The high temperatures used for the direct carbo-thermic synthesis of, for example, tungsten carbide (WC) or the sintering of alpha-SiC and a variety of other advanced non-oxide ceramics require relatively high energy input. Materials of construction are limited and, frequently, expensive. Additionally, the cost of cooling, though often overlooked by most furnace users, can be comparable to and sometimes greater than the cost of the energy used to heat the product.
Inert blanket gas consumption, often required for the quality of the product, if not for the protection of the refractories, is another significant operating cost.
Other thermal processing challenges include maintenance, downtime and rebuilds. Processes requiring or producing chemically aggressive atmospheres—such as the CO2 produced in the low temperature end of the WC synthesis, or the B2O3 off gassing in the production of BN and TiB2 or H2 used simply as a buffer—attack the furnace refractories as well as the conveying surfaces, the product containers and heating elements.
Material property advances, such as finer grains and good grain growth control and, in the case of synthesis, good chemical stoichiometry control, require a rapid and uniform approach to the maximum process temperature. Additionally, to make a uniform high temperature activated material, the time at temperature must be long compared to the time when some of the material was hot enough to react, sinter, etc. and some was not. The struggle to produce fine grain materials (powder synthesis or part sintering) therefore usually involves a battle between productivity and uniformity.
Conventional high-temperature continuous furnace designs typically seek to pack as much productivity into as small an envelope as possible to minimize energy consumption and purged volume. Load geometry is also frequently designed to minimize product-container surface area. For synthesis processes that generate reactive gases, a minimal surface reduces the replacement cost of the product containers. Together, these designs minimize operating and capital costs based on the stated principals. Unfortunately, this is sometimes at the cost of transient thermal uniformity.
Recently, an improved furnace design* was developed that simultaneously sets a new standard for all of these criteria. The furnace has been successfully used to synthesize an extremely unforgiving ceramic oxide powder, and it is now being adapted for continuous ultra-high temperature synthesis and sintering. The new system offers a minimized carrier mass, improved material handling and processing, and improved time-at-temperature uniformity, making advanced ceramic production more efficient and profitable.
Another reason that traditional load containers are relatively massive is because they are doing several things at once. For instance, they might carry a chemically aggressive reacting material while transmitting the force to overcome the friction drag on all of the downstream carriers. They, along with the hearth, also provide the stable conveyor behavior (stable, precise shapes to avoid buckling) required for continuous operation. Again, the traditional way to cope with these demands is additional cross-section (i.e., mass), but the costs can quickly add up in terms of energy use, part replacement, hearth wear and product uniformity.
The new furnace’s carrier breaks this job apart. The mass of carrier in thermal contact with the load is minimized to just that required to carry the load. The compression, abrasion and precision-geometry requirements of the task of transmitting the pushing force are handed off to the peripheral pusher blocks. In a variant called the “cold rail,” the pusher blocks are at least partially removed from the thermal/chemical environment (see Figure 1). Although massive in much the same way and for the same reasons as the traditional carriers, the carrier exerts a positive rather than negative influence on the economics of operation and quality of production.
The now simple job that the carrier components must handle allows the design to be modified to decrease replacement costs. Other maintenance costs are reduced by decreasing the wear and the criticality of wear of the load-force transmitting parts of the system.
Another advantage of this type of carrier is that it allows a wide, low profile load to be processed economically. To a good first approximation, every halving of diffusion distance cuts the time to reach a specific level of uniformity by a factor of four. With the new furnace, a 1-in. (25 mm) deep, 40-in. (1 m) wide load can be processed more economically and with better quality than a traditional 4-in. (100 mm) deep, 10-in. (250 mm) wide load. A thermal model comparing the production and quality of a traditional pusher tunnel kiln and the new furnace design is shown in Figures 2 and 3.
Most pusher furnace systems recirculate their load carriers in a horizontal plane, requiring a return track that consumes an additional footprint within the plant. The new system, however, recirculates in a vertical plane, allowing manufacturers to conserve plant space.
In the new system, a forward moving set rides on top of a backward moving set. The product on the carrier is pushed into the furnace and is heated as it moves through the first level. It is then lowered while in the hot zone and pushed back in the reverse direction. The incoming trays and product pass directly over the heated material coming back out in the opposite direction and are partially heated by the hot materials that are moving toward the exit (see Figure 4). This conserves large fractions of the energy normally required by almost any high temperature process while also substantially reducing cooling requirements. Additionally, the atmosphere is easier to control in a tube that is only open on one end than in a design (i.e., a conventional pusher tunnel kiln) that has openings on both ends.
A simple re-radiating-layers model was constructed to compare a low-profile, internally recirculated system with a once-through, double high system. The latter is analogous to a twin-hearth processing a load that is twice as tall as the new furnace system’s load. The model was based on the processing of a thermally conductive powder in a large number of layers and was a reasonable approximation of the temperature gradients experienced in sintering setter loads of relatively small parts.
The model is an explicitly integrated transient finite difference equation (see Equation 1, above), where i and j are the discretization indices in the z (length) and y (vertical) directions, respectively; mCp is the mass of a volume of load, with dimensions W, Dz and Dy multiplied by the heat capacity; and v is the velocity (push rate) of the load. Velocities less than zero indicate material being pushed back towards the entrance. Gradients across the width (W) are neglected.
The parameters used for the new furnace system are shown in Table 1. The raw temperature profile, shown in Figure 5, is hard to interpret because it folds back upon itself, it can be unfolded through a somewhat tedious rearrangement of cells addressing the original set (see Figure 6).
In comparison, the single pass but otherwise equivalent furnace, with the notable difference of having 50% the production rate of the previous case, gave the results shown in Figure 2. It’s interesting to note the obvious difference. Where the conventional furnace produced a large range of times-at-temperature, the new furnace system, with a little rephasing of the temperatures, gave a much better time-at-temperature uniformity, as shown in Figure 3.
Other applications currently being evaluated include non-oxide ceramic powder synthesis and part sintering (aluminum nitride, silicon nitride, silicon carbide, boron nitride), WC powder and other hard-metal and refractory-metal synthesis and sintering, mesophase carbon beads (MSCB), dielectric powder calcinations, base metal electrode (BME) multilayer ceramic capacitor sintering and soft ferrite sintering. (The last two take advantage of the fact that the low profile load and enhanced mass trasfer, combined with reliable load tracking, provides the capability to discretize the atmosphere along the length of the furnace more efficiently in terms of the degree of isolation versus gas consumed.) Higher temperature powder metal part sintering and even some heat-treating applications are also candidate applications—as well as any process that requires a uniform heating environment.