How metal foams are made
Follow these links to see short descriptions of all the processes used to make metal foams - and their results...

1 - Exploiting curious alloy systems
> Heating a metal with another (volatile) one
> Cooling a molten metal supersaturated with hydrogen (a.k.a. GASAR process)

2 - Joining loose material together
> Sintering together metal powders or fibres
> Sintering of hollow metal spheres
> Mixing a metal with a soluble powder, then dissolving the powder

3 - Using some mould of a cell structure
      (Small quantities, high quality)
> Making a soluble mould of the foam structure
> Mixing hollow spheres into molten metal
> Reproducing a plastic foam using several moulding steps
> Deposition of metal from the gas phase onto a disposable preform of the foam shape

4 - Industrial melt-based processes
      (Used for large scale production)
> Directly injecting gases to molten metals with increased viscosity
> Improving control over the cell structure by using a foaming agent
> The FORMGRIP and FOAMCARP processes - a two-step system involving a foamable precursor (both developed in Cambridge!)

© Dave Curran  . This is a general introduction to metal foams, principally aimed at people new to the subject. The content is not necessarily accurate or up to date! To see more in-depth content, look at the published papers...

A closed-cell foam (produced in Cambridge) in cross section. The cells are about 3mm wide.

A closed-cell foam with porosity of about 80% floating on water.

Fracture surface of a GASAR foam, showing the long parallel pores in cross section (here about 0.1 mm wide).

The main production methods
The ways of making foamed metals can be divided into four broad categories. The first is that of foams which are really just combinations of materials which happen to give a lot of open space - for example metal powders or fibres which have been compressed together. The foams made by these methods tend to be of fairly poor quality. A far more reliable type of process is infiltration, where some sort of solid mould of pore shapes is made and molten metal is poured in between them, with subsequent removal of the mould. Because you can spend as long as you like making a perfect mould, this sort of method tends to give highly uniform open-cell foam structures. The downside of these processes is that they are generally expensive and complicated, and only suitable for making fairly small quantities of foam. Another set of processes is based on powder processing - powdered metal and foaming agents (i.e. some sort of chemical which will react to give off a gas) are mixed together and compressed; subsequent heating the mixture above the metal melting temperature produces a foam.

By far the cheapest type of process is melt-route processing, where some sort of gas is injected into a viscous molten metal, and is trapped in to form pores, with the metal then cooling and solidifying to form a foam. Although cheap and relatively simple; consistently obtaining a cell structure with a reasonably high quality is difficult, which has given rise to various adaptations to control the cells more closely.

There are other methods, though they tend to produce foams with a lower porosity or of lower quality. For example the GASAR process (invented at the Ukraine State Metallurgical Academy in 1993) involves rapid cooling of metal saturated with hydrogen. When the metal solidifies, the hydrogen is rejected and forms pores. Because the cooling happens in a particular direction, the pores are very long and parallel in the direction of cooling (and it's arguable that these meterials are honeycombs rather than foams).

To find out more detail about these methods (including diagrams and pictures) follow the links to different processes near the top of this page.

Infiltration: Open-cell foams 1: Salt moulds

Click here for visitor statistics

The first metal foams were produced by filling a graphite-coated mould with rough rock salt (the sort that is put in grinders, with grains up to 4mm wide), and heating it so that the grains fused together. This produced a solid preform of connected salt grains. Molten aluminium was then poured into this, while shaking the mould to make sure it spread through the preform, and the mould was then cooled. The salt was then dissolved in water, leaving a network of aluminium struts (in the shape of the gaps between the salt grains). Follow this link for a full description opf the process.

1: Large grained salt is used as a starter material	2: It's heated under pressure to make the grains fuse together	3: Molten aluminium is forced into the gaps between grains	4: It's cooled, and the salt is dissolved out in water

This method was independently rediscovered 30 years later at the Swiss Federal Institute of Technology (EPFL), but with several improvements - in particular, the molten metal is now added to the salt under vacuum, meaning that the metal spreads more reliably and far smaller salt grains can be used (foams have been produced with cells as small as 0.3 mm wide). The quality of the foams produced using this method is very high, but it is difficult to scale up the process.

2: Plaster Moulds
A variant on this technique is to use another foam as a base to make a mould. The most common route is to infiltrate an open-cell plastic foam (which can be made fairly easily - the sort of thing used to stuff sofas) with something which can later be dissolved, but which will put up with high temperatures (plaster is the best example of these materials). This produces a solid lump with plastic struts encased in plaster. Heating it to high temperatures makes the plastic decompose - to reveal a network of hollow channels throughout the block of plaster. Molten metal is squeezed into this (usually using a combination of vacuum and high pressure), and the plaster is dissolved, to give an open-cell foam with exactly the same cell structure as the original polymer foam. For a diagram & more information click here.

1: An open-cell polymer foam is used as a starting material	2: It's filled with a soluble plaster	3: It's heated - the polymer burns out to leave hollow channels	4: Molten aluminium is squeezed into the hollow channels in the plaster	5: The plaster is dissolved out to give an open-cell foam

3: Electrodeposition
A fairly subtle way of making an open-cell foam is to start with an open-cell plastic foam as in the plaster route, but to expose it to a gas mixture which causes carbon to be deposited on it, and then (because the deposited carbon will conduct electricity) electroplate aluminium onto the surface. It can then be heated to melt the plastic, giving an open-cell foam with hollow struts. Follow this link for a diagram and details.

1: An open-cell polymer foam is used as a starting material	2: It's coated with carbon (from a gas or a liquid)	3: Aluminium is then electroplated on it	4: On heating the polymer burns - leaving hollow aluminium struts

Summary of advantages and disadvantages of infiltration methods:

Advantages: Very close control of the cell size - preforms can be made with good control of grain size, and for (1) even different sized salt grains in different areas are possible. Any metal or alloy can be used (provided it melts below the salt decomposition temperature) - there is no need to adjust the alloy or add particles to make it sufficiently viscous to avoid the foams collapsing during manufacture. The moulds can be close to the final shape required, minimising the need for machining.

Disadvantages: It can take a long time to dissolve the salt or plaster (particularly with small grain sizes or large moulds). These processes are not continuous, and quite complex, so the foams produced are on the expensive side. Foams 1&2 are difficult to machine once the salt/plaster has been removed. Method 3 is very expensive.

Gas evolution: Closed-cell foams
1: The Alcan Process (Alcan, Cymat)
The simplest way to make a closed cell foam is to bubble air into the bottom of a vat of molten aluminium, and skim off and cool the bubbles that form at the top. There are complications - in particular, the metal has to have fine ceramic particles in it to make it more viscous, or the air would simply escape at the top surface without forming bubbles. This process is used by Alcan™ to make foam in large quantities, and produces a low-cost but low-quality foam. For a process diagram click here.

2: The Alporas Process
More complex processes produce better foams, with finer and more uniform pore sizes, by using a solid foaming agent which can be mixed into the aluminium before giving off a gas rather than simply adding the gas directly. This gives more control over the location and size of the pores. Alporas™ foam, produced by the Shinko Wire company in Japan, mixes titanium hydride powder into molten aluminium. The hydride decomposes at the temperature of molten aluminium, to form titaniun (which is dispersed in the aluminium melt) and hydrogen gas. Because the hydride is mixed throughout the aluminium as a powder, it forms pores throughout the structure. Unlike the Cymat™ process, the entire volume of aluminium is foamed at once, rather than skimming a foam off the surface of molten metal. This gives finer pores (they do not have to reach a sufficient size to move up to the surface, and do not coalesce to form larger pores after reaching the surface either) and a more regular structure. As with the Cymat™ foam, the viscosity of the molten aluminium has to be increased - in this case by adding calcium to the metal, which encourages the formation of solid oxides. The cell structure of the Alporas™ foam is of a higher quality, but the foam produced is more expensive. Follow this link for more information about the Alporas process.

3: The Formgrip Process
The foaming process developed in the Composites and Coatings Group is called Formgrip (a.k.a. Foaming Of Reinforced Metal by Gas Release In Precursor). It improves the foam by incorporating an additional processing step: rather than adding the titanium hydride directly to the melt, which limits the time available for mixing, the hydride powder is heated before the addition to form a thin oxide layer. When it is mixed into the molten metal, the oxide layer protects the hydride and delays the evolution of the gas. This means that instead of foaming immediately, the hydride-and-aluminium mix can be cooled to form a solid precursor material, which can be cut up and stored.

In a second foaming step, a piece of the precursor is put in a mould (of any shape) and heated just above the melting temperature of the aluminium alloy. The hydride gradually decomposes to form a foam, which expands the precursor to fill the mould. The advantage of this method is that a longer stirring time is possible: because the hydride does not foam immediately when it is mixed with the aluminium, fine particles can be stirred until the distribution is uniform (which will in turn produce a more uniform structure of foam cells). This method is also capable of manufacturing final-shape castings: because no access is needed for a stirrer during the foaming step, pieces of the precursor can be placed in any shape of mould, or even be used to fill hollow cavities inside metal structures. For more information click here, or see Vlado Gergely's foam pages.

One area of current work concerns the ceramic particles (usually SiC) added to increase the viscosity: How important is the particle size and shape? What effect does it have on the final cell structure - and metal microstructure? What volume fraction of particles is necessary? Is there a cheaper or better alternative to the expensive and complicated process of adding ceramic particles to the melt? What effect do the solid particles have on the strength of the final foam? Right: Typical microstructure of an Al-Si foam. The white is aluminium, black lines are Si plates, and the grey shapes are SiC particles added to increase the viscosity.

For more information on the development of the Formgrip process, and experimental results concerning the mechanical properties of the Formgrip foam, see Vlado Gergely's page (on this website).