Understanding bottom wear in coreless induction furnaces.
In the late 1940s, large capacity, high-powered, line frequency, careless induction furnaces were introduced as the prime melters for melting iron. In the case of gray iron, the lining of the furnaces would last for about 10-16 weeks before they had to be replaced due to bottom lining wear.
The lining wear profile, which resembled an elephant's foot, suggested that whatever causes the wear is selective to the lower part of the furnace. The furnaces were lined with a dry vibrated silica ([SiO.sub.2]) refractory grain bonded with boron oxide ([B.sub.2][O.sub.3]), which at the time, was introduced in the form of boric acid ([H.sub.3][BO.sub.4]).
This article will look at the possible causes of coreless induction furnace lining wear, and explain why melting ductile iron will cause a more aggressive reduction of the silica lining than when melting gray iron.
Reduction of Silica by Carbon
Historically, lining wear has been attributed to the reduction of the silica lining by carbon dissolved in the melt. The reduction of a silica lining by carbon (C) dissolved in the molten iron produces silicon (Si), which dissolves in the molten iron and carbon monoxide (which leaves as a gas). Whether or not this reaction will take place is determined by the use of the free energy equation.
The free energy of formation ([DELTA]F) is equal to the free energy of formation at equilibrium ([DELTA][F.sup.o]) minus the multipicand of the gas constant (R), the temperature (T) and the natural log of the equilibrium constant (K), as shown in the equation:
[DELTA]F=[DELTA][F.sup.o] -RTIn (K)
The free energy of formation for the reduction of silica by carbon is shown in the equation below:
[DELTA]F=131300-73.96T + RTIn (%Si/%[C.sup.2])
If the free energy of formation calculates to a positive number, the reduction of silica will not take place. If it calculates to a negative number, the reaction of silica will occur. If it is zero, the reaction is at equilibrium. Thus, the temperature at which the reaction is at equilibrium represents the temperature above which the reduction of silica by carbon will begin.
The equilibrium temperature can be calculated by rearranging the above equation, setting F to zero and solving for T. Therefore, a gray iron with 3.5% C and 1.9% Si will reduce a silica lining when melting at 2775F (1524C) because the melt temperature is above the equilibrium temperature of 2584F (1418C). Likewise, a ductile base iron with 3.8% C and 1.8% Si will reduce a silica lining when melting at 2650F (1454G) because the melt temperature is above the equilibrium temperature of 2567F (1408C).
Although the entire lining is exposed to molten iron at temperatures above the equilibrium temperature during the latter stages of the melt, only the bottom portion of the lining wears. Since a coreless furnace is overfilled above the top of the power coil, there is more metal above the electrical center of the coil than below the electrical center of the coil. If the coil induces the same amount of power in each portion of the melt, it follows that the heating rate will be higher in the bottom half than in the top of the furnace.
For example, in a 7000kW, 35-ton coreless furnace, the metal in the bottom of the furnace will gain energy at 27.3F per minute while the metal in the upper part of the furnace gains energy at only 19.5F per minute.
Back charge to a Heel Melting
Consider what takes place during the melting of iron in a 35-ton, 7000kw coreless induction furnace with a 7000-lb back charge. As the charge submerges in the molten iron energy is transferred from the molten metal to the charge, causing the temperature of the charge materials to increase and the temperature of the molten metal to decrease. Since the charge materials tend to float in the molten bath, the bottom portion of the bath is totally liquid while the upper portion of the bath is a mixture of a liquid and solid.
As energy is applied to the lower portion of the bath, it will superheat the liquid molten iron, raising its temperature. As energy is applied to the upper portion of the bath, it can not superheat the molten iron until sufficient energy has been provided to accommodate the heat of fusion requirements to melt all of the charge. Only then will the temperature in the upper part of the furnace increase,
As the molten metal in the bottom of the furnace gets hotter than the molten metal in the upper part of the furnace, heat is transferred from the hotter iron to the colder iron so that the entire molten bath becomes the same temperature at the end of the melt cycle. Since the metal in the bottom of the furnace gets hotter than the tap temperature, preferential reduction of silica by carbon takes place in the bottom of the furnace.
Gray Iron vs. Ductile Base Iron
The free energy for gray iron with 3.5% C and 1.9% Si at a melting temperature of 2775F (1524G) is -8552 calories. The free energy for ductile base iron with 3.8% C and 1.8% Si at a melting temperature of 2600F (1427C) is -1439 calories. Since the free energy is negative in both cases, the silica lining will be reduced by carbon in the molten iron. Since the free energy is more negative with gray iron, gray iron should reduce a silica lining more than ductile base iron.
However, such is not the case. For example, in a 35-ton, 7000kW line frequency coreless induction furnace, the lining throughput was 8000 tons of gray iron and only 2600 tons of ductile base iron. Since this is contrary to the free energy predictions, either something is wrong with the free energy predictions or the wear of a silica lining by ductile base iron involved other wear mechanisms.
Previous studies have identified eight causes for bottom erosion in a coreless induction furnace with a silica lining. They include:
* installation of the working lining;
* sintering of the working lining;
* chemical reactions of the initial charge materials;
* design of the form's taper section;
* excessive, uncontrolled superheating of a low metal heel (including bridging of charge materials);
* altering the thermal gradient of the lower sidewall and floor;
* metal finning and localized metal saturation;
* presence of the nonferrous metals in the charge.
While all of these factors can influence lining wear, the reason why ductile base iron wears the lining more than gray iron continues to elude explanation. Thus, the search for the silica reduction mechanism, needs to look at the differences when melting ductile base iron and gray iron.
Both gray and ductile base iron use returns, scrap steel and pig iron in their charge makeup. Since the steel and pig iron are essentially the same for each iron, the returns must hold the answer. Since gray and ductile iron returns both contain the major elements of carbon, silicon and manganese in relatively the same amounts, these elements can be eliminated. This leaves the minor elements, which include sulfur and phosphorous in gray iron and copper and magnesium in treated ductile iron.
Because of its reactivity, attention is directed toward magnesium and its compounds. Treated ductile iron returns contain about 0.0300.045% Mg, which can be found in numerous status:
* alloyed with the iron in the returns;
* solidified in the returns as magnesium;
* suspended in the returns as magnesium sulfide;
* suspended in the returns as magnesium oxide.
Since the lowest free energy state for magnesium is magnesium oxide, the magnesium arid magnesium sulfide are constantly looking to react with something from which it can obtain oxygen. One source of oxygen is the silica lining.
Since the charge is preferentially melted while it is submerged in the molten bath, the magnesium that had solidified in the returns melts, vaporizes and becomes suspended in the molten iron. The magnesium sulfide that was trapped in the returns is freed and becomes suspended in the molten iron. Where do the magnesium and magnesium sulfides go when they are freed from the solid charge pieces?
The generally accepted metal flow within a coreless induction furnace is that the molten metal in the center of the furnace flows upward above the electrical center of the coil. Likewise, the molten metal in the center of the furnace flows downward below the electrical center of the coil.
A study made by the British Non Ferrous Metals Research Assn. on the magnetodynamics within a coreless induction furnace confirms that the generally accepted stirring pattern exists with a symmetrical load. A symmetrical load is one wherein the bottom of the melt extends below the bottom of the power coil in an amount equal to that which the melt extends above the top of the power coil.
However, a conventional coreless induction furnace is characterized by a non-symmetrical load, The bottom of the melt is at, or slightly above, the bottom of the power coil while the top of the melt is above the top of the power coil. Figure 1 is a plot of the British Non Ferrous Metals Research Assn.'s data showing the stirring pattern for a non-symmetrical load.
What is important about the stirring pattern is that when solid charge pieces are immersed in the molten iron, the liquid iron is directed downward and to the center of the furnace where the magnesium and magnesium sulfide are swept into the bottom stirring pattern. Herein, they are superheated and circulate within the bottom portion of the furnace. Then as the magnesium and magnesium sulfide come in contact with the lining in the bottom of the furnace, they reduce the silica.
Reduction of Silica by Magnesium
The free energy for the reduction of silica by magnesium in ductile base iron with 3.8% C and 1.8% Si at a melting temperature of 2600F (1427G) is -75,712 calories. Since the free energy for the reduction of a silica lining by the magnesium is more negative than the free energy for the reduction of a silica lining by the carbon in the gray iron, it follows that the silica lining wear will be more aggressive when melting ductile base iron than when melting gray iron.
The free energy for the reduction of silica by magnesium sulfide in ductile base iron of the same properties is -16,342 calories. Since the free energy for the reduction of a silica lining by the magnesium sulfide is more negative than the free energy for the reduction of a silica lining by the carbon, it again follows that the silica lining wear will be more aggressive when melting ductile base iron.
Silica Loss Calculation
Now that it has been shown that the magnesium in the returns can reduce the silica lining, it becomes necessary to determine if there is enough magnesium to account for the silica loss. To determine how much silica will be reduced by 0.035% Mg in the returns, consider the following assumptions:
* the back charge consisted of 55% returns;
* all of the magnesium that is oxidized gets the oxygen from the silica lining;
* 2600 tons of iron is melted.
By calculation, 2378 lb of silica would be reduced, which at a density of 134 lb per cu. ft, amounts to 1775 cu. ft of silica. Since the typical volume loss of silica in the bottom of the subject furnace is about 15-20 cu. ft, it appears that magnesium can account for a major portion of that silica loss.
Carbon Theory Challenged
The historical explanation that silica linings are reduced by carbon in the molten iron does not fully address the situation in a coreless induction furnace melting ductile base iron. While the reduction of a silica lining in a careless induction furnace melting ductile base iron involves the reduction of silica by carbon, the primary reduction mechanism appears to be the reduction of silica by magnesium and magnesium sulfide.
For More In formation
Visit www.moderncasting.com to read "Reducing Elephant's Foot Erosion in Careless Induction Furnaces," D.C. Williams and Y.H. Ko, MODERN CASTING, July 2000, p.22-25.
About the Author
William J. Duca is the president of Duca Manufacturing & Consulting, Inc., Boardman, Ohio. Jerry Beaird is the melt superintendent at Rochester Metal Products Corp., Rochester, Indiana.
|Printer friendly Cite/link Email Feedback|
|Comment:||Understanding bottom wear in coreless induction furnaces.|
|Date:||Jul 1, 2003|
|Previous Article:||Survey results reveal 'razor-thin' cost picture: AFS-NADCA 2003 Confidential Operational Cost Survey. (Executive Memorandum).|
|Next Article:||Do you make a difference? (2003 Charles Edgar Hoyt Memorial Lecture).|