In this chapter, new heat transfer enhancement technologies with unidirectional porous metal called “EVAPORON” and “Lotus’ Breathing” are introduced to remove and manage heat from high heat flux equipment. The unidirectional porous metals introduced here can be easily fabricated by unique techniques such as mold casting technique, explosive welding technique, and 3D printing technique. First of all, many kinds of porous media, which have been introduced by the author so far as a heat transfer promoter, are compared each other to clarify what kind of porous metal is more suitable for high heat flux removal and cooling by focusing on the permeability and the effective thermal conductivity. For the practical use of the unidirectional porous copper with high permeability and high thermal conductivity, at first, heat transfer performance of two-phase flow cooling using a heat removal device called “EVAPORON” is reviewed aiming at extremely high heat flux removal beyond 10 MW/m2. We have been proposing this device with the unidirectional porous copper fabricated by 3D printing technique as the heat sink of a nuclear fusion divertor and a continuous casting mold. Second, two-phase immersion cooling technique called “Lotus’ Breathing” utilizing “Breathing Phenomenon” is introduced targeting at thermal management of various electronics such as power electronics and high performance computers. The level of the heat flux is 0.1 MW/m2 to 5 MW/m2. In addition, as the other heat transfer enhancing technology with unidirectional porous metals, unidirectional porous copper pipes fabricated by explosive welding technique are also introduced for heat transfer enhancement of single-phase flow.
- unidirectional porous metal
- high heat flux
- effective thermal conductivity
- heat transfer
- phase change
- lotus’ breathing
Porous media are defined as “materials containing very large numbers of pores with various sizes.” Owing to their characteristic large surface area, the porous media are used in various fields such as chemical plants, architecture, agriculture, environment, medical care, biology etc. In heat transfer engineering, which is one of this book’s topics, heat pipes  and vapor chambers  that utilize independent liquid supply based on the capillarity in a porous wick play an active part in the thermal management of electronic devices. For instance, thin heat pipes of <1 mm in thickness are even used in smartphones whose usage has witnessed an exponential increase . In addition, in recent years, to cool electronic devices, two-phase immersion cooling with saturated pool boiling heat transfer has attracted attention for increasing the heat generation density and reducing the costs of thermal management. In particular, great numbers of studies have focused on improving the critical heat flux by loading a porous medium with a functional shape and structure onto the cooling surface, in addition to enhance the boiling heat transfer. It is recommended to refer to the review studies of Yuki , Mori [5, 6], and Kandlikar et al. ; however, there have been multiple studies that attempted to separate liquid supply and vapor discharge, such as the honeycomb porous plate of Mori et al. , the porous layers having a non-uniform spatial structure [9, 10, 11], and porous meshes provided with hydrophilic and hydrophobic functions [12, 13]. Note that the results of Bai et al.  and Yuki et al.  in which the critical heat flux of the saturated pool boiling of water in an atmospheric pressure environment exceeded 5 MW/m2, thus demonstrating the high potential of the porous media as a latent heat transport promoter. On the other hand, also for a single-phase flow, there have been a large number of attempts to utilize the porous media as a heat transfer promoter. The examples include heat transfer enhancement by metal foams [16, 17], utilization as a radiation converter [18, 19], and promotion of heat transfer using sphere-packed tubes [20, 21].
Focusing on high potential of the porous media as the heat transfer promoter as mentioned above, multiple studies have introduced porous metals for heat removal from high-heat flux equipment of a level of 10 MW/m2. For example, the divertor of a nuclear fusion reactor currently under research and development is exposed to a steady heat load of approximately 10 MW/m2 due to the inflow of α particles generated by nuclear fusion reaction. Sharafat et al. [22, 23] introduced metal foams as a heat transfer promoter of He gas flow pressurized to 40 MPa and demonstrated high cooling performance of >10 MW/m2. Furthermore, Joshi et al. proposed a cooling technology that makes use of the phase change of coolant in a pin-fins microchannel for high heat generation density electronic devices exceeding 10 MW/m2 [24, 25, 26]. Moreover, he attempted a technology that assists heat transfer in the promotion of evaporation in a nanoporous film via a gas impinging jet . Regarding additional applications of the porous media under high heat flux conditions, refer to the review article by Smakulski et al. . Obviously, in the thermal management of these high heat flux devices, in addition to securing the cooling performance, it is necessary to consider both economical efficiency and maintainability; therefore, thermal management with low flow rate and low pumping power is indispensable.
As an efficient heat removal technology at a level of 10 MW/m2, Toda and Yuki proposed a cooling device known as EVAPORON (Evaporated Fluid Porous-Thermodevice) to take maximum advantage of the latent heat of vaporization of the coolant [29, 30]. EVAPORON can remove extremely high heat flux by evaporating the cooling liquid that is fed inside the porous medium jointed to the heat transfer surface against a heat flow, using the vast surface area with microchannels of the porous media. Because rather than relying on capillarity such as in the heat pipes, a minimum amount of liquid corresponding to the heat flux level is gently pumped into the porous medium for cooling, a significant reduction in pumping power can then be expected. Furthermore, we verified that the important factor of this technological breakthrough is the quick discharge of considerable amounts of vapor generated in the porous medium . Subsequently, the authors proposed EVAPORON-2 [32, 33] loaded with subchannels for vapor discharge within sintered copper particles and succeeded in the heat removal of over 20 MW/m2. Targeting at a cooling performance exceeding 10 MW/m2 under much lower pumping power conditions, furthermore, the EVAPORON-3 equipped with a vapor–liquid separator on the upper part of the porous medium was also proposed .
To summarize, when applying the porous media in the two-phase cooling technology for high heat flux environment, the most important factor is to actively promote the vapor discharge that was generated in the porous medium. Moreover, it is necessary to optimize the porous structure to maximize the advantage of the latent heat of vaporization of the coolant. On the other hand, there are various porous media such as foams, open cells, sintered particle/fiber, meshes etc. (see Figure 1), even if all of these are simply called porous metals. Generally, it is impossible to generalize the thermophysical properties and mechanical properties of the porous media with these different structures, thus referring only on their porosity and pore size. However, to promote phase change heat transfer in the porous medium, it is an undisputed fact that a porous medium with high permeability, which enables rapid discharge of the vapor generated, should be introduced in which its thermal conductivity is also increased to expand the effective heat transfer area. In general, there is a trade-off relationship between increasing the effective thermal conductivity and reducing the flow resistance of the vapor flow (increasing the permeability). Thus, the porous media to be introduced should be selected based on the heat flux level and coolant used. In this chapter, the effective thermal conductivity and absolute permeability for various porous media are first discussed. Subsequently, the effectiveness of unidirectional porous coppers  proposed by the author is quantitatively evaluated as a cooling technique on a level of 10 MW/m2. Finally, two our innovative cooling technologies are introduced; “EVAPORON-4”, which is the newest EVAPORON that combines the unidirectional porous copper and a grooved heat transfer surface and “Lotus’ Breathing”, which makes use of two-phase immersion cooling in a saturated pool boiling environment, based on spontaneous liquid supply effect called “Breathing phenomenon” proposed by the author for the first time all over the world.
2. Thermal conductivity and permeability of various porous media
Figure 2 shows the porosity and pore size distributions of sintered-metal-particles/fibers and foamed metals the author used to date. For example, regarding the thermal conductivity and the permeability of isotropic porous media with a simple structure, their modeling is frequently performed based on the porosity and the pore size for each porous medium. As shown in Figure 2, each porous medium has its characteristic porosity and pore diameter. For instance, most foamed metals and open cells have a high porosity of >90%. As far as the author knows, although those with a microscale pore diameter are not commercially available, Unno et al. developed the porous media with high porosity and micropores and demonstrated that it is effective for promoting evaporative heat transfer as per the capillary limit theory . Multiple sintered-metal-particles have a porosity of 30%–50% because of the producing process and the packing structure; moreover, the pore diameter varies from the order of micron to millimeter. For sintered-metal-fibers (a kind of fibrous porous media), the porosity can be regulated over a wide range by the extent of pressing; however, it is difficult to control the orientation of the fiber. We will now discuss the effective thermal conductivity and permeability of these typical porous media from the view point of phase-change enhancement of coolant.
2.1 Effective thermal conductivity of porous media
To discuss the effective heat transfer area, it is important to examine the effective thermal conductivity of the porous media. The prediction formulae for various porous structures have been developed to date; however, in general, the effective thermal conductivity
Herein, as shown in Figure 3, we performed heat conduction simulation to estimate the effective thermal conductivity of the sintered-copper-particle, in which a cubic sintered-copper-particle with a side length of 5 mm is sandwiched between copper square rods of 5 × 5 × 50 mm3. At the lower end surface, the heat flux of 0.5 MW/m2 is exposed and a constant temperature of 100 °C is given to the upper end surface. Adiabatic conditions are attributed to the side surfaces of the rods and sintered-particles. The effective thermal conductivity of the sintered-copper-particle is evaluated using the Fourier law from the difference in the average temperature at the two interfaces between the copper square rod and the sintered particle. The particle size is 1.0 mm and packed structure of particles is a simple cubic structure. To reproduce the neck structure formed between the particles during sintering, a cylinder with a diameter of
On the other hand, to maximize the effective thermal conductivity of the porous media, a porous medium to which the parallel model of Eq. (1) can be applied is desirable. This indicates that the introduction of a porous medium with unidirectional pore structure is ideal. Its thermal conductivity differs between the direction of the pores (
The abovementioned equation is driven by adapting the Behrens’ thermal conductivity model for composite material to the unidirectional porous media. Because the fluid in the pore is assumed to be a gas and its thermal conductivity can be ignored, Eq. (2) corresponds to the parallel model.
2.2 Permeability of porous media
As mentioned above, to effectively use the latent heat of vaporization in a high heat flux environment, a large amount of vapor generated in the porous medium must be quickly discharged to the outside. In other words, it is indispensable to take a mean, including introducing the porous media with high permeability or modifying the spatial structure while considering the improvement in effective thermal conductivity discussed in the previous section. Here, for reference, a simple comparison of the absolute permeability
Figure 7 shows the permeability of the sintered-particles and the unidirectional porous media. The horizontal axis is the porosity with a characteristic length
2.3 Other heat transfer-related issues for introducing porous media
An important issue when introducing the porous media as a heat transfer promoter is the contact thermal resistance generated at the contact interface between the porous medium and the heat transfer surface. In general, with increasing heat flux, the temperature gap increases due to the increase in contact thermal resistance generated at the interface, so that the heat transfer coefficient is directly affected. For instance, when a sintered-particle is mechanically pressed to the heat transfer surface, a large contact thermal resistance is generated because of the point contact state between the heat transfer surface and the particles. In the point contact state, the heat conduction to the first particle is considerably deteriorated, so that boiling and evaporation occur mainly on the heat transfer surface or the first several layers of the particle, and thus, effective utilization of latent heat of vaporization inside the porous medium cannot be expected. The best measure is sintering the sintered-particle to the heat transfer surface. Remarkable research results are reported by Kibushi et al.  regarding the contact thermal resistance in a high heat flux environment exceeding 1 MW/m2 (for the experimental details refer to the reference). Figure 8 shows the temperature gap that occurs at the contact interface of two flat surfaces. The loads on a jointing surface of ϕ10 mm are 0.33, 1.71, and 3.03 MPa, and there are the two patterns of mechanical and solder joints. Focusing on the heat flux of 2 MW/m2, in the mechanical joint case, a temperature difference of ~40 K occurs under a load of 3.03 MPa, and a temperature gap exceeding 100 K takes place at the load of 0.33 MPa. However, it can also be confirmed that the contact thermal resistance is significantly improved by solder bonding. At the heat flux of 5 MW/m2, the temperature gap is ~10 K, and it is ~20 K upon linear extrapolation to 10 MW/m2. Accordingly, it is evident that when the porous medium is loaded on the heat transfer surface, the contact thermal resistance should be reduced by solder bonding or the like. Our heat transfer tests regarding a gas impinging jet flow into a sintered-particle verified that the heat transfer coefficient is spectacularly increased by soldering the porous medium to the heat transfer surface . Further, the pool boiling results by Peterson et al.  have also demonstrated that boiling heat transfer is dramatically enhanced by HIP bonding of the porous medium to the heat transfer surface.
For introducing the unidirectional porous media proposed this time, if the pores that serve as the flow path are parallel to the heat transfer surface, a perfect surface contact state can be achieved over the entire heat transfer surface. Furthermore, even if the pores are perpendicular to the heat transfer surface, the surface contact state can be maintained in the solid phase portion, which can maximally reduce the contact thermal resistance between the porous medium and the heat transfer surface.
2.4 Introduction of unidirectional porous copper
A majority of unidirectional porous media proposed in this study is generally fabricated by electric discharge machining or MEMS. As the production with these techniques is very expensive, it is desirable to introduce the unidirectional porous media with excellent mass productivity. Here, three unidirectional porous media are presented, which are currently introduced to the author’s laboratory. The first is the lotus copper shown in Figure 9(a), in which pores are formed utilizing different saturation degree of dissolved hydrogen existing in molten copper with temperature gradient . Chiba et al. already demonstrated its superiority in single-phase flow heat transfer . The boiling heat transfer enhancement technology that the author is currently developing in cooperation with Lotus Thermal Solutions Co., Ltd. will be introduced in the next section. Hokamoto et al. also proposed a technique for forming a group of thin metal tubes into a unidirectional porous tube by exploded welding technique (Figure 9(b) ). To date, the author has introduced unidirectional porous copper tubes as a heat transfer promoter of a gas flow in a joint research with Hokamoto et al. of Kumamoto University. Based on single-phase flow heat transfer tests, the unidirectional porous tubes of 21 mm in outer diameter shows a heat transfer performance that is up to approximately eight times higher than that of a smooth tube at the maximum as shown in Figure 9(c) . The pore size and the porosity of this porous tube can be adjusted by varying the inner diameter and wall thickness of the thin tube. The lastly introduced unidirectional porous copper is the one molded by a metal 3D printing technique . Recent advances in 3D printing technology enabled the molding of even porous media composed of copper and having microchannels. The adjustment of the pore diameter and the pore structure matching the heat transfer mechanism are possible, which is the greatest attraction of this technology.
On the other hand, when this type of unidirectional porous copper is applied as a promoter of boiling/evaporation heat transfer, particularly if the pores are perpendicular to the heat transfer surface, one of the advantages is that the coolant can surely be supplied to the heat transfer surface. The major limitation is how to rapidly discharge a large amount of vapor generated in the porous medium to the outside. As a solution, we propose to joint the unidirectional porous copper to the grooved heat transfer surface; thus, as shown in Figure 10, two cooling structures are enabled. The first structure on the left shows the liquid supply direction is opposite to the vapor discharge direction, whereas the second structure on the right shows a method of directly discharging the vapor outside the porous medium via these grooves. In the next section, we will focus on the cooling performance related to flow boiling heat transfer and pool boiling heat transfer using these cooling structures.
3. Enhancement of flow boiling and pool boiling heat transfers using unidirectional porous copper
3.1 Enhancement of flow boiling/evaporative heat transfer with EVAPORON-4
As mentioned in the Introduction, the author proposed EVAPORON, EVAPORON-2, and EVAPORON-3, cooling devices using porous metals [29, 30, 31, 32, 33, 34]. The porous metal is connected to the cooling surface as shown in Figure 11, the cooling liquid is supplied into the porous medium in a countercurrent to the heat flow, and the heat is removed by the vigorous phase change of the cooling liquid within. To smoothly discharge the large amount of vapor outside the porous medium, EVAPORON-2 has several subchannels inside the porous medium, and EVAPORON-3 has a liquid–vapor-separator on the porous medium. Here, the concept of cooling device EVAPORON-4, using a unidirectional porous media is realized as shown in Figure 12(a). As shown in Figure 12(b), 9 × 5 grooves for the vapor discharge are formed on the heat transfer surface, and the unidirectional porous copper shown in Figure 12(c) is joined by soldering. The groove is 1.0 mm in width and 0.5 mm in depth. The unidirectional porous copper with a diameter of 20 mm and a thickness of 10 mm is produced by machining and has 248 small pores of ϕ0.5 mm in diameter. In addition, the porous medium has large five ϕ2.6 mm holes for the vapor discharge. EVAPORON-4 is a once-through type cooling device in which the cooling liquid supplied from the upper part of the unidirectional porous copper through the small pores undergoes phase change in the pores and the grooves to be then discharged through the grooves and the vapor discharging large holes. Figure 13 shows some of the results in two cases that the unidirectional porous copper is jointed onto the heat transfer surface with/without soldering (“Without TIM (Thermal Interface Material)” shows the data of mechanical joint without bonding). For details of the experimental apparatus and various results, please refer to Reference . The inlet liquid subcooling of distilled water is 40 K, whereas the x mark on the plot shows that the obtained data represent the critical heat flux. From this figure, the data with soldering move to the lower wall-superheat side under each flow rate condition, which verifies that the phase change heat transfer is enhanced by solder bonding the porous copper to the grooved heat transfer surface. Focusing on the data obtained at the flow rate of 0.5 L/min, while the critical heat flux without solder bonding is ~3 MW/m2, the critical heat flux is improved to exceed 6 MW/m2 by solder bonding (1.8-fold enhancement). For the flow rate of 2.0 L/min, as the maximum heat flux (not the critical heat flux) without solder bonding is approximately 9 MW/m2, the critical heat flux by solder bonding is expected to exceed 10 MW/m2, though the experiment was stopped due to the temperature limit of the cartridge heaters we used. For reference, the measured critical heat flux of the impinging jet at the flow rate of 2.0 L/min is approximately 4 MW/m2 (the same flow rate is ejected from a hole of ϕ2 mm), so that the cooling performance of EVAPORON-4 is much higher compared with other cooling technologies. These results are attributed to the effective functioning of the phase change within and the vapor discharge outside the porous medium, because the heat transfer is remarkably improved under high heat flux conditions.
Currently, we are examining the flow rate distribution and the phase change in porous media by CFD and visualization experiments. From now onward, we intend to push forward discussions and optimization by introducing the porous copper fabricated by a 3D metal printer, as shown in Figure 14, thus comprising the un-uniform unidirectional pore and groove structures that can make maximum use of the latent heat of vaporization . Moreover, we are examining a small-sized cold plate having in mind an application to electronic devices, a theme to be referred to in Reference .
3.2 Enhancement of critical heat flux of saturated pool boiling using lotus coppers by “breathing phenomenon”
In the past heat transfer experiments using EVAPORONs, the author noticed that the inlet pressure of the fluid rapidly decreases when the phase change in the porous medium is considerable and the vapor is vigorously ejected to the outside, i.e., there must be spontaneous liquid supply phenomenon associated with the discharge of vapor (henceforth referred to as “Breathing Phenomenon”). As shown in Figure 15, the author has proposed to attach a porous plate called “lotus copper” with a unidirectional pore structure to the grooved heat transfer surface (Figure 9(a)), whereby the critical heat flux of saturated pool boiling is spectacularly increased [15, 52, 53, 54]. This two-phase immersion cooling technique we usually call “Lotus’ Breathing” offers the following features.
Spontaneous liquid supply phenomenon due to vapor blowout (Breathing Phenomenon)
No requirement of capillarity induced in the porous medium
Once-through type liquid supply and vapor discharge (Figure 15). Liquid supply and vapor discharge are separated and formation of coalesced bubbles is retarded.
Use of a unidirectional porous copper with high permeability and high thermal conductivity to enhance the breathing phenomenon and the heat transfer in the porous medium.
There are two possible breathing modes “Mode-G” and “Mode-L”, as in Figure 15(a) and (b), i.e., the Mode-G on the left in which the vapor is discharged from the groove and accordingly the liquid is supplied from the upper part of the porous medium as well as the inverse mode-L.
To demonstrate the increase in CHF by the breathing phenomenon, saturated pool boiling experiments have been conducted in an atmospheric pressure environment. For details of the experiment, please refer to Reference . The Lotus copper plate (10 mm x 10 mm, thickness = 2 mm, porosity = 65.9%, and average pore diameter = 0.49 mm) is attached to a 10 mm × 10 mm heat transfer surface provided with 0.5 mm-square or 1.0 mm-square unidirectional grooves as a boiling heat transfer surface. The results for water and FC72 are shown in Figure 16. In the case of water, the critical heat flux for the smooth surface in this device is 1.4 MW/m2, which is an adequate value for a boiling heat transfer experiment using the heat transfer block. The results also verify that the critical heat flux when utilizing the lotus copper jointed on the grooved heat transfer surface considerably improves by providing wider grooves. The 0.5 mm-square groove achieves the critical heat flux of 4.6 MW/m2 at the wall superheat of 58.9 K, and the one of 1.0 mm-square groove achieves the critical heat flux of 5.3 MW/m2 at the wall superheat of 111 K, which indicates that the groove size is an important factor for improving the critical heat flux by the breathing phenomenon. The boiling heat transfer is also remarkably enhanced even if compared to that for the smooth surface; this technology has proven to achieve both the enhancement of the boiling heat transfer and the critical heat flux. By visualizing the boiling phenomenon shortly before the critical heat flux, a large amount of vapor is discharged from the grooves. Thus, the breathing phenomenon in these experiments conceivably corresponds to the Mode-G of Figure 15(a). On the other hand, in the FC72 case, the maximum heat flux, which is not the critical heat flux, is 0.83 (MW/m2) that is 5.5 times higher than CHF of the smooth surface. This result verifies that the breathing phenomenon can work for dielectric fluid and increases the critical heat flux.
Currently, the mechanism of critical heat flux enhancing and the factors that further entails the breathing phenomenon are being discussed based on the visualization experiment and CFD simulation as shown in Figure 17. In addition, the effects of varying the porosity, the pore size of the lotus copper, and the groove structure have been evaluated [55, 56].
In this chapter, the thermal conductivity and permeability of various porous media were quantitatively evaluated; as one of the porous media appropriate for high heat flux heat removal, targeted by the author, the introduction of “porous metal with unidirectional pores” was strongly suggested. Furthermore, as an application of using unidirectional porous coppers, we demonstrated the cooling performance of the flow boiling/evaporation cooling device EVAPORON-4, and that of the pool boiling enhancing technology “Lotus Cooler” based on “Breathing Phenomenon”. From now onward, we intend to optimize the structure of the unidirectional porous media and the grooved heat transfer surface, thus aiming at improving and controlling the performance of these cooling technologies.