Refractory failure in induction furnaces can be limited.
Perhaps the most pivotal difference is that the electromagnetic field passes through the refractory, which is electrically nonconductive, except in the few instances where graphite-bearing crucibles are used. This means that the flux field of the coil will ignore the refractory as long as it remains an electrical insulator.
Additionally, a steep thermal gradient through the refractory is caused by water-cooled power coils together with other furnace components. Induced molten metal circulation and more depth than in most other furnace types also combine to provide a potential for intensified melting conditions. These can easily exceed those in ordinary furnaces. Most critical is the fact that an induction furnace is essentially an energy transfer system. The temperatures achievable by a conductor in its electromagnetic field are limited only by the furnace's ability to dissipate the heat generated. Often, an induction furnace can rapidly reach the failure temperature of even the highest melting point refractories without warning to someone unfamiliar with the conditions.
An induction melting furnace is perhaps the most exacting melting tool available. It is amenable to computer control, electrically the most efficient and, from an emission point of view. the most environmentally desirable. Using this type of furnace does, however, create refractory problems that reflect its unique physical characteristics.
These problems are diverse, but they can be organized under three major headings, which will be the topics for this series of articles: * build-up/clogging; * metal saturation/finning; * erosion.
These are not mutually exclusive problems. Quite the contrary, they are interdependent and, frequently, all are present in the same situation, often one causing the other. Buildup/Clogging
The phenomena of buildup in coreless and clogging in channel furnaces are often described as a strange and unexpected condition at first notice. However, it is impossible, over an extended period of time, to have a refractory surface exposed to a liquid metal environment without that surface wearing or collecting deposits.
Liquid metals are not necessarily "clean." They contain reactants that cause wear and stable materials that result in buildup. This mix consists of suspended particles and small liquid slag droplets, most of which are oxides. If the particle or droplet is of stable composition (such as silica, alumina or magnesia, which may be found in the refractory itself, it can stick on contact. This creates a film that, with time, can grow into a substantial deposit.
By way of analogy, this process is similar to the mechanism of a ceramic filter in removing inclusions from a melt. The ceramic filter, a refractory composition, acts as a substrate for particles that adhere by exactly the same physical mechanism that causes build-up in an induction furnace. Considering the enormous volume of metal circulated over the refractory surface in an induction furnace, it is not surprising that a coating will develop if the right particles are in suspension. Nature of a Buildup
Examination of hundreds of cases of deposit buildup in induction furnaces has shown that the deposits are composed of fine-grained particles. This is because, in most melting applications, stable particles that form deposits have a much lower specific gravity than the melt; thus, particles of any significant size tend to float out into slag. The particles that remain in suspension to be recirculated across the refractory surface must be small enough to be dominated by the circulation forces and thereby be kept in suspension.
Generally, the particles that satisfy these requirements are oxides formed either when the initial alloy was produced and then were released by secondary melting of its scrap or when either tramp or treatment elements in the foundry melting process were oxidized. Seldom found are pieces of deteriorated refractory or surface slag except when buildup develops to the point that it can entrap these larger particles before they rise to the surface. The composition of a buildup reflects the full gamut of the melting process. it will often show oxides resulting from minor trace element oxidation if those byproducts are refractory and have a high adhesion characteristic for the refractory surface. For example, aluminum-coated or killed steel used in cast iron melting will result in alumina or alumina complex mineral deposits such as hercynite (iron aluminate), mullite (silica alumina), etc. The addition of magnesium or calcium quickly brings in the complexes of spinel and calcium aluminate, resulting in a "thumb print" of the buildup process. With nonferrous alloys, the situation is similar. Aluminum bronze results in alumina deposits, silicon bronze in silica deposits and trace aluminum in brass results in the complex gahnite (zinc alumina spinel). Deposition Determinants
Deposition is a function of many factors. The principal ones are the refractory surface adhesion characteristics of the deposited particles and the relative balance of corrosives versus stable particles in suspension. The factors blend into one another. Corrosives in suspension will so affect the refractory surface that stable particles associated with them cannot stick. Conversely, if stable particles overwhelm the more reactive particles that would prevent their adhesion or would react and flush them, a coating is established. This is followed by a rapid buildup of deposits on a substrate of their own composition.
The physics and mechanics of deposition are the same for both channel and coreless furnaces. The location of deposits is governed by flow patterns, temperature, oxygen availability and the electromagnetic field of the furnace. Because of the significant difference in design, ways to control and eliminate buildup differ widely between channel and coreless furnaces. There are also differences between ferrous, nonferrous and certain alloy melts.
Whether or not buildup occurs in a particular production furnace depends upon the integration of a number of variables. Two apparently similar foundries may differ widely in the degree of the problems they encounter. On an experimental basis, for example, where subtle variables are masked by controlled conditions, a gray iron operation not experiencing buildup can induce it by the addition of aluminum coated scrap or by a shift to ductile-base iron where returns are remelted in the furnace. Refractory Buildup Problems
Buildup poses serious problems because it leads to furnace inefficiency and can create conditions leading to unexpected and sometimes disastrous refractory failure. The deposits entrap metal that superheats, causing penetration and even refractory melting. Coreless furnaces can build up in the midcoil area to the extent that the entire lower crucible becomes bridged. A similar scenario results in channel furnaces when the throat becomes clogged, bridging the inductor to the body of the furnace. When the bridge becomes complete, a run-out failure is almost certain to follow, particularly if the failure temperature for the throat refractory is lower than that of the inductor.
if heavy deposits can be removed or prevented from forming on the crucible walls of a coreless furnace or on the channel exits of a channel furnace, the severe overheating caused by metal entrapped in the buildup and/or constriction of flow will not develop.
The design and operating characteristics of coreless and channel induction furnaces are so different that they must be treated separately. However, techniques that focus on scrap control, additions and melting techniques have considerable common ground. Channel Furnace Buildup
Two important aspects of channel furnace design make the control and elimination of buildup a problem that can be handled during operation.
First, the power source for a channel furnace is isolated from the uppercase, or bath reservoir, of molten metal. The inductor can operate under conditions that are far different from the molten bath.
Second, buildup in the inductor is not directly connected to buildup in the uppercase, although the physical reasons for them may be similar. Isolation of the inductor from its associated metal reservoir makes it possible to create conditions that will prevent inductor refractory damage without adversely affecting production or furnace performance.
When buildup begins in a channel furnace inductor, deposition is generally over the entire channel surface until constriction raises temperatures and, in combination with the venturi effect, prevents further channel deposits. The deposits in the channel most often, but not always, will wash out leaving deposits that constrict inlet and/or exit openings. The degree of constriction becomes a function of the power applied to the inductor because this determines the temperature and flow characteristics in the channel.
When the power is high, the channel opening must be large enough to accommodate the metal flow, but because of the temperature generated, the opening can be eroded further. Conversely, at holding power, the channel exit can nearly close yet still maintain a balanced state because the flow is weak and the temperature is low. If a furnace is operated at low or holding power for an extended period, the constriction sinters and shrinks, allowing more deposition to occur. In this way, buildup becomes dense and strong. When the furnace is placed back on melting or superheating power, temperature in the channel will rise, causing severe refractory damage until the constriction is washed out sufficiently to drop the temperature.
If the throat refractory extends below the constriction and if, as in many cases, the throat refractory failure temperature is lower than that of the inductor, there is extreme risk of run-out in the lower throat or at the inductor/throat junction. This is a frequent reason for failure following a shutdown or idle period. Particle Adhesion
When particles first touch and stick to the refractory, their adhesion is generally weak, but the strength of the bond goes up quickly with time. This is particularly true if more particles are deposited over them at high temperature. If, however, the furnace is periodically "pulsed" to high power, these weakly adhering particles can be flushed into the uppercase. in this way, the constriction characteristic of a low-power operation can be prevented from forming and, when the furnace is placed on high power, severe over temperature in the inductor/lower throat will not result.
Refractories currently being used in channel furnace inductors are generally capable of handling the flow and temperature of normal inductor operation. None of them, however, are able to handle the conditions created by severe channel constriction. The life of inductor refractories can be increased substantially by preventing the severe over temperature "spike" that occurs when a severely blocked inductor is placed on high power. The consistency of refractory performance will improve markedly if the constriction can be prevented or eliminated.
The technique is simple, amounting to nothing more than automatically pulsing the furnace to high power for two to eight minutes every half hour. Power is correspondingly dropped during the balance of that time to maintain temperature control. (Ed. note: Those metalcasters not familiar with pulsing should contact their furnace manufacturer prior to implementing this practice.)
Once pulsing is done, further control can be accomplished by raising or lowering the median temperature during off-production periods. furnaces operated continually at high power may require other techniques such as those involving thermochemistry, e.g., adding silica or the formation of silica using silicon carbide to react with silica-sensitive deposits like spinel or calcium aluminate. In some shops, it is possible to change alloys, causing deposition of a more corrosive type. Examples would include changing from ductile-base iron that is causing deposition to gray iron that will, because of its acid nature, erode the buildup. With copper-base alloys that create clogging problems (such as aluminum-bronze, silicon-bronze and magnesium-deoxidized copper-nickel), there are many techniques that can be used depending upon the alloy schedule and the furnace characteristics.
In any case, pulsing should be a foundation technique for anyone encountering clogging in a channel furnace. Whether it is the initial clogging caused by moisture from the refractory oxidizing trace elements in suspension (a problem that usually fades after a few weeks), or general operational clogging from the nature of the alloy itself, pulsing will greatly improve control of the situation. Furnaces that are pulsed also derive other benefits such as more efficient inductor operation, more rapid temperature increase for the same power draw, less carbon loss in cast iron applications and improved inductor performance.
Adopted jointly with the pulsing technique should be a program of inductor instrumentation in order to be able to read a furnace's "vital signs." This involves a network of thermocouples placed in or against the inductor and throat refractories. There should also be thermocouples on the bushing whose readings are integrated into the normal electrical readout program to depict changes in inductor refractory condition. Running the inductor through a series of power settings shortly after startup to establish a database will enable furnace operators to assess changes caused by clogging, saturation, etc. This will allow them to take prompt corrective action.
Many other techniques for controlling buildup exist. Some are useful and/or complimentary while others must be undertaken cautiously. Rodding and/or green poling are effective against weakly consolidated buildup in the upper part of an inductor or in the throat. However, these measures often involve considerable risk to personnel who should be fully protected with approved protective clothing.
Although chemical flux may cause considerable damage to exposed refractory, it can be helpful in some instances, but patience is required because flux response is not always immediate and over application is almost always self-defeating. Coreless Furnace Buildup
In a coreless induction furnace, the entire bath is under power and circulates in a double torroidal circulation pattern resembling two doughnuts superimposed one over the other. Deposits occurring at the junction of the two patterns, at the slag line where radial flow intersects the sidewall and at the bottom-sidewall juncture cannot be isolated and treated separately as in the case of the channel furnace. Nevertheless, pulsing techniques can be helpful if used in conjunction with the following procedures: * Taper the lining (wide at the top and
narrow at the bottom) with the midcoil
point at the nominal lining thickness.
The degree of taper should be
worked out in conjunction with the
furnace manufacturer to avoid altering
the furnace's electrical performance. * Avoid buildup accumulation of more
than an inch on a furnace of 6-35 ton
iron capacity. increased thickness
gives the buildup mechanical stability
that resists removal. * With silica-lined furnacesmelting cast
iron, temperatures should be raised
to 2900-2950F (1590-1620C) and the
furnace alternately pulsed to high
power and then allowed to go quiet.
If successfully done, the buildup will
peel and lift free, rising into slag where
it can be removed.
If the procedure is unsuccessful in removing the deposits, the bath should be allowed to cool or should be charged with returns sufficient to cool it back to normal operating temperature. Otherwise, the deposit may slump when the furnace is poured, filling the lower part of the crucible with a difficult-to-remove material. Notching the deposited rings, particularly in the mid-coil area, before refilling and raising the temperature for pulsing treatment is very helpful in breaking the mechanical stability of the circle and will promote release of the deposit.
Deposits in furnaces melting light metal alloys must be cleaned mechanically because the melting temperatures are too low for thermochemical treatment. The refractories will not soften and release as in the case of silica, and the deposits are, for the most part, refractories themselves. In the case of aluminum and its alloys, scraping and cleaning is necessary at least once a shift and, with the use of thin, oxidized or coated scrap, after every pour. Failure to do this will entrap metal in the deposit causing it to superheat, penetrate the lining and force premature lining replacement.
Copper and copper-base alloys are usually handled in relatively small coreless furnaces of under three tons capacity where mechanical cleaning is the most effective technique and operation is intermittent. When buildup does become a problem, the use of flux should be used with great care and a full understanding of the chemistry of the alloy, the refractory and the deposit.
The use of flux in silica-lined coreless furnaces used for melting copper-base alloys or cast iron should be done with particular care because of the ease with which silica forms low melting point glasses with most commonly used fluxing compounds, such as borates and alkaline halides. Furnace Controls
Control of stable refractory oxide-forming elements is an excellent way to reinforce the techniques described previously with particular emphasis placed on aluminum, magnesium, calcium, silicon and zinc. In ferrous melting, the first three are most important, with zinc important only as a tramp element in general purpose scrap.
Aluminum is the most troublesome and most pervasive element found in buildup. it often provides the other oxide partner in mineral complexes, such as magnesia/alumina spinel, calcium aluminate, iron alumina spinel (hercynite) and zinc alumina spinel (gahnite). In the melting situations leading to this type of buildup, the aluminum source is often the most controllable since the magnesium, calcium, iron and zinc are all requisite to the finished product.
The type of refractory used has some bearing on the adhesion of a deposit, depending upon its chemistry, but its effect depends upon its ability to provide oxygen for a significant period after startup. This oxygen source role should also be extended to include structural refractories such as coil support rings and cast dome bottoms in coreless furnaces. Repair materials (primer, patch and/or gunning material) will also act as sources of oxygen for a period of time after restart, as will cause hydrogen pin hole defects where this defect is already borderline.
Finally, furnace operation itself can have a significant bearing on whether or not buildup forms. Proper practice is effective in eliminating certain types of deposits, e.g., using the reducing potential of carbon in cast iron to break down the silica component of an alumino-silicate deposit, allowing the weakened material to be flushed up into the slag.
Conversely, the introduction of weak oxide-bearing charge materials (such as rusty scrap with oxide-forming elements) can cause deposits to develop rapidly, reducing effective furnace operation in a matter of hours. An example of this is the charging of inexpensive byproduct steel with high residual aluminum together with hot rolled steel scrap that provides oxygen. As a result, the mid-coil ring deposit of iron aluminate (hercynite) leads to ineffective charging and superheat of the lower circulation pattern in about four hours, threatening the integrity of the silica lining. Controlling Buildup
The causes of buildup are complex, but procedures do exist for both control and elimination of most types. Byway of caution, operators must remember that a coreless furnace with heavy deposits or a channel furnace with a plugged channel is potentially dangerous, and pulsing should not be initiated without fully understanding the operations of each. Complete surveillance instrumentation should be applied before trying to clear a clogged inductor. The pulsing technique is intended for use with unclogged channels to prevent clogging. Pulsing already obstructed channels may precipitate a failure.
Next month: Avoiding metal saturation and finning problems.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Part 1 of 3|
|Author:||Stark, Ronald A.|
|Date:||Jun 1, 1991|
|Previous Article:||International meeting updates technology, stresses need to reach casting specifiers.|
|Next Article:||Dry ice blasting: a new technology for corebox cleaning.|