@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Materials Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Agarwal, Prakash K."@en ; dcterms:issued "2010-03-22T20:04:18Z"@en, "1979"@en ; vivo:relatedDegree "Master of Applied Science - MASc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """The spray cooling system of an operating billet caster has been redesigned with the aim of reducing the formation of mid-way cracks. These cracks are caused by tensile strain which is generated at the solidification front when the surface temperature of the strand rebounds owing to a sharp reduction in surface heat extraction. The objective of the design, therefore, was to achieve a cooling system that would minimize surface temperature rebound of the strand as it passes from one cooling zone to the next. A computer program based on the explicit finite difference method has been used for the design work. The spray design was implemented on one strand of an operating continuous casting machine which produced 10.8 cm square billets. Transverse sections were cut from the test strand and sulfur printed, then compared to sulfur prints of sections taken from an adjacent strand of the same heat but with unmodified sprays. It was shown that with empirical adjustment, the redesigned spray system reduced the severity of mid-way cracks in over 80% of the heats. It was also found that the carbon content and cast structure have a profound effect on the cracking tendency, whereas, the Mn/S ratio (up to 30%) is less effective. Finally, a new design method for sprays has been proposed which may result in a better temperature distribution and may be easier to adjust to suit specific operating conditions."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/22243?expand=metadata"@en ; skos:note "CASE STUDY OF SPRAY DESIGN FOR A CONTINUOUS BILLET CASTER by PRAKASH K.JAGARWAL B. Tech. (Met.), I.I.T. Bombay (1971) A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE i n THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF METALLURGICAL ENGINEERING We accept t h i s t h e s i s as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1979 (c) Prakash K. Agarwal, 1979 In presenting th i s thes is in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make i t f ree ly ava i lab le for reference and study. I further agree that permission for extensive copying of th is thes is for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th is thes is for f inanc ia l gain sha l l not be allowed without my written permission. Depa rtment The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date Q^&zw^&es/ 14-, 'H71 ABSTRACT The spray cooling system of an operating b i l l e t caster has been redesigned with the aim of reducing the formation of mid-way cracks. These cracks are caused by t e n s i l e s t r a i n which i s generated at the s o l i d i -f i c a t i o n front when the surface temperature of the strand rebounds owing to a sharp reduction i n surface heat extraction. The objective of the design, therefore, was to achieve a cooling system that would minimize surface temperature rebound of the strand as i t passes from one cooling zone to the next. A computer program based on the e x p l i c i t f i n i t e d i f f e r -ence method has been used for the design work. The spray design was implemented on one strand of an operating continuous casting machine which produced 10.8 cm square b i l l e t s . Trans-verse sections were cut from the t e s t strand and s u l f u r printed, then compared to s u l f u r p r i n t s of sections taken from an adjacent strand of the same heat but with unmodified sprays. I t was shown that with empirical adjustment, the redesigned spray system reduced the severity of mid-way cracks i n over 80% of the heats. I t was also found that the carbon content and cast structure have a profound e f f e c t on the cracking tendency, whereas, the Mn/S r a t i o (up to 30%) i s l e s s e f f e c t i v e . F i n a l l y , a new design method f o r sprays has been proposed which may r e s u l t i n a better temperature d i s t r i b u t i o n and may be easier to adjust to s u i t s p e c i f i c operating conditions. i i TABLE OF CONTENTS Page ABSTRACT i i LIST OF FIGURES .' v i LIST OF TABLES i x LIST OF SYMBOLS x i ACKNOWLEDGEMENT x i i i Chapter 1 INTRODUCTION 1 1.1 Scope of the Present Study 3 2 BACKGROUND WORK 6 2.1 Mechanism o f Formation of Internal Cracks 6 2.2 Factors Influencing the Formation of Internal Cracks 6 2.2.1 Solute Elements 7 2.2.2 Structure 9 2.2.3 E f f e c t of Carbon and Transformational Stresses 9 2.2.4 Thermal and Mechanical Stresses 10 2.2.5 Comments 12 2.3 High Temperature Mechanical Properties of Steel 12 2.4 Heat Transfer C h a r a c t e r i s t i c s i n Sprays 17 i i i Chapter Page 2.5 Mathematical Model 20 2.5.1 Mathematical Formulation 25 2.5.2 Characterization of Input Conditions 26 2.5.3 Computer Program 31 3 MODIFICATION OF SECONDARY COOLING SYSTEM 33 3.1 C r i t e r i a f o r Spray Design 33 3.2 Continuous Casting Machine 34 3.3 Design Constraints and the Spray Design Steps 38 3.4 Se l e c t i o n of the Surface Temperature D i s t r i b u t i o n 40 3.5 Ca l c u l a t i o n of Spray Heat Transfer C o e f f i c i e n t s 41 3.6 Determination of Spray Parameters 41 3.7 Outcome of the Spray Design Study 46 4 RESULTS OF DISCUSSION OF IN-PLANT TRIALS 61 4.1 Spray Evaluation - F i r s t Campaign 61 4.1.1 Comparison of Modified to Unmodified Sprays 64 4.1.2 E f f e c t of Chemistry on Crack Frequency 64 4.1.3 E f f e c t of Structure on Crack Frequency 69 4.1.4 Crack Depth 72 4.1.5 Surface Temperature Measurements 75 i v Chapter Page 4.1.6 Di s c u s s i o n of F i r s t Campaign 75 4.2 Spray E v a l u a t i o n - Second Campaign 81 4.3 Spray E v a l u a t i o n - T h i r d Campaign 84 4.4 Pressure Drop 88 4.4.1 F r i c t i o n a l Losses 88 4.4.2 Pressure Loss Due to G r a v i t y Head 89 4.4.3 Laboratory Spray T r i a l s 89 4.4.4 D i s c u s s i o n 91 4.5 Spray E v a l u a t i o n - Fourth Campaign 95 5 GENERAL DISCUSSION AND SUMMARY 103 REFERENCES 108 APPENDIX I HO V LIST OF FIGURES Figure Page 1.1 World-wide annual continuous-casting capacity. 2 1.2 Sulfur p r i n t showing mid-way cracks. 4 2.2.1.1 Scanning electron micrograph of surface of a crack formed near the solidus temperature. The inclu s i o n s are s u l f i d e s . 8 2.2.4.1 P r o f i l e s of (a) temperature and (b) stress and s t r a i n perpendicular to a side face at centerplane of 10.2 x 15.2 cm b i l l e t cast at 2.75 cm/s with 653 cm long sprays. 11 2.3.1 D u c t i l i t y and strength curves f o r various p l a i n carbon s t e e l s . 14 2.3.2 D u c t i l e - b r i t t l e t r a n s i t i o n temperature and solidus as a function of carbon content i n the s t e e l . 16 2.4.1 Relationship between the spray heat t r a n s f e r c o e f f i c i e n t , h , and water f l u x , W, as obtained by various i n v e s t i g a t o r s . 24 2..5.2.1 Average spray water f l u x as a function of spray distance and water pressure f o r 1/4 GG10 type nozzle. 29 2.5.2.2 E f f e c t of water pressure on the f i l m heat transf e r c o e f f i c i e n t . 30 2.5.3.1 Flow chart of the computer program 32 3.2.1 Mid-face surface temperature p r o f i l e s f o r the e x i s t i n g sprays. 37 3.4.1 Hypothetical mid-face surface temperature p r o f i l e s to obtain surface heat f l u x (10.8 x 10.8 cm square b i l l e t s ) . 42 3.5.1 Schematic representation of the method used f o r determining average heat t r a n s f e r c o e f f i c i e n t p r o f i l e i n the given zone. 43 v i Method of obtaining heat t r a n s f e r c o e f f i c i e n t p r o f i l e for 1/4 GG10 nozzles spaced 8.9 cm apart at a distance of 11.8 cm from the b i l l e t surface (water pressure 207 kPa). Mid-face temperature p r o f i l e s below the mould for d i f f e r e n t spray designs (10.8 x 10.8 cm square b i l l e t s ) . Mid-face temperature p r o f i l e below the mould for d i f f e r e n t spray designs (15.2 x 15.2 cm square b i l l e t s ) . Mid-face temperature p r o f i l e below the mould for d i f f e r e n t spray designs (20.3 x 20.3 cm square b i l l e t ) . Pool p r o f i l e s f o r 10.8 cm square b i l l e t s cast under d i f f e r e n t spray conditions. Pool p r o f i l e s f o r 10.8, 15.2 and 21.3 cm square b i l l e t s corresponding to 4-zone spray design. E f f e c t of speed on mid-face surface temperature p r o f i l e (10.8 x 10.8 cm square b i l l e t ) . E f f e c t of speed on mid-face surface temperature p r o f i l e (15.2 x 15.2 cm square b i l l e t ) . E f f e c t of speed on mid-face surface temperature p r o f i l e (20.3 x 20.3 cm square s e c t i o n ) . Reheating between the nozzles. Comparison of mid-face surface temperature p r o f i l e s f o r e x i s t i n g sprays (old), e x i s t i n g sprays (new) and redesigned sprays with 4-zone spray cooling. Grading standard f or the crack se v e r i t y i n the s u l f u r p r i n t s . E f f e c t of carbon on i n t e r n a l cracks. E f f e c t of cast structure on i n t e r n a l cracks (Heat no. 8). v i i Figure Page 4.1.4.1 S h e l l thickness at time of crack formation as a function of casting speed. Upper p l o t : modified sprays. Lower p l o t : e x i s t i n g sprays. 74 4.1.5.1 Predicted and measured mid-face temperature for a low-carbon b i l l e t cast on strand I. 76 4.1.5.2 Predicted and measured mid-face temperatures f o r a low-carbon b i l l e t cast on strand I I . 77 4.1.5.3 Predicted and measured mid-face temperatures f o r a medium-carbon b i l l e t c a st on strand I. 78 4.1.5.4 Predicted and measured mid-face temperatures f o r a medium-carbon b i l l e t cast on strand I I . 79 4.2.1 Arrangement of nozzles i n modification I and I I . 83 4.2.2 Comparison of s u l f u r p r i n t s f o r heat no. 14 (beginning of heat). 86 4.2.3 Comparison of s u l f u r p r i n t s f o r heat no. 14 (middle of heat). 87 4.4.2.1 Schematic representation of water flow rate and pressures at d i f f e r e n t l o c a t i o n s i n the spray pipe. 90 4.4.3.1 Arrangement of nozzles i n the upper spray pipe f o r laboratory spray t r i a l s . 92 4.4.4.1 Schematic representation of spray modifications I, I I , and I I I . 97 4.5.1 Comparison of s u l f u r p r i n t s f o r heat no. 16. 99 4.5.2 Comparison of s u l f u r p r i n t s f o r heat no. 18. 100 4.5.3 Comparison of s u l f u r p r i n t s f o r heat no. 24. 101 5.1 Schematic representation of the new spray design technique. 105 A . l Arrangement of nodes i n one-quarter of a transverse section. I l l v i i i LIST OF TABLES Table Page 2.3.1 Summary of Various Investigations on High Temperature Mechanical Properties of Steel 15 2.4.1 Summary of Studies of Heat Transfer i n Sprays 21 3.2.1 Present Spray System Design 36 3.2.2 Present Spray Water Conditions 36 3.2.3 Surface Reheating (°C) with E x i s t i n g Sprays 36 3.3.1 Design Casting Speeds 39 3.7.1 Secondary Cooling Design for 10.8 cm Square B i l l e t s 47 3.7.2 Secondary Cooling Design for 15.2 cm Square B i l l e t s 48 3.7.3 Secondary Cooling Design for 20.3 cm Square B i l l e t s 49 4.1.1 Modified Spray Design f o r 10.8 cm B i l l e t s 63 4.1.2 Results of Spray Evaluation - F i r s t Campaign 65 4.1.2.1 Mid-way Cracks and Carbon 68 4.1.2.2 Mid-way Cracks and Mn/S 68 4.1.4.1 Distance From B i l l e t Surface to Innermost Tip of Crack (mm) 73 4.2.1 Comparison of Second to F i r s t M odification of Spray Chamber 82 4.2.2 Results of Spray Evaluation - Second Campaign 85 4.4.3.1 Observations on Spray T r i a l s 93 i x Table Page 4.4.4.1 Corrected Pressures and Water Flow Rates for F i r s t and Second Modifications 94 4.4.4.2 Spray Arrangement and Water Flow Rates Modification IV 96 4.5.1 Results of Spray Evaluation - Fourth Campaign 98 x LIST OF SYMBOLS 2 cross-sectional area of the spray pipe, cm . correlation parameters in Equation 2.4.2. specific heat of the solid steel, kJ/kg-°C. constants in thermal conductivity equation (Equation A.l). acceleration due to gravity. 3 enthalpy of steel at time t and t+At, kJ/m . pressure head of water in the spray pipe, cm. 2 spray heat transfer coefficient, kW/m •°C. integers denoting nodal positions. thermal conductivity of steel, kW/m-°C. thermal conductivity averaged between nodes (l,j) and (2,j), and between (i , l ) and (i,2), kW/irr°C. pressures at different locations in the spray pipe (Figure 4.4.2.1), cm. pressure drop due to f r i c t i o n i n the spray pipe, cm. flow rate of water in the spray pipe at different locations (Figure 4.4.2.1), cm3/s. 3 flow rate of water in spray nozzle (Figure 4.4.2.1), cm /s. 2 instantaneous heat flux from the surface, kW/m . 2 instantaneous and average heat flux in the mould, kW/m . instantaneous heat^flux in the spray zone and the radiation cooling zone, kW/m . temperature, °C. x i temperature of the node ( i , j ) . pouring temperature, °C. temperature of the surface of the steel, of the spray water, and of the ambient environment, °C. time, s. mould dwell time, s. casting speed, cm/s. velocity of spray water droplets at the cooling surface, 2 water flux, 1/m -s. width of b i l l e t , cm. transverse direction, cm. thickness of b i l l e t , cm.. transverse direction, cm. axial direction, cm density emissivity of strand surface (?0.8). stefan-Boltzmann constant x i i ACKNOWLEDGEMENT I would like to sincerely thank my research supervisor Dr. J.K. Brimacombe for his assistance and guidance during the course of this research work. Thanks are also extended to Mr. R.W. Pugh of the Steel Company of Canada and Dr. F. Weinberg for their help and f r u i t f u l discussions. Financial support from the AISI in the form of a research assis-tantship i s gratefully acknowledged. x i i i Chapter 1 INTRODUCTION The development of continuous casting of steel has overshadowed the conventional ingot casting process during the last decade. The growth of the continuous casting process, which has been nearly exponential in nature (Figure 1.1) \\ is not surprising because capital and economical gains are made from the elimination of primary ro l l i n g mills and the achievements of higher product yield. The efficiency of the process is improved even further by sequential casting which is now extensively employed. The continuous casting process has not reached the present stage of acceptability without questions being raised over the quality of the resulting product. However, many of quality problems have recently been solved by operational and metallurgical research so that the process is now widely used for production of most of the steel grades. At U.S. Steel Corporation, for example, more than a thousand man-years were spent before their Gary-slab caster was put into operation in 1967.^ In spite of a l l these efforts and achievements, there are s t i l l many areas in the f i e l d of continuous casting which need further atten-tion. For example, various events taking place in the mould and in the sprays, which are directly related to many quality parameters, are not yet completely understood. One such event i s the surface temperature rebound of the strand which has i t s origin in faulty spray cooling, and leads to formation of mid-way cracks. Not many installations in the world, however, have their sprays designed with this parameter being considered; the general tendency in the b i l l e t machines is to use short spray cooling with 1 2 3 a high flow rate of water especially below the mould. This has encouraged the present study to be undertaken which aims at the prevention of mid-way cracks through modifications in the spray cooling system of an operating b i l l e t caster. Mid-way cracks are shown in Figure 1.2. These cracks run perpendicu-lar to the surface of the strand between parallel dendrites in the colum-nar zone, and have a high solute content of S, P etc. The need to control these cracks comes from the fact that they may not close even after signi-2 ficant r o l l i n g . Recently, Hawbolt et a l . have examined rolled b i l l e t s and shown that significant open defects can be seen in a rolled section even after a 4:1 reduction ratio. In laboratory studies, they found that reduction in excess of 6:1 was required to effectively close mid-way cracks in the continuously cast b i l l e t s . Earlier studies in this f i e l d 3 have been summarized by Ushijima which also indicates that a reduction ra-tio of 2 to 12 i s required before superior mechanical properties can be gained. These cracks i f not welded, can make the intermediate product unsuitable for further processing. In f i n a l products, unwelded cracks may act as internal stress raisers and in service, where cyclic stress conditions are involved, may lead to premature fatigue failure. 1.1 Scope of the Present Study The primary objective of this work is to reduce the incidence of mid-way cracks in continuously cast steel b i l l e t s . This requires a reduc-tion in surface temperature rebound in the strand as i t passes from one cooling zone to another. This has been achieved by designing spray cool-ing systems with the aid of a mathematical model for 10.8, 15.2 and Figure 1.2: S u l f u r p r i n t showing mid-way cracks. 5 20.3 cm square b i l l e t s cast through a straight mould machine. The industrial testwork consisted of casting the same heat with existing spray practice (which has been known to give mid-way cracks) on one strand and with the redesigned sprays on the other. The results obtained from these t r i a l s were then compared. During the course of this study various other parameters affecting the cracking tendency of steel have also been examined. 6 Chapter 2 BACKGROUND WORK This chapter contains the previous work which must be studied before tackling the spray design problem. This includes the mechanism of crack formation, the effect of various parameters on tendency to cracking and spray heat-transfer characteristics. 2.1 Mechanism of Formation of Internal Cracks For any crack to form and propagate, two conditions must be met simultaneously. There has to be a weak soli d i f i e d area and also s u f f i -ciently high tensile strain to tear apart the weak region. In the s o l i d i -fication of steel, such weak areas are provided by columnar crystal growth with the junction between dendrites containing segregated solute elements like sulfur and phosphorus. The second condition of tensile strain arises mainly from the tensile stresses imposed near the solidification front by surface temperature rebound and/or due to phase transformation in steel. Mechanical stresses from a misaligned strand or bending may add to these stresses. A crack then forms i f the combined effect of various stresses is sufficient to separate dendrites in the columnar zone. 2.2 Factors Influencing the Formation of Internal Cracks It i s quite apparent from the foregoing discussion that the parame-ters influencing the strength of steel near the solidus temperature and causing tensile stress at the soli d i f i c a t i o n front w i l l affect crack forma-tion in steel. Some of these factors are discussed below: 7 2.2.1 Solute Elements Sulfur and phosphorus segregate positively in steel. In & -iron, for example, sulfur and phosphorus have a relatively much lower distribu-tion coefficient, kQ, (0.02 and 0.13, respectively) than other elements 4 like C (0.20), Mn (0.90), Si (0.83) etc. This leads to higher concentra-tions of such elements in the residual liquid between adjacent dendrites during soli d i f i c a t i o n . Adams^ has shown in the case of phosphorus that the concentration of this element can be as high as 0.5% when the matrix level is of the order 0.02%. This phenomenon increases the risk of crack formation in steel, because sulfur and phosphorus lower the melting point of the liquid that remains between' the dendrite arms thereby giving rise to low strength and d u c t i l i t y near the solidification front. Evidence of the presence of liquid film i s seen in Figure 2.2.1.1^ which shows an SEM photograph of an open mid-way crack. The smooth dendrite branch that can be seen with no apparent deformation indicates the presence of resid-ual liquid during crack formation at high temperature. The high concen-tration of MnS inclusions on the surface also indicates segregation of sulfur. It is desirable to establish acceptable limits of these elements. However, this i s d i f f i c u l t owing to the strong influence of various other parameters on cracking. It is well known that higher Mn:S ratio can reduce the deleterious effect of sulfur considerably, whereas, as w i l l be discussed in the next section, a predominantly equiaxed cast structure can almost completely eliminate the possibility of mid-way crack formation. It may also be noted that to achieve very low levels of these elements to minimize cracking would be uneconomical for most commercial grades of Figure 2.2.1.1: Scanning e l e c t r o n micrograph of surface of a crack formed near the s o l i d u s temperature. The i n c l u s i o n s are s u l f i d e s . 6 9 s t e e l . Thus, i t appears that the l i m i t s suggested by Vom Ende and Vogt for s u l f u r and phosphorus (0.025% and 0.030% max., respectively) would be reasonable i n most cases, although i t may be required to modify them depending on the product and the operating p r a c t i c e . The effects of copper and t i n , although not studied, are expected to be s i m i l a r to s u l f u r and phosphorus. 2.2.2 Structure In contrast to the columnar structure discussed e a r l i e r , there i s 8 a l e s s e r tendency for segregation i n a predominantly equiaxed structure. I t can withstand adverse conditions l i k e thermal stresses which a r i s e 9 from surface reheat of up to 200°C without i n t e r n a l cracking. Therefore, i t i s advantageous to have a predominantly equiaxed structure i n the cast s t e e l since t h i s parameter alone can co n t r o l the mid-way cracks most e f f e c t i v e l y . I t i s possible to achieve t h i s structure by casting with low s u p e r h e a t . 1 1 This, however, i s not often a prac-t i c a l s o l u t i o n since i n most cases, the length of casting time (around 1 hour) r e s u l t s i n cold s t e e l at the end of a heat. Many operational problems can a r i s e i n t h i s s i t u a t i o n v i z . formation of s k u l l s i n the tundish and the l a d l e , n o z z l e blockage etc., which may lead to return of the heat to the furnace. I t , therefore, remains advantageous to control other parameters to reduce mid-way cracking. 2.2.3 E f f e c t of Carbon and Transformational Stresses 7 According to Vom Ende and Vogt, s t e e l with carbon contents between 0.17% and 0.24% or greater than 0.6% i s more prone to cracking than other 10 carbon steel grades. In medium carbon steel, this phenomenon is obvious-ly associated with the additional tensile stress produced due to volume change arising from the peritectic reaction at the solidus. During this reaction, the b.c.c 6 structure changes to f.c.c. y structure which is 12 accompanied by a linear shrinkage of about 0.38%. The effect of carbon on the transformation characteristics i s dealt with in detail in the next section on high temperature mechanical properties. It might, however, be mentioned that stresses generated due to transformation alone are normally insufficient to cause cracking, but when combined with other stresses, they can enhance the cracking tendency considerably. 2.2.4 Thermal and Mechanical Stresses Thermal tensile stresses are generated close to the solidification front whenever the surface temperature of a b i l l e t or slab rebounds owing to a sharp reduction in heat extraction. There are a number of instances during casting when such changes in heat-transfer characteristics occur viz. when the strand passes from mould to top zone sprays, from one nozzle to the next and from the sprays to the radiation cooling zone. The stresses introduced are compressive in nature at the surface and tensile near the solidification front during surface reheating. This stress 13 pattern has been obtained by Sorimachi and Brimacombe for the 10.2 x 15.2 cm casting with 653 cm long sprays at a casting speed of 2.75 cm/s (Figure 2.2.4.1). According to these authors, the halfway cracks form in the region shown by the shaded area where the tensile strain exceeds 0.2%. Among other stresses, those imposed due to the non-uniformity of shell thickness, which lead to rhomboidity, are important. This condition 11 1600 <-> 1400 CM to CO CO 1000 A = before surface reheating B = after surface reheating (a) 40 Strain / / / / Stress Halfway Cracks (b) 0 1 2 3 4 Distance From Surface (cm) 0-3 0-2 01 0 -01 -0-2 -0-3 -0-4 CM +o c o *-CO Figure 2.2.4.1: Profiles of (a) temperature and (b) stress and strain perpendicular to a side face at centerplane of 10.2 x 15.2 cm b i l l e t cast at 2.75 cm/s with 653 cm long 13 • -sprays. x 12 may a r i s e as a r e s u l t o f m i s a l i g n m e n t o f t h e s t r a n d , mould d i s t o r t i o n , improper stream p o s i t i o n i n g , non-uniform c o o l i n g i n t h e s e c o n d a r y - c o o l i n g zones, improper p o s i t i o n i n g o f t h e g u i d e r o l l s e t c . S t r e s s e s a r e a l s o g e n e r a t e d due t o r a p i d c o o l i n g o f the s t r a n d , f e r r o s t a t i c p r e s s u r e , bend-i n g o r s t r a i g h t e n i n g f o r c e s e t c . These s t r e s s e s l e a d t o v a r i o u s i n t e r n a l d e f e c t s which may enhance t h e mid-way c r a c k i n g tendency by a d d i n g t o t h e a l r e a d y e x i s t i n g s t r e s s e s . In a w e l l o p e r a t e d p l a n t , however, t h e e f f e c t o f t h e se parameters may not be v e r y c r i t i c a l . 2.2.5 Comments I t appears from t h e d i s c u s s i o n on t h e v a r i o u s parameters a f f e c t i n g i n t e r n a l c r a c k i n g t h a t c o n t r o l s on s u l f u r and phosphorus o r s u p e r h e a t i n the t u n d i s h a r e n o t p r a c t i c a b l e due to ec o n o m i c a l and o p e r a t i o n a l c o n s i d e r a -t i o n s , r e s p e c t i v e l y . T h i s l e a v e s one r e m a i n i n g c o n t r o l parameter, t h a t i s , the c o n t r o l o f r e h e a t i n g o f t h e s t r a n d s u r f a c e , e s p e c i a l l y i n s t e e l s which undergo t h e 6—>y t r a n s f o r m a t i o n . T h e r e f o r e , the p r e s e n t work i s f o c u s s e d on t h e d e s i g n o f a sp r a y c o o l i n g system which r e s u l t s i n minimal r e h e a t i n g o f t h e s t r a n d d u r i n g i t s passage from one c o o l i n g zone t o t h e ne x t . 2.3 High Temperature M e c h a n i c a l P r o p e r t i e s o f S t e e l R e c o g n i z i n g the f a c t t h a t c r a c k s form a t t h e s o l i d i f i c a t i o n f r o n t , i t i s e s s e n t i a l t o know t h e m e c h a n i c a l p r o p e r t i e s o f s t e e l c l o s e t o i t s m e l t i n g p o i n t . These p r o p e r t i e s a r e a f f e c t e d by s e v e r a l v a r i a b l e s v i z . 6 temperature, s t e e l c h e m i s t r y , s t r a i n r a t e , and t h e r m a l h i s t o r y . Many l a b o r a t o r y s t u d i e s ^ ' \" ^ 1 6 have been c a r r i e d o u t u s i n g d i f f e r e n t a p p a r a t a 13 and methods to simulate the conditions during casting. The results are usually given in terms of reduction of area such as shown in Figure 2.3.1 where i t can be seen that there i s a sudden drop in ducti l i t y at a particular temperature (referred to as ductile-brittle (D-B) transition temperature) for a given steel. Steel is b r i t t l e between this tempera-ture and the solidus, and i s prone to cracking in this region when strained. The outcome of these studies has been summarized in Table 2.3.1, and is represented graphically in Figure 2.3.2. A wide difference i s observed in the D-B transition temperature reported by various investiga-tors. Also, observations regarding other variables like structure and solute content are not consistent with each other and with practical obser-vations. These discrepancies are not surprising i f one considers the number of variables involved and the problems faced in duplicating the casting situation. These studies, thus, may have d i f f i c u l t i e s in direct application in., the stress models; but have importance in revealing a qualitative picture of the mechanical behaviour of steel at high tempera-ture . It i s interesting to note the cracking tendency in steel in the light of the above studies and the 6—*-y transformation. With reference to Figure 2.3.2 the following observations can be made: In low-carbon steels (<0.09% C), the transformation occurs in the temperature range where steel i s ductile. Around 0.09% C, the liquid steel f i r s t transforms to 6 completely, and then to y in the temperature range where steel i s b r i t t l e . Thus, this should be the ideal range for cracking. In practice, however, this i s not the case. It has been shown by Nilles et a l . 1 5 that equilibrium conditions 14 100 80 g 4 0 I-o P, 20| • O\" Ductility Steel Grades O - C 1 0 1 2 a - C 1025 o - C 1043 0 - C 1095 .^ -fpm-pfn-pA-rrn-1800 2000 2200 TEMPERATURE, °F 2400 2600 Figure 2.3.1: Ductility and strength curves for various plain carbon steels.^ Table 2.3.1 Summary of Various Investigations on High Temperature Mechanical Properties of Steel Investigator Sample Position Apparatus S t r a i n Rate, s -1 Observations and Conclusions C.J. Adams C h i l l Zone Gleeble Machine P.J. Wray 14 2.5 2.8(10 _ 5) to 2 . 3 U 0 \" 3 ) Four p l a i n carbon s t e e l from 0.12 to 0.95% C, a resulfurized s t e e l and a stainless s t e e l were studied. Heating-cooling cycles (sensitizing) had no e f f e c t . Decreasing s t r a i n rate lowers the strength curve, but has l i t t l e e f f e c t on d u c t i l i t y curve. Resulfurized steels have similar curves as normal steels of same composition so long as Mn i s high. In C-1025 and C-1095 steels, grain-boundary enrichment of phosporus was ob-served by probe analysis to be between 0.2 and 0.5% (from 0.020% base level) which lowers the melting point of steel considerably. Thus, the major cause of f a i l u r e i s phosphorus segregation and not sulfide s melting. Studied d e l t a - f e r r i t e iron containing up to 0.044% C. Strain rate a f f e c t s the stress l e v e l . F a i l u r e i n d u c t i l e manner when deformed i n tension up to the melting point. P. K i l l e s et a l . 1 5 Gleeble Machine Weinberg 16 0.3 minimum Across the columnar zone Instron Machine 0.017 Three p l a i n C steels containing 0.09, 0.185, and 0.4% C were studied. Increasing sulfur content reduces the d u c t i l e - b r i t t l e t r a n s i t i o n temperature appreciably i n low carbon steels. In 0.014% S and 0.025% S st e e l s the D-B tr a n s i t i o n temperature was 40°C and 70°C below the solidus respectively. In 0.1% C steels, 6 - ytransformation takes place much below the equilibrium transformation temperature, whereas, i n 0.175% C steels, the p e r i t e c t i c trans-formation i s completed within a short temperature range near the solidus. Various steels with carbon varying from 0.05 to 1.0% were studied. D u c t i l e - b r i t t l e t r a n s i t i o n temperature was found to be independent of cast structure and residual l e v e l other than carbon. Increasing s t r a i n rate up to 0.33 s--*- had no e f f e c t on the D-B t r a n s i t i o n temperature. Fracture surface of low carbon steels _(<0.06%) shows large facets i n d i c a t i n g r e c r y s t a l l i z a t i o n and grain growth. Grain boundary movement could be the cause of high phosphorus concentration. Up to 0.2% C showed smaller facets of r e c r y s t a l l i z e d grains. Above 0.2% C, there i s no r e c r y s t a l l i z a t i o n . Failure occurs due to i n t e r -dendritic melting. 16 Figure 2.3.2: Ductile-brittle transition temperature and solidus as a function of carbon content in the steel. 17 are not observed in this case during cooling, and that the reaction takes place at temperatures much below the equilibrium transformation tempera-ture (1493°C) between 1443° and 1390°C, i.e. below the low duct i l i t y range. The cracking tendency i s , therefore, not affected by the transfor-mation stresses in this case. Around 0.19% C, the peritectic reaction (6+L—>- y ) takes place at 1493°C. Near the peritectic point, this reaction i s no longer slow, and takes place in a short temperature range at and near the solidus tempera-ture where the steel i s b r i t t l e . Higher carbon steels undergo less transformation of L—>-<5, and after 0.53%C, no <5 phase i s encountered; thus, with increasing carbon, the transformational stresses become less important. It i s apparent from the foregoing discussion, that the only carbon range where the steel i s most susceptible to the influence of stresses due to the 6 — t r a n s f o r m a t i o n i s around the peritectic composition. This coincides with observations made in the plant, and confirms the measure-, . . 31 ments of Morozenskn. It must be mentioned here that based on these studies, there is no apparent reason for steel containing more than 0.6% C to crack as cited 7 in the literature. In fact, the present author has not experienced-any problem of mid-way cracks in casting 100 mm square b i l l e t s with 0.6% to 0.8% C steel with <0.03% sulfur and high superheat. 2.4 Heat Transfer Characteristics in Sprays Heat extraction in water sprays is very d i f f i c u l t to predict theoretically owing to a number of parameters affecting the heat-transfer 18 conditions. These parameters have been enumerated by Baptista as 1. water flux 2. water impact velocity 3. angle of impingement 4. size of water droplets 5. surface temperature 6. surface roughness 7. type of scale on the surface 8. interaction among spray nozzles 9. angle of the sprayed surface to the horizontal. Several studies ranging from single droplet to in-plant measurements have been carried out to determine the effect of these parameters on the heat transfer coefficient, h., which characterizes heat extraction in the s sprays. 18 Pedersen, for example, has investigated single water droplets and shown that as the surface temperature increases, there i s a marked transi-tion from wetting to a non-wetting regime. In the wetting regime, the droplet wets a considerable area of the surface and conventional nucleate boiling takes place. In the non-wetting regime, which i s normally encoun-tered in continuous casting, the droplet breaks up into a number of small droplets and vaporizes suddenly. In this case the spray heat transfer coefficient, h g, increases with increasing approach velocity. Also, the transition temperature for the wetting to non-wetting regime i s affected by the approach velocity and the surface condition. Several laboratory studies have been conducted to correlate the heat transfer coefficient to operating variables, e.g. Muller and 19 20 21 Jeschar, Mitsutsuka, and Mizikar. Muller and Jeschar have measured the electrical energy required to maintain a constant surface temperature on a resistance-heated steel plate. They correlated h to the impact 2 velocity of water V, m/s, and water flux W, 1/m -s, by the following equation (effect of radiation excluded): h = 0.01 V + (0.107 + 0.00068 V) • W kW/m2-°C (Eq. 2.4.1) s Mitsutsuka found that h <* Wa (1 - b-T ) (Eq. 2.4.2) s w where a i s between 0.5 and 0.8, and for a water flow rate of around 2 10 1/m -s, b is between 0.005 to 0.008. This equation has been employed 22-26 27 wxth modification by several workers. Bolle and Moureau have found a similar relationship as Eq. 2.4.2 in a study involving a horizon-t a l plate sprayed from the top and the bottom. They found that the heat transfer coefficient has a value about 15% lower when the plate is sprayed from the bottom in comparison to when i t i s sprayed from the top. These results may be useful when designing sprays for the curved-mould continuous casting machine. Another classical example of laboratory work on sprays i s that by 21 Mizikar. He obtained a linear relationship between the heat transfer coefficient and the water flux given by h = 0.0776 W + 0.22 (at 276 kPa) (Eq. 2.4.3) s h = 0.1 W + 0.22 (at 620 kPa) (Eq. 2.4.4) s where the constant 0.22 is due to heat transfer by radiation. 28 Alberni has studied spray heat transfer directly on a centrifugal continuous-casting machine by measuring surface temperatures. His investi-gations indicate that at high water flux, the heat transfer coefficient 20 reaches a nearly constant value. This phenomenon of \"saturation\" of h s has not been reported by other authors in laboratory studies. These investigations have been summarized in Table 2.4.1, and plotted in Figure 2.4.1. In the case of the study of Alberni, there are 3 curves which correspond to different spray settings. It may be noted that the correlation between the heat transfer coefficient and water flux i s not consistent from one study to the next. Also the effect of other parameters like impact velocity of the droplet, surface temperature of the steel or water temperature etc. have not been included in a l l the equations. Wide scatter in h V W relationship s s (Figure 2.4.1) i s , therefore, seen in the normal working range of water 2 flux (0 to 16 1/m -s). Thus, when choosing a particular equation, one must ensure that i t has been obtained from the study which i s closest to the operating system under investigation. At the time of starting the present work, Mizikar's study was found to be most suitable since i t pertains to 1/4 GG10 nozzles which were also used on the billet-caster under investigation. His data was, there-fore, used in this study. 2.5 Mathematical Model In order to design optimum spray conditions, one needs to have a thorough knowledge of the heat extraction rates in the different cooling zones and also a mathematical model which i s capable of producing a three-dimensional temperature distribution in the strand. This work has been 9 described in detail by Brimacombe. Table 2.4.1 Summary of Studies of Heat Transfer i n Sprays Method used for -obtaining data for Equation for the the heat transfer heat transfer Reference c o e f f i c i e n t c o e f f i c i e n t Comments 19 Muller and Jeschar Measured e l e c t r i c a l h = 0.01 V + (0.107 + 0.00068 V)*W S ± 12% - h changes only s l i g h t l y i n the temperature energy to maintain range studied (700 - 1200*C). a constant surface temperature on a resistance-heated (radiation effect excluded) l i m i t s : 0.3 - ~ — < W < 9.0 — - Effect of impact velocity, V, has been accounted for. steel plate. m * s m • s 11 51 < V < 32 51 s ~ s 20 Kitsutsuka Experiments on p l a i n h * wa (1 - bT ) s w - a i s between 0.5 and 0.8. carbon stee l raised to 930°C; then cooled by water spray (38°C) 2 - For water flow 10 to 10.3 1/m *s, b i s between 0.005 to 0.008. from both sides. Shimada and h = 1.57 W ° * 5 5 (1-0.0075 T ) s w Mitsutsuka22 - This formula i s l i k e l y to give very high heat transfer coefficent. It probably needs an adjustment parameter as given by Nozaki et a l . 2 3 23 Kozaki et a l . Used Shimada and h = 1.57 W ° ' 5 5 (1-0.0075 T ) / a s w - This equation accounts for the e f f e c t of Mitsutsuka's equation water temperature, T^, on h^. - Changing water temperature from 5°C to 30°C w i l l , thus, decrease h g by 20%. with adjustment c o e f f i c i e n t a. where a = 4 from surface tempera-ture measurements at the straightner. Ishiguro et a l . ^ 4 Used Mitsutsuka's h = 0.5813 W0*451 {1-0.0075 T ) s w - Similar case as above. equation. Table 2.4.1 Summary of Studies of Heat Transfer in Sprays (Continued) Kawakazu et al.25 Used Mitsutsuka's equation 36.2 W Neglected effect of water temperature Units of W are not clear. Probably l/m2*min. Relationship between h and W is not consis-tent with the work of other investigators. Sasaki et al Steel plate heated in a furnace. Temperature measure-ments by optical pyrometer. h = 708 W°* 7 5 T \" 1* 2 + 0.116 s 0 limits: 700 < TQ S 1200°C 1.67 S W < 41.8 This equation accoutns for the effect of surface temperature T^ . Effect of water temperature has been ob-served but not included in the equation. Constant 0.116 in the equation is smaller than the heat transfer coefficient for radiation alone at 1100°C (0.184 kW/m2-s). 21 Mizikar Used stainless steel a) h = 0.0776 W (at 276 kPa) s - h increases with water flux s plate. b) h s = 0.1 W (at 620 kPa) - h s increases with droplet velocity Temperature measure-ment by two thermo-couples (Radiation effect is excluded in both equations) - h s is not affected by surface temperature - h is not affected by angle of impingement. s Bolle and Used horizontal plate\" a) Sprays from the top - This work is specifically useful in curved Moureau27 h s = 0.423 rt°*556 ± 17% (Radiation excluded) 627 < T < 927°C mould machines. - Sprays from the bottom give about 15% lower value of h than for sprays from the b) s ~ Sprays from the bottom top. h s = 0.360 W0,556 ± 17% (Radiation excluded) This work was done for relatively lower temperatures and lower water flux. 0.8 < W < 2.5 1/nT-s to Table 2.4.1 Summary of Studies of Heat Transfer in Sprays (Continued) Alberni Used actual plant data from a centri-fugal continuous casting machine. Surface temperature measurement by two-colour optical pyrometer. At high water flux, saturation of h is observed beyond a certain point. hg depends on actual impact of droplets and also on the homogeneity in the dis-tribution.of droplets. 19 Muller & Jeschar Nozaki et a l . 2 ^ 24 — • — Ishiguro et a l . — o — o — S a s a k i et a l . 2 6 Mizikar 2 1 (at 276 kPa) Mizikar 2 1 (at 620 kPa) — A — A — Bolle & Moureau2^ (horizontal plate, top spray) /j£ — A — A — Bolle & Moureau2^ (hori- / zontal plate, bottom spray) / -^ 1 #-— • — • - Alberni 28 uO^^T^ 3? f i t * / •-•-•-o-a - D - D-a-o - D-n - D-o i- D _ a\" D -•-•-•- D _ 20 30 40 Water flux, l / m s Figure 2.4.1: Relationship between the spray heat transfer coefficient, h g, and water flux, W, as obtained by various investigators. 4^ 25 2.5.1 Mathematical Formulation The heat flow can be easily described by the following three-dimensional steady-state equation: 3 ( k | T } 3 ( k 9T = 3 T = | | 2 > 5 _ 1 - 1 } 3x 3x 3y 3y 3z 3t where u = withdrawal rate and H = p-C-T. It may be noted that in deriving this equation, axial conduction i s neglected since i t is much smaller than the bulk flow of heat in the withdrawal direction. The two-dimensional form of the equation permits the temperature to be predicted in the trans-verse plane of a slice descending through the casting machine. This equation requires four boundary conditions and an i n i t i a l condition for i t s solution. These conditions are given below: Surface x = 0, z > 0, 0 < y < |, -k |^ = q Q (Eq. 2.5.1.2) X 3T y = 0, z > 0 , O ^ x ^ - , -k — = q (Eq. 2.5.1.3) z 3y u where q„ is the instantaneous heat flux from the surface. Centre plane, assuming symmetrical heat flow y Y 3T x = f ' z - ° ' ° - y - 2 ' ~ k 3 ^ = 0 ( E q \" 2 - 5 - 1 - 4 ) y = |, z > 0, 0 < x < |, -k = 0 (Eq. 2.5.1.5) I n i t i a l Condition X Y z = 0, 0 < x < — , 0 < y < — , T = T (Eq. 2.5.1.6) 26 Equations 2.5.1.1 to 2.5.1.6 give a complete mathematical description of the heat flow in continuous casting. The solution of this problem could be obtained in several different ways. It would, however, be impracticable to seek the solution by analyti-cal or integral profile methods owing to the complexities arising from the variations in thermophysical properties and heat transfer coefficients with temperature, and also from the evolution of latent heat during s o l i d i -fication. Numerical methods (finite difference or f i n i t e element) are, therefore, more suitable for predicting the heat flow characteristics in continuous casting. Various methods of solving the partial differential 29 . . equations have been described by Carnahan et a l . , a l l of them giving virtually the same results. In the present study, the explici t f i n i t e difference method has been used because of i t s simplicity in solving the partial differential equation; the disadvantage of this method i s that a stability criterion must be f u l f i l l e d to obtain a convergent solution. The f i n i t e difference approximation of Equations 2.5.1.1 to 2.5.1.6 i s given in Appendix I, which also includes the solution for the inverse boundary conditions to obtain the heat transfer coefficient from a specified surface temperature condition. Before proceeding to the computer program, i t i s necessary to review the treatment given to various parameters in the equations. 2.5.2 Characterization of Input Conditions a) Thermal Conductivity of Liquid: Turbulent mixing imparted by the input metal stream makes i t d i f f i c u l t to assign proper values to this parameter. An arbitrary value, five to ten-fold larger than normal, i s 27 used in the model, which, fortunately, has a minor effect on the thickness of and temperatures in the solid shell. b) Heat Extraction in the Mould: The most complex mode of heat extraction i s in the mould since i t i s influenced markedly by the formation of a gap between the solid shell and the mould face. L i t t l e i s known about the nature of this gap, the magnitude of i t s variation and the resultant effect on the rate of heat extraction. In the absence of such mechanistic data, efforts have been made to obtain useful heat flow data in:the mould.from a heat balance on the mould cooling water. Such a rela-30 tionship has been obtained by Lait et a l . , wherein, the average mould heat flux, q , arid the mould dwell time, t , are related by m m q\\ = 2675 - 221 V€ (Eq. 2.5.2.1) in m Z m The instantaneous heat flux, q^, is then given by = 2675 - 334 (Eq. 2.5.2.2) It should be emphasized that i s a function only of t, or distance below the meniscus (=z/u), and that i t s variation in the transverse direction from mid-face to corner i s not accounted for. c) Heat Extraction in Sprays: The complex mode of heat transfer in sprays can be characterized by a spray heat transfer coefficient, h s, which i s related to heat flux by v = h s ( T o - v ( E q - 2 - 5 - 2 - 3 ) m As discussed earlier (Section 2.4) among the various studies carried out 21 on correlating h g to water flux, Mizikar's data was chosen for the 28 present work because the type of nozzle employed in both studies is the same. His study i s mainly divided into three categories: spray charac-terization, heat transfer studies, and engineering of a spray pattern. Figure 2.5.2.1 shows the water flux at different locations on the b i l l e t surface as a function of spray distance and water pressure, whereas, Figure 2.5.2.2 shows the relationship between the heat transfer coeffi-cient and the water flux. Mizikar also observed that the normal tempera-tures in continuous casting are above the c r i t i c a l temperature where film boiling prevails, and that the heat transfer coefficient in this region is relatively independent of surface temperature. The heat transfer coefficient, however, i s affected by the dynamics of droplet collisions as shown in Figure 2.5.2.2 where i t can be seen that h g i s higher at higher pressure for the same water flux. d) Heat Extraction in the Radiant Cooling Zone: Perhaps the easiest of a l l the surface heat transfer zones to characterize mathemat-i c a l l y i s the radiant cooling zone. Here the heat extraction from the surface is given by the Stefan-Boltzmann equation (Eq. 2.5.2.4) The value of emissivity, e, i s usually taken as 0.8 for the oxidized strand surface. In the basic equations, q^ may be replaced by q^, q^ or q^ depending on the position of the slice in the strand, whereas, the tempera-ture and the thermophysical properties of liquid and solid can be denoted by subscripts 1 and s respectively. Figure 2.5.2.1: Average spray water flux as a function of spray distance and water pressure for 1/4 GG10 type nozzle.21 1 gal/min.ft = 0.67 1/m -s 1 psi = 1 inch = 2.54 cm 6.9 kPa WATER FLUX (gal./min., ft 2) Figure 2.5.2.2: Effect of water pressure on the film heat transfer coefficient. 2-'-1 psi =6.9 kPa 1 BTU/hr.ft2•°F = 5.67 x 10 3 kW/m2-°C 2 2 1 gal/min.ft. = 0.67 1/m -s CO o 31 2.5.3 Computer Program The computer program, already developed in the Department of Metallurgical Engineering, U.B.C., was used for the purpose of further calculations. This program is based on the principles described earlier, and has been verified by comparing computer-predicted and measured pool profiles. With input parameters like section size, pouring temperature, carbon content, working mould length, casting speed and details of sprays, the computer program gives the three dimensional steady-state temperature distribution of the strand as output. Conversely, i f surface temperatures are used as input condition, the program gives surface heat transfer coefficients as output. The flow chart of the computer program is given in Figure 2.5.3.1. 32 Figure 2.5.3.1: Flow Chart of the Computer Program Calculate thermophysical properties, solidus and liquidus temperatures, enthalpies at these and the pouring temperatures etc. pertaining to the given carbon content. T Calculate node size, and set the i n i -t i a l and the boundary conditions. Check the st a b i l i t y criterion; change the specified time step i f necessary. Time step k = 1, kcount Calculate time, t = k x At Calculate the enthalpy of the interior nodes for the current time step with appropriate values of the thermophysical properties. Decide the position of the cast, calculate the heat flux from the specified heat transfer co-efficient or the specified surface temperature. i Calculate the current enthalpy of' surface and corner nodes. Calculate the new temperatures of a l l the nodes from the current enthalpies. \\ : z z Write the results 33 Chapter 3 MODIFICATION OF SECONDARY COOLING SYSTEM With the help of the studies carried out in Chapter 2 the cooling system on an operating b i l l e t caster was modified to reduce the extent of reheating at different positions of the strand. The internal quality of the b i l l e t s produced on this machine was not consistently good which made i t possible to see the effect of the modification on the present quality parameter. This chapter, thus, defines the spray design c r i t e r i a , describes the caster, and explains the method of spray design. 3.1 Criteria for Spray Design There are mainly two c r i t e r i a that need to be satisfied for the design study: (i) the extent of permissible reheating of the surface during casting, and (ii) the temperature at which the midTSurface region should'be maintained in-the sprays. This quantification i s very essential, but at the same time, an extremely d i f f i c u l t task, especially in regard to the f i r s t criterion. As discussed in the earlier sections, the cast structure and the composi-\" tion are very important parameters which change the degree of reheat that can be sustained by the steel without cracking. Also depending on the heat transfer conditions, thickness of casting, and the rate of reheating, the same amount of reheat may result in different stress-strain conditions at the solidification front and alter the extent of permissible reheat. A somewhat simplified approach was, therefore, taken in deciding this criterion. It is based on the fact that the common 34 grades of steel have a ducti l i t y to fracture of 0.2 to 0.3% at high temper-ature; thus, the reheat which gives rise to expansion in excess of 0.2% may cause a crack to form. From the thermal expansion coefficient of steel i t can be deduced that a 100°C:rise in temperature i s required to produce 0.2% expansion; and hence the f i r s t criterion i n the present study i s that the maximum reheating at the surface should be less than 100°C. It may be stated here that this criterion i s not based on the strain at the solidification front, and that the approach taken to quantify i t is not completely satisfactory. However, in the absence of more r e a l i s t i c information in this f i e l d , specification of a maximum 100°C reheat seems to be adequate. The second criterion regarding the strand temperature in sprays i s relatively easy to define. Normally, the higher the temperature of the strand in the sprays, the lower i s the reheating when the strand passes from the sprays to the radiant cooling zone. However, in order to achieve efficient operation of the continuous casting process, i t is necessary to maintain a high rate of solidification, which can be obtained i f the surface temperature of the strand is kept relatively low. In the present study, a temperature of 1150°C in the sprays was found to be a satisfactory compromise satisfying both conditions. 3.2 Continuous Casting Machine The continuous casting machine under study i s being run to produce 10.8, 15.2 and 20.3 cm square b i l l e t s . The machine i s a straight mould type caster with no r o l l support of the b i l l e t between the mould and the 35 pinch r o l l . The present casting rate for these sizes was 4, 2.75 and 1.7 cm/s respectively. The spray system for the caster is comprised of three zones, details of which are given in Table 3.2.1; the spray water flow rates and pressures are given in Table 3.2.2. Under conditions of high superheat and high levels of P and S, mid-way cracks are observed in b i l l e t sections from this machine. An example i s shown in Figure 1.2. The need for spray redesign on the machine to improve this situation is clear i f one examines the surface temperature profiles that are presently obtained with the existing sprays. These have been calculated with the computer model and are presented in Figure 3.2.1. As can be seen, there is a large amount of reheating from the bottom of Zone I through the early part of the radiation cooling region for a l l three b i l l e t sizes. The values of surface reheating are summarized in Table 3.2.3. The surface reheating has been broken into i t s zonal constituents in Table 3.2.3 to show that for each b i l l e t size even a single reheating event gives rise to a surface temperature rebound exceeding or approaching the \" c r i t i c a l \" value of 100°C. It may be noted that the effect of i n d i v i -dual reheats may not be totally additive because the stresses generated due to one reheat may be relieved by the time the strand experiences a further reheat. For example, in the case of the surface temperature pro-f i l e in 10.8 x 10.8 cm b i l l e t (Figure 3.2.1), the time over which reheating occurs following the third spray zone i s about 30 seconds, which may be long enough to relieve stresses from earlier reheating. Addition of reheats may be valid when the two reheats immediately follow one another. 36 Table 3.2.1 Present Spray System Design Zone Length (cm) Nozzle Type Number Nozzle Spacing (cm) I 15.2 Fan Type 1 1. 3 cm from mould 1 7. 6 cm from mould II 26.7 1/4 GG10 4 8.9 III 142.2 1/4 GG10 8 17.8 Table 3. 2.2 Present Spray Water Conditions B i l l e t Size Zone I Flow Zone II and III Nozzle Pressure (cm) Rate per Face Flow Rate per in Zone II and (1/s @ 414 kPa) Face (1/s) III (kPa) 10.8 x 10.8 0.227 1.060 138 15.2 x 15.2 0.227 1.483 276 20.3 x 20.3 0.232 1.521 304 Table 3. 2.3 Surface Reheating (°C) with Existing Sprays B i l l e t Size Zone I Zone II -»- Zone III (cm) Zone II Zone III Radiation 10.8 x 10.8 90 82 110 15.2 x 15.2 19 90 170 20.3 x 20.3 19 72 159 Figure 3.2.1: Mid-face surface temperature profiles for the existing sprays. 38 3.3 Design Constraints and the Spray Design Steps As might be expected in an industrial situation, the spray redesign was subject to a number of constraints: (i) the modification to the existing spray system should be minimal. This meant that insofar as was possible, the spray length and the nozzle type should be l e f t unchanged; (ii) a maximum water flow, determined by existing capacity should not be exceeded; ( i i i ) spray conditions within a given zone should be uniform; (iv) the spray design, particularly nozzle types and zone lengths should be similar for a l l three sizes of b i l l e t . One of the constraints that could be relaxed, i f i t proved desirable, was an increase in the overall length of the sprays by the addition of a fourth spray zone. The maximum length of the fourth zone which was limited by machine layout, was 137 cm. The design casting speeds were constrained within the practicable range shown in Table 3.3.1. With these constraints, and the aid of the heat flow model, the spray design can be broken down into the following steps: (i) specification of the surface temperature map desired in the sprays; (ii) generation of spray heat transfer coefficient map using the heat flow model; ( i i i ) determination of most suitable length for each zone of sprays; (iv) specification of nozzle type, nozzle-to-nozzle and nozzle-Table 3.3.1 Design Casting Speeds B i l l e t Size Minimum Casting Maximum Casting (cm) Speed (cm/s) Speed (cm/s) 10.8 x 10.8 4.0 6.35 15.2 x 15.2 2.75 3.39 20.3 x 20.3 1.7 2.12 40 to - b i l l e t spacing and pressure using experimental spray data. These steps are explained in detail in the following sections. 3.4 Selection of the Surface Temperature Distribution The f i r s t step in the design is specification of the surface temperature distribution. Four hypothetical cases were considered for each b i l l e t size. These cases were: 1) steady change of mid-face temperature to 1100°C from mould bottom; 2) steady change of mid-face temperature to 1150°C from mould bottom; 3) change of mid-face temperature to 1100°C at end of Zone I sprays and constant thereafter to the bottom of the sprays; 4) change of mid-face temperature to 1150°C at end of Zone I sprays and constant thereafter to the bottom of the sprays. The temperature distribution across the face for each case was determined in a separate computer run using a radiative heat flux for the sub-mould, surface boundary condition. The surface temperature profile was then taken from the resulting calculated temperature f i e l d at a point where the mid-face surface temperature equalled a pre-set value (1100°C or 1150°C in the above cases). This a r t i f i c i a l procedure was adopted to ensure the surface temperature profile near the bottom of the sprays was reasonably close to the distribution due to radiation alone. The choice of the above cases i s not entirely arbitrary. If surface temperatures much below 1100°C are sought, surface reheating becomes 41 increasingly serious below the sprays. On the other hand, surface tempera-tures above 1200°C result in a weak shell and slightly lower solidification rate. The mid-face temperature profile for the four cases is presented in Figure 3.4.1 for the 10.8 cm square b i l l e t . This shows that reheating below the sprays is least (79°C) with case 4. Therefore, this case was used in the subsequent design calculations. 3.5 Calculation of Spray Heat Transfer Coefficients Given the surface temperature distribution to define the surface boundary condition, the heat flow model was employed to compute the two dimensional, spray heat transfer coefficient distribution. It was decided at the outset to divide this region up into three zones corres-ponding to the existing spray zones, and to maintain spray conditions constant within each zone. To transform the constantly decreasing (with distance) heat transfer coefficients into a single profile within each zone the following procedure was adopted. The heat transfer coefficients at a given position relative to the mid-face were simply averaged as shown schematically in Figure 3.5.1. The heat transfer coefficient pro-f i l e for the zone was then obtained by assembling the averaged values. 3.6 Determination of Spray Parameters To arrive at a spray arrangement that w i l l give the desired heat transfer coefficient profile in each zone, the experimental data of 21 Mizikar has been employed. As mentioned in Section 2.5.2, he has related the spray heat transfer coefficient to spray water flux for the Mid-face Temperature (°C) 1000 1100 1200 1300 Figure 3.4.1: Hypothetical mid-face surface temperature profiles to obtain surface heat flux (10.8 x 10.8 cm square b i l l e t s ) . Beginning of Zone End of Zone 0 © © o h] hghjhj,. hfh 2 2h 2 3h 2 4 1 I-i 1 1 1 I 1 rH n rri n 1 1 i | 1 1 1 1 l i t ! 1 1 1 1 1 1 h > m n-i n m m Heat Transfer Coefficient Contour \\^ Corner Midface Heat Transfer! Coeff. Corner Figure 3 . 5.1: Schematic representation of the method used f or determining average heat transfer c o e f f i c i e n t p r o f i l e i n the given zone. 44 1/4 GG10 nozzle as was shown in Figure 2.5.2.2. He has also measured the water flux distribution for this nozzle at different pressures (Figure 2.5.2.1). Thus, using Figure 2.5.2.1, the water flux profile can be determined and, from Figure 2.5.2.2 the most suitable spray pressure and nozzle-to-nozzle spacing obtained. It should be noted that the nozzle-to-billet distance has been fixed such that the spray does not extend beyond the corners. The choice of spray pressure and nozzle-to-nozzle spacing involves some iterations obviously because both i n f l u -ence the water flux and spray heat transfer coefficient. In this work, the nozzle-to-nozzle spacing i s retained as in the existing arrangement and average water flux distributions and corresponding heat transfer coefficient profiles are obtained for different pressures. These pro-f i l e s are then compared with the desired spray heat transfer coefficient profile (as predicted by computer), and the most suitable arrangement is adopted. Figure 3.6.1 (a-d) illustrates the procedure in greater detail for the specific case of the 15.2 cm square b i l l e t with 1/4 GG10 nozzles oper-ating at 207 kPa pressure. The nozzle-to-billet and nozzle-to-nozzle distances are 11.8 and 8.9 cm respectively. Figure 3.6.1(a) shows the spray pattern, from two adjacent nozzles, broken down into concentric 2 annuli. The number indicates the water flux in 1/m -s (obtained from Figure 2.5.2.1) while the shaded area shows the element of two-fold sym-metry between the nozzles. A more detailed picture of the water flux in the region of symmetry can be seen in Figure 3.6.1(b). The average water flux i s presented in Figure 3.6.1(c) together with a straight line approxi-mation of the water flux profile. Figure 3.6.1(d) gives the corresponding (a) Figure 3.6.1: Method of obtaining heat transfer coefficient pr o f i l e for 1/4 GG10 nozzles spaced 8.9 cm apart at a \" distance of 11.8 cm from the b i l l e t surface (water pressure 207 kPa) 2 (a) Spray pattern on b i l l e t (Figures in 1/m >s) 2 (b) Approximate water flux (1/m -s) in different regions of the symmetric section. (c) Water flux distribution from center to the corner of the cast. (d) Heat transfer coefficient profile. Center Corner 10-7 6-7 4-7 — ) — 6-7 10 \\ 30 8 8 / 6 7 9-3 9-2 7-5 6-9 30 (b) 10-0 W Water Flux 1/m s 50 -j. ^ Averaged from (7b) Linearized ^ _ Profile _ i i (c) 46 heat transfer coefficient profile (using Figure 2.5.2.2) and the profile calculated by the computer. As can be seen there is a reasonably good match at the mid-face and corner but in between there i s less coincidence. Slightly better agreement can be obtained by increasing the pressure. 3.7 Outcome of the Spray Design Study Using the procedure described in the last section, heat transfer coefficient profiles were obtained for different cooling zones, and several alternative spray designs were examined for the 10.8, 15.3 and 20.3 cm square b i l l e t s . These and the existing design are summarized in Tables 3.7.1 to 3.7.3. The effect of these designs on the mid-face temperature profile is shown in Figures 3.7.1 to 3.7.3 for the three sizes respectively. Figure 3.7.4 shows the pool profiles for 10.8 cm square b i l l e t s cast under different spray conditions (for existing sprays, f i r s t design and the third design in Table 3.7.1), whereas Figure 3.7.5 gives the pool profiles for 10.8, 15.2 and 21.3 cm square b i l l e t s corresponding to the 4-zone spray design (third design in Table 3.7.1 to 3.7.3) The f i r s t design had the same spray length with three zones, as the existing sprays; the speed, however, was higher. As can be seen for a l l the three sizes the new design is a distinct improvement over the existing sprays, but excessive reheating, particularly in the radiant cooling zone cannot be avoided. Further reduction in surface reheating can only be achieved by extension of the sprays with the addition of a fourth zone. The second and third designs, thus, have included the fourth zone and are clearly an improvement. It was also found, at this point, to be advan-tageous to sub-divide zone III into two zones (this was also done for the Table 3.7.1 Secondary Cooling Design for 10.8 cm Square Billets Existing Industrial Practice Design for 3-zone Cooling -First Design Design for 4-zone Cooling -Second Design Modified Design for 4-Zone Cooling - Third Design Zone (Speed 4.02 cm/s) (Speed 5.08 cm/s (Speed 5.08 cm/s) (Speed 5.08 cm/s) Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating I H 1/4 U6515 (1.3 cm from mould) H 1/4 U6515 (7.6 cm from mould) 415 kPa — H 1/4 U6510 (1.3 cm from mould) 1/8 HH 4.8 SQ. (7.6 cm from mould) 415 kPa — H 1/4 U6510 1/8 HH 4.8 SQ. 415 kPa — H 1/4 U6515 H 1/4 U6515 172.5 kPa — II 1/4 GG10 (8.9 cm spacing, 4 nozzles) 138 kPa 90°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 138 kPa 9°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 138 kPa 9°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 9°C III (a 1/4 GG10 (17.8 cm spacing, 8 nozzles) 138 kPa 1/4 GG10 (17.8 cm spacing, 4 nozzles) 138 kPa 83°C 3/8 GG15 (17.8 cm spacing, 3 nozzles) 138 kPa 16°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) '138 kPa IIKb 82°C 1/4 GG10 (17.8 cm spacing, 5 nozzles) 42°C 1/4 GG10 . (17.8 cm spacing, 6 nozzles) 57°C IV — — ~ — — — 1/4 GG10 (17.8 cm. spacing, 7 nozzles) 103.5 kPa 9°C 1/4 GG10 (17.8 cm spacing, 7 nozzles) 103.5 kPa 4°C Radiant Coolinu - — 110°C — — 102°C ~ 89°C — — 90°C Table 3.7.2 Secondary Cooling Design for 15.2 Square B i l l e t s Zone 1 Existing Industrial Practice (Speed 2.75 cm/s) Design for 3-zone Cooling -F i r s t Design (Speed 3.39 cm/s) Design for 4-zone Cooling -Second Design (Speed 3.39 cm/s) Modified De Cooling -(Speed sign for 4-zone Third Design 3.39 cm/s) Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating I H 1/4 U8015 (1.3 cm from mould) H 1/4 U8015 (7.6 cm from mould) 415 kPa — H 1/4 U6510 HH 10 SQ. 415 kPa -H 1/4 6510 1/4 HH 10 SQ. 415 kPa -H 1/4 U8015 H 1/4 U8015 172.5 kPa — II 1/4 GG10 (8.9 cm spacing, 4 nozzles) 276 kPa 19°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 276 kPa 4°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 276 kPa 4°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) | 207 kPa 24-°C III(a) 111(b) 1/4 GG10 (17.8 cm spacing, 8 nozzles) 276 kPa 90°C 3/8 GG15 (17.8 cm spacing, 3 nozzles) 138 kPa 70°C 3/8 GG15 (17.8 cm spacing, 3 nozzles) 207 kPa 54°C 1/4 GG10 (8. 9 cm spacing, 4 nozzles) 1/4 GG10 (17.8 cm spacing, 5 nozzles) 48°C 1/4 GG10 (17.8 cm spacing, 5 nozzles) 39°C 1/4 GG10 (17.8 cm spacing, 6 nozzles) 73°C IV — — — — — 1/4 GG10 (17.8 cm spacing, 7 nozzles) 103.5 kft 23°C 1/4 GG10 (17.8 cm spacing, 7 nozzles) 103.5 kPa 23°C Radiant Cooling — — 170°C — — 104°C - — 81°C — — 83°C 00 Table 3.7.3 Secondary Cooling Design for 20.3 cm Square Billets Zone Existing Industrial Practice (Speed 1.69 cm/s) Design for 3-zone Cooling -First Design (Speed 1.91 cm/s) Design for 4-zone Cooling - Modified Des Second Design Cooling -(Speed 1.91 cm/s) (Speed ign for 4-zone Third Design 1.91 cm/s) Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating Nozzle Type Pressure Reheating I H 1/4 W11015 (1.3 cm from mould) 1/8 VV730770 (2 nozzles 7.6 cm from mould S 5.1 cm from edge of billet) 415 kPa — H 1/4 U6510 1/8 GG6SQ. 415 kPa --1/4 U6510 1/8 GG6SQ. 415 kPa ~ H 1/4 W11015 1/8 W730770 (2 nozzles) 172.5 kPa — II 1/4 GG10 (8.9 cm spacing, 4 nozzles) 303 kPa 19°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 207 kPa 7°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 207 kPa 7°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) > 207 kPa 3°C III(a) IIKb) 1/4 GG10 (17.8 cm spacing, 3 nozzles) 303 kPa 72°C 1/4 GG10 (17.8 cm spacing, 8 nozzles) 207 kPa 55°C 1/4 GG10 (17.8 cm spacing, 8 nozzles) 276 kPa 46°C 1/4 GG10 (8.9 cm spacing, 4 nozzles) 1/4 GG10 (17.8 cm spacing, 6 nozzles) 42°C IV — — ~ — — — 1/4 GG10 (17.8 cm spacing, 7 nozzles) 103.5 kPa 16°C 1/4 GG10 (17.8 cm spacing, 7 nozzles) 103.5 kPi 10°C Radiant Cooling — -- [ 159°C — ~ 133°C . — — 108°C — — 108°C 50 Mid-Face Temperature (°C) 1000 1100 1200 1300 i • 1 1 n r r Figure 3.7.1: Mid-face temperature profiles below the mould for different spray designs (10.8 x 10.8 cm square b i l l e t s ) . 1000 Mid-Face Temperature (°C) 1100 1200 4 0 r -Existing Sprays Design with 3-zone spray cooling(3/8 6GI5 type nozzles in \\\\ zone Ula) \\ Design with 4-zone spray cooling (3/8 GGI5 type nozzles in zone ma) Design with 4-zone spray cooling (I/40G10 nozzles in zone I,m tI7,and same pressures in zone 2 f lU, i.e 207 KPa Radiant Zone 5 0 l Figure 3.7.2: Mid-face temperature profile below the mould for different spray designs (15.2 x 15.2 cm square b i l l e t s ) . 52 Mid-Face Temperature °C 1100 1200 1300 Figure 3.7.3: Mid-face temperature profile below the mould for different spray designs (20.3 x 20.3 cm square b i l l e t ) . . 4 0 h-6 0 r -Shell thickness (cm) 2 4 6 T Figure 3.7.4: Pool profiles for 10.8 cm square b i l l e t s cast under different spray conditions. 54 Shell thickness (cm) 0 2 4 6 8 10 Figure 3.7.5: Pool profiles for 10.8, 15.2 and 21.3 cm square b i l l e t s corresponding to 4-zone spray design. 55 15.2 cm b i l l e t in the f i r s t design). The difference between the latter two designs is essentially in the use of nozzle types. The latter speci-fies 1/4 GG10 nozzles only while the other design also includes 3/8 GG15 nozzles. The two designs give very similar results from the standpoint of reheating, but owing to the uniformity of nozzles in the latter design i t was f i n a l l y adopted for the in-plant t r i a l s . A l l designs are corresponding to higher casting speeds which has resulted into higher pool depths. In a l l cases, the pool depth extends be-yond the sprays and exists even at the pinch r o l l s . This phenomenon, however, i s not important in mid-way cracks study, and is not likely to- pose any problems in b i l l e t casting, except possibly centre cracks. A number of further computer runs were undertaken to check on the effect of variations i n casting speed on surface reheating, and also to calculate the reheating between successive sprays. Figures 3.7.6 to 3.7.8 show the mid-face temperature profile at different casting speeds for the three sizes of the b i l l e t respectively. These clearly indicate that changes in casting speed with constant spray conditions affect the level of the sur-face temperature profile, but leave the magnitude of reheating virtually unchanged. Figure 3.7.9 shows the reheating that occurs between the last nozzle in the Zone III(a) and the f i r s t nozzle in Zone III(b) for the 10.8 cm square b i l l e t cast at 5.08 cm/s. The reheating i s 63°C which, relative to the values computed elsewhere in the strand, i s small and should not lead to cracking problems. At 4.02 cm/s, the reheating rises to 75°C whereas at an increased speed of 6.35 cm/s i t is 60°C. Reheating between the nozzles i s 51°C for 20.3 cm square b i l l e t at 1.91 cm/s. Thus, for a l l speeds that Figure 3.7.6: Effect of speed on mid-face surface temperature profile (10.8 x 10.8 cm square b i l l e t ) . Figure 3.7.7: Effect of speed on mid-face surface temperature profile (15.2 x 15.2 cm square b i l l e t ) . Figure 3.7.8: Effect of speed on mid-face surface temperature profile (20.3 x 20.3 cm square section). Reheating Between The Nozzles Mid-Face Surface Temperature (°C) 1120 1140 M 6 0 1180 1200 I 1 1 1 1 I ! Figure 3.7.9: Reheating between the nozzles. Ln 60 are l i k e l y to be encountered, reheating between nozzles should not be a source of mid-way cracks. 61 Chapter 4 RESULTS AND DISCUSSION OF IN-PLANT TRIALS For the purposes of evaluating the spray design, in-plant t r i a l s were conducted on 10.8 cm square b i l l e t s . By the time the t r i a l s were conducted some changes had been made to the existing practice. The water pressure in Zone I had been reduced to 172.5 kPa from 415 kPa, whereas, the pressure in Zone II and III had been raised to above 276 kPa instead of 138 kPa. The casting speed had been raised to 5.08 cm/s from 4.02 cm/s. With the result, the surface temperature rebound changed to 69° and .. 166°C as the b i l l e t passed successively from Zone II to III, and III to radiation cooling (Figure 4.1). Details of the newly designed spray system involved in the f i r s t campaign are shown in the last column in Table 3.7.1. The major changes to the existing sprays were: - subdivision of Zone III into two sections - addition of fourth zone, thereby lengthening the spray chamber to 3.09 m. With this arrangement, the calculated surface reheating was reduced to 9, 57, 4 and 90°C due to the transition between successive spray zones. This temperature profile and that for the old and the new existing sprays have been plotted on Figure 4.1 for comparison. 4.1 Spray Evaluation - F i r s t Campaign A spray chamber was constructed to the specifications set out in Table 4.1.1, and duly installed on strand II of the multistrand casting 62 Mid-face temperature (°C) 100 1200 300 Figure 4.1: Comparison of mid-face surface temperature p r o f i l e s f o r e x i s t i n g sprays (old), e x i s t i n g sprays (new) and redesigned sprays with 4 zone spray cooling. 63 Table 4.1.1 Modified Spray Design for 10.8 cm B i l l e t s * Zone I Zone II Zone I I i a b Zone III Zone IV Length (cm) 15.2 26.7 35.6 106.7 124.5+ Nozzle Type H 1/4 U6515 1/4 GG10 1/4 GG10 1/4 GG10 1/4 GG10 Number of Nozzles* 1 2 .3 cm from 4 4 6 7 Nozzle Spacing (cm) 7 mould .6 cm from mould 8.9 .8.9 17.8 17.8 Pressure (kPa) 172.5 138 138 138 103.5 Flow (1/s) 0.151 0.353 0.353 0.530 0.574 *Per Face tLength measured to centerline of last nozzle 64 machine. The existing spray system was l e f t unchanged on other strands for comparative purposes. Ten heats with a range of chemistries were then cast into 10.8 cm square b i l l e t s , and transverse sections taken from strands I (unmodified) and II (modified) at the beginning, middle and end of each heat. The transverse sections were machined f l a t and surface ground on both sides, then were sulfur printed. Finally a comparison was made between equivalent sections from strands I and II with respect to the presence of mid-way cracks in the sulfur prints. During some of the test heats the surface temperature below the spray chamber was measured with an optical pyrometer. The chemistry, superheat, casting speed and observations of mid-way cracks are summarized in Table 4.1.2. The cracking severity observed in the sulfur prints i s graded as good (G), good-medium (G/M), medium (M), medium-bad (M/B), and bad (B). Examples of this grading are shown in Figure 4.1.1. 4.1.1 Comparison of Modified to Unmodified Sprays As can be seen in Table 4.1.2, mid-way cracks were present in only 5 of the 10 heats sampled: heats 1, 3, 7, 8 and 10. In these no s i g n i f i -cant improvement in the severity of mid-way cracks was observable with the modified sprays. Thus, some aspect of the design was faulty. The probable culprit is discussed later; but clearly further t r i a l s were necessary. 4.1.2 Effect of Chemistry on Crack Frequency a. Carbon. The relationship between carbon in b i l l e t s and the presence of cracks can be seen in Table 4.1.2.1. Thus, with one minor exception, the heats with significant cracking a l l f a l l in the medium Table 4.1.2 Results of Spray Evaluation - F i r s t Campaign Final Heat Chemistry Beginning of Heat Middle of Heat End of Heat C Mn S P Superheat/ Structure Strand I Strand 11 Superheat/ Structure Strand I Strand II Superheat/ Structure Strand I Strand II General Trend Speed Grade Speed Grade Speed Grade Speed Grade Speed Grade Speed Grade 1 .33 .64 .034 .012 ^ = 18.8 s 17°C Equiaxed 's' p r i n t on the other side 4.66 B 4.87 G }-»• M G 17°C(?) Columnar 4.66 B 4.66 B B M 17°C Equiaxed 4.66 G 4.87 G G G + 2 .52 .81 .030 .016 ' ^ = 2 7 s 33°C Columnar 5.08 G 4.66 G 28°C Columnar 5.16 G 4.87 G 22°C Equiaxed 4.87 G 4.66 G 3 .49 .78 .044 .023 17.7 s 28°C Columnar 's' p r i n t on the other side 4.87 G 4.87 M }-v G M 39°C Columnar 5.08 G 5.08 G G G 22°C Partly Columnar 4.87 G 5.08 G G G/M 4 .60 .88 .035 .023 S i - 25.1 s 44°C Columnar 5.16 G 4.87 G 50°C Columnar 5.16 G 4.66 G 39°C Columnar 5.16 G 4.87 G 5 .08 .54 .036 .003 ^ = 1 5 s ~ 33°C Fine Columnar 5.29 G 5.16 G 42°C Fine Columnar 5.50 G 5.16 G 28°C Fine Columnar 5.50 G 5.42 G (Cont'd) Table 4.1.2 (Continued) Final Heat Chemistry Beginning of Heat Middle of Heat E nd of Heat General Trend C Mn S P Superheat/ Structure Strand I Strand II Superheat/ Strand I Strand II Superheat/ Structure Strand I Strand II Speed Grade Speed Grade Structure Speed Grade Speed Grade Speed Grade Speed Grade 6 .14 .60 .035 .013 £52- - 22.8 s 44°C Fine Columnar 5.16 G 5.50 G 28°C Equiaxed 5.16 G 5.50 G 6°C Equiaxed 5.08 G 5.50 G 7 .33 .63 .043 .019 ^ = 14.7 s 33°C Columnar 's* print on the other side 5.00 M 4.66 B }->- B M/B 28°C Columnar 5.08 M 4.87 B M B 11°C Fine Columnar 5.00 M 5.08 M M \" M 8 .29 .65 .037 .019 552- = 17.6 s 28°C Columnar 's' print on the other side 5.16 M 5.00 B M M 28°C Fine Columnar 5.16 M 5.00 M G/M B 17°C Equiaxed 5.16 G 5.00 G G/M G — 9 .13 .74 .029 .016 ! S 1 - 25.5 s 38°C Fine Columnar 5.16 G 5.29 G 44°C Fine Columnar 5.42 G 5.59 G 33°C Fine Columnar 5.16 G 5.29 G 10 .33 .63 .042 .016 15 s 39°C Columnar 's' print on the other side 5.00 M 4.87 M }->• M G 22°C(?) Equiaxed 5.16 G 5.00 G G G 11°C(?) Equiaxed 1 5.16 M 5.50 B M M + 1 + ON Good Good/Medium Medium Medium/Bad Bad Figure 4.1.1: Grading standard for the crack severity in the sulfur prints. Table 4.1.2.1 Mid-way Cracks and Carbon Carbon Range No. of Heats No. of Heats with Cracks 0.08 - 0.14 3 0 0.29 - 0.33 4 4 0.49 - 0.60 3 1 (only one sample) Table 4.1.2.2 Mid-way Cracks and Mn/S Mn/S Range No. of Heats No. of Heats with Cracks < 20 6 5 > 20 4 0 6 9 carbon range. Normally, such cracks were seen to be associated with a columnar structure. In low and high carbon steels, cracks were absent irrespective of structure. This point i s illustrated in Figure 4.1.2.1. b. Manganese/sulfur. The influence of Mn/S ratio on cracking i s summarized in Table 4.1.2.2. Thus, even with the small sample size the cracks can be correlated to Mn/S. It may, however, be noted that a l l four medium carbon heats had Mn/S < 20, and that, as observed in later campaigns, medium carbon heats with Mn/S >> 20 also give rise to cracking. It, there-fore, appears that, of the two, the carbon content is a more important parameter than the Mn/S ratio. 4.1.3 Effect of Structure on Crack Frequency It i s well known that cast structure has a profound influence on the formation of mid-way cracks, a coarse columnar structure being more suscep-tible to cracking than an equiaxed or fine columnar structure. Thus chemistry effects cannot be viewed in isolation. Indeed the results shown in Table 4.1.2 confirm the earlier experience. In heats where a fine dendritic or equiaxed structure was observed in mid-way regions, cracks were rarely present. Note in this regard that the b i l l e t structure toward the end of a heat was usually equiaxed or fine dendritic owing to the low superheat, and crack severity was reduced correspondingly. Such improve-ments in the fraction of the equiaxed structure and the corresponding re-duction in the cracking tendency can be seen in Figure 4.1.3.1. This shows sulfur prints of the samples from strand II at the beginning of the heat (columnar structure, superheat 28°C, presence of internal cracks), and at the end of the heat (predominantly equiaxed structure, superheat 17°C, no Beginning of casting End of casting columnar structure equiaxed structure Figure 4.1.3.1: Effect of cast structure on internal cracks (Heat no. 8) 72 cracks). The superheat i n Table 4.1.2 were taken from the heat sheets and are approximate. Thus, they may not correspond to the samples collected which may explain why the superheat and the structures do not always match. Normally, the relationship between the structure and superheat i s well defined. Also in heats 1 and 8 some cracks were seen in samples from strand I even though the structure was predominantly equiaxed. This i s because in strand I, where the sprays are shorter, the cracks formed closer to the surface where the structure was fine columnar. 4.1.4 Crack Depth The distance between the inside tip of' the cracks and the b i l l e t surface was measured in the sulfur prints because i t gives an indication of the position of the solidification front at the time of crack formation. The results of these measurements are shown in Table 6 where N, W, S, E denote the orientation of the b i l l e t face. The average distance to the inner crack.tip and the o limits for each heat are shown on the right of Table 4.1.4.1 for strands I and II. These average values, which are taken to be the average shell thickness at the time of crack formation are plotted with + o\" limits against casting speed in Figure 4.1.4.1. The upper set of data corresponds to the modified spray system (strand II) while the lower set was found with the existing sprays (strand I). Thus i t can be seen that the modified spray system, with i t s increased length, cuases the mid-way cracks to form further from the surface than in the un-modified case. In both cases, the position of the cracks, i.e. the shell thickness at the time of crack formation as predicted by the computer model, indicates that the cracks are forming mainly below the spray chamber. The Table 4.1.4.1 Distance From Billet Surface to Innermost Tip of Crack (mm) Heat Beqinninq of Heat Middle of Heat End of Heat Stand I x,0,speed (av.) Stand I I x,a,speed (av.) No. I* I I * I* I I * I* I I * 1 ' N , w C-.33 s E Av. 26,25,24,20 22,23,24 24 25 -26,27 24,23,31 25 27 27,21,22 28 24,26 24,25 35 28,30 30 31 . 32 -24.7,2.39,4.66 31,2.37,4.76 23.8 23.5 24.75 26 31 31 3 N W C.-49 g E - 27,28 29,29,28 31 30 30 27,28 - ; ; - - -7 N W C.-33 s E Av. 19 19 30,21 27,19,19 19 18,20,22 18 25,30 27,30 25 26,26 28,30 23 22 30 24,20 27 29 24.25 27 25.26 30 20 26 26 25 23 21.7,3.80,4.99 26.7,2.11,4.87 21.3 20.4 26.75 28 30 22.25 25.8 28 20 26 24.7 8 N E Av. 20,21,25,22 28,21 20,20 25,20 20,22,24 21 26,35 29 35 31,31 28,33 30 26 20 20,22 21 26 30 25 31 27 26,30 28 21 -21.9,2.37,5.16 29.5,3.04,4.97 21.4 23 30.6 30.2 20.7 23.5 29.25 26.7 21 10 N C.-33 ^ E Av. 21 20,20 28,22 22,19 21,22 28 17,19,23,30 29,31 ---24,27 21,23,31 14 21 22,30 23 24,24 22,23 18,22,21 32,32 20,22 22.8,4.34,5.11 24.5,4.59,5.08 24.75 21.3 30.0 - f - 24 23.5 26.4 21.6 i 1 each billet sample, sulfur prints were taken from both sides of 74 Speed (cm/s) Figure 4.1.4.1: Shell thickness at time of crack formation as a function of casting speed. Upper plot: modified sprays. Lower plot: existing sprays. 75 computer-predicted trends of shell thickness with casting speed is in good agreement with the data. 4.1.5 Surface Temperature Measurements Surface temperatures were measured at three to four locations below the spray chamber during the casting of two heats - a low carbon and a medium carbon. The measurements are compared to computer-prediced mid-face temperature profiles for the low-carbon heat in Figures 4.1.5.1 (strand I) and 4.1.5.2 (strand II), and for the medium-carbon heat in Figures 4.1.5.3 (strand I) and 4.1.5.4 (strand II). Turning f i r s t to the low-carbon heat, the measured temperatures can be seen to be slightly higher than the calcu-lated profile but agreement i s reasonable for both strands I and II. How-ever, in the case of the medium-carbon heat, the measured temperatures are significantly lower than predicted. This would indicate that the medium-carbon b i l l e t was being cooled more in the spray chamber than was intended. 4.1.6 Discussion of F i r s t Campaign This t r i a l was not a success in terms of reduction of the incidence of mid-way cracks in the strand-cast b i l l e t s . However, i t did establish a number of important points such as the effect of structure and steel composi-tion. Earlier i t was mentioned that 0.17 - 0.24% C steels tend to crack due to additional stresses from the 6—>y transformation. In the present cam-paign there were no heats available in this range; however, a l l the four heats cast with 0.29 to 0.33% C range showed cracks. This suggests that the effect of the 6—>y transformation is also extended to carbon levels of at least 0.33%. As expected, heats in the low or high carbon range did not show cracks irrespective of the structure. A single heat cast with 76 Temperature °C 900 1000 1100 1200 1300 Figure 4.1.5.1: Predicted and measured mid-face temperature f o r a low-carbon b i l l e t cast on strand I. Figure 4.1.5.2: Predicted and measured mid-face temperatures for a low-carbon b i l l e t cast on strand II. 78 1000 Temperature (°C) 1100 1 1200 Strand I Tundish temperature 1521 °C Casting temperature 1520 °C Casting speed 4 . 6 6 cm/s Water pressure : 310-5 KPa (Zone 2 a 3) O Observed temperature 3 _ e 22 o E 5 o a) o c o H8 Figure 4.1.5.3: Predicted and measured mid-face temperatures for a medium-carbon b i l l e t cast on strand I. 79 Temperature (°C) Figure 4.1.5.4: P r e d i c t e d and measured mid-face temperatures f o r a medium-carbon b i l l e t c a s t on strand I I . 80 0.60% C also did not crack. The most important finding i s from the crack-depths measurements which clearly indicate that the cracks moved towards the centre as the spray length was increased, and that the cause of cracking was reheating in the radiant cooling zone. It i s also apparent that the b i l l e t s were overcooling in the sprays which would result in higher reheats than predicted. Spray conditions, therefore, needed to be reassessed for the medium carbon heats. It w i l l be recalled that a crucial part of the design is the relationship between spray heat transfer coefficients and spray water flux which must be measured. Thus, the use of Mizikar's data i s called into question. It can be seen in Figure 2.4.1 that in the operating range of water flux (0 to 2 16 1/m -s), the heat transfer coefficients obtained by other investigators are higher than that predicted by Mizikar at 276 kPa. Obviously, the water flux predicted on the basis of Mizikar's data was more than actually needed. Other circumstantial evidence that this data may be inapplicable i s the high specific water flow that obtains in the modified sprays, 1.7 1 water/kg steel. This i s significantly higher than the \"rule-of-thumb\" value of about 1 1/kg which is applied widely throughout the industry. Clearly then, the sprays operating in the b i l l e t caster are more efficient in extracting heat than was measured in a laboratory environment; and in the second campaign the spray water would have to be reduced. Before moving on to the second set of t r i a l s , i t i s interesting to speculate on what caused the discrepancy in spray cooling rates between laboratory and plant. There are at least a couple of differences between the two sets of conditions: i) In the laboratory a single spray was employed to cool a 81 stainless steel plate, whereas, in the caster, a sequence of sprays is used through the secondary cooling chamber. Thus, in the latter case, water from sprays higher up, flows through the lower sprays. i i ) Scale properties on the surface of the stainless steel are different from those on the b i l l e t s in question which are plain-carbon steel. Although both differences could be important, the question of water fa l l i n g through the lower sprays i s especially significant. The possible effects of differences in steel type may also be important since different cooling patterns were observed with medium- and low-carbon heats. One explanation might be differences in scale formation which could contribute to dissimilar cooling rates. Another justification i s the lower rate of heat extraction in the mould in case of low-carbon steels due to formation 32 of ripples on the b i l l e t surface. Overall heat extraction in the mould and the sprays w i l l , thus, be lower, and the b i l l e t coming out of the sprays w i l l be hotter than predicted. 4.2 Spray Evaluation - Second Campaign It was clear then that the second campaign of spray evaluation t r i a l s would have to be performed with a reduced specific water flow. Thus the water flow was decreased from 1.7 to 1.3 1/kg by: i) reducing the water pressure in Zone I to 138 kPa, and in zones II and III to 103.5 kPa. i i ) plugging the third nozzle in Zone I l i a and the second, fourth and sixth nozzles in Zone IV. Details of this second modification are compared to the f i r s t four zone spray systems in Table 4.2.1 and Figure 4.2.1. It should be emphasized 82 Table 4.2.1. Comparison of Second to F i r s t Modification of Spray Chamber Modification I Modification II Zone No. of Nozzles Pressure (kPa) 1/s No. of Nozzles Pressure (kPa) 1/s I 2 172.5 0.151 2 138 0.114 II 4 138 0.353 4 103.5 0.328 a III 4 138 0.353 2 103.5 0.164 I I I b 6 138 0.530 7 103.5 0.574 IV 7 103.5 0.574 4 103.5 0.328 1/s 1.962 x 4 1.533 steel.cast at 5.08 cm/s 4 .617 kg/s 4.617 kg/s 1/kg 1.7 1.3 FT I X - X — m(a) m(b) Modification! IE HI (a) HUb) 121 Modification IT Figure 4.2.1: Arrangement of nozzles i n m o d i f i c a t i o n I and II. 84 that these changes were arrived at by essentially empirical means. It was equally clear from the f i r s t campaign that attention should be focussed on medium carbon heats with high superheats. Accordingly, five medium carbon heats were tested in the second campaign. Again trans-verse sections were cut from b i l l e t s cast on strand I (unmodified) and strand II (modification II) at the beginning, in the middle and at the end of a heat. Following surface grinding and sulfur printing, the sections were compared for mid-way cracks. The results of the tests are summarized in Table 4.2.2 while comparative sulfur prints from one heat are shown in Figure 4.2.2 and 4.2.3. The most important finding to emerge from the sulfur prints i s the reduction in crack severity observed with the modification II sprays in four out of the five heats (Table 4.2.2). Thus the conclusions reached earlier from the f i r s t campaign with respect to excessive cooling appear to be correct. Similarly the effect of structure appear to be consistent with the f i r s t campaign. Cracks were, however, seeen when Mn/S ratio was more than 20. Temperatures measured with a spectray unit, 7 cm below the spray chamber, were again lower than the predicted values by about 80°C, and indicate excessive cooling in sprays. 4.3 Spray Evaluation - Third Campaign In an attempt to improve the performance of the sprays further a third campaign involving medium carbon heats was planned. The third modi-fication reduced the specific water flow from 1.3 to 1.1 1/kg by plugging the f i r s t nozzle of Zone I l i a and the f i f t h and last nozzles of Zone I l l b . Table 4.2.2 Results of Spray Evaluation - Second Campaign Fir.al Heat Chemistry Beginning of Heat Middle of Heat End of Heat General Trend C Mn S P Superheat/ Structure Strand 1 Strand II Superheat/ Structure Strand I Strand II Superheat/ Structure Strand I Strand II Speed Grade Speed Grade Speed Grade Speed Grade Speed Grade Speed Grade 11 .30 .50 .029 .014 17.2 s 39°C Columnar 5.72 M 5.50 G 39°C Columnar 5.72 M 5.50 M 39°C Columnar 5.72 G/M 5.50 G/M . + (Improved) 12 .32 .65 .040 .020 ^2.= 16.3 s 22°C Equiaxed Fine Columnar 6.14/ G/M 6.14/ G 6.35 6.35 28°C Equiaxed (II ) / Columnar (I) 6.35/ M 6.35/ G. 6.56 6.56 6-10°C Equiaxed 6.35/ G 6.35/ G 6.56 6.56 + 13 .32 .63 .027 .018 ^ 1 = 25.2 s 44°C Columnar 5.50 M/B 5.29/ M 5.50 39°C Columnar 5.50 M 5.29/ G/M 5.50 33°C Columnar 5.72 G/M 5.72/ G/M 5.93 + 14 .32 .66 .029 .019 Mr. — = 22.8 s 50°C Fine Columnar 5.72 M 5.50 G/M 36°C/ 50°C Columnar 5.72/ G/M 5.50 G 5.93 15 .32 .65 .038 .023 ^ - 1 7 . 1 s 39°C Columnar 5.50/ B 5.50/ B 5/72 5.72 33°C Columnar 5.50/ B 5.50/ B 5.72 5.72 No Change 00 Existing Sprays Modified Spray Design Figure 4.2.2: Comparison of sulfur prints for heat no. 14 (beginning of heat) Existing Sprays Modified Spray Design Figure 4.2.3: Comparison of sulfur prints for heat no. 14 (middle of heat). oo 88 Results obtained from this t r i a l were not consistent with the earlier t r i a l s . However, upon investigating the spray-risers, i t was found that owing to an in-plant error the f i f t h nozzle on one side of the sprays was missing and two nozzles were blocked. These results, therefore, have not been included in the present study. It may be noted that so far in this study, the pressure drop in the spray pipe has not been considered, i.e. a l l the sprays were assumed to be operating at the same base pressure. Since the b i l l e t caster under study had a f a i r l y long single spray-pipe for Zone II and III, and the pressure gauge was about 4 to 5 feet below the spray pipe, the pressure loss due to gravity could be an important parameter. With these questions in mind, the spray pipes for Zone II and III, and for Zone IV were obtained for further study. 4.4 Pressure Drop The pressure drop in the pipe was estimated theoretically, and then compared to the experimental observations. 4.4.1 Frictional Losses The spray pipe under consideration had 1.1 i n . (2.8 cm) internal dia. and 56 i n . (142 cm) length. Taking a flow rate of 0.82 1/s, the f r i c t i o n a l loss would be approximately 12.0 feet/100 feet. In the present calculation, for the entire spray pipe length, this loss amounts to 1 2 f t X — 5 6 i n ' X 0.43 p s i / f t = 0.24 psi = 1.66 kPa 100 f t 12 in.'/ft This i s an overestimate since the drop in flow rate along the length 89 of the pipe has not been considered. For operating pressures of around 103.5 kPa, the fricti o n a l loss of 1.66 kPa can be considered as negligible. However, in the upper region of the sprays the f r i c t i o n a l losses could be high due to obstructions from the clustering of nozzles. 4.4.2 Pressure Loss Due to Gravity Head Assuming a spray system as shown in Figure 4.4.2.1, and using Bernoulli's equation we have A „. , A A 2 P + - = P' + — — + H + P^ . . (Eq. 4.4.2.1) w f r i c t i o n 2g 2g (Point a) (Point b) Pressure at point c = P' + — s + H + P 2g w f r i c t i o n Since 6 = 6 ' + q_ , we have W W T \\ 7 P' = P - H - P,, . . (Eq. 4.4.2.2) w f r i c t i o n P' i s the pressure recorded by the pressure gauge associated with the spray nozzle. Neglecting losses due to f r i c t i o n (as discussed in Section 4.4.1), the static pressure recorded at the nozzle would indicate the pressure loss due to the gravity head only, i.e. this loss would be independent of the flow rate (at reasonably low flow rates) and the starting pressure. 4.4.3 Laboratory Spray Trials In order to check the flow characteristics, a few spray t r i a l s were conducted. In the f i r s t t r i a l , the upper spray pipe (Zone II and III, 90 T H 1 Q' w P' (Pressure recorded at the nozzle) 9. spray [ — Q P (Base pressure) I Q w Figure 4.4.2.1: Schematic representation of water flow rate and pressures at different locations in the spray pipe. 91 Figure 4.4.3.1) was fed at 82.8 kPa from the bottom. The pipe was then inverted and fed from the top at 48.3 kPa. Static pressures were measured at each nozzle during both t r i a l s and also calculated using Eq. 4.4.2.2. P1 (in kPa) = P (in kPa) — 0.0981 x H (in cm) w The results are given in Table 4.4.3.1. It may be noted that, in general, there is a good agreement between the calculated and observed pressures except in Zone II, where fri c t i o n a l losses could be significant.. In another t r i a l , a l l nozzles were plugged except nozzle no. 2, 4 and 8 and several pressure conditions between 69 kPa to 345 kPa were observed to see whether the pressure difference between the base pressure gauge and at 10th nozzle was in the v i c i n i t y of 17 to 21 kPa. Within the accuracy of the pressure gauges used, the readings confirmed the difference. Since the pressure drop due to gravity alone is expected to be 16.6 kPa, this t r i a l confirms that for a l l pressures in the practical range, the pressure drop due to f r i c t i o n i s negligible. 4.4.4 Discussion With the help of these observations, the flow rates and water flux in the f i r s t two campaigns were corrected for the pressure drop and are presented in Table 4.4.4.1. An additional drop of 3.5 kPa due to f r i c t i o n was incorporated in Zone II. The pressure drop in the spray pipe gives rise to an undesirable situation, because the pressure i n the upper region of the pipe i s lower than that at the bottom. It was, therefore, necessary to redistribute the water. Thus, for the fourth and last t r i a l , the base 92 8;-9cm 10 9 8 7 T T n L.4 rn(o) T I78cm -L . T 35-6cm IE(b) Base -+--Figure 4.4.3.1: Arrangement of nozzles in the upper spray pipe for laboratory spray t r i a l s . 93 Table 4.4.3.1 Observations on Spray Trials Spray Pipe in Normal Position Spray Inverted Pipe in Position Nozzle No. Distance from the Base Nozzle (cm) Observed Pressure kPa Calculated Pressure kPa Observed Pressure kPa Calculated Pressure kPa Base 0 82.8 82.8 48.3 48.3 1 17.8 77.6 81.0 50.0 50.0 2 53.3 75.2 77.6 53.1 53.5 3 71.1 73.1 75.8 54.5 55.3 4 88.9 — 74.1 — 57.0 5 106.7 70.4 72.3 57.3 58.8 6 124.5 69.0 70.6 58.7/59.3 60.3 7 142.2 68.0 68.9 60.0 62.2 8 151.1 63.8 68.0 60.0 63.1 9 160.0 62.1 67.1 62.3 64.0 10 168.9 60.7 66.2 62.3 64.9 Table 4.4.4.1 Corrected Pressures and Water Flow Rates for First and Second Modifications Modification I v. Modification II Zone Type of Nozzles No. of Nozzles Base Pressure kPa Anticipated Pressure at the Nozzle kPa 1/s No. of Nozzles Base Pressure kPa Anticipated Pressure at the Nozzle kPa 1/s I H 1/4 U6515 2 172.5 131 0.139 2 138 97 0.114 II 1/4 GG10 4 138 93-100 0.316 4 103.5 59-65 0.240 III 1/4 GG10 10 138 100-107 (f i r s t 5 nozzles) 107-114 (remaining 5 nozzles) 0.410 0.416 9 103.5 65-72 (firs t 4 nozzles) 72-79 (remaining 5 nozzles) 0.252 0.347 IV 1/4 GG10 7 103.5 83-86 (f i r s t 4 nozzles) 90-93 (remaining 3 nozzles) 0.303 0.227 4 103.5 83-86 (first 2.nozzles) 90-93 (remaining 2 nozzles) 0.151 0.151 1/s 1 .811 x 4 1.255 x 4 steel cast at 5.08 cm/s 4 .617 kg/s 4.617 kg/s 1/kg 1.57 1.09 U3 95 pressure in Zone II and III was restored to 138 kPa. In addition, as in the third t r i a l , three nozzles were plugged in Zone III. To account for the higher water pressure in the bottom of the sprays, the last two nozzles in Zone IV were replaced by small capacity nozzles, 1/4 GG6.5. This arrangement gives rise to a reduced specific water flux in the last zone which should reduce reheat when the strand passes from the sprays to the radiation cooling zone. This arrangement is shown in Table 4.4.4.2 and Figure 4.4.4.1. 4.5 Spray Evaluation - Fourth Campaign Eleven heats were cast in this campaign and samples at the beginning of the cast (after 10 minutes of casting) were taken for sulfur printing. These samples were expected to correspond to a high superheat normally observed at the beginning of the cast in which columnar or fine columnar structure and cracks would normally show up. In this t r i a l , samples from the third strand, which also has the existing sprays, were studied. The results have been tabulated in Table 4.5.1. As can be seen, out of the eleven heats cast, eight show improvement by application of the modified sprays (strand II). In two cases, the cracks were completely removed. Figures 4.5.1 to 4.5.3 show examples of the reduction in cracking tendency. In this t r i a l , cracks were observed even when the Mn/S ratio exceeded 20, and also in the case when carbon was 0.37%. It, therefore, appears that the adverse effect of the 6—>y transformation i s extended up to 0.37% C steels, and that in order to reduce the cracking tendency by controlling sulfur, one may have to increase Mn/S ratio or achieve very low levels of Table 4.4.4.2 Spray Arrangement and Water Flow Rates Modification IV 96 Zone Base Anticipated Pressure Type of No. of Pressure at the Nozzle Nozzles Nozzles kPa kPa 1/s H 1/4 U6515 138 97 0.114 II III 1/4 GG10 1/4 GG10 138 138 93-100 100-107 (f i r s t 3 nozzles) 0.316 0.246 107-114 0.252 (remaining 3 nozzles) IV 1/4 GG10 1/4 GG6.5 103.5 103.5 83-86 90-93 0.151 0.095 1/s 1.174 x 4 steel cast at 5.08 cm/s 4.617 kg/s 1/kg 1.02 97 rr » K K X - X -HI (a) nr(b) Modification! • — 9 n UKa) HL(b) 121 Modification n K K -9 — 9 i n (a) TH(b) 32: I — x — Modification JUa iZ F i g u r e 4.4.4.1: Schematic r e p r e s e n t a t i o n o f s p r a y m o d i f i c a t i o n s I> I I , and I I I . Table 4.5.1 Results of Spray Evaluation - Fourth Campaign F i n a l Chemistry. Superheat/ I II III General Heat C Mn S P Mn/s Structure Speed Grade Speed Grade Speed Grade Trend 16 .32 .68 .032 .014 21.3 25°C Fine Columnar 125/130 M 125/130 G 125/130 M f 17 .30 .70 .030 .012 25.0 42°C Fine Columnar 125/130 M 125/130 M 125/130 B t 18 .31 .67 .040 .021 16.8 25°C Columnar 125/130 B 120 G/M 120/125 M/B: T 19 .32 .68 .042 .027 16.2 19°C Columnar 120/125 B 115/120 G/M 125/130 G/M + 20 .37 .70 .032 .012 21.9 16°C Fine.Columnar 120/125 B 120/125 G — t 21 .34 .74 .028 .017 12.4 28°C Fine Columnar 125/130 G/M 120/125 M 125/130 G/M + 22 .30 .70 .030 .012 23.3 53°C Fine Columnar 125/130 G/M 125/130 G/M 125/130 G/M 23 .29 .68 .039 .018 17.4 58°C Columnar 125/130 B 125/130 M 125/130 B T 24 .30 .69 .047 .020 13.0 44°C Columnar 125 B 115/120 M 120/125 B i 25 .32 .50 .024 .011 20.8 36° Fine Columnar 120/130 M 120/125 G/M 125/130 M 26 .34 .64 .021 .011 30.5 47°C Fine Columnar 125/130 M 125/130 G/M 125/130 — Strand I Existing Sprays Strand II Modified Sprays Strand III Existing Sprays Figure 4.5.1: Comparison of sulfur prints for heat no. 16. 10 Strand I Strand II Strand III Existing Sprays Modified Sprays Existing Sprays Figure 4.5.2: Comparison of sulfur prints for heat no. S t r a n d I S t r a n d I I S t r a n d I I I E x i s t i n g S p r a y s M o d i f i e d S p r a y s E x i s t i n g S p r a y s Figure 4.5.3: Comparison of sulfur prints for heat no. 24. sulfur. This t r i a l has clearly demonstrated the correctness of the spray design, although for complete removal of cracks, a larger spray chamber may be needed. 103 Chapter 5 GENERAL DISCUSSION AND SUMMARY This study has been successfully completed in terms of reducing the incidence of mid-way cracks in continuously cast steel b i l l e t s . As i s evident from the outcome of the second and fourth t r i a l s , about 75% of the heats have shown improvements in this quality parameter. However, a problem did arise with the relationship between spray heat-transfer coeffi-cient and the water flux. The heat transfer taken from Mizikar's study, which are the basis of the spray design, appear to be too low, and, there-fore, lead to an overestimate of the water flux. It i s , therefore, not surprising that the fi n a l results had to be obtained by empirical adjust-ment of the sprays. New measurements of spray heat transfer coefficients have been recently reported in the literature, and a design based on them may overcome this problem. The design method described in Chapter III i s the f i r s t attempt of i t s kind and has proven to be satisfactory in the present work. However, another and possibly better approach would be to divide the b i l l e t surface into vertical strips and to obtain the heat transfer coefficient profile as a function of distance below the mould for each strip. Sprays would, then, be set up to match these profiles as closely as possible. For such a design, one of the restrictions imposed in the present work, i.e. distance between the two nozzles and the distance of the nozzles from the b i l l e t can be relaxed. Spray-box construction for such a design is not l i k e l y to pose any problem; however, one should restr i c t oneself to choosing 104 a single type of nozzle since in day-to-day operation, mix-ups may occur. The new technique i s schematically shown in Figure 5.1. The steps in this method are as follows: a) Divide the b i l l e t surface into a number of strips (which w i l l be 'nodes' in the computer program); b) Obtain the desired heat transfer coefficient profile (computer-predicted) for these strips which would give acceptable reheating below the sprays (in the radiant cooling zone). c) Construct a spray cooling pattern as shown in Figure 4.1a, average the water flux between successive nozzles in the strip under con-sideration and find the heat transfer coefficient for i t . d) Plot the values of heat transfer coefficients as steps (Figure 4.1c) and draw a smooth curve through them. e) The sprays are then adjusted such that the heat transfer co e f f i -cient profiles for sprays for different strips are as close to the desired profiles as possible, especially for the strips at and near the mid-face of the b i l l e t . Such adjustments in sprays are possible by varying nozzle-to-nozzle, and nozzle-to-billet distances. More than one spray pipe may be used, in which case, advantage of varying pressures can also be ut i l i z e d . It might be more meaningful to use the correlation between the heat transfer 23 26 coefficient and the water flux given by Nozaki et al.. or by Sasaki et a l . f) Obtain a temperature profile for these sprays and compare i t with the ideal one. In this system, water may be fed from the top. This design has two distinct advantages: i) reheating within the sprays can be avoided, and i i ) adjustment of water flux, based on practical Heat Transfer Coefficient, h s Temperature °C (a) (b) (c) (d) Figure 5.1: Schematic representation of the new spray design technique. o Ul 106 observations, can be made by altering pressures rather than plugging nozzles. SUMMARY: This work has been concerned with the reduction of mid-way crack formation caused by surface reheat in cast steel through the redesign of sprays on an operating b i l l e t caster. The study has resulted in improve-ments in mid-way cracks in over 80% of the heats cast with redesigned sprays. The basic principles of the spray design were i) to maintain surface temperature rebound of the strand below 100°C, and i i ) to achieve a mid-face temperature of the strand through the sprays of about 1150°C. In this way the sprays reduced the tensile strain at the solidification front caused by the surface temperature rebound, and consequently, reduced the cracking tendency. The spray design consisted of the following steps. F i r s t , heat transfer coefficient distribution was obtained to maintain the mid-face strand temperature in the sprays at 1150°C. Second, heat transfer coeffi-cients were averaged along the length of the strand in the respective spray cooling zones, and the spray water distribution coreesponding to them was obtained. Third, various combinations of nozzle arrangements and water pressures were studied to achieve an optimum set. At the time of this 21 study, Mizikar's data on spray heat transfer coefficient and water flux was found to be most suitable; hence i t was used in the calculations. Four industrial campaigns were conducted with existing sprays on one strand and the modified sprays on the other. Empirical adjustments were made to the modified sprays in the latter t r i a l s since the strand was found to be overcooled in the sprays. Apparently, this arose due to an over-estimate of the water flux from Mizikar's data. 107 As mentioned earlier, the main aim of the project, i.e. reduction of mid-way cracks through modified spray design, has been achieved in over 80% of the heats cast in the second and fourth campaigns. In addition, the present work has also shown that the mid-way cracks were most effec-tively controlled by having a predominantly equiaxed cast structure, and that, of the three carbon ranges studied (0.08 - 0.14, 0.29 - 0.37 and 0.49 -0.60), mid-way cracks were observed only in the 0.29 - 0.37% range. Mn/S ratios of less than 30 had l i t t l e effect on cracking tendency. Finally, a new method for designing the sprays has been proposed which may give a better temperature profile and could be easier to implement in plant. 108 References 1. W.T. Lankford, Met. Trans., vol. 3, June 1972, 1331. 2. E.B. Hawbolt, F. Weinberg, J.K. Brimacombe, Met. Trans. B., vol. 10B, June 1979, 229. 3. K. Ushijima, Trans. ISIJ, vol. 15, no. 7, 1975, 380. 4. J. Chipman and J.F. E l l i o t , Electric Furnace Steelmaking, vol. II, AIME Publication, 1963, 99. 5. C.J. Adams, Proc. Open Hearth Conference, Pittsburgh, vol. 54, 1971, 290. 6. J.K. Brimacombe and K. Sorimachi, Met. Trans. B., vol. 8B, 1977, 489. 7. H. Vom Ende and G. Vogt, JISI, vol. 210, December 1972, 889. 8. T. Ohashi et a l . , Trans. ISIJ, vol. 15, no. 11, 1975, 571. 9. J.K. Brimacombe, Can. Met. Quart., vol. 15, 1976, 163. 10. T. Kawawa et a l . , Tetsu-to-Hagane, vol. 60, no. 5, 1974, 486 (HB Translation No. 9292) . 11. A. Suzuki, Tetsu-to-Hagane, vol. 60, no. 7, 1974, 36 (HB Translation No. 9354). 12. A. G r i l l and J.K. Brimacombe, Ironmaking Steelmaking, vol. 3, no. 2, 1976, 76. 13. K. Sorimachi and J.K. Brimacombe, Ironmaking and Steelmaking, vol. 4, no. 4, 1977, 240. 14. P.J. Wray, Met. Trans. A, vol. 7A, November 1976, 1621. 15. P. Nilles et a l . , Proceedings of the NOHjBOS Conference, Chicago, vol. 61, 1978, 399. 16. F. Weinberg, Met. Trans. B, vol. 10B, June 1979, 219. 17. L. Baptista, M.A.Sc. Thesis, Dept. of Met. Eng., Univ. of B.C., 1979. 109 18. CO. Pedersen, Int. J. Heat Mass Transfer, vol. 13, 1970, 369. 19. H. Miiller and R. Jeschar, Arch. Eisenhuttenwes., vol. 44, no. 8, 1973, 589. 20. N. Mitsutsuka, Tetsu-to-Hagane, vol. 54, 1968, 1457. 21. E.A. Mizikar, Iron Steel Eng., vol. 47, June 1970, 53. 22. M Shimada and M. Mitsutsuka, Tetsu-to-Hagane, vol. 52, 1966, 1643. 23. T. Nozaki et a l . , Trans. ISIJ, vol. 18, 1978, 330. 24. M. Ishiguro et a l . , Tetsu-to-Hagane, vol. 60, no. 11, 1974, 464 (HB translation no. 8735). 25. T. Kawakazu et a l . , Tetsu-to-Hagane, vol. 60, no. 4, 1974, 103. 26. K. Sasaki et a l . , Tetsu-to-Hagane, vol. 65, 1978, 90. 27. L. Bolle and J.C. Moureau, International Heat and Mass Transfer seminar, Dubrovhik, Yugoslavia, 1979. 28. R. Alberni et a l . , Circulare d'Informatio-techniques, vol. 3, 1973, 763 (BISI translation no. 11633). 29. B. Carnahan et a l . , \"Applied Numerical Methods,\" 1969, John Wiley & Sons, Inc., New York. 30. J.E. Lait et a l . , Ironmaking Steelmaking, vol. 1, 1974, 35. 31. L.I. Morozenskii et a l . , Stal', vol. 4, 1965, 272. 32. S.N. Singh and K.E. Blazek, Proceedings of the NOH-BOS Conference, Atlantic City, vol. 57, 1974, 16. 110 APPENDIX I NUMERICAL SOLUTION OF THE HEAT CONDUCTION EQUATION The node arrangement in one quarter of a transverse cast section i s as shown in Figure A.l. The enthalpy and temperature of a l l nodes are i n i t i a l l y assigned values corresponding to the enthalpy and temperature of the incoming steel. A linear relationship between the temperature-dependent thermal conductivity, k. ., and the temperature, T. ., has been assumed as i 1 3 i , 3 given by i/3 D T. . + E 1 / J (Eq. A.l) The enthalpies of the nodes are then recalculated over the next time step, At, using an explicit form of the finite-difference approximation of Eq. 2.5.1. Interior Node: H' ( i , j ) = H ( i , j ) + At i/3 (T. . - 2T. . + T. . ,) 1,3+1 1,3 1,3-1 (Ay) (T. n . - 2T. . + T. .) 1+1,3 1,3 1-1/3 D 4 (T ( A x ) ' (T. . 2 - 2T. T,. . + T. . 2) i/3+l 1,3+1 i , : - l i , ] - l (Ay) 2 2 — 2T T + T 2 } ^ i+l,j X i + l , j V l , j + X i - l , j } (Ax) (Eq. A.2) (1,1) • • 1 • 6 ( i > • • • • 1 < (i.l) • • • • • • • < 1 > • • • • < > I • — —A A- 1 y = Y / 2 x=X/2 Figure A.L: Arrangement of nodes in one-quarter of a transverse section. 112 Surface Nodes for the Face x = 0: H' ( l , j : H(l,j) + At ( T l , j + l - 2 T l , . j + T l f i - l > (Ay) 2 + — D ( T l , j + l 2 - 2 T l , 1 + l T l , ^ l + T l , 1 - 1 2 ) (Ay) + 2 (Ax) 7 { T„ . 2 2,j qQAx T . } 1/D (Eq. A.3) Surface Nodes for the Face y = 0: H' ( i , l ) H (i,l) + At i , l (T. — \"2T. . + T. , ,) - 1+1,1- - 1,1 1-1,1 (Ax) 2 2 2 (T — 2T T + T 1 D i + l , l z x i + l , i i-1-,1 J 4 2 (Axp + 2 (Ay) 2 1,2 - T. . } 1,1 (Eq. A.4) Corner Node: H' (1,1) H (1,1) + 2At - { T 2 1,2 (Ax) (Ay) - { T 2 1 2,1 T } 1,1 qQAx T l , l } (Eq. A.5) In these equations, k^ and are the average thermal conductivi-ties between T. and T. ., and between T. . and T. , respectively. It i/2 l,j i,2 i , l may be noted that the equation for the centerline nodes i s the same as 113 Equation A.2 except that T. . . i s set equal to T. , ., for the x = — 1+3,3 1-1,3 2 Y plane and T. i s set equal to T. for the y = — plane. 1,3 +1 i , 3 l ^ From the knowledge of enthalpy-temperature relationship, the newly calculated enthalpies H'(i,j) can be converted to temperatures, and the calculation can be repeated over successive time intervals. Since the temperatures are fixed at any particular level (steady-state condition), the three dimensional temperature f i e l d i s obtained by this procedure. The shell thickness i s then obtained by the locus of nodes at the solidus temperature. When the surface temperature condition i s specified (as in Section 3.4), the surface heat flux can be obtained by rearranging Equations A.3 to A.5 as follows: Surface Nodes for the Face x = 0: q 0 d,j) Ax 2 H(l,j) - H'(l,j) At k . (T . — 2Tn . + Tn . n ) 1,3 1,3+1 1,3 1,3-1 (Ay) 2 T 2 — 2T T +T 2 + P- I 1,3+1 l,j+l 1,3-1 1,1-1 4 * 2 (Ay) 2 2k, (Ax' ~-r { T . - T . } 2 2,3 1,3 (Eq. A.6) 114 Surface Node for the Face y = 0: H(i ,1) - H'(1,1) q Q (1,1) 2 At k ' ! ( T - j . i i ~~ 2 T - i + T - n i ) 1,1 1+1,1 1,1 1-1,1 ( A x ) 2 Corner Node: 2 2 T — 2T T + T + E / i+1,1 i+1,1 i-1,1 i-1,1 -, 4 2 \" (Ax)^ + 2 (Ay) T (T. - T. ) 2 1,2 1,1 (Eq. A.7) q Q (1,D Ax Ay 2 (Ax + Ay) 2 (Ax)' H(l,l) - H'(1,1) At ( T 2 , 1 - T 1 , 1 ) + (Ay) — (T 2 V 1,2 — T 1,1' (Eq. A.8) The spray heat flux values can be converted to heat transfer coefficient profiles by dividing q^ by the surface temperature driving force at each nodal point. "@en ; edm:hasType "Thesis/Dissertation"@en ; edm:isShownAt "10.14288/1.0078972"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Materials Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Case study of spray design for a continuous billet caster"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/22243"@en .