A study of throat clogging in the channel furnace.

A Study of Throat Clogging in the Channel Furnace

Many iron foundries have changed the methods by which they melt and/or hold molten iron. In many cases the channel induction furnace (vertical or drum-type) has proven effective for primary melting, often during off-peak electrical periods, and for holding large quantities of metal for added flexibility. Homogeneous metal chemistry and temperature control are characteristic of the channelfurnace. As production demands changed, a variety of operational problems have been resolved. One problem that has not been resolved is the clogging of the throats of the inductors of channel furnaces.

An AFS research sub-committee was developed to formalize a program to evaluate buildup and clogging of inductor throats. Approved in December 1987, the project was awarded to the University of Missouri at Rolla to perform basic testing.

Project Parameters

Clogging samples were obtained from the throats of the inductors of a variety of channel furnaces, and questionnaires were completed on melt conditions for each sample. They were then assigned an AFS 8D number and sent to the University of Missouri at Rolla, Ceramic Engineering Dept.

At this time physical tests have been completed. Initially, the chemical analysis of each buildup was obtained by using an X-ray spectrometer. The magnetic iron was separated out and the various oxides were detected. Sulfur and phosphorus were tested by wet chemistry methods.

Cathodoluminescence and optical microscopy were used to help classify each buildup. When viewing a polished sample section, the microscope was used to identify grain size and orientation. Cathodoluminescence uses a stream of electrons that are shot at the refractory buildup surface at a specific angle. This caused the various grains and mineralogical compounds to luminesce or glow different colors. These colors were classified as specific mineralogical phases such as spinel or periclase. By using cathodoluminescence, a gradual change in the buildup chemical composition could be detected. This was the most revealing test in explaining the buildup growth.

An electron probe analyzer was used to identify elemental concentrations in a particular zone or grain. These results will be used to verify the findings in the cathodoluminescence study and observe any possible nucleation sites where the buildup might begin growing.

Sample Results

In this case the ceramic buildup was a composite of magnesia and alumina in the form of spinel. Other compounds present were forsterite and periclase.

Sources of possible insoluble constituents in this sample include: * Type and sizing of steel scrap ([AI.sub.2][O.sub.3], [Fe.sub.2][O.sub.3]); * Amount of remelted treated ductile iron gates (MgO) and risers which affect the amount of MgO available; * Erosion of a refractory wall from the continuous variation of the molten metal bath line ([AI.sub.2][O.sub.3], [SiO.sub.2]); * Ferroalloys, such as calsifer, may contribute to the CaO and [SiO.sub.2] contents.

In one case, for example, it was determined that buildup occurred as a result of residual MgO carried over from melting treated ductile returns, combining with silica [SiO.sub.2] from the molten metal and ferroalloy additions, and iron oxide (FeO) from the rusty steel returns to form the forsterite. The spinel was more predominant and formed as a result of the MgO combining with the alumina [AI.sub.2] [O.sub.3] from the eroding condition of the sidewalls or from gunning operations.

Buildup Factors

In one particular study, samples were obtained from a 30-ton vertical channel furnace equipped with a water-cooled single loop inductor used to melt ductile base and gray iron. The inductor refractory was a 92% MgO dry ram, the floor refractory was a 88% alumina wet ram with a 9.0% [Cr.sub.2][O.sub.3] and the sidewalls were of a 90% [Al.sub.2][O.sub.3] brick refractory. Table 1 shows the chemical analysis and ceramic phases present in this particular sample. This study also demonstrated the following: * As the concentration of metallic oxides and sulfides within the molten metal increases, there is an increased propensity to form a ceramic buildup or clog in the inductor channels or throat openings.

Table : 1 Chemical Analysis of Sample Buildup

  [SiO.sub.2]   [Al.sub.2][O.sub.3]   [Fe.sub.2][O.sub.3]
  31.0%         17.4%                 3.5%
  MgO           CaO                   MnO
  45.2%         1.6%                  0.29%

Magnesia (MgO), alumina ([Al.sub.2][O.sub.3]) calcia (CaO), silica [SiO.sub.2], and iron oxide all contribute to the clogging condition, as well as the CaS and [MgS.sub.2]. * A change in the scrap quality in the primary melting unit, regardless of the furnace type--coreless, channel, cupola, arc or a gas-fired--will affect the generation of more oxides. This change could be related to a chemistry change or may represent a physical size difference (e.g. increased surface area). * Buildups form in furnace locations that experience significant heat losses. Typically, these areas correspond to the throat opening and the inductor channels. These severe heat losses may be due to lack of insulation behind the working lining, or because of the cooling effect of the water tracings or bushings. As superheated molten metal is discharged from the channels, the most severe temperature gradient is created between the refractory hot face and the steel shell cold face. To help alleviate the heat loss in the throats, thicker refractory walls may be needed. * Metallic oxide carryover may occur as contaminants oxidize when additions of various alloys are made to the molten metal in the transfer vessel, prior to filing the channel furnace. Undissolved alloy additions may also contribute to the clogging if the temperature in the transfer vessel is not high enough. * Increased production of ductile iron will influence clogging. As ductile iron is treated with magnesium or cerium, a resultant by-product (MgO or CeO) is insoluble. When more treated ductile returns are remelted, the concentration of MgO or CeO increases. MgO and CeO will tend to form complex oxide combinations with [Al.sub.2][O.sub.3] or [SiO.sub.2]. * A low average molten metal temperature tends to increase the buildup as well, especially over a weekend.

Sources of Clogging

As might be expected, a variety of constituents can contribute to the clogging of the throats in channel furnaces. Residual oxides from remelting solid charges can be a source of buildup. For example, as steel is remelted, iron oxide ([Fe.sub.2][O.sub.3]) and alumina ([Al.sub.2],[O.sub.3]) are generated. As treated ductile iron is remelted, magnesia (MgO) is often present as spinel (MgO [Al.sub.2][O.sub.3] or olivine (Mg, [Fe.sub.2] [SiO.sub.4], and are often formed as a result of these residuals. With larger variations in the molten metal levels in the furnace, the potential for pulling oxides and sulfides down into the throats and inductor channels is possible.

Spent foundry sand can also be a possible source for the silica ([SiO.sub.2]) content in the buildup. Mullite (3Al.sub.2][O.sub.3], [2SiO.sub.2]) is frequently deposited as a result of sand contamination. It is believed that a sand addition in some circumstances can actually soften the buildup, allowing for easier removal. This remains to be proven.

Another constituent is rust or iron oxide ([Fe.sub.2],[O.sub.3]) which is detrimental to the refractory in the uppercase area. The iron oxide will react with the bonding matrix of the refractory and cause erosion. As this erosion occurs, the refractory base such as alumina ([Al.sub.2],[O.sub.3]) becomes available to combine with the FeO iron oxide. Many different buildup compositions are generated from this source.

The premature erosion of refractories in molten metal transfer equipment or the deterioration of refractories in the primary melt unit, (i.e. cupola or coreless furnace), can cause a large quantity of refractory oxides to be carried over into the channel furnace. Often 60% and 70% alumina refractory products are used in transfer ladles, runner systems and fill bowls. As erosion progresses, more oxides are transferred to the channel furnace over a period of time. So the fact that the residual oxides are being dumped into the channel furnace may not be readily apparent. It may be necessary to consider higher alumina products, but this will require more insulation due to an inherent increase in thermal conductivity.

The size of the scrap or charge material will also contribute to the throat buildup. The smaller the scrap size, the more surface area that is available to oxidize. As the bath level drops, there is the possibility of pulling these oxidized particles into the inductor channels.

To minimize oxidation the atmosphere above the molten metal bath must be kept isolated from the surrounding air. This will help reduce the amount of oxygen available to oxidize the bath. But, regardless of the preventive measures used, a certain amount of oxidation of the molten metal will take place.

To reduce this source of buildup, the use of calcium-bearing compounds such as fluorspar, calcium carbide or limestone will also contribute to the buildup as these cause an erosion of either the refractory in the molten metal transfer equipment or the channel furnace uppercase refractory. As this occurs, the amount of CaO and other refractory oxides will float within the metal.

Various ferroalloys can also cause trace amounts of aluminum, magnesium or rare earth metals to readily oxidize and remain in the slag/metal. These often combine to form MgO [Al.sub.2][O.sub.3] spinel with another rare earth oxide chemically attached. The presence of sulfides, whether inherent to the molten metal or as a by-product of any treatment step, will influence the buildup. Typically, oxy-sulfides have been known to float within the molten metal.

Clogging Solutions

The first step in solving a clogging problem is to identify and minimize the use of buildup constituents, all sources including steel scrap which must be used in the production of ductile base iron, deserve careful scrutiny.

The addition of silica sand to a low metal heel may be advantageous on certain types of buildups containing FeO, but the overall results are marginal. The addition of silica might "soften" the buildup by reacting with the iron oxide present. But there is a definite possibility that silica may accelerate the buildup by reacting with the alumina or magnesia within the matrix. The addition of ferrosilicon will be a slight improvement over the addition of silica sand, but similar results should be expected.

One approach to minimizing throat clogging may be periodic superheating of inductor channels to help control certain buildup applications. When the molten iron is being maintained at a lower temperature such as 2400-2550F, this periodic "spiking" of the inductor power will cause the molten iron temperature to increase, which then allows the buildup to soften or melt. The lower metal temperature condition buildup which contains levels of iron oxide (FeO) and silica (Si[O.sub.2]) that exceed 10%; therefore periodic superheating may help.

Currently, individual foundries may elect to superheat for a 5-10 min period every hour. Other foundries may have a control unit built into the power cabinet that cycles to a "superheat" or "melt" mode. Yet other foundries will only superheat the inductor channels during the weekend hold period. In all cases, periodic superheating is beneficial in keeping the channels open during continuous operation, but it should be noted that whatever superheat method is followed, it should be initiated at the beginning of the inductor refractory campaign. Periodic superheating will not allow the molten iron to remain idle, which will reduce the amount of oxides and sulfides depositing on the refractory surface. Superheating at higher temperature applications when molten iron exceeds 2800F, is not as effective when compared to superheating at lower temperatures.

Carbon additions in the form of silicon carbide, or carbon pick-up from introducing a "green" wood pole into the throat, may reduce the amount of silica in the buildup and create a carbon boil. It must be noted that extreme caution be taken during this procedure since the reaction is very violent. For this reason, the carbon additions can be considered as a mechanical removal method. This treatment should be performed on a minimum molten metal heel level and is not recommended unless the furnace manufacturer has granted approval.

It has also been observed that build-ups occur in areas of the furnace where the most heat loss occurs. Typically, this would correspond to the throat or inductor locations. In the inductor, a refractory with high thermal conductivity is often recommended due to thin walls between the channel and the bushing. In the throat, the thickness of the wall is much less than in the uppercase sidewalls due to the furnace design.

Alternative measures to consider for reducing inductor buildup may include the following: * Consider switching refractories from a magnesia base to an alumina/mullite forming mix. On lower kilowatt-rated inductors (<750 KW), this exchange could prove beneficial depending on the given melt conditions and the temperature of the molten metal in the furnace. * Another alternative is to modify the channel loop form by enlarging its cross-sectional dimensions. Doing this will require that the power rating be increased thus, increasing the metal flow through the channel. These changes should be done only after consulting with the furnace manufacturer, as increased demand will be placed on the equipment.

This will not prevent buildup from occuring, but will allow for extended inductor operating times. Typically, throat walls, have little (if any) insulation used as backup to the working refractory. The additional cooling from the water tracings cause a severe temperature gradient which allows the insoluble oxides and sulfides to solidify as they pass over this area. By adding insulation and/or reducing cooling water, some foundries have helped alleviate the clogging problem by reducing heat loss. * The use of fluxes may also help reduce buildup but it is important to control these additions. Fluxes cannot distinguish between buildup components and refractory oxides. Consequently, it is imperative to add control measures to limit the amount of flux added to a molten metal heel.

PHOTO : Cathodoluminescence (CL) micrograph of a buildup showing a large crystal of corundum (dk red to black), bladed crystals of hibonite or [Ca.sub.6] (dk green) crystallized on corundum. (12.5x)

PHOTO : CL micrograph of spinel + melilite-rich buildup showing zoned isometric common spinel crystals (green) in a melilite (blue) matrix. (25x)

PHOTO : CL of periclase (blue) and oldhamite (yellow) in MgO rich sample. Oldhamite may be formed in highly sulfurizing conditions. (25x)

PHOTO : CL micrograph of a buildup sample showing anorthite (yellow), corundum (red) and hercynite (black) crystals. (25x)

PHOTO : CL micrograph of MgO-Si[O.sub.2] rich buildup showing fosterite (red to orange) and spinel (green) crystals. (25x)

PHOTO : MgO rich buildup showing internally zoned periclase crystals with a minor amount of oldhamite (yellow).(25x)

PHOTO : Cathodoluminescence of unusually zoned periclase crystals in MgO-rich buildup. (25x)

PHOTO : CL micrograph of anorthite fibers (yellow) and corundum (red) crystals. This texture indicates supercooling or undercooling of CaO-[Al.sub.2][O.sub.3]-Si[O.sub.2] melt or liquid resulting rapid crystallization of anorthite. (25x)

PHOTO : A CaO-[Al.sub.2][O.sub.3-]Si[O.sub.2] rich sample buildup showing hibonite (green) corundum (red), anorthite (It blue) crystals enclosed in glassy (It brown) material. (12.5x)