Casting Protocols for the Production of Open Cell Aluminum Foams by the Replication Technique and the Effect on Porosity

The basic method described here has been used in different forms by other researchers. Some of the key variants that allow foams of different types to be created are discussed. In characterizing these foams we have measured the porosity, as this is a quick and easy assessment to make, but characterization of other structural characteristics, such as pore size, specific surface area or strut thickness might be required to obtain a full understanding of foam characteristics for different applications. In practice, for production of foams by replication, the pore size is well-controlled by the particle size of the NaCl used, and links between this, the density and the other structural characteristics can be made.

Densified preforms

In this present contribution we have described a NaCl preform that is made by tipping NaCl grains into a chamber. While, as discussed, some degree of control over the density can be achieved by vibrating the sample, the range accessible remains rather restricted, due to the limited range of packing fractions of the NaCl that can be achieved. In order to produce foams of higher porosity, the density of the preform can be increased by compacting it mechanically (for example in Cold Isostatic Pressing to preserve isotropic structure), or by sintering where densification is driven by surface area reduction. Both of these methods would be expected to be more effective for smaller NaCl particle size (sub millimeter), as smaller grains are less liable to crack and have a larger surface area to volume ratio. As the pore size used in the experiments described in this paper is larger, and either process would require additional equipment, with the focus on a simple and easily implemented process, they have not been used.

Shaped preforms

In Goodall and Mortensen¹⁴ a method is introduced to control the pore size and shape further than is possible using single NaCl grains. In this method fine NaCl powder is mixed with a binder (for simplicity, flour and water can be used), then shaped into the desired form before a heat treatment is used to thermally remove the binder. Although not experimentally complex, this method has not been used in our experiments as it is not essential to produce a foam and requires somewhat more precise pressure control to ensure the fine scale porosity in the preform itself is not infiltrated.

Alternative preforms to NaCl

Although NaCl displays many desirable features as a preform material (including relatively high melting temperature, high degree of solubility in water and low toxicity and cost), it is not always suitable. One particular case is when higher melting point metals are to be processed, and in this situation it can be replaced with other materials, such as sodium aluminate¹⁶. These materials improve the temperature capability, but are generally more expensive and challenging to dissolve, and are not required to process foams from relatively low melting point metals, such as aluminum, the most common metal from which foams are made.

Cold wall / partial cold wall pressure chambers

To infiltrate metal into finer particle size preforms, to produce smaller pore size foams, higher pressures will be required. The test rig described in this work is suitable for use up to 6 atm pressure, but as the pressure is increased the probability of leaks from the seals goes up. This can be solved by alternative designs of pressure chamber, where the sealed regions are separated from the heated zone, usually protected by cooling water. While the capability of such equipment is increased over that described here, the design and fabrication is significantly more complex, and so has not been implemented in this version.

Characteristics of this embodiment of the process

While casting the molten aluminum into the bar mold, a pipe defect will form at the top, due to solidification shrinkage. The best results are obtained with fully solid ingots, so this part should be discarded or recycled.

It has been found that in order for the process to work, the seal between the molten metal and the mold wall must be good (otherwise argon will bypass the metal and no infiltration will occur). For this reason when applying infiltration pressures of 3 bars or higher, the best results are obtained with a large amount of aluminum, sufficient to fill the mold, even if the aim is obtaining short foams, as this increases the pressure of the liquid metal around the mold at the top of the preform and improves the seal. A small gap of half a centimeter has been found to be the ideal height between the aluminum piece and the mold lid for the current equipment. For infiltration pressures of 2.5 bars or less the size of the gap is irrelevant, the only amount of aluminum needed is the enough to completely fill the preform.

When tightening the nuts on the studs use a star pattern (tightening opposite pairs in a stepwise fashion) to ensure that the pressure around the gasket is even and a seal is obtained. To avoid damage to the valves on closing, this is always done manually.

Occasionally, there can be defects or regions of poor infiltration. These are most likely to form at the bottom, where the molten metal has furthest travel, or at the top, near the interface with the dense metal. Therefore the most consistent part of the sample is in the center of the region occupied by the NaCl preform. The top and bottom parts of the foam may be cut away and discarded. Whenever it is required to cut the foam to produce a sample, it is best to do so with the NaCl still present in it. If the cuts are made after leaching, wherever the cut is made, it will damage and block the structure of the foam. Where cutting the sample after leaching is required, a successful method is to use a non-loading technique such as Electro-Discharge Machining (EDM, also called spark erosion).

There are many variables in the process that can be modified for different effect, but in order to change the porosity the most suitable control variables are either the preform density or the infiltration pressure used.

The purpose of using different protocols (A, B, C and D) is to produce foams with different porosities, from 61% to 77%. Applying Protocol A will produce samples with 63% porosity on average; Protocol B produces samples with 66% porosity; Protocol C produces samples with 70% porosity and Protocol D produces samples with 76% porosity. By adding the fine NaCl at the bottom of the mold in Protocols B, C and D it creates a refuge for air trapped in the preform during infiltration if the evacuation of the chamber is not perfect. The NaCl being much finer will resist infiltration by the aluminum until higher pressures are reached, ensuring that the preform is completely infiltrated. Without this any air present will be compressed, not eliminated and unwanted additional porosity will be present, most likely as uninfiltrated regions. Protocols C and D were developed to allow infiltration to be achieved with much lower pressures. For the samples represented in Figure 6 a different particle size preform was used, it can be noted that this change does not have a significant effect when compared to the protocol used.

By not using the bottom gasket in Protocols C and D a small flow of gas through the bottom of the mold is possible, meaning that trapped gas from the preform can be evacuated without having to be compressed to higher pressures. If this was done without the fine NaCl then aluminum may also be forced out, but as this layer resists penetration by the liquid aluminum at the pressures applied it will prevent aluminum escape.

In Protocol D, by vibrating the preform, a higher porosity foam can be achieved; roughly 9–10% more porous when compared to Protocol C. This happens since the NaCl grains in the preform are closer together, leaving less space to be filled by the aluminum. The ceramic sheet is added in Protocol D to prevent the fine NaCl to mix with the infiltration NaCl during vibration, no significant effect was found in the final products when adding the ceramic sheet to Protocol C.

The principal limitation for the foam processing technique described is the porosity of the foams; the lowest achieved so far with the rig and protocols described here are around 61% and the highest close to 77%. However it is a cheap and easy to use technique compared to more complicated and expensive methods such as investment casting, sintering or additive manufacturing. Another limitation is the metals that can be used; any metal having a melting point too close or above the NaCl melting point (801 °C) cannot be infiltrated with this preform. Aluminum, magnesium and tin have been processed using this technique.

Equipment and several successful protocols for the production of aluminum foams are presented in detail. Using this method it is possible to create aluminum open celled foams with porosities of 61–77% (corresponding to density in the range of 1,053 to 621 kg/m³) and with pore sizes in the range 1 to 2.36 mm diameter. Furthermore, it is known that with variations in the conditions used, some of them relatively minor, these ranges can be extended significantly, and other variables, such as pore shape can be changed. The replication technique is highly suitable for research laboratory use for metal foam production.