Manipulating encapsulation.

The encapsulation of a metal saturation network within an induction furnace lining can actually help prolong refractory life.

As covered in the first article of this two-part series, the metal saturation of a channel induction furnace lining causes major damage and seriously shortens the life of the refractory. Following the path of a precursor of slag, metal seeping into the interconnected, sponge-like pores of the refractory aggregate can lead to superheating and breach of the lining.

As a newly formed saturation network diffuses into the inductor channel refractory, it couples with the electromagnetic flux from the coil. Accordingly, metal temperature increases, cooled only by the circulating metal in the channel behind it and the steep thermal gradient it encounters as it moves toward the shell and bushing.

At this point, however, the refractory becomes its own worst enemy. With its low thermal conductivity, it cuts off the saturating metal network from the cooling of both the channel alloy and the shell and bushing, effectively trapping metal that is becoming ever hotter with the aid of the electromagnetic flux. In addition, the metal on the leading edge of the network is not only farthest from the bath, it is also closest to the power coil.

As a result, the network begins to stabilize along an isotherm within the refractory because it has become hot enough to sinter the refractory ahead of itself, preventing it from advancing. This sintering continues until virtually all the refractory around the network has been sintered and the invading metal is effectively encapsulated, or completely surrounded by impassable refractory. Encapsulation, then, is essentially a pocket of metal saturated refractory trapped within the electromagnetic field of the power coil.

Using Encapsulation

Once encapsulation has occurred, it can actually be beneficial to furnace operation. Since it is going to occur anyway, it can be manipulated to prevent further deterioration of the refractory, provided conditions in the inductor don't become so abnormal that it is made to fail. In fact, in some cases where encapsulation was effectively developed and sustained, furnace performance was exceptional and lining life well above normal. A properly encapsulated inductor channel will take more abuse and accept higher sustained power than one where encapsulation is imperfect or underdeveloped.

Optimization of encapsulation is a matter of understanding and modifying the inductor, refractory and operational practice to avoid breaches in the encapsulation. Encapsulation creates a fused or highly recrystallized refractory "boot" in front of the naturally occurring saturation network that prevents its further advance and thus stabilizes it. Fused refractories must be considered to be brittle or "hot short" structures, highly subject to tensile or shear stress even at elevated temperatures.

If a highly sintered/fused encapsulation zone is to be formed around the channel and kept in tact, differential stress must be kept to a minimum. Otherwise, cracking and finning will lead to failure. This means that the refractory must not be an entirely rigid body and the backup should remain flexible enough to accommodate differential stress by shifting, thereby avoiding a fracture of the encapsulation zone.

However, wet rammed castable or plastic refractories are unsuitable for this purpose. Only dry vibrated (DVC), heat-setting monolithic refractory compositions will meet these mechanical requirements.

Improving Effectiveness

To facilitate encapsulation's benefits, a change in bushing design to eliminate the normal "hot spot" created by uncooled insulator strips is desirable to avoid repeated metal leaks caused by encapsulation failure in this area. This can be accomplished in two ways. First, a ceramic bushing composed of nitride-bonded or recrystallized silicon carbide can be used. These bushings are suitable for air-cooled inductors and have been used successfully to prevent early shutdown--particularly in copper-based alloy melting where random finning would work through a bushing and sort out the coil. The uniform ceramic composition also effectively eliminates the hot spot, as well as the differential stress that causes encapsulation failure and metal leaks in that critical area. Also, conventional water-cooled bushings can be upgraded by covering the bushing with thin silicon carbide plates cemented in place with a silicon carbide mortar.

The second approach involves a re-design of the bushing to incorporate a water-cooled insulator strip to thermally mask the hot spot.

Once the bushing's thermal anomaly is corrected, the same has to be done for the inductor case. This means ensuring that water cooling is complete, or that it at least produces a uniform smooth curving isotherm around the inside of the shell without hot and cold spots. These irregularities lead to variation in the depth of sinter, which in turn allows differential stress to focus locally and cause encapsulation failure.

A good way to determine if case cooling modifications are needed is to examine a used refractory to see if sinter depth was uniform. In some cases, existing water-cooled areas can be buffered by using insulating papers to achieve a more uniform thermal profile.

It is also important to keep the refractory from "locking" into irregular backup structures. In other words, because of the expansion or detraction that temperature changes cause in normal operation, the inside of the case should be smooth sided, without anything that would tend to mechanically hold the refractory and create a focus of differential stress. Among others, these items might be case patches, large steps in side plates or open case separations.

The channel cross section should also be modified to reduce stress, a circle being the preferable cross-sectional shape. For cast iron melting, the channel form should be made of heavy gauge, low carbon cast or welded steel, fabricated with matching resistivity filler rod with a nominal wall in excess of 1/2 in. thick. Normally, the form would be hollow, but when torch firing will not be used or no liquid iron is available for priming, the inside may be filled with gray iron.

Refractory Control

The amount of porosity in a fully compacted DVC refractory will determine the percentage of pores filled and thus the electromagnetic coupling of the network. As power increases, the tolerable network density decreases, but there is a reasonable range in which satisfactory encapsulation occurs and the unit will run well. When the critical coupling maximum is exceeded, the network internally superheats and circulates, destroying the bonded refractory structure and rapidly eroding the lining.

Sometimes, an interim condition known as "worm" or "rat" holes may develop in the network area. These secondary channels typically follow isotherms and often reconnect to the channel itself or penetrate into and through portions of the throat. They will unerringly follow an isotherm of thermal balance between the electromagnetic flux/channel on one side and the cooled case or bushing on the other.

If, however, they find an anomaly in the thermal balance--a slight hot or cool spot--they may turn and bore a hole right through the case or bushing. Therefore, a uniform thermal profile is even more critical to good refractory performance than to the electrical, mechanical or efficiency considerations of the equipment.

Secondary Effects

Last month, it was suggested that the application of a low vacuum could help reduce saturation formation and its subsequent damage by drying the refractory. The method is imperfect, however, and the tiny amount of moisture necessary to form oxides may still reside in the refractory.

The stabilization and full encapsulation of the metal saturation network that occurs within the same time frame as the recommended dryout of the refractory will actually do as much or, in some cases, more to prevent further damage. A channel that is effectively sealed by a fused refractory is impervious to the passage of any gas, including water vapor.

In addition, the metal-saturated refractory between the encapsulation zone and the inside channel surface is nearly impermeable because of the liquid and solid metal filling the pores. What appears to be a dryout process may in fact be an indicator of encapsulation of the network, and may make the application of the low vacuum necessary only during the initial stages of network development.