Below are the questions asked during the event, along with their respective answers.
Q: How uniform is the foam after compression?
A: As long as the compression ratio is greater or equal to 2x, the result is very uniform. Initially, when you start compressing the foam you tend to see one side of the piece start to compress first, but by the time you reach 2x compression, there is no discernible variation through the thickness. We often compress larger blocks of foam as starting stock for cutting and machining of smaller pieces, and those smaller pieces do not vary in density any more than the natural variability found in uncompressed foams.
Q: How strong is the foam in tension, and what kind of pressure can a foam heat exchanger withstand?
A: Tensile strength depends on relative density and level of mechanical compression. Typical values of yield strength for uncompressed foam range between 100-500 psi for most applications. Compression increases mechanical strength proportionally, so 1000+ psi yield strength is common for compressed foam.
Q: What is the variability in the foam manufacturing process?
A: Foam is categorized by its nominal PPI (pores per inch) and its relative density. There is about +/-1 percentage point in relative density variation across each large block of starting material, so when smaller pieces are machined out, the final density can vary about +/-1 percentage point. So if we are targeting 7% density we would typically specify an acceptable range of 6-8%. The PPI value is also nominal, and there is some natural variability in actual pore size. This is a bit harder to quantify since rigorous measurements based on CT data are still being accumulated. However, pressure drop, which is a strong function of pore size, varies on the order of +/-10% if using a consistent measurement setup. At ERG we continue to build our own data sets, and we have yet to have any surprises between predicted and as-built performance.
Q: How is foam applied to 2 Phase cooling?
A: In pool boiling applications, a layer of foam brazed or soldered to the heated surface can greatly increase the surface area for boiling, while also disrupting vapor film formation and thereby extending the critical heat flux limit. These advantages are amplified even more in flow boiling applications, such as pumped two-phase, where active pumping allows for the use of smaller pore sizes and compressed foams.
Q: porosity range of metal foam
A: The porosity can be controlled by thickening the ligaments and/or mechanically compressing the foam. For uncompressed foam, the porosity can range from 0.97 – 0.86 (3-14% metal volume fraction, a.k.a. relative density). Mechanical compression can further extend the porosity range to 0.5 (50% relative density).
Q: Can the foam be made out of two different metals? Or, with one metal filling the voids in a different parent metal foam?
A: Metal foam can be plated to achieve a bi-metallic foam. Aluminum foam with nickel plating is a common combination. As for filling the voids of a parent metal foam with another alloy – we have not seen such an application, but as long as the melting point of the parent foam is higher than the infusing foam, it should be possible. The process would be the same as for making a PCM heat exchanger, where the phase change material (PCM) is infused into the foam in a molten state and is subsequently solidified.
Q: Do you see any applications in EV battery packs?
A: There are numerous studies in the open literature that demonstrate the use of metal foam based PCM heat exchangers for thermal management of Li-ion batteries in EV applications. The PCM approach is attractive because it can achieve passive cooling for limited duration trips (e.g. 30 min commute, followed by a period of inactivity during which the PCM can cool and re-solidify). The high thermal conductivity of metal foam can greatly reduce thermal gradients in such PCM solutions.
Q: Both compressed and uncompressed foam shows significantly lower thermal conductivity comparing to that of solid material in the same material. Where is this difference coming from?
A: When we talk about the thermal conductivity of foam, we are considering the effective thermal conductivity of the bulk material, not the parent alloy. When you have a porous material that is mostly empty space, the effective conductivity will be lower. The effective conductivity of finned cores is also lower than the parent material. In both cases, if you look locally at the thermal conductivity within each ligament or fin, then that would be the same as the aluminum or copper they were produced from.
Q: Can foam be used as a straight heatsink without forced air?
A: Yes – metal foam has been used successfully in natural convection heat sinks, especially in situations where there are challenging constraints on the form factor, such as a need for a very low profile heat sink. Typical applications for these heat sinks are cooling LED lights. Our ERG Overview video has a case study on an LED heatsink used in drone lighting.
Q: What considerations should be made to minimize fabrication and layup efforts (and therefore part cost), when it comes to brazed interfaces?
A: ERG has been successfully brazing foam to metal housings for decades and has developed techniques to ensure good braze contact and a strong braze joint. In fact, brazing with foam can be easier to braze than traditional fins because the foam is compressible; what this allows is the foam ligaments to be compressed into the housing without bending the entire fin out of the way which can happen if the tolerance on the fins and the housing are off even slightly. Metal foam can also be locally compressed on the surface, allowing for more surface area at the braze joint.
Q: What is the recommendation for filtering of liquids to avoid clogging in the metal foam?
A: In general we recommend filtering down to 1/2 the average pore size. There is a natural distribution of pore sizes, and 50% below the mean corresponds to about two standard deviations, so >95% of the pores will be bigger than that and therefore pass those particle sizes without an issue. The mean pore size for a 40 PPI foam is ~1mm, so we would recommend at least a 500-micron filter for that foam. If the foam is compressed, that will require more stringent filtering requirements. This can be assessed on a case by case basis, considering both the compression level and the direction of compression.
Q: What are some examples of multifunctional applications utilizing metal foams for heat exchangers and something else?
A: This is a great question and helps highlight some advantages of metal foam’s multifunction capabilities. We’ve found that metal foam can be applied at the system level to perform several functions at once. This presentation was focused on heat transfer, but metal foam has a lot of other applications for designs. For example, the metal foam could be applied as a structural element that is also a heat sink and further offers some EMI shielding, vibration dampening, and impact attenuation capabilities. We’d be happy to talk more about these capabilities in detail.
Q: How does thermal performance compare with normal micro-channels?
A: In single-phase flow, metal foam outperforms virtually any competing technology, especially in situations where pressure drop is a secondary concern. A quantitative comparison would depend on the specific microchannels being compared with, as well as the geometry and pressure drop constrains. In two-phase flow (where microchannels are most commonly used), there is still a dearth of test data using metal foam, and at ERG we are actively pursuing two-phase testing to help fill that knowledge gap. We anticipate that the combination of extremely high surface area and small hydraulic diameter will be highly advantageous in two-phase cooling applications.
Q: What is the range of the pour size?
A: For aluminum and copper, the common pore sizes are 5, 10, 20, and 40 PPI. The PPI (pores per inch) is a nominal value, and actual pore sizes range from 1-2.5 mm.
Q: What is the min pour size for heat transfer application?
A: Most metal foam heat transfer applications use either aluminum or copper foams, which range from 5 to 40 PPI. These are nominal Pores Per Inch (PPI) values, and the actual pore size for 40 PPI foam is ~1 mm in diameter. However, the effective pore size can be reduced dramatically by compressing the foam by a factor or 4-6X, while simultaneously increasing surface area per unit volume proportionally. In certain specialized applications, we have also used ceramic foams (e.g. SiC) for heat transfer, where > 200 PPI is possible.
Q: Can you share data on copper foam ligament shapes and thickness for various PPI and porosities?
A: The ligament shape and thickness measurements from aluminum foam are applicable to copper foam as well (for the same PPI and relative density). There is a fair amount of published literature documenting aluminum foam morphology, both using optical and CT measurements. We would be happy to reference some of that literature, and we may be able to share some in-house data as well, depending on your application and need.
Q: Slide 11: Are similar data for cell breadth and pore breadth available at other PPI for copper and aluminum?
A: The data on slide 11 in the presentation is applicable to both copper and aluminum foams. While more data exists, the range of PPI is still limited to the 5-40 PPI that ERG offers for aluminum and copper foams.
Q: How we control the pore size?
A: This is part of the proprietary process for producing metal foam. However, it is useful to point out that pore size and cell size are controlled independently from ligament thickness and porosity (i.e. void or metal volume fraction), which is a significant and unique advantage of Duocel metal foam.
Q: Can these foams be used as wicks in heat pipes and vapor chambers?
A: Although this has not been verified experimentally, compressed foams are expected to develop the capillary forces necessary for them to serve as wicks. While there could be some packaging challenges in the heat pipe form factor, we think vapor chambers are a very promising application.
Q: Have you created or analyzed options where metal foam is used as a “thicker” core portion for vapor chambers at the vaporizer and condenser with other means to move liquid between the ends? Does this configuration seem to have any merit at first glance?
A: If I understand your question correctly, you are asking if the foam can serve as a permeable structural element separating the hot and cold faces of the vapor chamber, with the wicking function performed by another porous material. We believe this would indeed be a very good application for metal foam. While closed-cell metal foam is an excellent sandwich core material for purely structural applications, the permeable nature of open-celled metal foam would allow it to serve the same structural function while also allowing free flow of the liquid/vapor mixture throughout. Moreover, there could be significant manufacturability advantages in using foam, as it can be cut, machined, formed, and bonded with great ease and flexibility.
Q: How would you model this in CFD?
A: Resolving the actual ligament structure with mesh would be prohibitive for anything other than fundamental research-oriented CFD efforts. However, a very practical and efficient way to do design-oriented CFD modeling is to use a porous media approach, where the foam is modeled as a permeable block of material with user-specified properties, such as bulk thermal conductivity, permeability, specific surface area, and heat transfer coefficient. These properties must be obtained experimentally through fundamental testing, but once measured, they can be applied to the analysis of any new designs with a high degree of confidence. An alternative approach for obtaining flow and heat transfer properties is to run “numerical experiments” using high fidelity CFD (such as LES) on a periodic “unit cell” geometry resolved with mesh. In this case, the main features of the ligaments and cells are retained while the structure is simplified and idealized (e.g. as a tetrakaidecahedron) to allow for the use of periodic boundary conditions. This is just another way to obtain fundamental properties, but those would still be applied to a porous media model when solving macro-scale design problems (e.g. a full cold plate or heat exchanger).