Additives such as metallic silicon accelerate the breakup of zirconia grains but also appear to protect the carbon bond. Dissolution of the oxide phase is the predominant rate determining step and the relationship between diffusion of refractoryox ides in mould flux as a function of viscosity is consistent with the relationship between attack and flux viscosity.
The requirements for continuous casting nozzles are particularly onerous. They require to stand up to the
thermal shock caused by the introduction of molten steel at 1530'C and to resist chemical attack from the molten steel and the mould flux. A nozzle maybe required to last for a twenty or twenty two hours casting sequence without failure which could lead to downgradingof the cast steel or worse to a breakout which could be extremely costly in terms of lost production.
Increases in the lifetimes required from submerged nozzles have resulted in the widespread use of zirconia graphite "slag bands" to protect the part of the nozzle that is most vulnerable to attack. Whilst zirconia graphite is more resistant to attack than alumina graphite, and has solved problems with attack for most routine casting operations, current trends towards higher casting speeds and thin slab casting, with the lower viscosity fluxes employed in such operations, are increasing the requirements of slag bandmaterials.
There is always a slag film between the metal and the refractory in the eroded zone. So it is supposed that the oxide phase dissolved in the flux and the carbon was attacked by oxidation via iron oxide in the fiux. This iron oxide was believed to have been replenished by oxygen dissolved in the metal which originated in the impure argon used in the experiment. or
Cyclic contact between refractory and metal and refractory and slag (due to fluctuation at meniscus level). First the carbon is dissolved in the steel exposing oxide. The oxide is not wet by the steel but is wet by the slag, therefore the slag penetrates between the steel and the refractory and dissolves the oxide phase. This will expose more graphite which will alter the balance of interfacial tension so that the slag will recede allowing the metal to contact the refractory and dissolve the carbon and then the cycle is repeated.
Also observed that oxidation influenced the rate of attack but suggested that this was more related to agitation caused by the liberation of carbon monoxide and eliminated this by aluminum killing their steel. A decrease in wear rate with increasing carbon content and this decrease becomes much more significant as carbon saturation is approached. However, in the aluminum killed case there is no decrease in rate until very high carbons.
The Effect of Refractory Composition
In addition to carbon and partially stabilized zirconia, SEN slag bands can contain fluxes, and metallic silicon
and have a surface glaze to protect the carbon bond from oxidation during preheat.the effect of additives
and found that certain species increased the rate of attack and this effect decreased in the order Si >Si02 >
Fe203>Al203.the additives reacted with the calcium oxide in the zirconia and caused destabilisation leading to formation of mono clinic zirconia, which resulted in the breakup of the zirconia grains. This mechanism can not be entirely correct as monoclinic zirconia is not stable at casting temperatures regardless of composition.
found that the use of pure zirconia, which is not stabilised, gave greater. Comparison of zirconia graphite after corrosion testing for 240min a) with silicon and fluxes and b) without silicon and fluxes.14) Phases shown: zirconia, Z and calcium silicate, C corrosion resistance. This is because there is no calcium present to form low melting films at the grain boundaries.There also found that stabilized zirconia did not suffer from problems with additives, again because there was no tendency to form low melting silicates.
The effective diffusivity is a linear function of slag fluidity indicating that diffusion obeys either the Stoke-Einstein relation, Eq. (1), or the Eyring relation, Eq. (2).
D = KT/(6*pie* r*n) ....................(1)
D = KT/ n*lemda
(1) The attack of zirconiagraphite and alumina graphite requires attack of both the oxide and carbon phase.
(2) The rate of attack is controlled mainly by dissolution of the oxide phase in the fiux, however, oxidation
of the carbon appears to play an important role.
(3) Additives such as silicon metal and internal fluxes react with the zirconia and lead to breakup of the zirconia particles by forming low melting grain boundary films. Destabilization of the zirconia is irrelevant in this respect, and in fact it may be better to use un-stabilized zirconia in the first place.
(4) Attempts to eliminate additives improve the resistance of the zirconia phase but seem to accelerate the attack of the carbon bond.
(5) The diffusion of alumina in mould fiuxes is inversely proportional to flux viscosity which is consistent
with the observed behaviour of SENsas a function of flux viscosity.
The requirements for continuous casting nozzles are particularly onerous. They require to stand up to the
thermal shock caused by the introduction of molten steel at 1530'C and to resist chemical attack from the molten steel and the mould flux. A nozzle maybe required to last for a twenty or twenty two hours casting sequence without failure which could lead to downgradingof the cast steel or worse to a breakout which could be extremely costly in terms of lost production.
Increases in the lifetimes required from submerged nozzles have resulted in the widespread use of zirconia graphite "slag bands" to protect the part of the nozzle that is most vulnerable to attack. Whilst zirconia graphite is more resistant to attack than alumina graphite, and has solved problems with attack for most routine casting operations, current trends towards higher casting speeds and thin slab casting, with the lower viscosity fluxes employed in such operations, are increasing the requirements of slag bandmaterials.
Mechanism of erosion :
There is always a slag film between the metal and the refractory in the eroded zone. So it is supposed that the oxide phase dissolved in the flux and the carbon was attacked by oxidation via iron oxide in the fiux. This iron oxide was believed to have been replenished by oxygen dissolved in the metal which originated in the impure argon used in the experiment. or
Cyclic contact between refractory and metal and refractory and slag (due to fluctuation at meniscus level). First the carbon is dissolved in the steel exposing oxide. The oxide is not wet by the steel but is wet by the slag, therefore the slag penetrates between the steel and the refractory and dissolves the oxide phase. This will expose more graphite which will alter the balance of interfacial tension so that the slag will recede allowing the metal to contact the refractory and dissolve the carbon and then the cycle is repeated.
Also observed that oxidation influenced the rate of attack but suggested that this was more related to agitation caused by the liberation of carbon monoxide and eliminated this by aluminum killing their steel. A decrease in wear rate with increasing carbon content and this decrease becomes much more significant as carbon saturation is approached. However, in the aluminum killed case there is no decrease in rate until very high carbons.
The Effect of Refractory Composition
In addition to carbon and partially stabilized zirconia, SEN slag bands can contain fluxes, and metallic silicon
and have a surface glaze to protect the carbon bond from oxidation during preheat.the effect of additives
and found that certain species increased the rate of attack and this effect decreased in the order Si >Si02 >
Fe203>Al203.the additives reacted with the calcium oxide in the zirconia and caused destabilisation leading to formation of mono clinic zirconia, which resulted in the breakup of the zirconia grains. This mechanism can not be entirely correct as monoclinic zirconia is not stable at casting temperatures regardless of composition.
found that the use of pure zirconia, which is not stabilised, gave greater. Comparison of zirconia graphite after corrosion testing for 240min a) with silicon and fluxes and b) without silicon and fluxes.14) Phases shown: zirconia, Z and calcium silicate, C corrosion resistance. This is because there is no calcium present to form low melting films at the grain boundaries.There also found that stabilized zirconia did not suffer from problems with additives, again because there was no tendency to form low melting silicates.
The effective diffusivity is a linear function of slag fluidity indicating that diffusion obeys either the Stoke-Einstein relation, Eq. (1), or the Eyring relation, Eq. (2).
D = KT/(6*pie* r*n) ....................(1)
D = KT/ n*lemda
Conclusion :
(1) The attack of zirconiagraphite and alumina graphite requires attack of both the oxide and carbon phase.
(2) The rate of attack is controlled mainly by dissolution of the oxide phase in the fiux, however, oxidation
of the carbon appears to play an important role.
(3) Additives such as silicon metal and internal fluxes react with the zirconia and lead to breakup of the zirconia particles by forming low melting grain boundary films. Destabilization of the zirconia is irrelevant in this respect, and in fact it may be better to use un-stabilized zirconia in the first place.
(4) Attempts to eliminate additives improve the resistance of the zirconia phase but seem to accelerate the attack of the carbon bond.
(5) The diffusion of alumina in mould fiuxes is inversely proportional to flux viscosity which is consistent
with the observed behaviour of SENsas a function of flux viscosity.
No comments:
Post a Comment