Composite Metal Foam
( 17 October 2008 )
Foamy Invention Could Save Energy and Lives
by Nate DeGraff,
North Carolina State University Engineering Communications
Dr. Afsaneh Rabiei has invented an ultra-strong and lightweight composite metal foam. Credit: College of Engineering at North Carolina State University
Doctoral student Lakshmi J. Vendra (left) and senior Judy Brown (right) help Dr. Rabiei with her research. Credit: College of Engineering at North Carolina State University
She's fluent in three languages. She's studied everywhere from Tehran to Tokyo to Cambridge, Mass. And she's invented a space-age material so light and strong that it could revolutionize everything from vehicle bumpers to armor to biomedical devices.
The ultra-high-strength composite metal foam created by Afsaneh Rabiei is a highlight of a well-traveled career during which the researcher has tried to learn everything she can about advanced materials. The result: a brand new material that can save energy and lives.
"Basically, it is a new material for all sorts of safety devices," said Rabiei, associate professor of mechanical and aerospace engineering at North Carolina State University.
Rabiei's invention isn't the first metal foam, but she says it's the strongest. The main weakness of existing metal foams is the varying sizes of their cells --- tiny pockets of space inside the material. Instead, Rabiei used cells of standard sizes and combined them with a metallic matrix to support the cell walls.
A light, strong material
Rabiei has been working on high-performance materials for more than 20 years. She studied materials science and engineering while getting her undergraduate degree at Sharif University of Technology in Tehran, Iran, in 1986. After spending some time in industry --- getting more experience in casting, welding and materials testing --- she returned to academia and obtained her Ph.D. in 1997 from The University of Tokyo. Her work on the metal foams began when she was a post-doctoral researcher at Harvard University.
Along the way, she's become fluent in English, Japanese and Persian and written four books and dozens of other publications. The National Science Foundation awarded her a CAREER award in 2003, and her goal with the funding was to develop a light, strong material that could be used saving lives and energy in the aerospace, medical, automotive and other industries.
After five years of work --- with a group of her students --- she has the results. Rough traffic accident calculations show that by inserting two pieces of her composite metal foam behind the bumper of a car traveling 28 mph, the impact would feel the same to passengers as impact traveling at only 5 mph.
Applications: airplanes, boats
The results are most striking when the material is tested in a lab. The test itself is exciting: a high-powered machine smashes a piece of steel foam straight down into the base plate of the machine, and then does the same thing with a piece of bulk steel.
When she examines the base plates under both samples, there’s a clear indentation left under the bulk steel sample, while the plate under the foam shows no indentation. The test shows how the foam absorbed the energy and protected the plate, while the steel simply transferred it to the base plate with no protection.
And since the bulk steel is three times heavier than the steel foam, it's easy to see how the foam could attract car manufacturers looking for a bumper that will improve safety and gas mileage.
Rabiei sees plenty of uses for her invention, including in airplanes, boats, and structures that need impact protection with maintaining low weight. It's this high strength-to-density ratio --- defining a material that's both strong and light --- that makes Rabiei's foam unique.
"This material showed a much higher strength-to-density ratio than any metal foam that has ever been reported," she said.
High Performance Composite Metal Foam
by Terry Bray
Researchers at NCSU are joining forces with innovative companies to introduce new and enhanced products to the marketplace. Research and the commercialization of emerging technologies are key synergies of NCSU's mission. In support of that mission, the Office of Technology Transfer (OTT) promotes collaboration with industry partners to move university discoveries from the laboratory to the marketplace. We are pleased to promote and foster beneficial relationships between academia and industry. NCSU is currently seeking an industry partner to license a series of new methods for creating high-strength, ultra-light composite metallic foams that show 5 to 6 times greater strength to density ratio and over 7 times higher energy absorption than that of currently available metallic foams.
Researchers at NC State have developed, processed, and tested a new high-strength ultra-light material that combines the advantages of metal matrix composites with metallic foams. Dr. Afsaneh Rabiei has produced a new generation of metal foams showing 5 to 6 times greater strength to density ratio and over 7 times higher energy absorption than that of currently available metallic foams. As a result, the energy absorption of these materials is estimated to be over 80 times greater than the bulk material from which the foam is made. Dr. Rabiei was interested in maintaining the advantages of metallic foams (excellent rigidity/ weight ratio, durability, isotropic absorption of energy at low and constant stress) while improving the mechanical properties under cyclic compression loading. The performance advantages of this metal foam are based on improving foam cell structure and reinforcing the cells with a metallic matrix. The resulting novel, closed-cell, metallic foam composite is made from preform hollow metallic spheres and exhibits a strength of over 130 MPa in compression. The densification for the new foam occurs at strains of approximately 50-65%.
- Improved strength (over 5 to 6 times higher strength to density ratio than that of currently
- Greater energy absorption (at 50% strain over 65 MJ/m3, which is at least 7 times greater than
that of other foams made from the same materials through different techniques)
- Greater uniformity of cell structure, leading to controlled, predictable product deformation under loading
- Potential for further strengthening the foam through adding reinforcements into the matrix
- Higher capability for manufacturing through machining, welding, and joining without the need for adding sandwich panels.
- Vehicle crumple zones/collision absorbers
- Lightweight aerospace vehicle structures
- Fire retardant structures
- Boat hulls
- Military vehicle/structure armor
- Oil-well drilling platforms
- Sound absorptive material
- Volumetric thermal or electric conductive material applications
- Electromagnetic shielding material
- Implantable biomedical devices
- Structural shock absorbers to protect against earthquake vibration
Dr. Afsaneh Rabiei serves as an Assistant Professor of Mechanical and Aerospace Engineering and as anAssociate Faculty Member of Biomedical Engineering at NC State since Aug. 2000. Dr. Afsaneh Rabiei received her Ph.D. in advanced materials at Research Center for Advanced Science and Technology, The University of Tokyo, Japan in 1997 within the area of mechanics and nondestructive evaluation of metal matrix composites. Her prior working experience includes over 8 years of industrial experiences in materials science and processing including casting, welding and nondestructive testing. She received her B.S. from the Department of Metallurgy and Material Science at Sharif University of Technology in Tehran, Iran in 1986. She worked at Harvard University as a post doctoral researcher from 1997 until 2000.
MATERIALS TRANSACTIONS 47 (9): 2148-2153 SEP 2006
Rabiei A, Vendra L, Reese N, et al.
Processing and characterization of a new composite metal foam http://apps.isiknowledge.com/WoS/CIW.cgi?SID=1CeADGoK8IcMM9LLoLc&Func=
MATERIALS SCIENCE AND ENGINEERING A-STRUCTURAL MATERIALS PROPERTIES MICROSTRUCTURE AND PROCESSING 465 (1-2): 59-67 SEP 15 2007
Vendra LJ, Rabiei A
A study on aluminum-steel composite metal foam processed by casting http://apps.isiknowledge.com/WoS/CIW.cgi?SID=1CeADGoK8IcMM9LLoLc&Func=
MATERIALS SCIENCE AND ENGINEERING A-STRUCTURAL MATERIALS PROPERTIES MICROSTRUCTURE AND PROCESSING 404 (1-2): 159-164 SEP 15 2005
Rabiei A, O'Neill AT
A study on processing of a composite metal foam via casting http://apps.isiknowledge.com/WoS/CIW.cgi?SID=1CeADGoK8IcMM9LLoLc&Func=
Dr. Rabiei's patent:
Technical Insights division of Frost & Sullivan (www.ti.frost.com) has an article in their High Tech Materials Alert on Dr. Rabiei's new material published in their October 2006 issue.
journal Materials Science Forum Vols. 539-543, March 2007, pp. 1868-1873
Afsaneh Rabiei, Brian Neville, Nick Reese and Lakshmi Vendra, "New Composite Metal Foams Under Compressive Cyclic Loadings"
Materials Transaction, Japan Institute of Metals 47-49, Sep. 2006,
A. Rabiei, L. Vendra, N. Reese, N. Young, B.P. Neville "processing and characterization of a new composite metal foam"
BP Neville, A. Rabiei, "Composite Metal Foams processed through Powder Metallurgy", Materials and Design, in press http://dx.doi.org/10.1016/j.matdes.2007.01.026
Materials Science and Technology 2005
A. Rabiei, B. P. Neville, N. Reese and L. Vendra "Mechanical Properties of New Composite Metal Foams Under Compressive Monotonic and Fatigue Loadings",
MetFoam, 2005, pp 369- 374
A. Rabiei, L. Vendra, N. Reese, N. Young, B.P. Neville
"processing and characterization of a new composite metal foam", fourth International conference on porous metals and metal foaming technology, Japan,
A. Rabiei, Processing and Development a New Ultra-Light High-Strength Material, NSF-DMII awardee conference, St. Louis, Missouri, 2006
A. Rabiei, Processing and Development a New Ultra-Light High-Strength Material, NSF-DMII awardee conference, Tampe Arizona, 2005 5. A. Rabiei, Processing and Development a New Ultra-Light High- Strength Material, NSF-DMII awardee conference, Tampe Arizona, 2007
A. Rabiei, A .O'Neil and B. Neville, "Processing and Development of a New High Strength Metal Foam", MRS fall meeting 2004, NN.11.4, vol 851, pp517-526
A. Rabiei, L. Vendra, "Elastic and Plastic Behavior of Composite Metal Foams under Compressive Loading", Metfoam 2007 in press AfsanehMORE INFORMATION:
Office of Technology Transfer
920 Main Campus Drive
Campus Box 8210
Raleigh, NC 27695-8210
Composite Metal Foam and Methods of Preparation Thereof
Classification: - international: B22F3/11; B22F3/11 - European: B22F3/11B4
Abstract --- The present invention is directed to composite metal foams comprising hollow metallic spheres and a solid metal matrix. The composite metal foams show high strength, particularly in comparison to previous metal foams, while maintaining a favorable strength to density ratio. The composite metal foams can be prepared by various techniques, such as powder metallurgy and casting.
FIELD OF THE INVENTION
 The present invention is directed to composite metal foams and methods of preparation thereof. The composite metal foams generally comprise hollow metallic spheres and a solid metal matrix.
 Metallic foams are a class of materials with very low densities and novel mechanical, thermal, electrical, and acoustic properties. In comparison to conventional solids and polymer foams, metal foams are light weight, recyclable, and non-toxic. Particularly, metal foams offer high specific stiffness, high strength, enhanced energy absorption, sound and vibration dampening, and tolerance to high temperatures. Furthermore, by altering the size and shape of the cells in metal foams, mechanical properties of the foam can be engineered to meet the demands of a wide range of applications.
 Various methods are presently known in the art for preparing metallic foams. According to one method, metal powders are compacted together with suitable blowing agents, and the compressed bodies are heated above the solidus temperature of the metal matrix and the decomposition temperature of the blowing agent to generate gas evolution within the metal. Such "self-expanding foams" can also be prepared by stirring the blowing agents directly into metal melts. Metallic foams can also be prepared as "blown foams" by dissolving or injecting blowing gases into metal melts. Metallic foams can also be prepared by methods wherein gasses or gas-forming chemicals are not used. For example, metal melts can be caused to infiltrate porous bodies which are later removed after solidification of the melt, leaving pores within the solidified material.
 Metallic foams have been shown to experience fatigue degradation in response to both tension and compression. Plastic deformation under cyclic loading occurs preferentially within deformation bands, until the densification strain has been reached. The bands generally form at large cells in the ensemble, mainly because known processes for producing these materials do not facilitate formation in a uniform manner. Such large cells develop plastically buckled membranes that experience large strains upon further cycling and will lead to cracking and rapid cyclic straining. As a result, the performance of existing foams has not been promising due to strong variations in their cell structure (see Y. Sugimura, J. Meyer, M. Y. He, H. Bart-Smith, J. Grenstedt, & A. G. Evans, "On the Mechanical Performance of Closed Cell Al Alloy Foams", Acta Materialia, 45(12), pp. 5245-5259).
 In the production of closed cell metallic foams, one obstacle is the inability to finely control cell size, shape, and distribution. This makes it difficult to create a consistently reproducible material where the properties are known with predictable failure. One method for creating a uniform closed cell metallic foam is to use prefabricated spheres of a known size distribution and join them together into a solid form, such as through sintering of the spheres, thereby forming a closed cell hollow sphere foam (HSF).
 Hollow metal spheres, such as those available from Fraunhofer Institute for Manufacturing and Advanced Materials (Dresden, Germany), can be prepared by coating expanded plastic spheres (e.g., polystyrene) with a powdered metal suspension. The spheres are then placed into a mold and are heated to pyrolize the polystyrene and powder binder, and to sinter the metal powder into a dense, solid shell. Metal foams previously prepared through sintering of such hollow metal spheres are plagued by low strength. Foams prepared by sintering metal spheres made of stainless steel, when under compression, have been shown to undergo densification at a stress of approximately 2 to 7 MPa, reaching a strain of over 60%.
 Accordingly, it is desirable to have metallic foams wherein cell size, shape, and distribution are controllable, and wherein high strength is exhibited. Such goals are achieved by the composite metal foams of the present invention and the methods of preparation thereof.
SUMMARY OF THE INVENTION
 The present invention is a composite metallic foam comprising hollow metal spheres and a solid metal matrix. The foam exhibits low density and high strength. Generally, the composite metallic foam is prepared by filling in the spaces around the hollow metallic spheres, thus creating a solid matrix. Such preparation can be by various methods, including powder metallurgy techniques and casting techniques. The composite metallic foams of the invention have unique properties that provide use in multiple applications, such as marine structures, space vehicles, automobiles, and buildings. The foams are particularly useful in applications where weight is critical and vibration damping, as well as energy absorption, are useful, such as blast panels for military applications and crumple zones for automotive crash protection. The application of the foams can also be extended into biomedical engineering as medical implants and even to civil engineering for earthquake protection in heavy structures.
 The composite metal foams of the invention, partly due to their controlled porosity (through use of preformed hollow metallic pieces) and foam cell wall support (through addition of a metal matrix surrounding the hollow metallic pieces), exhibit highly improved mechanical properties, particularly under compression loading. Accordingly, the strength of the inventive composite metal foams is many times higher than other metallic hollow sphere foams. Furthermore, the energy absorption of the inventive foams is much greater than the bulk material used in the foams (on the order of 30 times to 70 times greater), while the inventive foams also maintain a density well below that of the bulk materials.
 In one aspect of the invention, there is provided a composite metal foam comprising a plurality of hollow pieces (preferably hollow metallic pieces) and a metal matrix generally surrounding the hollow pieces. The hollow pieces and the matrix can be comprised of the same or different materials. In one embodiment, the hollow pieces are metallic spheres comprising steel, and the metal matrix comprises steel. In another embodiment, the metal matrix comprises aluminum, while the hollow spheres comprise steel.
 According to another aspect of the invention, there is provided a method of preparing a composite metallic foam comprising placing a plurality of hollow metallic pieces in a mold and filling the spaces around the hollow metallic pieces with a metal matrix-forming material. The method can be carried out through the use of various techniques, such as powder metallurgy or metal casting.
 In one particular embodiment according to this aspect of the invention, the method comprises the following steps: arranging a plurality of hollow metallic pieces in a mold; filling the spaces around the hollow metallic pieces with a matrix-forming metal powder; and heating the mold to a sintering temperature, thereby forming a solid metal matrix around the hollow metallic pieces. Various packing techniques, such as vibrating the mold according to a specific frequency, or varying frequencies, can be used for maximizing packing density of the metallic pieces within the mold. Further, such techniques can also be used during the step of filling the spaces around the hollow metallic pieces to facilitate movement of the metal powder through the mold and around the hollow metallic pieces.
 The method can further comprise applying pressure to the hollow metallic pieces and the matrix-forming metal powder within the mold, as would commonly be done in powder metallurgy techniques. Such compression within the mold can be carried out for the duration of the sintering step of the method.
 According to another embodiment of the invention, the method comprises the following steps: arranging a plurality of hollow metallic pieces in a mold; casting a matrix-forming molten metal into the mold, thereby filling the spaces around the hollow metallic pieces; and solidifying the liquid metal, thereby forming a metal matrix around the hollow metallic pieces. As noted above, various packing techniques, such as vibrating the mold, can be used for maximizing packing density of the metallic pieces within the mold.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is an optical image providing a cross-sectional view of a 3.7 mm hollow metallic sphere useful according to the present invention;
FIG. 2 is a cross-sectional view of a composite metal foam of the invention comprising hollow steel spheres surrounded by a metal matrix formed by powder metallurgy using steel powder;
FIG. 3 is a detailed cross-sectional view of a composite metal foam formed by powder metallurgy, according to one embodiment of the invention, comprising hollow steel spheres surrounded by a steel matrix;
FIG. 4 is a cross-sectional view of the composite metal foam shown in FIG. 3 providing an even greater detailed view of the metal matrix;
FIG. 5 is a SEM image of the composite metal foam of FIG. 2 showing a cross-section of two steel spheres in contact with each other and the steel matrix filling the spaces around the spheres;
FIG. 6 is another SEM image of the composite metal foam of FIG. 2 showing a cross-section of two spheres not in contact and the steel matrix filling the spaces between and around the spheres;
FIG. 7 is a three-dimensional drawing of a permanent casting mold useful in one embodiment of the invention;
FIG. 8 is a cross-sectional view of a permanent casting mold useful in one embodiment of the invention;
FIG. 9 is a cross-sectional view of a composite metal foam of the invention formed by casting molten aluminum around hollow steel spheres;
FIG. 10 is a SEM image of a cross-section of a composite metal foam of the invention showing an aluminum matrix between two hollow steel spheres;
FIG. 11 is a detailed view of the SEM image from FIG. 10 showing the interface between the aluminum matrix and the steel wall of the hollow sphere;
FIG. 12 is a SEM image of a cross-section of a composite metal foam of the invention formed by casting an aluminum matrix around hollow steel spheres and shows (a) four spheres embedded in the matrix with a visible void at the interface of two spheres, and (b) a detail view of the aluminum matrix showing the different phases present;
FIG. 13 is a cross-sectional optical image of three composite metal foams prepared according to various embodiments of the invention;
FIG. 14 is a chart of the stress-strain curves of composite metal foams according to various embodiments of the invention under monotonic compression;
FIG. 15 shows a stainless steel composite metal foam according to one embodiment of the invention both before and after compression testing with 60% strain;
FIG. 16 is a chart showing a curve of strain versus number of cycles during a compression fatigue test of a cast composite metal foam according to one embodiment of the invention; and
FIG. 17 shows optical images of a cast composite metal, according to one embodiment of the invention, before and during a compression fatigue test.
DETAILED DESCRIPTION OF THE INVENTION
 The invention now will be described more fully hereinafter with reference to the accompanying drawings, in which preferred embodiments of the invention are shown. These embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Indeed, the invention may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will satisfy applicable legal requirements. As used in the specification, and in the appended claims, the singular forms "a", "an", "the", include plural referents unless the context clearly dictates otherwise.
 The composite metallic foam of the present invention combines the advantages of metal matrix composites with the advantages of metallic foams to provide higher strength metallic foams of controlled porosity. The inventive metal foams generally comprise a plurality of hollow metallic pieces and a metal matrix filling the spaces around the metallic pieces.
 Metal matrix composites are generally understood to be metals that are reinforced with various materials. Such materials can include natural or synthetic fibers or particulate matter. Materials particularly useful include boron, silicon carbide, graphite, alumina tungsten, beryllium, titanium, and molybdenum. Fibers may be continuous filaments or discontinuous fibers. Examples of natural discontinuous fibers include hair or whiskers. The reinforcements, of which the above are only non-limiting examples thereof, can be chosen for specific purposes, such as increasing stiffness, strength, heat resistance, and wear resistance. Combining the advantages of metal matrix composites with the advantages of metal foams results in the composite metal foams of the invention, which exhibit increased strength, as well as additional beneficial properties as discussed herein.
 The composite metal foams of the invention comprise hollow pieces. In a particular embodiment of the invention, the hollow pieces are spherical in shape (i.e., "hollow spheres"). While such a shape is particularly useful, the hollow pieces comprising the metal foam can also take on other geometric shapes as could be beneficial for imparting improved properties to the finished composite metal foam. For simplicity, the hollow pieces used in the invention are described herein by the particular spherical embodiment. However, description of the hollow pieces as spheres is not intended to limit the scope of the hollow pieces, which can take on other shapes.
 The hollow spheres used in the composite metal foams of the invention can comprise any material that would be useful for providing strength in an overall composite foam of the invention and can withstand the preparation process, such as powder metallurgy or casting, as described herein. In one preferred embodiment, the hollow spheres are metallic.
 Hollow metallic spheres, according to the invention, can comprise any metal generally recognized as being useful for preparing metal foams, metal matrix composites, or other metal components useful in various industries, such as automotive, aerospace, construction, safety materials, and the like. Particularly useful are metals commonly used in applications where lightweight materials, or materials exhibiting relatively low density, are desirable. For example, the hollow spheres can comprise iron (and alloys thereof), aluminum, titanium, nickel, ceramics, such as alumina and silica carbide, and the like. The metals comprising the hollow spheres can be a single, essentially pure metal; however, the term metal, as used herein to describe the components of the composite metal foams of the invention, can also refer to metal mixtures, including alloys, intermetallic compounds, such as titanium aluminide, and the like. Further, the metals can include trace components as would be recognizable as being beneficial, as well as non-detrimental trace impurities. In one particularly preferred embodiment, the hollow metallic spheres are comprised of steel, such as stainless steel or low carbon steel. The composition of one type of low carbon steel and one type of stainless steel (316L stainless steel) useful in particular embodiments of the invention are provided in Table 1. TABLE-US-00001 TABLE 1 Exemplary Metal Compositions for Hollow Metallic Spheres Type Composition Low Carbon Steel: <0.007% carbon and 0.002% oxygen; remaining balance iron 316L Stainless Steel: 0.03% carbon, 0.3% oxygen, 17% chromium, 13% nickel, 0.9% silicon, 0.2% manganese, 2.2% molybdenum, and remaining balance iron
 The average size of the hollow metallic spheres can vary depending upon the desired physical properties of the finished composite metal foam. Average size of the spheres is generally evaluated in terms of sphere diameter. When considering the physical properties of the finished composite metal foam, though, sphere wall thickness must also be considered. Accordingly, assuming sphere wall thickness remains unchanged, the use of spheres having a greater average diameter would be expected to result in a finished composite metal foam of lower density than if spheres of smaller average diameter are used. The average diameter is also limited by the size and dimensions of the finished composite metal foam. For example, if the desired finished composite metal foam is a metal sheet having a 25 mm thickness, the hollow metallic spheres would necessarily need to have an average diameter of less than 25 mm.
 The hollow metallic spheres used in the invention generally have an average diameter of about 0.5 mm to about 20 mm. Preferably, the spheres have an average diameter of about 1 mm to about 15 mm, more preferably about 1.5 mm to about 10 mm, still more preferably about 2 mm to about 8 mm, most preferably about 2.5 mm to about 6 mm. In one particular embodiment, hollow metallic spheres having an average diameter of about 3 mm to about 4 mm (nominally about 3.7 mm) have been used to prepare the composite metal foam of the invention. Depending upon the desired properties of the composite metal foam, other sphere sizes can also be used.
 As noted above, sphere size is also described by the sphere wall thickness, which similarly affects the physical properties of the finished composite metal foams. For example, assuming sphere average diameter is unchanged, the use of spheres having a lesser average wall thickness would be expected to result in a finished composite metal foam of lower density than if spheres of greater average wall thickness are used. Accordingly, in one embodiment of the invention, it is desirable to minimize wall thickness. If wall thickness is too minimal, though, strength of the finished composite metal foams can be compromised. It is therefore beneficial to use spheres wherein the ratio of wall thickness to average sphere diameter is in a range where density of the finished composite metal foam is minimized but overall strength of the composite metal foam is not appreciably sacrificed.
 The hollow metallic spheres of the invention generally have an average wall thickness that is about 1% to about 30% of the average diameter of the spheres. Preferably, the average sphere wall thickness is about 1% to about 15% of the average sphere diameter, more preferably about 1.5% to about 10%, still more preferably about 2% to about 8%, and most preferably about 2.5% to about 7% of the average sphere diameter. In one particular embodiment, the average sphere wall thickness is about 5% of the average sphere diameter. A cross-section of a hollow metallic sphere, such as useful according to the invention is shown in FIG. 1 (note that the sphere in the figure has not been cut through the diameter of the sphere). As seen in the Figure, the sphere walls have a generally uniform thickness. This is particularly advantageous in that composite metal foams, according to the invention, can be prepared to uniform porosity, said porosity being easily adjustable by use of hollow metallic spheres of a desired average diameter and average wall thickness.
 Preferentially, the percentage and size of porosities in the sphere walls are minimized to increase stability of the spheres during processing of the foams. For example, when casting techniques are used in preparing the composite metal foams, minimizing sphere wall porosity decreases the possibility of the matrix-forming molten metal penetrating the cavities of the spheres. Such penetration should be avoided as filling of the cavities could reduce the overall pore volume of the composite metal foam, unnecessarily increasing the overall density of the foam. In one embodiment, sphere wall porosity is less than about 12%. Preferably, sphere wall porosity is less than about 10%, more preferably less than about 8%, most preferably about 5% or less.
 In addition to the hollow metallic pieces, the composite metal foam of the invention also comprises a matrix surrounding the hollow metallic pieces. The matrix generally comprises a metal, and the type of metal comprising the matrix can be varied depending upon the technique used in preparing the composite metal foam of the invention.
 According to one embodiment, the metal comprising the matrix can be the same metal type comprising the hollow metallic pieces. According to another embodiment, the metal comprising the matrix is a different metal type than that comprising the hollow metallic pieces. Preferably, the metal matrix includes a metal that is generally lightweight but still exhibits good strength attributes. The use of such metals is beneficial for maintaining a high strength to density ratio in the finished composite metal foam of the invention. As before, the metal comprising the metal matrix can be an essentially pure single metal or can be a mixture of metals. In one particular embodiment, the metal matrix comprises steel. In another embodiment, the metal matrix comprises aluminum.
 Matrix composition may at least partially be dependant upon the method of preparation of the composite metal foam. The composite metal foams of the invention can be prepared through various techniques known in the art. While the use of such techniques would not be readily apparent for preparing metal foams, one of skill in the art, with the benefit of the present disclosure, could envision similar techniques which could be used in preparing the composite metal foams of the invention. Such further techniques are also encompassed by the present invention.
 According to one embodiment of the invention, there is provided a method for preparing a composite metal foam by powder metallurgy. According to this method, the hollow metallic spheres are first placed inside a mold. At this point, it should be noted that the composite metal foam can be prepared directly in the final desired shape through use of a mold designed to provide the desired shape. Alternately, the composite metal foam may be prepared as a "stock" piece (e.g., a large rectangle) and then be cut to the desired final shape. The size of the composite metal foam prepared according to this embodiment of the invention is generally limited by the size of the mold.
 The metallic spheres are preferentially arranged in the mold to be in a specific packing arrangement. Desirably, the packing arrangement is such that the metallic spheres are in their most efficient packing density (i.e., most closely packed). As such, the open space between the spheres is minimized, and the number of spheres arranged in the mold is maximized. In this packing arrangement, the porosity of the finished composite metal foam is maximized, which correlates into a minimized density.
 The arrangement of the metallic pieces in the mold can be facilitated through mechanical means, such as vibrating the mold. In embodiments where metallic spheres are used, vibration is particularly useful as the spheres tend to "settle" into a most preferred packing density. For example, such vibration can be performed using an APS Dynamics model 113 shaker and an APS model 114 amplifier with a General Radio 1310-B frequency generator. Vibrating at specific frequencies may be beneficial for facilitating a closest packing density or for minimizing the time necessary to obtain such a packing density. Vibrating time may vary depending upon the size of the mold, the average size of the hollow metallic pieces, the average size of the metal powders, and the frequency of the vibration. Generally, vibrating can be performed for a period of time up to about 12 hours, although longer or shorter time periods may be necessary or sufficient. In one particular embodiment, vibrating is performed for a period of time ranging between about 30 seconds and about 4 hours, preferably about 1 minute to about 3 hours, more preferably about 5 minutes to about 2 hours.
 Similarly, other mechanical means can be used for facilitating packing density. Computer modeling could also be used to determine optimum packing techniques, including establishing sphere sizes most useful for optimum packing densities in light of mold size and shape. Where computer modeling is used, automated packing of the spheres could be beneficial for arranging the spheres in a maximized packing density. Further, in embodiments where hollow metallic pieces having a non-spherical geometry are used, different mechanical means could be used for establishing the most efficient packing density of the metallic pieces given their geometries.
 Once the hollow metallic spheres have been arranged in the mold, a matrix-forming metal powder is introduced into the mold, filling the spaces around the hollow metallic spheres. Again, mechanical means, such as vibrating, can be used to facilitate movement of the metal powder around the hollow metallic spheres, preferentially completely filling any voids within the mold. Multiple rotations of adding powder and applying mechanical means to move the powder into the voids between spheres within the mold could be used to ensure complete filling of the mold. Further, it may be beneficial, particularly when filling large molds, or molds of complex shape, to alternate introduction of the spheres and the matrix forming powder into the mold to ensure complete filling of the spaces between the hollow metallic pieces.
 As previously noted, the metal powder used in the powder metallurgy process can comprise various different metals, the metal being the same metal type or a different metal type as the metal comprising the hollow metallic spheres. In one embodiment, the composite metal foam comprises hollow steel spheres and aluminum powders. In another preferred embodiment, the composite metal foam comprises hollow steel spheres and steel powder. Particular, non-limiting examples of materials useful as a metal matrix, according to certain embodiments of the invention, are 316L stainless steel, Ancorsteel-1000C steel, and aluminum 356 alloy, the compositions of which are provided in Table 2. TABLE-US-00002 TABLE 2 Exemplary Metal Powder Matrix-Forming Compositions Type Composition 316L stainless steel: 0.03% carbon, 0.3% oxygen, 17% chromium, 13% nickel, 0.9% silicon, 0.2% manganese, 2.2% molybdenum, and remaining balance iron Ancorsteel-1000C steel: 0.003% carbon, 0.006% phosphorus, 0.007% sulfur, 0.002% silicon, 0.005% oxygen, 0.003% nickel, 0.02% molybdenum, 0.1% manganese, 0.05% copper, 0.02% chromium, and remaining balance iron Pure Aluminum: 99.9% Al
 Choice of metal powder can depend upon the desired physical properties of the composite metal foam. Further, choice of metal powder can be limited by such characteristics as particle size and flow characteristics. For example, electrostatic interactions can limit the flow of some powder types leading to agglomeration and incomplete filling of the voids between the hollow metallic spheres.
 Choice of metal powder can also be limited by chemical and physical changes in the matrix material brought about by sintering. For example, it is known that the strength of sintered steel increases with increasing carbon content, up to a bout 0.85% carbon (see ASM Metals Handbook, 9.sup.th Edition, Vol. 7, "Powder Metallurgy", American Society for Metals, 1984, which is incorporated herein by reference). Beyond this, a network of free cementite begins to form at the gain boundaries. Additionally, it has been shown that for similar sintering conditions, shrinkage decreases with increasing carbon content up to 8%, at which no shrinkage was noted (see, N. Dautzenberg, Powder Metallurgy International, vol. 12, 1971 and Dautzenberg and Hewing, Powder Metallurgy International, vol. 9, 1977, both of which are incorporated herein by reference).
 Further considerations in choosing the metal matrix-forming powder arise from the possible formation of unsuitable intermetallic compounds during sintering. Such formation can be prevented, to some extent, by controlling sintering conditions. For example, when using an aluminum powder matrix-forming material with hollow steel spheres, diffuision of matrix material into the spheres and the formation of a brittle intermetallic phase may occur, particularly with slow process and prolonged exposure of the combination of iron and aluminum at higher temperatures.
 The metal powder is preferentially of a particle size capable of achieving a favorable packing system for maximizing matrix density. For example, in one embodiment, aluminum powder is used, the powder being a 98% pure mixture of the following components: 75% H-95 Al powder (about 100 micron particle size); 14% H-15 Al powder (about 15 micron particle size); and 11% H-2 Al powder (about 2 micron particle size). Such powders are available commercially from vendors, such as Valimet, Inc. (Stockton, Calif.). A powder composition, such as described above, is close to the ideal 49:7:1 ratio to achieve an optimum trimodal packing system for greater matrix density. In another embodiment, Ancorsteel-1000 C iron powder is used. The powder is sieved to 81.3% -325 mesh (44 micron) powder and 18.7% -400 mesh (37 micron) powder. Ancorsteel-1000 C powder is commercially available from ARC Metals (Ridgway, Pa.). In further embodiments, powders of an essentially uniform particle size, or of various particles sizes, can be used for maximizing matrix density. For example, powders having particle sizes most favorable for achieving an optimum bimodal packing system could also be used.
 In one embodiment, the metal powders used as a matrix-forming metal powder in the invention have an average particle size of about 1 .mu.m to about 200 .mu.m. Preferably, the metal powder has an average particle size of about 10 .mu.m to about 100 .mu.m, more preferably about 15 .mu.m to about 75 .mu.m, most preferably about 20 .mu.m to about 50 .mu.m.
 Metal powders, such as those described above, can be used alone as the matrix forming metal powder. Alternately, further additional components can be combined with the metal powder. For example, in one embodiment of the invention, the metal powder further includes carbon in the form of -300 mesh crystalline graphite to further increase the strength of the low carbon steel matrix, as described above.
 Further reinforcement agents can also be added to the metal powder prior to introduction of the powder into the mold. For example, natural or synthetic fibers or particulate matter could be mixed with the metal powder, or added into the mold, to provide additional benefits, such as increased strength or heat resistance.
 Once the spaces around the hollow metallic spheres have been filled with the matrix-forming metal powder, the mold is heated to a sintering temperature, the appropriate temperature being dependant upon the composition of the powder and the composition of the metallic spheres. Where the metal powder has a sintering temperature well below the sintering temperature of the metallic spheres, the lower temperature may be used, thereby sintering the metal powder and forming a solid metal matrix around the hollow metallic spheres. When the metallic spheres and the metal powder comprise the same metal, or different metals having similar sintering temperatures, the metal powder and, to some extent, the metallic spheres are sintered, thereby forming a solid metal matrix around the hollow metallic spheres.
 In one embodiment, the metal powder is sintered with the contents of the mold under pressure, such as in a hot press. Pressure values can vary depending upon the mechanical and physical properties of the spheres. Further, sphere size can also affect the applied pressure range. Acceptable pressure ranges can be calculated based upon the yield strength of the hollow sphere and the permissible load that can be applied to the spheres without any permanent deformation of the spheres.
 According to one embodiment, sintering is conducted without application of external pressure. In this embodiment, thermal expansion of the spheres during sintering and the resulting localized pressure around the spheres were used to facilitate pressing of the powder into the spaces between the spheres. The results show minimal porosity in the matrix of the composite metal foam after sintering.
 Sintering temperature can vary depending upon materials used in the spheres and, particularly, in the matrix-forming metal powder. In one embodiment, where hollow steel spheres and aluminum powder are used, the sintering is performed at a temperature of about 630.degree. C. In another embodiment, where hollow steel spheres and steel powder are used, the sintering is at a temperature of about 1200.degree. C. Preferably, sintering is performed at a temperature sufficient to exceed the solidus temperature of the metal matrix-forming powder but remain below the liquidus temperature of the powder. Further, preferably, the sintering temperature does not exceed the solidus temperature of the hollow metallic spheres. In one particular embodiment, sintering is performed at a temperature of between about 500.degree. C. and about 1500.degree. C., preferably between about 550.degree. C. and about 1400.degree. C., more preferably between about 600.degree. C. and about 1300.degree. C.
 Sintering time can also vary depending upon the materials used in the hollow metallic spheres and the metal matrix-forming powder. Sintering time also varies, however, based upon the relative size of the mold (and therefore the size of the sample being prepared). Larger molds obviously require a longer sintering time to ensure sintering completely through the thickness of the sample. Likewise, smaller molds requires a lesser sintering time. Size considerations in relation to sintering time generally follow guidelines similar to those previously provided in relation to powder metallurgy processes.
 Sintering conditions are preferably optimized to achieve improved mechanical properties. In one preferred embodiment, a duplex cycle is used to provide improved mechanical properties due to different sintering mechanisms taking place at each temperature. Such a method generally comprises cycling temperature increase phases with temperature hold phases. In one particular embodiment, where a composite metal foam is prepared using steel spheres and steel powder, the sample is heated at 10 .degree. C./minute, held for 30 minutes at 850.degree. C., heated at 5.degree. C./minute, held for 45 minutes at 1200.degree. C., and cooled to room temperature at 20.degree. C./minute. In such cycles, the lower temperature portion assists in the removal of oxides and impurities and helps bring the mold to thermal equilibrium to avoid gradient properties. Surface transport effects are most prevalent at lower temperature, so the bonds between particles are strengthened without densification of the matrix. At higher temperatures, strength is increased greatly as a result of the higher sintering rate due to greater atomic motion. For both temperatures, rapid increases in strength are noted for times up to 30 minutes, where the rate begins to decrease.
 FIG. 2 provides an optical, cross-sectional image of a hollow metallic foam according to the invention prepared by powder metallurgy using hollow steel spheres of 3.7 mm average diameter and a sintered steel powder matrix. FIGS. 3 and 4 provide scanning electron microscopy (SEM) images of a composite stainless steel foam prepared using a powder metallurgy technique, as described above. FIG. 3 shows the cross-section of intact spheres, and FIG. 4 shows the sintered powder matrix completely filling the space between the spheres. The bonding between the spheres and the matrix is seen to be strong with no voids at the interface.
 Returning to FIG. 2, the hollow metallic spheres show some signs of uniform packing; however, it is desirable to further increase the uniformity and density of the packing of the spheres to create metal foams exhibiting more uniform properties and even lower densities. The benefits of improving uniformity and density of packing are further illustrated in FIGS. 5 and 6.
 FIG. 5 provides a SEM image of a cross-section of two spheres in contact with one another. FIG. 6 provides a SEM image of a cross-section of two spheres not in contact, but with the metal matrix filling the space between the spheres. Increasing packing density of the spheres increases the contact between the spheres reducing the amount of free space between the spheres. Consequently, increased packing density reduces the amount of metal matrix present in the foam, which generally leads to lower densities, without sacrificing strength. FIGS. 5 and 6 further illustrate the ability to reduce the density of the composite metal foam by using hollow metallic spheres having lesser wall thicknesses. This is particularly illustrated in FIG. 5, wherein the sphere in the lower portion has a noticeably thinner wall than the sphere in the upper portion of the figure. The presence of the metal matrix surrounding the hollow metallic spheres allow for reducing the wall thickness to lower density of the composite metal foam without sacrificing strength.
 According to another embodiment of the invention, there is provided a method for preparing a composite metal foam by casting. In one embodiment, which is described below, the composite metal foam is prepared by permanent mold gravity casting; however, various other casting methods, as would be recognizable by one of skill in the art, could be used. Accordingly, the present invention is not limited by the permanent mold casting method described herein but rather encompasses all casting methods that could be recognizable as useful.
 In one embodiment of a casting method according to the invention, the hollow metallic spheres are first placed inside the mold. The hollow spheres are preferably arranged inside the mold, such as through vibrating, to pack the spheres into a best attainable close packed density. Vibration methods and apparatus, as described above in relation to powder metallurgy methods, would also be useful according to this aspect of the invention. Once the spheres are packed in the mold, a matrix-forming liquid metal is cast into the mold, filling the spaces around the hollow metallic spheres. The liquid metal is then solidified to form a solid metal matrix around the hollow metallic spheres.
 One embodiment of a mold useful in the casting method of the invention is illustrated in FIGS. 7 and 8, which show a three-dimensional view and a cross-sectional view, respectively, of an open atmosphere gravity feed permanent mold. The mold incorporates a sprue, runner, melt filter, and overflow riser. Carbon steel is a particularly preferred material for the mold allowing for repeated exposure to molten metal and high pre-heating temperatures.
 In a mold, such as shown in FIGS. 7 and 8, liquid metal is poured into the sprue. The liquid metal then travels through the runner, rises up through a slide gate and melt filter, fills the spaces between the hollow metal spheres, and flows up into the over-flow riser. Such a "bottom-up" filling approach allows the liquid matrix-forming metal to push out the air as the metal fills the interstitial space between the hollow metallic spheres. The slide gate allows for easy de-molding after solidification, and the melt filter serves to remove any solid impurities or oxides in the melt. The overflow riser feeds any shrinkage during aluminum solidification.
 In one particular embodiment, prior to introduction of the matrix-forming liquid metal, the mold and hollow spheres are pre-heated. Preferably, the pre-heat temperature is at least about equal to the casting temperature of the matrix-forming liquid metal. For convenience, the matrix-forming metal can be liquefied in the same furnace used for pre-heating the mold and spheres. The temperature of the mold and spheres should be at least about equal to the casting temperature of the matrix-forming liquid metal in order to prevent premature solidification of the matrix before complete filling of the mold, including the spaces between and around the spheres. The pre-heat temperature can be greater than the casting temperature of the matrix-forming liquid metal so long as the temperature does not approach the solidus temperature of the spheres.
 In this method, the hollow metallic spheres and the matrix-forming metal comprise different metal compositions, the compositions being distinguished by a difference in their melting temperatures. Since the matrix-forming metal is introduced to the mold in a molten state, it is necessary that the hollow metallic spheres comprise a metal composition having a melting temperature greater than the melting temperature of the matrix-forming metal composition. This avoids the possibility of melting of the hollow metallic spheres during pre-heating or during introduction of the matrix-forming liquid metal melting into the mold.
 Where the metal compositions comprise essentially pure single metals, the transition from solid to liquid generally can be described as a single melting temperature. Where metal mixtures are used, however, the state transition becomes more complex and can be described with reference to the solidus temperature and the liquidus temperature. When an alloy is heated, the temperature at which the alloy begins to melt is referred to as the solidus temperature. Between the solidus and liquidus temperatures, the alloy exists as a mixture of solid and liquid phases. Just above the solidus temperature, the mixture will be mostly solid with some liquid phases therein, and just below the liquidus temperature, the mixture will be mostly liquid with some solid phases therein. Above the liquidus temperature, the alloy is completely molten.
 The metal composition used as the matrix-forming liquid metal of the invention should have a melting point (or a liquidus temperature) that is below the melting point (or solidus temperature) of the metal composition comprising the hollow metallic spheres. Preferably, the melting temperature of the matrix-forming liquid metal is at least about 25.degree. C. less than the solidus temperature of the metal composition comprising the hollow metallic spheres, more preferably at least about 40.degree. C. less, most preferably at least about 50.degree. C. less than the solidus temperature of the metal composition comprising the hollow metallic spheres.
 In one embodiment, the hollow metallic spheres are comprised predominately of steel and the matrix-forming liquid metal is aluminum or an aluminum alloy. For example, the hollow metallic sphere could comprise low carbon steel or 316L stainless steel, such as according to the compositions exemplified in Table 1. Likewise, the matrix-forming liquid metal could comprise aluminum 356 alloy, such as according to the composition exemplified in Table 2. Aluminum 356 alloy is particularly useful due to its low density, high strength and stiffness, and ease of casting of the material.
 Further reinforcement agents can also be added to the matrix-forming liquid metal prior to casting. For example, natural or synthetic fibers or particulate matter could be mixed with the liquid metal, or added into the mold, to provide additional benefits, such as increased strength or heat resistance.
 Preferentially, the matrix-forming liquid metal is cast into the mold in such a manner as to facilitate complete filling of the voids around the hollow metallic spheres while avoiding disturbance of the hollow metallic spheres arranged within the mold. In some embodiments, it may be useful to use screens, or other similar means, for maintaining the arrangement of the spheres within the mold. In addition to gravity casting, the mold may be subject to pressure differentials during the cast process. For example, in one embodiment, the mold may be pressurized. In another embodiment, the mold may be under a vacuum.
 Once the matrix-forming liquid metal has been cast into the mold, the liquid metal is solidified to form a solid metal matrix around the hollow metallic spheres. Such solidification is generally through cooling of the mold, which can be through a.mu.mospheric cooling or through more controlled cooling methods.
 A composite metal foam, according to one embodiment of the invention, prepared by casting an aluminum metal matrix around hollow low carbon steel spheres, is shown in FIG. 9. As can be seen in the figure, the closest packing arrangement of the hollow spheres is somewhat disturbed by the inflow of the liquid metal matrix. Nevertheless, strong bonding between the metal matrix and the hollow spheres is achieved.
 Bonding between the foam components is more clearly evident in FIGS. 10-11 which provide SEM images of a cast metal foam according to the invention comprising hollow low carbon steel spheres surrounded by an aluminum metal matrix. As can be seen in FIG. 10, the aluminum metal matrix fills the interstitial space between the hollow steel spheres with consistent bonding to the surfaces of the spheres. FIG. 11 provides a detailed view of the sphere wall interaction with the aluminum matrix. Very little evidence of influx of aluminum matrix into the walls of the hollow steel spheres is seen in FIG. 11 indicating low porosity in the wall of the hollow steel spheres.
 SEM images of a cast metal foam according to another embodiment of the invention are provided in FIGS. 12(a) and 12(b). While it is preferable for the interstitial space between the spheres to be completely filled by the metal matrix, as can be seen in FIG. 12(a), voids can be present, particularly at an interface between two spheres. Using geometrical calculations, the void space at the interface of two spheres can be calculated according to an estimated void angle, and the resulting void volume (V.sub.void) per contact point of two spheres can be calculated according to the following equation:V.sub.void=3.16.times.10.sup.-2(R.sup.3) (1) in which R is the outer radius of the spheres used in the foam (see Sanders, W. S. and Gibson, L. J., Mater. Sci. Eng. A A347, 2003, p. 70-85). Projecting this into a face-centered cubic (FCC) arrangement of spheres with four spheres per unit cell, and knowing that in a random loose condition, there are three contacts per unit cell, the total volume percentage of voids per unit cell can be estimated as: V.sub.vf=(3.16 .times.10.sup.-2R.sup.3.times.12)/22.627R.sup.3 (2) in which V.sub.vf, is the volume fraction of voids. In one particular composite metal foam according to the invention prepared by casting molten aluminum around hollow low carbon steel spheres, the volume percentage of voids calculated according to equation (2) was 1.68%. However, the actual matrix porosity is expected to be even less, given there are less than four spheres in each unit cell of the random arrangement and not all contacts have a void space. In one embodiment of the invention, the void volume percentage is less than 1.5%, preferably less than 1.25%, more preferably less than 1%.
 As can be seen in FIG. 12(b), multiple different phases are present in the aluminum matrix. Scanning electron microscope energy dispersive x-ray spectroscopy (SEM EDX) compositional analysis was performed on an Al--Fe composite metal foam according to one embodiment of the invention prepared by casting, and the compositional analysis is provided below in Table 3. TABLE-US-00003 TABLE 3 Compositional Analysis (atomic %), t = trace Al Si Fe Mg Mn Cu Al matrix generally 97.9 1.5 t t t t Light gray phase 65-75 10-20 15-25 t 3.5 1.9 Dark gray phase 97.1 .ltoreq.3 t t t 1.0
 As can be seen in FIG. 12 and Table 3, the Al matrix typically comprises three different phases. The Al matrix generally comprises approximately 98% Al. The phase designated the light gray phase is a ternary alloy of Al, Si, and Fe (estimated to be Al.sub.4FeSi) and is typically found in two different shapes, plates and needles. The phase designated the dark gray phase has a composition that is close to the composition of the Al matrix generally but also includes copper.
 The composite metal foams of the invention (whether prepared through casting or powder metallurgy) are particularly characterized in that they exhibit high compressive strength and energy absorption capacity while maintaining a relatively low density. Of course, actual density of the finished composite metal foam can be calculated using the measured sample weight and structural dimensions. It is also possible, however, to determine an estimated density based on component properties and packing properties of the spheres in the mold.
 Sphere packing density is a measure of the relative order of the arrangement of spheres, such as in a mold. It is desirable to achieve a maximum density of spheres in order to maintain a lowest possible density for the prepared composite metal foam and have a uniform distribution of spheres, thus contributing to isotropy of mechanical properties. It is generally recognized that there are three types of packing arrangements for spheres: ordered packing, random dense packing, and random loose packing (See, German, Particle Packing Characteristics, Metal Powder Industries Federation, Princeton, N.J., 1989). The highest order is represented by the face-centered cubic (FCC) or hexagonal closed packed (HCP) structure, with a 74% packing density of spheres, assuming mono-sized spheres. A random dense arrangement has a packing density of 64%. This is achieved by vibrating an initially random arrangement to the best attainable packing density. Random loose packing has a fractional density of 56%-62.5%, with an average reported fraction density of 59% and a three-point contact per sphere.
 As previously noted, in one preferred embodiment of the invention, after the hollow metal spheres are loaded into a mold (in either a casting or powder metallurgy technique), the mold with the spheres is vibrated to achieved increased packing density, which ultimately leads to reduced overall density for the prepared composite metal foam of the invention. In one test, hollow spheres were poured in bulk into an acrylic box. Isopropyl alcohol was then poured into the box as a testing replacement for the matrix material to determine the volume needed to fill the box. With this random placement, sphere packing density was measured as 56%. In a second run, the spheres were manually vibrated prior to introduction of the isopropyl alcohol. The vibrated spheres exhibited a packing density of 59%.
 The density of a composite metal foam according to the invention can be estimated as a function of component density according to the following equation: .rho..sub.CF=.rho..sub.SV.sub.fs+.rho..sub.mV.sub.fm (3) in which .rho..sub.CF is the density of the composite metal foam, .rho..sub.s is the density of spheres, V.sub.fs is the volume fraction of spheres (the packing density of the spheres), .rho..sub.m is the density of the matrix, and V.sub.fm is the volume fraction of the matrix. Considering the effect of porosity in the wall thickness of metal spheres, equation (3) can be altered to: .rho..sub.CF=.rho..sub.metal[1-(V.sub.in/V.sub.out)]V.sub.fs(1-V.sub.fp)+- .rho..sub.mV.sub.fm (4) in which .rho..sub.metal is the density of the metal used in the hollow metal spheres (e.g., steel), V.sub.in/V.sub.out is the ratio of inner volume to outer volume of the metal spheres, and V.sub.fp is the volume fraction of porosities in the wall thickness. As previously noted, the porosity of the walls of the hollow metal spheres can vary and is preferably less than about 12%.
 As previously noted, the composite metal foams of the invention are particularly useful in that they provide a material that combines strength with light weight. In particular, the composite metal foams generally have a density that is less than the density of the bulk materials used in the composite metal foams. For example, steel is generally recognized as having a density in the range of about 7.5 g/cm.sup.3 to about 8 g/cm.sup.3. A composite metal foam prepared according to the present invention using hollow steel spheres and a steel metal matrix would exhibit a density well below these values.
 The composite metal foams according to the present invention preferably have a calculable density of less than about 4 g/cm.sup.3. Preferably, the composite metal foams of the invention have a density of less than about 3.5 g/cm.sup.3, more preferably less than about 3.25 g/cm.sup.3, and most preferably less than about 3.0 g/cm.sup.3. In one embodiment of the invention, there is provided a composite metal foam comprising hollow steel spheres surrounded by an aluminum metal matrix, the composite foam having a density of less than about 2.5 g/cm.sup.3. In another embodiment of the invention, there is provided a composite metal foam comprising hollow steel spheres surrounded by a steel metal matrix, the composite foam having a density of less than about 3.0 g/cm.sup.3.
 The metal foam of the invention can also be evaluated in terms of relative density. By analysis of this parameter, it is possible to compare the level of porosity of the metal foam (or the level of foaming) with the level of porosity of the bulk material. According to one embodiment of the invention, the inventive composite metal foam has a relative density (compared to bulk steel) of between about 25% and about 45%.
 The usefulness of the composite metal foams according to the invention is particularly characterized by their favorable strength to density ratio. As used herein, strength to density ratio is determined as the plateau stress of the metal foam under compression (measured in MPa) over the density of the metal foam. The composite metal foams of the invention typically exhibit a strength to density ratio of at least about 10. Preferably, the composite metal foams of the invention exhibit a strength to density ratio of at least about 15, more preferably at least about 20, still more preferably at least about 25, and most preferably at least about 30.
 The composite metal foams of the invention are further characterized by improved energy absorption. Energy absorption capability can be characterized in terms of the amount of energy absorbed by the material over a given level of strain. As used herein, energy absorption is defined as the energy absorbed (in MJ/m.sup.3) up to 50% strain. The composite metal foams of the invention typically exhibit energy absorptive capability of at least about 20 MJ/m.sup.3. Preferably, the composite metal foams of the invention exhibit energy absorptive capability of at least about 30 MJ/m.sup.3, more preferably at least about 50 MJ/m.sup.3, most preferably at least about 75 MJ/m.sup.3.
 As detailed above, the composite metal foams of the invention are further particularly beneficial in that they provide improved mechanical properties under cyclic compression loading. Further, microstructural, mechanical, and physical properties show noticeable improvement over previous metal foams through analysis by optical microscopy, scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX), and compression test and strain mapping during monotonic compression loading.
 Any method or apparatus recognizable as useful by one of skill in the art for obtaining and analyzing the above-noted properties could be used and is fully envisioned by the present invention. For example, SEM images can be obtained through use of a Hitachi S-3200N environmental SEM equipped with energy dispersive X-ray spectroscopy. Of course, other SEM equipment, as would be recognized as suitable by the skilled artisan, could also be used in accordance with the invention.
 One particular method of analysis of the mechanical properties of the composite metal foams according to the invention is through monotonic compression testing and compression fatigue testing. Exemplary equipment useful in such testing is a MTS 810 FLEXTEST.TM. Material Testing System (available from MTS Systems Corporation). According to one testing procedure, monotonic compression testing is performed using a MTS 810 loading machine with a 220 kip load cell. According to another testing procedure, compression fatigue testing is performed using a MTS 810 loading machine with a 220 kip load cell having a fixed R-ratio (R=.sigma..sub.min/.sigma..sub.max) of 0.1 at a frequency of 10 Hz and an applied maximum stress of 37.5 MPa.
 FIG. 13 provides a side-by-side comparison of three foams prepared according to the invention. FIG. 13(a) is a cross-section of cast metal foam comprising hollow low-carbon steel spheres (3.7 mm mean diameter) and an aluminum matrix. FIG. 13(b) is a cross-section of a metal foam prepared by powder metallurgy comprising hollow low-carbon steel spheres (3.7 mm mean diameter) and a metal matrix prepared from powdered low carbon steel. FIG. 13(c) is a cross-section of a metal foam prepared by powder metallurgy comprising hollow low-carbon steel spheres (1.4 mm mean diameter) and a metal matrix prepared from powdered low carbon steel.
 Monotonic compression testing of all three samples from FIG. 13 demonstrated the typical behavior of an elastic-plastic foam under compression. There is an initial linear elastic region, which is followed by an extended region of deformation at a relatively constant level of stress. Unlike most foams, however, the foams prepared according to the present invention do not exhibit a level plateau stress. Rather, the material densifies at a slowly increasing rate, and there is no distinct point at which full densification occurs. As used herein, plateau stress is understood to refer to the average stress between the yield point and the point at which 50% strain (i.e., deformation) has been achieved. All three composite metal foams shown in FIG. 13 reached a minimum of 50% strain before reaching a point of full densification.
 FIG. 14 shows stress-strain curves of composite metal foams according to various embodiments of the invention under monotonic compression. Sample 1 is taken from an embodiment formed through powder metallurgy using 3.7 mm hollow low carbon steel spheres and low carbon steel powder. Sample 2 is taken from an embodiment formed through casting an aluminum matrix around 3.7 mm hollow steel spheres. Sample 3 is taken from an embodiment formed through powder metallurgy using 1.4 mm hollow low carbon steel spheres and low carbon steel powder. Sample 4 is taken from an embodiment formed through powder metallurgy using 2.0 mm hollow stainless steel spheres and stainless steel powder. After 50% strain, the composite metal foams begin to approach densification as the hollow spheres are completely collapsed and the material begins to heave like a bulk material.
 The compression test results, as well as further physical properties of the embodiments of the inventive composite metal foam corresponding to the four samples discussed above in relation to FIG. 14, are shown below in Table 4. As a comparative, the physical properties of previously known metal foams made of hollow spheres alone are also provided. Comparative HSF 1 is a steel foam described by Anderson, O., Waag, U., Schneider, L., Stephani, G., and Kieback, B., (2000), "Novel Metallic Hollow Sphere Structures", Advanced Engineering Materials, 2(4), p. 192-195. Comparative HSF 2 is also a steel foam described by Lim, T. J., Smith, B., and McDowell, D. L. (2002), Behavior of a Random Hollow Sphere Metal Foam", Acta Materialia, 50, P.2867-2879. TABLE-US-00004 TABLE 4 Sample Sample Sample Sample HSF HSF 1 2 3 4 1 2 Sphere OD (mm) 3.7 3.7 1.4 2.0 2-3 2 Sphere wall 0.2 0.2 0.05 0.1 0.25 0.1 thickness (mm) Density (g/cm.sup.3) 3.2 2.4 2.7 2.9 1.4 1.4 Relative Density 40.7 42 34.2 36.8 -- -- (%) Plateau Stress 42.3 67 76 136 23 4.8 (MPa) Densification 55 50 50 50 60 65 Strain (%) Strength/Density 13.2 28 29.5 47 16 4 ratio Energy Absorbed 21 32.3 37.6 123 -- -- up to 50% strain (MJ/m.sup.3)
 Samples 1 and 3 above are powder metallurgy foams comprising low carbon steel. Sample 2 is a cast Al--Fe foam. Sample 4 is a powder metallurgy foam comprising stainless steel. The comparison in Table 4 indicates the composite foams of the invention have a noticeably increased strength while maintaining a comparable strength to density ratio. Further, the inventive composite foams show improved energy absorptive properties making the composite foams particularly useful in the various applications described herein.
 The present invention is more fully illustrated by the following examples, which are set forth to illustrate the present invention and are not to be construed as limiting.
Composite Metal Foam Prepared by Powder Metallurgy
 A composite metal foam was prepared using stainless steel spheres and stainless steel powder according to the specifications provided in Tables 1 and 2, respectively. The stainless steel spheres had an outside diameter of 2.0 mm and sphere wall thickness of 0.1 mm. The spheres were cleaned in a solution of 3.0 mL HCl and 97 mL water to remove oxides, rinsed in acetone, and dried. A permanent mold made of 304 stainless steel and having interior dimensions of 5.1 cm.times.5.1 cm.times.10 cm was used. The mold was prepared by coating its surfaces with a boron nitride mold release. The spheres were placed in the mold and vibrated for 5 minutes using an APS Dynamics model 113 shaker and an APS model 114 amplifier with a General Radio 1310-B frequency generator. The powder was added and the mold was further vibrated to completely fill the spaces between the spheres. Total vibration time was 30 minutes at 15-20 Hz.
 The mold was placed in a vacuum hot press during sintering. Although no pressure was applied, the mold cap was held in place by the press, and the thermal expansion of the spheres was used to create internal pressure to aid in the sinter of the powder. The powder and spheres were sintered using a 10.degree. C./minute heating rate, held for 30 minutes at 850.degree. C., further heated at a rate of 5.degree. C./minute and held for 45 minutes at 1200.degree. C. The mold was then cooled at a rate of 20.degree. C./minute. The finished composite steel foam was then removed from the mold for testing.
 Optical microscopy was performed using a Buhler Unitron 9279 optical microscope equipped with a Hitachi KP-M1 CCD black and white digital camera. SEM images were taken with a Hitachi Ss-3200N environmental SEM equipped with EDX to determine the microstructure and chemical composition of the metal foam. Monotonic compression testing was performed using an MTS 810 with a 980 kN load cell and a crosshead speed of 1.25 mm/minute.
 The composite metal foam had a density of 2.9 g/cm3 (37% relative density) and reached a plateau stress of 113 MPa from 10-50% strain and began densification at around 50% strain. These and further analytical results are provided in Table 4 (wherein the metal foam from this example is shown as Sample 4). FIG. 15 shows a comparison of the stainless steel composite metal foam (a) before compression testing and (b) after compression testing with 60% strain.
Composite Metal Foam Prepared by Casting
 A composite metal foam was prepared by casting using low carbon steel hollow spheres and a matrix-forming liquid aluminum 356 alloy according to the specifications provided in Tables 1 and 2, respectively. The steel spheres had an outside diameter of 3.7 mm and sphere wall thickness of 0.2 mm. An open atmosphere gravity feed permanent mold casting system made of carbon steel was used, the mold cavity having dimensions of 121 mm.times.144 mm.times.54 mm. The mold was partially preassembled after coating with a boron nitride powder spray to prevent oxidation to mold surfaces during preheating and for providing easy release of the sample after cooling. The spheres were placed in the mold with a stainless steel mesh to hold them in place and vibrated for 10 minutes to pack the spheres into a random dense arrangement. The mold used was similar to that illustrated in FIGS. 7 and 8.
 The aluminum alloy was melted in a high temperature furnace (3300 series available from CM Furnaces) at a temperature of 700.degree. C. At the same time, the mold with the hollow spheres inside was pre-heated in the furnace to 700.degree. C. to prevent instant solidification of the aluminum upon contact with the spheres while casting. The fully liquid aluminum alloy was then poured in the sprue tube of the heated mold. The liquid aluminum fills out the cavity while pushing the air out from the cavity. The filled mold was allowed to air cool, and the mold was disassembled and the composite metal foam removed. Testing was performed on the cast composite metal foam as described in Example 1.
 The cast composite metal foam had a density of 2.4 g/cm.sup.3 (42% relative density). During monotonic compression, the composite metal foam reached an average plateau stress of 67 MPa up to 50% strain before it began densification at around 50% strain. These and further analytical results are provided in Table 4 (wherein the cast metal foam from this example is shown as Sample 2). Optical and SEM observation indicated the Al matrix had nearly filled all of the interstitial spaces between the steel spheres (see FIG. 12(a)). The void space due to incomplete filling of the interstitial space at the sphere point contacts with the Al matrix was calculated to be less than 1%.
 The cast composite metal foam was tested to calculate maximum cyclic stress. Testing methods are fully described by Banhart, J. and Brinkers, W., "Fatigue Behavior of Aluminum Foams", J. Material Science Letters, 18(8), 1999, p. 617-619, and Lehmus, D., et al., "Influence of Heat Trea.mu.ment on Compression Fatigue of Aluminum Foams", Journal of Material Science, 37, 2002, which are incorporated herein in their entirety. The average yield strength calculated, using the 0.2% offset method, was 29 MPa. The maximum stress was chosen to be 85% of the reference strength. The fatigue was continued with this maximum stress for 250,000 cycles with no apparent deformation. The maximum stress was then increased to 37.5 MPa (the stress at 5% strain from the stress-strain curve). FIG. 16 shows the curve for the compression fatigue test as the relation between strain and the number of cycles.
 The cast composite foam deformed by an initial progressive shortening, followed by collapse of the spheres starting at certain regions with possible defects like holes or cracks, causing the subsequent failure of the neighboring spheres leading to the formation of macroscopic collapse bands. Visual observation of the deformed fatigue cast foam revealed that extensive fatigue failure had occurred within the crush bands. The S-N curve (FIG. 16) shows the initial progressive fatigue damage at the onset of an abrupt strain jump. The cast foam sample had endured 1,440,000 cycles before the end of the fatigue test. Optical images of the cast foam taken before and during the fatigue test are shown in FIG. 17.
 Many modifications and other embodiments of the invention will come to mind to one skilled in the art to which this invention pertains having the benefit of the teachings presented in the foregoing descriptions and associated drawings. Therefore, it is to be understood that the invention is not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.