Ceramic-metal composites: bulletproof strength.
Boron carbide is the third hardest material known to man. It's also lightweight, extremely stiff, and less expensive than diamonds and cubic boron nitride--the leading entries on the hardness scale. So it's not surprising that weapons engineers would want to make armor from it. But boron carbide has the weakness common to all hard covalent ceramics: it is quite brittle, making even slightly flawed components fracture relatively easily when stressed.
In the early 1980s, a trio of materials researchers at the University of California at Los Angeles, Ilhan A. Aksay, Danny C. Halverson, and Aleksander J. Pyzik, made great strides in understanding how ductile aluminum could be combined with brittle boron carbide to boost the latter's low fracture toughness. Their research project was funded by the Defense Advanced Research Projects Agency or DARPA (Washington) and administered by the Lightweight Armor Program at Lawrence Livermore National Laboratory (Livermore, Calif). The project's goal: to produce a low-mass, low-cost, high-performance armor material for military vehicles.
After several years of continued development, that fundamental research has yielded a charcoal-gray ceramic-metal composite--or cermet--stronger for its weight than steel, somewhat less dense than aluminum, and several times more resistant to fracture than conventional structural ceramics. More accurately, the work has resulted in a family of composites ranging from 50 to 85 percent boron carbide that can be tailored by processing and, later, heat treatment for desired mechanical and physical properties. For example, the higher the aluminum content, the greater the composite's fracture toughness.
Substantial further study at Livermore, the University of Washington (Seattle), and The Dow Chemical Co. (Midland, Mich.) has produced several practical and presumably economic manufacturing techniques for the cermet, which generally contains less than 50 percent by volume metal phase. In two of these processing procedures, molten aluminum is infiltrated into sintered porous "sponges" of boron carbide or into boron carbide powder compacts at temperatures below 1200[degrees]C and then heat-treated at lower temperatures to achieve desired properties. The fabrication procedures differ in that one employs temperatures above 2000[degrees]C to form the boron carbide sponge, while in the other, chemical treatment of the powder compact is followed by a low-temperature aluminum infiltration step. Another technique rapidly consolidates aluminum and carbide powders with high impulse pressures and the heat that those pressures generate. In the most recent work, a procedure has been developed in which sandwiches of thin boron carbide-polymer tapes and aluminum foil are layed up like carbon fiber composites into laminated architectures whose design was suggested by the study of anomalously tough, layered microstructures found in certain shellfish.
Though little can be stated about the U.S. Defense Department's classified research on armor performance, the cermet is said to have successfully withstood multiple hits from projectiles in firing tests, proving the material's ability to take the intense shock of ballistic impact. Efforts to perfect the tank armor are reportedly continuing.
In the meantime, commercial licenses concerning at least one of the now-patented processing technologies have been issued to several corporations to develop proprietary products expected to reach market in 1993-94. Among the proposed uses: lightweight police body armor; low-inertia, dimensionally stable, and low-vibration rotating components for computer hard-disk drives; long-life bearings, races, and other wear parts; premium-performance sporting goods; high efficiency, electron-emission devices; cutting tools for hard-to-machine, silicon-aluminum alloys; and nuclear shielding (boron carbide is an excellent neutron absorber).
An Old Idea
"The idea to make armor from boron carbide was first considered following the end of World War II," said Ilhan Aksay, who is now professor of materials science and engineering at the University of Washington. "Boron carbide's extreme hardness makes it useful in blunting projectiles, while its low density means you don't have to carry around a lot of weight. Though many tried, including specialists at the Norton Co., the U.S. Army, and the U.S. Air Force, nobody knew how to process it cost-effectively."
But that didn't deter William E. Snowden, who was then leader of Livermore's Lightweight Armor Program and now directs the Washington, D.C., office of Failure Analysis Associates, an engineering and scientific services firm. In 1981, Snowden had been looking at several high performance materials in an attempt to meet the emerging threats to armor, when one day, as he and Aksay were driving back from a business trip, they began to discuss why U.S. industry had yet to develop lower density (lower specific gravity) analogues to the heavy cermets, such as titanium carbide-nickel-molybdenum and tungsten carbide-cobalt, it had used so successfully for decades. In particular, the two former school-mates focused on the potential of boron carbide and silicon carbide as the main composite constituents with low density metals such as aluminum serving as densification agents. "We concluded that part of the reason was the lack of stressful enough performance requirements, accompanied by the challenge of finding compatible systems and developing processing techniques," recalled Snowden. Traditional techniques such as hot pressing would not be successful, they felt, but perhaps a sintering technique would do the job.
"By the end of the ride back," said Snowden, "we'd decided that if we paid strict attention to the wettability, capillarity thermodynamics, reaction kinetics, and densification rates in these systems, there was no fundamental reason that we couldn't make them. Besides achieving high strength, high hardness, and hopefully, reasonable fracture toughness, we were also interested in finding systems that would lend themselves to affordable mass-production techniques."
When Snowden returned to Livermore, he undertook a series of experiments that established that rapid heating to more than 1100[degrees] C could promote the
wetting of boron carbide by aluminum, and that the formation of deleterious aluminum carbide phases could be avoided. He approached his boss, Richard L. Landingham, leader of Livermore's ceramic and composites section, with the idea of issuing a research contract to Aksay and his UCLA graduate students, Alek Pyzik and Danny Halverson, to conduct preliminary processing studies on the proposed composite materials. Soon after, the contract was awarded.
"At the start of the project in 1982, our model cermet system was tungsten carbide-cobalt, which the cutting-tool industry has been making for decades using a process called liquid-phase sintering," said Halverson, now vice president for research and development at Synergetic Materials Inc., a start-up consulting firm in Auburn, Calif. In this procedure, said Halverson, tungsten carbide and cobalt powders are mixed, placed in refractory molds, and heated to full density. "The reason liquid-phase sintering works in the tungsten carbide-cobalt system is that the cobalt wets [reacts with] the carbide, but at a slow enough rate that full, uniform densification takes place."
However, the investigators soon found that similar sintering of the boron carbide-aluminum system is another matter entirely. The difference lies in the difficulty in establishing proper reaction conditions. "Liquid-phase sintering of these materials yielded a final product that looked like Swiss cheese," said Halverson. "From previous reactivity tests we conducted, we knew that aluminum would wet the boron carbide."
These experiments involved placing drops of molten aluminum onto substrates of hot-pressed boron carbide and then measuring the contact angle between the upward curving side of the sessile metal blobs and the horizontal ceramic surface. Contact angles less than 90 deg indicate reactivity. But, according to Halverson, "The kinetic issue of how fast it would wet the ceramic remained. When we put the ceramic powder with the molten aluminum, the metal would wick around the carbide grains and react so rapidly that the metal phase was depleted before the sample could fully densify. The result was gross porosity."
Study by Pyzik and Aksay of liquid-phase sintering in the boron carbide-aluminum system several years later revealed that the just-melted metal tends to rearrange itself to more densely packed regions due to capillary suction, leaving the volume originally occupied by the metal powder as a pore or void. These dense ceramic-metal agglomerates support the formation of binary and ternary compounds as solid bridges between boron carbide grains, so that densification stops ("locks up") due to the formation of a solid skeleton.
Still short of the mark, the team had by now developed a fuller understanding of the fundamental chemistry of the system. "At that stage," said Halverson, "we knew we could hot isostatically press (HIP) the boron carbide-aluminum mix to full density, and also that we could sinter it to a porous state, so we figured that a pressureless route to full consolidation could be accomplished. The key was to find a cheap way to accelerate the densification kinetics externally. But how?"
"Our main contribution," said Pyzik, now project leader at Dow Chemical Co.'s Advanced Ceramics Laboratory, "was that we realized the system could be wetted, that the wetting could be changed by heating, and that material could be consolidated to high densities. But nobody had any idea how to do it in a practical manner. We tried all kinds of tricks, but we didn't realize that the mechanisms for liquid-phase sintering were not present in this system."
Though their overall goal to produce the armor material went unattained, the trio's work had proved the concept of boron carbide-aluminum composites. Halverson, Pyzik, and Aksay received a process patent for the material in 1986, which is considered the basic composition patent for the new cermet. It was assigned to the University of California.
Soon after they completed the research in 1983, the three split up. Aksay and Pyzik moved to the University of Washington, and Halverson was eventually hired by Landingham to head the Lightweight Armor Program at Livermore after Snowden left for a new position as a DARPA program manager.
At the University of Washington, Aksay and Pyzik worked under a series of DARPA research contracts administered by the U.S. Air Force Office of Scientific Research to refine the processing techniques they had developed at UCLA. Pyzik started investigating the sintering of boron carbide with carbon. It was found that when they sintered boron carbide with no additives at high temperatures, the result was a reticulated network of boron carbide grains with interconnected porosity.
The problem with the earlier densification attempts, reasoned Pyzik, was that there was no solid rearrangement stage in the liquid-phase sintering of boron carbide-aluminum as there is in the tungsten carbide-cobalt model system. So, he thought, why not just eliminate the need for it by freezing the microstructure through the process of sintering the boron carbide grains into a rigid network? The next step would be to infiltrate the molten aluminum into the sintered sponge preform. "The key trick," said Pyzik, "was to realize that rather than mixing the metal and ceramic at the beginning and getting densification from shrinkage due to solid rearrangement, we said, start with a porous ceramic preform and allow liquid rearrangement to cause the densification."
Following this preparatory work, Aksay and Pyzik heated a green boron carbide preform to about 2000[degrees]C in an argon environment and succeeded in producing a porous sponge of bonded carbide grains. They then submerged the sponge in a bath of molten aluminum and infiltrated it according to the temperature and time indicated by the wetting data they had developed earlier, which turned out to be about 1200[degrees]C for a few minutes in a vacuum. The final product was the fully densified boron carbide-aluminum cermet. They had solved the problem.
Later it was determined that the wetting reaction rates were substantially reduced when the sintering process was conducted in the presence of a small amount of free carbon, preferably in the form of graphite. This step allows greater control of the reaction kinetics, thus ensuring full densification. Aksay and Pyzik also investigated post-infiltration heat treatment regimes to achieve desired phase compositions and microstructures. The pair received a patent for the work in 1987.
"After Aksay and Pyzik told us about their success, Landingham and I conducted repeat experiments and duplicated their results," said Halverson. "However, we decided that we should look for a way to get similar results at lower temperatures, because heating to 2000[degrees]C is expensive and we wanted a lower-cost procedure."
Landingham told him about an infiltration technique he had patented years earlier in which he had heated metal powders at low pressure to remove an oxide film that hindered wetting. He suggested that an oxide film might be causing the same problem in the ceramic powders. Landingham explained, "Most hard ceramic powders are crushed to size in iron mills. To remove the residual iron and oils that inevitably enter the powder, an acid such as hydrochloric acid is introduced to leach out these impurities. Unfortunately, the acid leach oxidizes the surface of the boron carbide powder, forming boric oxide, which prevents good metal infiltration. We realized that by cleaning the powder with an alcohol such as methanol, we could remove the boric oxide on the surface and get essentially the same kind of effect at low temperatures that the Washington team was getting with high temperatures."
"I did contact-angle tests like we did at UCLA," said Halverson, "and it turned out we could indeed control the reactivity as a function of time and temperature by treating the carbide substrate with alcohol. When we tried chemically treating the carbide powders, we found that we could consolidate the infiltrated preform at 1200[degrees] C.
"The alcohol wash removes the residual boric oxide on the surface," said Halverson. "When you heat the carbide as our Washington colleagues did, you exceed the vapor phase of the surface boric oxide and it leaves, permitting full consolidation. When you wash boric acid in alcohol, you form trimethyl borate, which washes away. In addition, the residual hydrocarbon forms an aliphatic hydrocarbon when heated, and that is converted to a carbon-rich residue on the surface, which allows more control of the reaction kinetics." The key, he said, is controlling the surface chemistry of the starting constituents.
Landingham described their procedure, "We place the packed, treated powder into a vacuum [or inert atmosphere] furnace with the aluminum, heat it to 1200[degrees] C and infiltrate the metal. Afterwards we drop the temperature to about 800[degrees] C to anneal or soak. The process produces aluminum borides and aluminumboro-carbides which fuse the ceramic particles together into a bonded matrix. The aluminum acts as a toughener. If you heat the sample longer, you get less aluminum and hence, less toughness."
Halverson and Landingham patented this lower temperature, chemical treatment process in 1987 and received an IR (industrial research) 100 Award from Research & Development magazine for it that same year. Currently, the chemical process has been used to fabricate components in lots of about a hundred. The largest item that has been made is about 6 inches square and a couple of inches thick.
As with all these processes, the final product is so hard that grinding and machining, even with diamond tooling, is to be avoided, said Halverson. "You don't want to grind it after it's consolidated, but because it is electrically conductive, you can shape it with wire or electrode EDM [electrical discharge machining]."
Forming to near-net shape is more desirable, so the Livermore researchers have developed a two-step injection molding process. In this procedure, small highly configured geometries can be molded using organic binders containing the appropriate chemical pretreatment agents. Then the binder is volatized from the precursor at about 200[degrees] to 300[degrees] C, making it ready for subsequent infiltration.
In related work, Halverson, Landingham, and Livermore researcher J. Birch Holt have developed a combustion synthesis or self-propagating exothermic boosting process in which they make boron carbide from boron oxide. The idea is produce the powders more cheaply--a key to the eventual application of the cermet.
Rapid Compaction Method
Following his work with Aksay at the University of Washington, Pyzik moved on to lead a group of researchers at Dow Chemical working on cermets in general. Pyzik's team has applied Dow Chemical's proprietary Rapid Omnidirectional Compaction (ROC) technique to boron carbide-aluminum cermets.
A paper written by Pyzik and Dow colleague Alexander Pechenik in 1988 explains that the process is based on the concept that the rapid application of sufficiently high pressure to ceramic-metal greenware causes reduction of the powder compact, which generates a temperature spike due to the nearly adiabatic nature of the process. This sudden temperature increase in the material, when controlled, can produce a sufficient amount of molten metal to collapse the rigid structure of interlocked ceramic grains and to help in the rearrangement. ROC's major densification mechanism is plastic deformation. "With ROC, we can do the same things as the chemical method without washing the powders by using the proper heating schedules," said Pyzik.
In the first step, the powders are mixed and consolidate at high pressure to full density, said Pyzik. A fluid die containing the sample is heated to the required temperature in an argon-purged furnace. After heating, the glass fluid die is removed from the furnace and placed into a press, which is activated for a few seconds. "As the pressure increases," he said, "the temperature rises, rapidly melting the aluminum which is injected--almost extruded--into the pores. The advantage of the ROC process is that everything happens at lower temperatures. You can still take it to higher temperatures to optimize the material's chemistry and microstructure."
Parts up to 5 in. in diameter and 2.5 in thick have been made by ROC. Gradient structures with up to ten 1/16-in layers of varying phase composition, where the gradation is accomplished by a proprietary technique, have also been fabricated. Pyzik noted that Dow Chemical's ROC process is adaptable to large-scale production.
Over the last few years, the Dow Chemical team has also used ROC to produce hollow objects of the cermet with complex internal shapes. In one variant, an aluminum mandrel is machined to the internal shape, and the precursor powders are compacted around it using essentially the same processing as before. The metal melts, enters the ceramic, and forms the desired hollow, near-net-shape part.
Meanwhile at the University of Washington, Aksay's research group has been "trying to extend the cermet work to a laminate fabricating approach, using the layered structure of seashells as a model." The nacre section (inside wall) of abalone shells consists of laminated crystalline calcium carbonate (aragonite) as thin as 0.25 [micrometer] or less, and 20-nm thick layers of organic materials, which were mostly proteins, said Aksay. This microstructure is said to be responsible for the shell's unusually high fracture toughness and fracture strength. When this "brick and mortar" microstructure is impacted, the hard aragonite slides along the slip planes formed by the layers by riding on the ductile ligament proteins. The dislocation dissipates the force of impact by rearrangement, said Aksay. "The result is a big boost to impact properties.
"Using the shell structure for a model," he added, "we lay up ceramic tapes with aluminum foil in a sandwich structure like graphite fiber prepreg materials. The flexible boron carbide-polymer tape we employ is produced by machinery used in the electronics industry to produce multilaminate ceramic substrates for computers. After lay up, we heat the laminate at about 1200[degrees] C to bake out the polymer and consolidate.
"The unique structure gives the laminate flexural strength up to 1 GPa and a fracture toughness up to 19 [MPa-m.sup.1/2]." These are some of the best property values yet obtained for boron carbide-aluminum cermets.
The University of Washington group believes that the lay up process is well suited to mass-production scale-up. A patent disclosure was filed several months ago and they are awaiting a patent. Aksay said that the researchers are now concentrating on refining the laminate architecture.
"The applications for this material are not obvious," said Pyzik. "This is a classic problem for all new materials, especially those that don't directly replace another material. After all, boron carbide-aluminum offers a unique combination of properties, so direct replacement is probably not the way to go. Finding the niches in which it would best be exploited will take some time."
Fighting the battle for engineering acceptance of the cermet, based on the University of California's compositional and chemical treatment patents, is Candy Voelker, licensing manager for chemistry and advanced materials in the Patent, Trademark and Copyright Department of the University of California (Alameda). "So far, we've had about 250 inquiries from companies all over the world. We've let four licenses, but unfortunately we can't disclose the names of the companies nor what they plan to do with the material," she said.
"We did talk to six or seven automakers. They're looking at high-speed, high-friction environments such as reciprocating parts in engines; anything that could take advantage of the material's low mass and high wear-resistance." Flying parts of engines, including connecting rods, piston parts, even automotive turbochargers have been investigated because anything that moves so much could improve fuel consumption if it were lightened. One major U.S. automaker is said to have considered making wrist pins, but the concept has not been realized.
"Probably the best applications for boron carbide-aluminum are wear parts," said Halverson. "Simple geometric components that must resist high wear, such as pump seals, or ball bearings and races."
Landingham agreed, "It has tremendous wear resistance, so it might be appropriate for something like aircraft brake shoes, because it would carry away frictional heat well. Other components of aircraft landing gear could also be converted to the cermet. Precision machinery might be another application for the cermet. For instance, you could make it into runners that don't wear much, perhaps for use in producing high-precision parts."
It has been reported that the Soviets poured tons of boron carbide over the Chernobyl nuclear reactor meltdown because it is a high neutron absorber; it has high capture cross section. Add that the composite conducts heat well, and it would make sense to use the material as part of the shield wall for a fusion reactor, which currently uses boron carbide-copper composites. Boron carbide-aluminum is much stronger. Another possible use is nuclear shielding for the electronics compartments of robots designed for work in high neutron-flux environments or as a container for nuclear materials.
"One interesting application might be to use the cermet in cutting tools for difficult-to-machine, high-silicon aluminum alloys," said Halverson. "At Livermore, we performed preliminary machining tests on one such alloy and it performed as well as tungsten carbide." Although machining heats the tool tip to temperatures higher than the melting point of aluminum, "it looks as if the cutting edge is being microheat-treated as it cuts, forming hard ceramic phases that ablate off and so sharpen the edge," he said.
Other applications might take advantage of the fact that the aluminum matrix has a vibration dampening effect, which is important for controlling acoustic emissions. The cermet also has high secondary electron-emission characteristics, making it a good emitter--one with a high ratio of electrons out to those in.
"The cost of making these materials is determined in part by the cost of boron carbide, which is about $20 per pound in large volumes, depending on the grade," said Landingham. "Of course, until the demand goes up for boron carbide, the price will remain high and the supply low. It's something of a vicious cycle. Dow Chemical, for example, is said to be ready to convert a large production facility to boron carbide should the demand for it rise. It's possible that greater production volumes could drop the price to $5 or $6 per pound. If that happens, the cermet cost could go to $10 to $15 per pound. When people tell me that price is still too high, I remind them that the cermet is one-third the weight of steel, for instance, so you get three times the material for a given weight, which makes it more competitive."
But for now, all that is still in the future. Fully qualified and reliable mechanical and physical properties figures are needed to convince engineers to consider the cermet for use. Said Pyzik, "We've been spending a lot of time trying to characterize the properties of samples of different compositions, made by different processing techniques. We're finally nearing the stage that when somebody comes to us with an application, we'll know enough to tell them right away whether it is possible or not."