WO2005120747A1

WO2005120747A1 - Computer-controlled method for the continuous casting of a metal bar

Info

Publication number: WO2005120747A1
Application number: PCT/EP2005/004568
Authority: WO
Inventors: Christian Chimani; Karl Mörwald; Guoxin Shan
Original assignee: Voest-Alpine Industrieanlagenbau Gmbh & Co
Priority date: 2004-06-11
Filing date: 2005-04-28
Publication date: 2005-12-22
Also published as: ATA10072004A; AT413951B

Abstract

The invention relates to a method for the continuous casting of a metal bar, whereby a bar (1) is removed from a cooled continuous mould (3), supported in a bar support device (7, 11) that is located downstream of the continuous mould (3), cooled by means of a coolant and optionally reduced in viscosity. To form a specific structure in the cast bar, the continuous casting process is carried out on the basis of a thermo-mechanical computer model that describes the stress on the metal during the casting process and during the solidification process, said model permitting the stress condition of the bar to be calculated online and the variables of the continuous casting process that affect the stress on the material, such as for example the specific quantity of coolant that is provided for cooling the bar, to be set dynamically online, i.e. during the casting process To form a specific structure that is devoid of cracks, the invention uses a computer model that describes the susceptibility of the structure to cracks and the crack-forming energy that is stored in the structure.

Description

COMPUTER-CONTROLLED METHOD FOR CONTINUOUSLY casting a METAL STRAND

The invention relates to a method for continuous casting of a metal strand, in particular a steel strand, wherein a strand is pulled out of a cooled continuous mold, supported in a strand support device downstream of the continuous mold and cooled with coolant and optionally reduced in thickness, the continuous casting being used to form a specific structure in the cast strand on the basis of a thermomechanical calculation model describing the load on the metal during casting and during the solidification process taking place, with which the load state of the strand is calculated on-line, and the variables of the continuous casting process influencing the material load, such as those for cooling the strand, for example envisaged specific coolant quantity, on-line dynamic, ie while casting is in progress.

A method of this type is known from AT 409.352 B. With this method, it is possible to specify the formation of a desired structure of the metal as a target, specifically for metals of different chemical composition and for continuous steel casting for all steel qualities or steel grades to be cast. In particular, it is possible to set a certain ferrite-pearlite structure and / or to avoid precipitates, such as aluminum nitride, at the grain boundaries.

In the work of K. Schwertfeger; Susceptibility to cracking of steels during continuous casting and hot forming, steel and iron, 1994 it is shown that two crack forms predominate in the formation of cracks in continuous casting products. An interdendritic crack and an intercrystalline crack in the austenite structure.

In the work of Chimani et.al .; Micromechanical Investigation of the Hot Ductility Behavior of Steel, ISU Int. Vol. 39, No. 11, 1999, it is shown that for the intergranular crack type there is only a very strongly localized plastic deformation of the material before and during the crack formation. This forms the basis for the fact that a fracture flow can be assumed for a fracture mechanical description of the crack formation, see e.g. J. Lemaitre; A Course on Damage Mechanics, Springer-Verlag, 1992

So far it has been state of the art that only the statistical influence of process parameters is used for the evaluation of the product quality produced, such as for the evaluation of the formation of surface cracks. The invention aims to further develop a method according to AT 409.352 B, namely that it is possible to specify the suppression of crack formation as a target or to identify the formation of cracks, for metals of different chemical composition during continuous casting, in particular for all continuously cast steel qualities or steel grades. Furthermore, it should be possible to set a certain structure and avoid superimposing critical loads and crack-sensitive structural states.

This object is achieved in a method of the type described in the introduction in that a computational model describing the crack sensitivity of the structure and the crack formation energy stored in the structure of the structure is used to form a specific crack-free structure.

The on-line dynamics of the thermomechanical loads that act on the material and vary in time and place are calculated and the energy release rate G released when cracks grow is determined. The calculated energy release rate is compared with the crack resistance R of the material, which is characteristic of the given material and its structure, and thus local damage is predicted. If a critical energy release rate is calculated, you can adjust e.g. the cooling and / or the structure and / or the mechanical loads react dynamically. These measures can be taken in addition to the measures taken in accordance with AT 409.352 B.

A preferred method for the continuous casting of a metal strand, in particular a steel strand, wherein a strand is pulled out of a cooled continuous mold, supported in a strand support device arranged downstream of the continuous mold and cooled with coolant and optionally reduced in thickness, the continuous casting being carried out under on to form a specific structure in the cast strand -line calculation is carried out on the basis of a thermomechanical and metallurgical calculation model describing the formation of the particular structure of the metal and the variable of the continuous casting process influencing the material load and the structure formation, such as the specific coolant quantity provided for cooling the strand, on-line dynamic, ie during ongoing casting, is characterized in that in order to form a specific crack-free structure, the crack sensitivity of the structure and that in the S structure of the microstructure that describes the stored cracking energy is used. A model based on the fracture mechanics of metals, in particular a model for crack growth under small area flows, is preferably integrated into the calculation model.

The concept of small-area flow is a general description of crack growth in which only a small proportion of the volume around the crack banks is plastically deformed. By taking into account the equations originating from this concept, the driving forces determining crack growth, namely the energy release rate, can be determined from the thermomechanical loads. The crack resistance that opposes the energy release rate is a material parameter that depends on the material structure, the mechanical properties and the phase proportions and that can be determined from the stereographic evaluation of fracture surfaces. By using this model, one can react appropriately to the casting process when a critical energy release rate is reached and either by reducing the mechanical loads or by adjusting a structure with a higher crack resistance to avoid crack formation. If the critical energy release rate is exceeded, the location of the crack propagation can be determined and forwarded to a quality control system.

With the calculation model to be used according to the invention, the quantities necessary for describing crack formation, such as the energy release rate and the crack resistance, can be calculated on the basis of a predetermined steel analysis, the phase volume fractions, the applied deformation and the temperature history.

For this purpose, the stress-strain states on the surface of the strand and in the interior of the strand are first calculated from the applied loads. With basic continuum mechanical methods, a basic equation for the energy release rate for crack growth can be estimated for small area flows:

E where G is the energy release rate, E the modulus of elasticity and a an existing defect size, σ is the largest principal normal stress and Y is a parameter that takes into account the influence of a multi-axis stress state and the component geometry.

The resistance of the material to crack propagation can be determined using the following equation: R = 2y ₀ + 2y _pl , where R is the crack resistance, γ _{0 is} the specific surface energy and γ _p ι are the specific plastic deformation work required to produce a crack surface.

The condition for crack formation results from this as G ≥ R

The determining factors for the crack resistance are the phase volume fractions.

The specific plastic deformation work γ _pl essentially depends on the material and can be determined from the experimental investigation of fracture surfaces. The dimple fracture surfaces characteristic of the crack formed in small-area flow are measured, and from this the specific one

Deformation work calculated. γ _pl ≡ 2S - σ _m - h where σ _{m is} the mean yield stress of the material, S is a parameter that results from the shape of the dimple, and h is the dimple depth. It is therefore possible to determine the crack resistance purely experimentally.

Since γo can be neglected compared to γ _p ι, the crack resistance R is determined.

The changes in the crack resistance of the material due to changing phase volume fractions of e.g. Austenite and ferrite, temperature differences and amounts of precipitates can advantageously be used to estimate the dimple height and the yield stress.

According to a preferred embodiment, the computational model constantly changes the thermodynamic state of the entire strand, such as changes in the surface temperature, the mean temperature, the shell thickness, by solving the heat conduction equation and solving an equation that describes the phase conversion kinetics, and the cooling of the strand is dependent on the calculated one Value of at least one of the thermodynamic state variables set, the strand thickness and the chemical analysis of the metal and the continuously measured casting speed being taken into account for the simulation. By coupling the calculation of the temperature of the strand according to the invention with the computing model, which includes the formation of a specific time and temperature-dependent structure of the metal, it is possible to determine the variables of the continuous casting process that influence the microstructure, such as the amount of coolant to be applied to the strand surface , the chemical analysis of the metal and the local temperature history of the strand. In this way, a desired microstructure in the broadest sense (grain size, phase formation, excretions) can be achieved in the region of the strand near the surface.

A continuous phase conversion model of the metal is preferably integrated into the computing model, in particular according to Avrami.

In its general form, the Avrami equation describes all diffusion-controlled conversion processes for the respective temperature under isothermal conditions. By taking this equation into account in the calculation model, ferrite, pearlite and bainite fractions can be set in a targeted manner in steel continuous casting, etc. also taking into account a holding time at a certain temperature.

Preferably, the method is characterized in that thermal computational changes in the entire strand, such as changes in surface temperature, mean temperature, shell thickness, are solved by solving the heat conduction equation and solving an equation describing the excretion kinetics, in particular non-metallic and intermetallic precipitations, and the cooling of the strand is set as a function of the calculated value of at least one of the thermodynamic state variables, the strand thickness and the chemical analysis of the metal and the continuously measured casting speed being taken into account for the simulation, the excretion kinetics due to free phase energy and nucleation and use of thermodynamic basic variables being advantageous , in particular Gibb's energy, and the germ growth according to Zener is integrated in the calculation model.

Structural quantity relationships in equilibrium states are expediently also integrated into the computational model according to multi-component system diagrams, in particular according to the Fe-C diagram.

Grain growth properties are preferably integrated in the computing model, in particular taking into account recrystallization of the metal. Here, a dynamic and / or delayed and / or post-recrystallization, ie a recrystallization that takes place later in an oven, are taken into account in the calculation model.

Preferably, the variable of the continuous casting which influences the microstructure formation, a thickness reduction which takes place during the conveying of the strand before and / or after solidification of the strand is set on-line in addition to the specific amount of coolant acting on the strand, so that thermodynamic rolling, for example high-temperature rolling, also takes place during the continuous casting. thermodynamic rolling with a surface temperature greater than A _c3 can be taken into account.

Furthermore, the mechanical state, such as the deformation behavior, is preferably also constantly included in the calculation model by solving further model equations, in particular by solving the thermal conductivity equation.

A preferred embodiment is characterized in that phase components defined in terms of quantity are set by applying specific strand coolant quantities calculated on-line before and / or after the strand has solidified.

Furthermore, a defined structure is expediently set by applying an on-line strand deformation calculated before and / or after the strand has solidified, which causes the structure to recrystallize.

An advantageous variant of the method according to the invention is characterized in that the specific strand coolant quantity calculated for the final casting of the continuous casting with setting of a phase component of the strand defined in terms of quantity is set after solidification of the strand in the end region of a secondary cooling zone in a cooling zone causing increased cooling.

The invention is explained in more detail below for continuous steel casting. The method according to the invention can be used for other metals analogously to the explanations below.

Based on a predetermined chemical analysis of the steel, the austenite grain size and the temperature history of the strand, the calculation model to be used according to the invention allows all transformation temperatures and data to be used for the prediction and Calculate the description of the transformation processes for the phase fractions ferrite, pearlite, bainite and martensite are necessary.

For this, a carbon equivalent is first calculated for the individual alloy components. This results in analysis-dependent starting temperatures for ferrite conversion, pearlite conversion, bainite formation and martensite formation (based on the iron / carbon diagram).

Based on the Avrami equation, which describes in its general form all diffusion-controlled conversion processes for the respective temperature under isothermal conditions, basic equations for the conversion curves can be determined.

X = 1 - exp (-bt ⁿ )

where X is the proportion of the converted phase and b and n are parameters which are dependent on the nucleation and the growth of the phase formed. These parameters b and n are dependent on the analysis and can be determined by dilatometer tests. In connection with ZTU diagrams, the Avrami equation can be used to calculate the start and end times as well as the temperature for the ferrite, pearlite and bainite transformation under isothermal conditions.

In order to take into account non-isothermal conversions, i.e. to be able to fully take into account the cooling of the strand taking place in the continuous casting plant - possibly also unevenly taking place - the proportion is shown as a function of time based on the ZTU diagrams stored in the computer and the dependence of the temperature converted material, etc. by integrating the Avrami equation over the cooling time of the strand (see TT Pham, EB Hawbolt, JK Brimacombe: "Preciding the onset of transformation under non continuous cooling conditions. II Application to austenite - pearlite transformation", Met. Mat. Trans A, 26A, pp. 1993-2000, 1995).

X (t) = J _{s (} τ [l-exp (-bt ⁿ )] -dt

where t _s (T) means a virtual start time of the conversion at a temperature T in accordance with the amount actually converted.

For this calculation algorithm, the temperature is defined as a function of time. Since the calculated conversion or excretion share according to Avrami no information about gives the actual microstructure / quantity ratios, but only shows whether and how the equilibrium state is reached, the conversion fractions are related to the equilibrium lines from the iron / carbon diagram and also taken into account in the calculation model to determine the microstructure proportion.

Nucleation processes are taken into account in the calculation model based on the chemical Gibb's energy or phase energy (shown below for aluminum nitrides).

ΔG _chem = Δ G ⁰ _A1N - R - T - (ln χ ln;)

where G ° _{AIN is} the standard Gibb energy for the formation of A1N,] ^° _A ^° _[ the molar fraction of aluminum in the austenite volume and] ζ ~ mean nitrogen content. The nucleation rate can be calculated as follows:

ΔG "I = S ^• DAI ^• XAI ^• exp? \

where S is the density of nucleation in austenite.

reflects the condition for nucleation. Herein σ is the austenite / AIN interface energy. k _ß is the Boltzmann constant and D _{AI is} the expansion capacity of aluminum in austenite.

The germ growth is taken into account according to Zener (e.g. dealt with in J.S. Kirkaldy, "Diffusion in the Condensed State", The Universities Press, Belfast, 1985).

The calculation process takes place in two main stages. In the first stage the number of currently formed germs is determined and in the second stage the growth of all previously formed excretions is calculated.

The accompanying FIGS. 1 and 2 serve to explain the invention further. 1, a steel strand 1 is formed from a molten steel 2 with a specific chemical composition by casting in a continuous mold 3. The molten steel 2 is poured from a ladle 4 via an intermediate vessel 5 and one from the intermediate vessel 5 into the continuous mold 3 by means of a pouring tube 6 extending below the casting level formed in the continuous mold 3. Below the continuous mold 3, strand guide rollers 7 are provided for supporting the steel strand 1, which still has a liquid core 8 and initially only a very thin strand shell 9.

The steel strand 1 emerging from the continuous mold with a straight axis is deflected in a bending zone 10 into a circular arc path 11 and is also supported in this by strand guide rollers 7. In a straightening zone 12 following the circular arc path 11, the steel strand 1 is again straightened and conveyed out via an outfeed roller table or directly reduced in thickness, e.g. by means of an on-line mill stand 13.

To cool the steel strand 1, the latter is cooled directly or indirectly - via strand guide rollers 7 provided with internal cooling, so that a certain temperature can be set on its surface to a certain depth range.

The steel strand 1 is supplied with the amount of coolant necessary for the desired structure of the steel strand 1 via a closed or open control circuit by means of a computer 14. Machine data m, the format f of the steel strand 1, material data, such as the chemical analysis St _{Ch of} the molten steel 2, the pouring state z, the pouring speed v, the molten steel temperature t _f i at which the molten steel 2 enters the continuous mold 3, as well as the desired structure α / γ and, if appropriate, a deformation w of the steel strand 1, which on the way of Strand guidance is performed, entered. This deformation can also be given, for example, by straightening the steel strand 1 in the straightening zone 12.

In the computer 14, based on a fracture mechanical calculation model that takes into account the crack sensitivity of the structure and the cracking energy stored in the structure of the structure, and on the basis of a thermo-mechanical calculation model that enables the temperature analysis based on the solution of the heat conduction equation, a target water quantity Q _s calculated, etc. due to the current, already applied amount of water Q _A , which is also entered into the computer. The current temperature T _A calculated by the thermomechanical calculation model and the calculated current temperature Stress field G is fed to the fracture-mechanical calculation model, which continuously calculates the target temperature Ts, a crack formation indicator D, the desired casting speed v _s , and any other desired values that influence the continuous casting.

A solution of the heat conduction equation using a process computer is state of the art and e.g. dealt with in detail in DE-C2 - 44 17 808 for continuous casting. The finite difference method with Lagrangian description is given as one way of solving the heat conduction equation.

According to the embodiment shown in Fig. 2, a metallurgical is additionally

Calculation model used for the computer 14, the phase transition kinetics and

Nucleation kinetics taken into account. Furthermore, the metallurgical takes into account

Calculation model the current steel analysis St _eh in order to _cope with different material behavior. The current one calculated by the thermo-mechanical calculation model

Temperature TA is fed on-line to the fracture-mechanical calculation model, and this continuously calculates the desired phase components a _s , γ _s for the metallurgical one

Calculation model that calculates the target temperature Ts, based on which the thermo-mechanical calculation model calculates and automatically sets the target water quantity Q _s for the individual strand cooling sections.

Claims

claims:

1. A method for continuous casting of a metal strand, in particular a steel strand (1), a strand (1) being pulled out of a cooled continuous mold (3), supported in a continuous support device (7, 11) downstream of the continuous mold (3) and cooled with coolant and if necessary, the thickness is reduced, in order to form a specific structure in the cast strand, the continuous casting is carried out on the basis of a thermomechanical computing model describing the load on the metal during the casting and during the solidification process taking place, with which the load state of the strand is calculated on-line, and the variable of the continuous casting process influencing the material load, such as, for example, the specific coolant quantity provided for cooling the strand, on-line dynamic, ie during ongoing casting, characterized in that a computational model describing the crack sensitivity of the structure and the cracking energy stored in the structure of the structure is used to form a specific crack-free structure.

2. Method for continuous casting of a metal strand, in particular a steel strand (1), wherein a strand (1) is pulled out of a cooled continuous mold (3), supported in a strand support device (7, 11) downstream of the continuous mold (3) and cooled with coolant and if necessary, the thickness is reduced, the continuous casting being carried out using on-line calculation based on a thermomechanical and metallurgical calculation model describing the formation of the specific structure of the metal, and the material loading and microstructure-influencing variables of the continuous casting method, such as, for the formation of a specific structure in the cast strand for example the specific coolant quantity provided for cooling the strand, on-line dynamic, ie during ongoing casting, characterized in that a computational model describing the crack sensitivity of the structure and the cracking energy stored in the structure of the structure is used to form a specific crack-free structure.

3. The method according to claim 1 or 2, characterized in that a model based on the fracture mechanics of metals, in particular a model for crack growth under small area flows, is integrated into the computing model.

4. The method according to claim 1, 2 or 3, characterized in that with the computing model based on a predetermined metal analysis, preferably steel analysis, and on the basis of the phase volume proportions, the applied deformations and the temperature history, the quantities necessary for describing crack formation, such as the energy release rate and the crack resistance, are calculated.

5. The method according to one or more of claims 1 to 4, characterized in that with the computing model, the stress-strain states on the surface of the strand and inside the strand are first calculated from loads applied to the strand from the outside, whereupon using a basic equation for the energy release rate for crack growth with small area flows is estimated, whereby the basic equation is:

G = Y ^ - E here is:

G the energy release rate,

E the modulus of elasticity, a an existing defect size σ the largest principal normal stress and

Y is a parameter that shows the influence of a multiaxial stress state and the

Building geometry considered; that the resistance that the metal offers to crack propagation is determined according to the

equation

R = 2γ ₀ + 2γ _pl

it is determined where

R the crack resistance γ ₀ the specific surface energy and γ _pl the specific plastic necessary to create a crack surface

Deformation work is what G is compared to R and if the R value is exceeded by the G value or if these values are the same, corrective intervention is made in the continuous casting process.

6. The method according to claim 5, characterized in that to determine the specific plastic deformation work γ _pl experimental studies of

Dimpel fracture surfaces can be used by measuring cracks formed in small-area tiles, etc. according to the equation γ _pl ≡ 2S - σ _m - h

where σ _{m is} the mean yield stress of the material, S is a parameter resulting from the dimple shape, and h is the dimple depth.

7. The method according to claim 6, characterized in that the dimple depth due to changing phases volume fractions of e.g. Austenite and ferrite, is estimated on the basis of temperature differences and on the basis of precipitation quantities.

8. The method according to one or more of claims 1 to 7, characterized in that with the computing model thermodynamic changes in state of the entire strand, such as changes in the surface temperature, the mean temperature, the shell thickness by solving the heat conduction equation and solving an equation describing the phase conversion kinetics are constantly included in the calculation and the cooling of the strand is set as a function of the calculated value of at least one of the thermodynamic state variables, the strand thickness and the chemical analysis of the metal and the continuously measured casting speed being taken into account for the simulation.

9. The method according to claim 8, characterized in that a continuous phase conversion model of the metal is integrated in the computing model, in particular according to Avrami.

10. The method according to one or more of claims 1 to 9, characterized in that with the computer model thermal changes in the state of the entire strand, such as changes in the surface temperature, the mean temperature, the shell thickness, by solving the heat conduction equation and solving the excretion kinetics, in particular non-metallic and intermetallic precipitates, the descriptive equation are constantly included in the calculation and the cooling of the strand is set as a function of the calculated value of at least one of the thermodynamic state variables, the strand thickness and the chemical analysis of the metal and the continuously measured casting speed being taken into account for the simulation.

11. The method according to claim 10, characterized in that the excretion kinetics due to free phase energy and nucleation and use thermodynamic Basic parameters, in particular the Gibb energy, and the Zener germ growth are integrated in the calculation model.

12. The method according to one or more of claims 1 to 11, characterized in that structural relationships in equilibrium states according to multi-material system diagrams, in particular according to Fe-C diagram, are integrated into the computing model.

13. The method according to one or more of claims 1 to 12, characterized in that in the computing model grain growth properties, in particular taking into account recrystallization of the metal, are integrated.

14. The method according to one or more of claims 1 to 13, characterized in that as the structure forming variable of the continuous casting takes place during the removal of the strand thickness reduction before and / or after solidification of the strand in addition to the specific amount of coolant acting on the strand is set.

15. The method according to one or more of claims 1 to 14, characterized in that with the computing model, the mechanical state, such as the deformation behavior, is constantly included in the calculation by solving further model equations, in particular by solving the thermal conductivity equation.

16. The method according to one or more of claims 1 to 15, characterized in that quantitatively defined phase fractions are set by applying on-line calculated specific strand coolant amounts before and / or after the strand has solidified.

17. The method according to one or more of claims 1 to 16, characterized in that a defined structure is set by applying an on-line calculated strand deformation before and or after the solidification of the strand, which causes recrystallization of the structure.

18. The method according to one or more of claims 1 to 17, characterized in that a final phase change, optionally taking into account a subsequent reverse conversion, is set after solidification of the strand in a cooling zone causing increased cooling.