Continuous casting of steel is a procedure with many factors involved. If any of those factors are non controlled it can hold inauspicious effects to both the steel produced and the equipment bring forthing it. This paper will briefly travel over the demands and equipment needed to adequately command the degree of steel in a uninterrupted casting cast every bit good as jobs that arise if non decently controlled.
Reason For Control
The steel ‘s semilunar cartilage ( top of liquid ) needs to be monitored and controlled for a figure of different grounds. The most obvious grounds are to forestall the steel from overruning the cast or from the cast going drained. Both of those instances would ensue in liquefied steel falling onto the equipment below the cast, doing harm and halting of the procedure. Steel over the cast will slop out and it will besides forestall lubrication from come ining the cast pit. In the instance of excessively small steel nowadays in the cast, the sum of clip the steel is in contact with the cast will be excessively short. If the outer part of the steel does non pass adequate clip near the cast walls, a shell of equal thickness to incorporate the internally liquid steel will non be formed. This would do a jailbreak, leaking molten steel and holding production every bit good.
If it was merely those two extremes mentioned above, the allowable degree in the cast would be big and merely necessitate a simple method to keep steel degree within that scope. But it has been noticed industrially that fluctuations in the semilunar cartilage degree can hold noticeable effects on the surface quality of the ensuing product1,2. The formation of laps and bleeds on the steel surface every bit good as deepened oscillation markers have been accredited to limitations on lubrication flow caused by traveling semilunar cartilages. Laps and bleeds are proposed to be caused when lubrication is blocked and the outer steel shell is allowed to touch the cast wall1. That contact bonds the two surfaces together doing cryings to organize in the steel tegument let go ofing liquefied steel. Even the cause of lubrication loss can be its ain job. If the limitation is caused by the semilunar cartilage making the scoria rim, the rim can do denting in the steel surface and can be broken off onto the steel surface through scratch ( Fig. 1 ) .
The deepness of oscillation Markss formed by the cast ‘s oscillation is related to the fluctuation way of the meniscus2. Since oscillation Markss are a byproduct of the speed differences between the cast and steel during denudation, a alteration in the semilunar cartilage degree would alter the comparative speed between the two. So, when the semilunar cartilage rises, more steel progresss per oscillation shot which increases the distance between markers but besides increasing the deepness of the marker ( Fig. 2 ) . On the other manus, when the steel degree in the cast lowers, the antonym is seen and the markers are closer together but each is shallower because of it.
Equipment To Monitor
The basic mold degree control systems relied merely on feeling devices at the cast to supervise the degree of the steel within it. Two chief options for the finding of steel degree are thermally supervising the cast or optical review of the meniscus1,3. The thermic method requires the installing of thermocouples within the cast above the coveted steel degree. The temperature at the thermocouples can be used to find their distance from the liquefied steel. The job with this method was complication in graduating it, since the temperature-to-distance computation would necessitate the thermocouples ‘ exact distances up the cast and within the cast every bit good as the heat transportation rate through the cast, its chilling rate and the initial temperature of the steel. A presently preferred method is an optical detector positioned above the cast that straight observes the steel degree.
Other countries that require monitoring are the usher axial rotations, the tundish, and the province of the tundish-to-mold flow regulator ( either a stopper rod or a slide gate valve ) ( Fig. 3 ) 4. The set projecting velocity of the steel is monitored through the usher rolls below the cast. The tundish is monitored to find the deepness of steel in the tundish since its ensuing metallostatic force per unit area caput will impact the flow velocity into the cast. The place of the flow regulator is required so that the cross sectional country of the tundish-to-mold flow can be calculated.
Many parts are needed to interact with each other to travel a flow regulator3. At the regulator itself there are motion limitation forces to see. For slide gate valves both clash with its holder and force per unit area from the fluxing steel can impact motion, while for a stopper rod clash against the liquid steel every bit good as perkiness within that liquid demand to be considered. About all flow regulators are moved by usage of a hydraulic device. The force per unit area within a hydraulic cylinder is non ever of a degree to do motion, so the hold clip to construct up equal force per unit area to travel the flow regulator should be noted. The other beginning of response hold is from the electrical detector commanding the hydraulic system. All electrical switches have a deadband value which is the minimal electromotive force signal needed to trip a response. Taken together, the holds caused from electromotive force and force per unit area build up and get the better ofing clash can show about half a 2nd hold between computing machine bid and physical accommodation.
Computer Control Methods
All the centripetal readings are inputted into a dynamic control system that determines what accommodations are needed to command the steel degree in the cast. The exact type of control system used presently is a closed-loop automatic control regulator ( Fig. 4 ) 5. This type of accountant does non necessitate human input, regulates the end product ( in this instance steel degree ) to a set value, and mentions prior end product readings for future determinations. The accountant is divided into two chief countries of concern: bid response and perturbation rejection3.
Command response trades with doing major accommodations to the system to make the programmed steel level3. This is accomplished by executing a mass-flow balance for the mold4. This portion of the accountant modus operandi is dominate when the cast is ab initio make fulling up or when the degree of steel demands to be changed to a new set point. Once the coveted degree is reached in the cast, all that is required of this portion of the control plan is to maintain the incoming volume rate equal to the outgoing volume rate. The outgoing rate is equal to the projecting velocity from the usher axial rotations multiplied by the face country of the casting. The ingoing rate is based on the tundish-to-mold flowing velocity caused by the deepness of steel in the tundish and its corrected flow country. Adjustment is needed for computation of the ingoing volume rate since frictional forces curtailing flow alteration based on flow speed and turbulency.
Disturbance rejection trades with minimising the mistake difference between the current steel degree and the mark steel level3. When the steel degree is near to the mark value, the bid response is used maintain the current mold degree while the perturbation rejection everyday Acts of the Apostless to rectify for minor fluctuations in steel degree. These fluctuations from the mark are normally the consequence of nose clogging, but can besides be caused by other things. There are two major methods for minimising mistake, through a formulated response or through an expert system.
A expression modus operandi at the simplest contains a relative invariable that is multiplied by the mistake to rectify the end product system4. While this corrects for the mistake it besides can bring forth extra fluctuation in the steel degree, unless a differential invariable is besides used. By multiplying a changeless by the rate of steel degree fluctuation the rectification response is damped and will settle to the mark degree bit by bit. By utilizing this, the mark degree will non be overshoot ensuing in changeless gap and shutting of the flow regulator and seting unnecessary wear on it. Additional expression parts, like a fuzzed control term, can be used to accelerate response4,6.
An expert system everyday uses pre-programmed responses for outputs7,8. It takes the input degrees, and based on anterior technology cognition, picks a predetermined action. While this everyday type is faster and simpler to make than a formula modus operandi, it does non hold the same degree of preciseness.
The surface quality of uninterrupted dramatis personae steel can be greatly affected by motion of the steel semilunar cartilage in the cast. To acquire the best surface quality possible, the fluctuation scope of the semilunar cartilage should be minimized every bit much as possible. Through supervising the volume flow rates into and out of the cast, every bit good as the alteration in steel degree, the steel degree in the cast can be controlled. In add-on, jobs with steel flow into the cast can be corrected for by usage of computing machine modus operandis.
1. S. Kumar, B. N. Walker, I. V. Samarasekera, and J. K. Brimacombe: “Chaos at the Meniscus – The Genesis of Defects in Continuously Cast Steel Billets” , 1995, 13th PTD Conference Proceedings, pp.119-141
2. B. G. Thomas, M. S. Jenkins, and R. B. Mahapatra: “Investigation of Strand Surface Defects Using Mould Instrumentation and Modelling” , 2004, Ironmaking and Steelmaking, Vol.31, No.6, pp.485-494
3. H. L. Gilles, J. A. Stofanak, I. W. Whiteman, and J. W. Brunicon: “Dynamic Modeling of Slab Caster Mold Level Control” , 1995, 13th PTD Conference Proceedings, pp.263-277
4. D. Lee, Y. Kueon, and S. Lee: “High Performance Hybrid Mold Level Controller for Thin Slab Caster” , 2004, Control Engineering Practice 12, pp.275-281
5. G. F. Franklin, J. D. Powell, and A. Emami-Naeini: Feedback Control of Dynamic Systems, 4th Ed. , 2002
6. E. J. Saarelainen, P. Lautala, M. Inkinen, and J. Johansson: “Steel Caster Mold Level Control Using Fuzzy Logic” , 1995, 13th PTD Conference Proceedings, pp397-402
7. Y. S. Kueon, and S. Y. Yoo: “Mold Level Control in Continuous Caster Via Nonlinear Control Technique” , 1997, Automation in the Steel Industry
8. T. Watanabe, K. Omura, M. Konishi, S. Watanabe, and K. Furukawa: “Mold Level Control in Continuous Caster by Neural Network Model” , 1999, ISIJ International, Vol.39, No.10, pp.1053-1060