume. Therefore, a flow distribution chamber has been added at the regenerator inlet and outlet, and the geometry of this space has been adjusted to have the most uniform flow along the regenerator. The effect of the different chamber shapes on the flow distribution has been verified in FLUENT and is shown in the Figure below. Because the distribution chamber acts as dead volume for the regenerator, we tried to keep it as small as possible. Moreover, we also focused on how to decrease the parasitic losses to the environment. This process has been done by insulating the regenerator from the environment and by installing some passive heat sink on the iron core of the electromagnet. After initial validation tests and debugging of the system, we were able to obtain the first experimental results that will be presented in this paper.

Urban Tomc(a), Matej Šadl(b), Katja Klinar(a), Hana Uršič(b), Andrej Kitanovski(a),

(a) Faculty of mechanical engineering, University of Ljubljana, Ljubljana, 1000, Slovenia, urban.tomc@fs.uni-lj.si (b) Electronic Ceramics Department, Jožef Stefan Institute, Ljubljana, 1000, Slovenia

Abstract

Today’s state-of-the-art magnetocaloric technology is based on the so called Active Magnetic Regeneration (AMR) principle. The AMR is based on the reciprocating movement of the fluid through a porous magnetocaloric (MC) structure. Such a system usually comprises a large amount of caloric material and a fairly complex hydraulic system, which is more suitable to be implemented in large cooling, refrigeration or heat pump devices. On the other hand, miniaturized electronics also produce vast amounts of heat that need to be efficiently managed. In this manner, an alternative research approach is emerging in the fields of MC technology. It involves new concepts of devices, which would apply so called thermal switches. The application of thermal switches could lead to drastic improvements in the heat transport from/to the MC material and consequently to the miniaturization of MC devices. An interesting domain, to look for thermal switch mechanisms, is microfluidics, which has enabled the development of integrated lab-on-chip devices. Although most microfluidic devices are based on a continuous flow of liquids in microchanells, there has been an increasing interest for the past couple of years in devices that rely on manipulation of discrete droplets using surface tension effects. One such technique is ElectroWetting On Dielectric (EWOD), which is based on wettability of liquids on a dielectric solid surface by varying the electrical potential. In this contribution we will present a new concept of a magnetocaloric device which couples MC effect and EWOD droplet actuation as a thermal switch mechanism. We will show different potential designs of such devices and their operation. For example, Figure 1 presents CFD simulations (Ansys Fluent) of the operation of such a potential device. Furthermore, the materials and its properties which constitutes the whole device will be discussed.

Keywords: magnetocaloric, digital microfluidics, electrowetting, thermal switch

Luka PORENTA (a), Parham KABIRIFAR (a), Andrej ŽEROVNIK (a), Matjaž ČEBRON (a), Borut ŽUŽEK (b), Miha BROJAN (a), Jaka TUŠEK* (a)

(a) University of Ljubljana, Faculty of Mechanical Engineering, Ljubljana, 1000, Slovenia, *jaka.tusek@fs.uni-lj.si

Abstract

Elastocaloric cooling technology is being considered as one of the most promising alternatives to vapor-compression cooling in the recent years. This technology is based on elastocaloric effect (eCE), which is closely related to the superelasticity of shape memory alloys. When a superelastic shape memory alloy (elastocaloric material) is mechanically loaded, an exothermal martensitic transformation occurs and consequently a latent heat is released, which heat up the material under adiabatic conditions. Upon unloading, the reverse transformation occurs and the latent heat is absorbed, which cool down the material under adiabatic conditions.

Elastocaloric cooling or heat-pumping devices need to work continuously and require long-term cyclic loading/unloading of elastocaloric materials, therefore, the fatigue life of the elastocaloric material is their crucial limiting factor. In tensile loading, lower transformation stresses and the possibility of utilizing geometries with good heat transfer properties (thin wires or sheets) are advantageous. Nevertheless, an acceptable fatigue life in tension can be achieved only through the application of small strains (and therefore small eCE) [1]. On the other hand, compressive loading prevents crack growth and consequently increases the fatigue life of the elastocaloric material [2, 3], but, thin elastocaloric elements that have good heat transfer properties are prone to buckle under compression forces (strains) that are required for a complete phase transformation. Currently, thin-walled tubes seem to present the best compromise between buckling stability and the required heat transfer properties. Therefore, this study aims to evaluate the buckling stability, eCE (manifested through adiabatic temperature changes) and structural and functional fatigue behavior of thin-walled Ni-Ti tubes (with outer diameter of 3 mm and wall thickness of 0.25 mm) with austenitic finish temperature (Af) around 0 °C. Initially, buckling stability of the tubes with different gauge lengths (between 10 mm and 20 mm) were tested under compressive stresses of up to 1600 MPa at a strain rate of 1.3×10-3 s-1. Six tubes of 10 mm gauge length, which showed a stable response through the entire transformation plateau, were used in subsequent fatigue life and eCE studies. All six tubes were first subjected to 100 training (stabilization) cycles at a strain rate of 2.3×10-3 s-1 between 10 MPa and 1150 MPa. Four tubes were further subjected to fatigue cycling between 185 MPa and 1015 MPa (full transformation plateau) at a frequency of 5 Hz. All four tubes survived 106 cycles (runout) without enduring any noticeable structural damage. The eCE (i.e., adiabatic temperature changes) of all six tubes (four fatigued and two only trained tubes) were measured at a strain rate of 6.8×10-2 s-1 (adiabatic conditions) up to the strains that corresponded to the end of the transformation plateau. All six tubes generated adiabatic temperature changes of 27 K and 20 K upon being loaded and unloaded, respectively. The results demonstrate that eCE of thin-walled tubes does not degrade during compressive fatigue cycling and thus durable and stable adiabatic temperature changes, which are prerequisites for practical applications of eCE, can be achieved.

Acknowledgement

The work was supported by European research Council (ERC) under Horizon 2020 research and innovation program (ERC Starting Grant No. 803669).

[1] J. Tušek et al., Acta Mater.,150, p. 295-307 (2018); [2] H. Hou et al., MRS Bulletin, 43, p. 285-290 (2018); [3] J. Chen et al., Appl. Phy. Lett, 115, 093901 (2019).

Keywords: shape memory alloys, elastocaloric effect, fatigue life, compressive loading, structural stability

1. Masche1, L. Ianniciello1, J. Tušek2, K. Engelbrecht1

1 Department of Energy Conversion and Storage, Technical University of Denmark – DTU, Anker Engelunds Vej B301, 2800 Kgs. Lyngby, Denmark

2 Faculty of Mechanical Engineering University of Ljubljana Aškerčeva cesta 6, SI-1000 Ljubljana, Slovenia

Conventional refrigeration technologies based on the compression and expansion of greenhouse gases with high global warming potential (GWP) are a risk for environment and human health. Hence, the advent of alternative refrigeration technologies is a welcome development. Caloric cooling is a technology based on the caloric effect in solid-state materials driven by an externally applied field. Depending on the nature of the driving field, four main solid-state cooling technolgies can be distinguished: magnetocaloric, elastocaloric, electrocaloric, and barocaloric. Caloric cooling that employs solid-state materials with zero GWP as refrigerants is an environmentally friendly cooling alternative and may solve problems associated with conventional refrigeration technologies. The most promising caloric refrigerants are first-order phase transition (FOPT) materials that yield large adiabatic temperature changes when an external field is applied, but also experience a hysteresis effect. One big concern manufacturers have about novel caloric cooling systems is the cyclability of the solid-state refrigerant through its FOPT, which can be impeded due to the existence of hysteresis. In this work1, we quantify the hysteretic behavoir of six modeled caloric materials (with different isothermal entropy changes and specific heats) and its impact on their cooling performance. Understanding the hysteretic behavior can help to maximize the efficiency of solid-state cooling devices.

A 1D active regenerator model with hysteresis term (treated as a entropy generation) was used to quantified the impact of realistic hysteresis values (0%, 0.04%, 0.5%, 1%, 2%, 6% and 12%) on the cooling performance of caloric materials. The model shows that hysteresis has a larger impact on the coefficient of performance (COP) than on the caloric cooling power. Caloric materials with higher specific heat and lower isothermal entropy change (HiCLoS) show a greater reduction of the cooling performance, and performance reductions become more substantial at higher hysteresis levels and higher working frequencies. For instance, a hysteresis entropy generation value (qhys) of 0.04% will lead to a large performance reduction for HiCLoS materials, while materials with low specific heat and low isothermal entropy change (LoCLoS) show positive COP and cooling powers up to a qhys of 2% (see Figure 1). Modeling results are of great importance when designing a solid-state cooling system and selecting the most suitable caloric refrigerant.

Keywords: Hysteresis; caloric materials; 1D model; solid-state cooling

References:

1 M. Masche, L. Ianniciello, K. Engelbrecht, L. Ianniciello, and K. Engelbrecht, Int. J. Refrig. (2020).

Lucia Ianniciello(a), Jaka Tušek(b), Kurt Engelbrecht(a)

(a)Department of Energy Conversion and Storage, Technical University of Denmark, Anker Engelunds Vej, 2800 Kgs. Lyngby, Denmark, luciann@dtu.dk, (b) Faculty of Mechanical Engineering, University of Ljubljana, Aškerčeva cesta 6, SI-1000 Ljubljana, Slovenia

Abstract

Elastocaloric cooling could be an alternative to classic cooling systems, as they represent a way to cool more efficiently and being more environmentally friendly. Several shape memory alloys have been investigated for this purpose, either Cu-based, Fe-based or Ni-based. Ni-Ti was selected in this study to build an elastocaloric regenerator. A stack of nine plates has been tested in a cooling system to evaluate its performances. The plates are dog-bone-shaped. The Ni-Ti alloy is composed 49.1 at.% Ti and has an austenite finish temperature of approximately 270 K. The edges of the plates were hand polished with sandpaper and wax compounds, and a buffing wheel. Microscopy imaging was done on the edges of the plate after polishing. The sample was trained up to 120 cycles at 6% of strain, a strain rate of 1.10-4 s-1 and 60 s of waiting time between loading and unloading using a Zwick/Roell testing machine to stabilize the material behavior. The flow system is composed of two loops, one to extract the heat and the second to cool the load. Water was used as the heat transfer fluid. The hot loop contains a heat exchanger and a hot reservoir whereas the cold loop contains a heater to simulate a cooling load. The adiabatic temperature change of the regenerator were measured using an IR camera for different strain amplitudes. To assess the performances of the system different strain amplitudes, different waiting times at the end of loading and unloading and different straining speed were tested.

Keywords: elastocaloric materials; shape memory alloys; caloric devices; cooling; regenerator.