ORNL Results Help Firm Decide to Market Silicon Nitride

The Norton Company, a ceramic manufacturer in Worcester, Massachusetts, had a problem. The company had developed a new silicon nitride material that might be more suitable for use in high-temperature gas turbines than its current product. Its ceramicists had systematically adjusted the chemistry of its commercial silicon nitride product, NT154, and produced a new ceramic, NT164. However, the Norton researchers were not sure if they had developed a product that was good enough to be marketed. So Norton turned to ORNL for help through the user program at the High Temperature Materials Laboratory (HTML). Norton engineers in collaboration with ORNL researchers proposed through a user project to test the mechanical properties of NT164 at high temperatures and compare them with the results they had obtained earlier on NT154. In addition they wanted to analyze the two ceramic materials using the HTML's powerful microscopes. The challenge was to find differences in the microstructure of the two materials that might account for differences in mechanical properties.

As a result of the ORNL findings, Norton decided to commercialize NT164, and Michael Jenkins, Matt Ferber, and Ted Nolan, all of ORNL's Metals and Ceramics Division, received a 1992 Martin Marietta Energy Systems Technical Achievement Award. They were cited "for significant materials characterization and analysis contributions to the development and commercialization of a high-performance silicon nitride ceramic."

Silicon nitride is the preferred material for components of high-temperature gas turbines because of its combination of important properties. It is very strong, hard, and highly resistant to wear, oxidation, and decomposition at high temperatures. It is incredibly resistant to thermal shock--large changes in temperature, such as a drop from 1200 to 20 degrees Celsius in a matter of seconds--that would cause ceramics such as alumina and silicon carbide to shatter.

Because the ceramic gas turbine could operate at higher temperatures than the nickel-based superalloy engine, it would use fuel more efficiently and produce less pollution. However, such an engine has not been produced commercially yet because of problems in fabricating dense, precisely shaped components that are reliable at high temperatures. To overcome this problem Norton is developing new silicon nitride materials by adjusting the chemistry of the silicon nitride (Si3N4) powders and sintering aids (e.g., oxides of yttrium and aluminum) used to form the material and make it dense.

The ORNL researchers tested the Norton ceramics for high-temperature creep deformation--a gradual change in length, or strain, in a material as a result of prolonged exposure to stress and high temperatures. They also evaluated each ceramic for static fatigue--the time it takes for a material to fail under a constant stress--to determine its long-term reliability.

The ORNL researchers subjected both materials to tensile tests, applying stresses of 100 to 200 megapascals (MPa) at temperatures of 1260 and 1370 degrees Celsius, the temperature that turbine components must endure. Dumbbell-like tensile specimens of each material held through SupergripTM couplers were pulled at each end and heated to high temperatures at the center. ORNL results showed that, under the same conditions of 100 MPa and 1370 degrees Celsius, Norton's commercial ceramic deformed at a higher rate and failed after 1200 hours, whereas the new material survived for 4800 hours.

"We found that NT164 lasted four times as long yet accumulated three times as much strain as NT154," says Jenkins, now with the University of Washington in Seattle. "The new ceramic clearly was more resistant to creep degradation and static fatigue and more reliable in the long term than the already commercialized material."

To determine the reason for the mechanical superiority of the new material, the ORNL researchers characterized the microstructure of both ceramics using transmission electron microscopy.

"What we found was that NT164 had very little intragranular cavitation," Jenkins says. "It had little of the Swiss-cheese-like appearance of the NT154."

Commercial silicon nitride ceramics can also be processed to be self-reinforced rather than reinforced with silicon carbide whiskers (which may pose a health hazard if they are inhaled and deposited in the lungs). By using sintering aids such as oxides of rare earths (e.g., yttria, ytterbia, and scandia) and applying proper processing temperatures and pressures, long columnal grains are grown among the uniformly sized grains. The long grains act like whiskers, bridging cracks and toughening the material. During this processing, amorphous, or noncrystalline, material may form between the ceramic's crystalline grains, areas called grain boundaries.

When silicon nitride is exposed to high enough temperatures, this glassy material softens, allowing creep deformation by mechanical and diffusional mechanisms. In mechanical deformation, the silicon nitride grains slide relative to each other; creep cavities or holes may develop, and the ceramic becomes deformed. In diffusional deformation, elemental material (e.g., silicon and nitrogen) may dissolve into the glassy material, forming holes or cavities in the silicon nitride grains, and redeposit or unite with other grain-boundary elements. This elemental transport took place at junctions between two grains but not at three-grain junctions where enough glassy material was trapped and crystallized. The dissociated elements cannot move through the crystalline regions in these "triple points," which are formed during processing.

"By controlling the chemistry of the starting material and sintering process for NT164, Norton researchers almost eliminated the formation of the glassy material at the two grain boundaries," Jenkins says. "We found that the glassy regions in NT164 were only about one nanometer thick compared with several nanometers in NT154. Norton researchers made grain boundaries so thin that the bulk of the glassy material was forced into triple points where it becomes crystalline."

This collaborative work between ORNL and Norton, says Jenkins, is a good example of how the diverse and unique user facilities and personnel available at the HTML can help industrial firms solve problems.--Carolyn Krause

Microwave-Processed Silicon Nitride is Cost-Effective

Using microwaves, three Oak Ridge researchers have developed a cost-effective method of making ceramic parts for advanced engines for transportation. The Oak Ridge technique produces silicon nitride parts that cost less and are denser than parts made by conventional processes under ordinary conditions. The denser the material, the stronger and usually more fracture resistant it is.

According to Terry Tiegs of ORNL, one of the developers of the technique, applications include components for engines operated at high temperatures, such as turbocharger rotors, valves and valve parts, and pump seals. Other uses could include tools to cut metals and dies for forming aluminum beverage cans.

Silicon nitride is the ceramic material of choice for components of high-temperature engines being developed to improve the fuel efficiency of cars and trucks. It is highly resistant to wear, deformation, oxidation, and decomposition at high temperatures, and it is also incredibly resistant to thermal shock--large changes in temperature that would shatter other ceramics. In fact, the latest silicon nitride materials have been shown to have outstanding characteristics for rotors and stators in gas turbines for cars and trucks and for valve trains in diesel- and gasoline-powered engines.

Some silicon nitride parts that meet the requirements for use in engine applications have been made, but because of the processes used, these components are much more expensive than metal parts. The Oak Ridge process using microwave heating could produce ceramic parts that are economically competitive with metal components. The chief reasons are that the process uses a combination of low-cost raw materials (about one-fourth that of the materials used in other processes) and a simplified processing route made possible by the microwave heating. The process was developed by Tiegs and James Kiggans, both of ORNL's Metals and Ceramics Division, and Cressie Holcombe, a researcher in the Development Division of the Oak Ridge Y-12 Plant.

In the ORNL process, a silicon nitride ceramic is fabricated in a microwave field. Silicon powder mixed with additives in a preformed shape is reacted with a nitrogen-containing gas as the ceramic part is heated to 1200 to 1400 degrees Celsius by microwave power. As a result, nitridation of the silicon (Si) to silicon nitride (Si3N4) occurs. Without removing the parts from the microwave furnace or cooling them down, the parts are then heated to 1750 to 1825 degrees Celsius, making them extremely dense.

With conventional heating, the nitridation and densification steps have to be done in two different furnaces. By using microwave heating to accomplish both tasks, the fabrication times and labor costs are significantly reduced.

According to the developers, microwave heating offers several advantages over conventional heating. Nitridation begins at a lower temperature and occurs at a faster rate. Nitridation and sintering (heating) are accomplished in one continuous process. Densification rates are increased. Finally, thicker parts can be made because nitridation proceeds from the inside out.

Microwave heating of silicon nitride parts has been done in furnaces in Building 4508 at ORNL. The process has been successfully tested on silicon nitride parts containing sintering aids in a cooperative research and development agreement (CRADA) with the Norton Company, the ceramic manufacturer in Worcester, Massachusetts.--Carolyn Krause

Where to?

[ Next article | Comments | Contents | Review Home Page | ORNL Home Page ]