Formation and Prevention of Nozzle Clogging during the Continuous Casting of Steels: A Review

Abstract

Nozzle clogging is a common phenomenon in the continuous casting of steels, especially for aluminum killed steels, and has detrimental effects on the continuity of casting and the steel cleanliness. The formation and prevention of nozzle clogging during the continuous casting of steels was reviewed in the current study. The characterization of nozzle clogging was firstly summarized, finding that the composition, morphology, and structure of the nozzle clogging significantly depended on the steel grade, casting condition, and clogging degree. Then three factors affecting the formation of nozzle clogging was briefly summarized as the physical adhesion of high-melting-point inclusions, the temperature drop resulted from the insufficient preheating of nozzles and the temperature fluctuation of molten steel, and the chemical reactions occur during the continuous casting. The prevention methods of nozzle clogging could be roughly divided into 5 types, including the cleanliness improvement of steel, the liquid modification of inclusions, the fluid flow control inside the nozzle, the coating treatment on the inner surface of nozzles, and the external field treatment. Moreover, the challenge of the nozzle clogging control in the future was proposed. Multiple methods should be coordinated, especially the external field treatment should be developed to achieve a better inhibitory effect on the nozzle clogging.

1. Introduction

Owing to the higher production efficiency, higher iron yield and lower energy consumption, continuous casting (CC) process has played a leading role in the world steel production. Continuous casters could be classified into slab, bloom and billet casters according to the cross sectional dimension of products. The refined molten steel in the ladle is poured into the tundish through a shroud nozzle, and was then casted into the mold through a submerged entry nozzle (SEN), finally, was solidified to slabs after the primary and secondary cooling. The flow rate of the molten steel from the ladle to the tundish is controlled by slide gate and that from the tundish to the mold is usually controlled by stopper or slide gate. There are several nozzles in the CC system, including the ladle nozzle, the shroud nozzle, the tundish nozzle, and the SEN.¹⁾ The continuity of casting depends largely on the condition of nozzles. However, clogging would be formed inside the aforementioned nozzles. And according to statistics, clogging is more likely to form in the SEN in most industrial CC productions, which is detrimental to the castability and steel quality.

The detailed detrimental effect of nozzle clogging has been investigated by many researchers and is briefly summarized and illustrated in Fig. 1. Due to the characteristics of high temperature and non perspective and the development of computer technology, most of the mold related research was carried out by numerical simulation. It was proposed by Bai et al.²⁾ that the initial clogging might enhance the steel flow rate but great clogging would restrict the flow channel in the nozzle. The SEN clogging could also affect the flow pattern in the mold. It was reported through numerical simulations^3,4,5,6) that the nozzle clogging would break the symmetry of the flow field in the mold by increasing the jet stream velocity magnitude on the side with weaker clogging, leading to the higher velocity and level fluctuation at the steel-slag interface as well as the increment of the occurrence frequency, top-diameter, rotational speed, and penetration depth of vortices. The asymmetrical fluid flow pattern in the mold caused by the SEN one-sided clogging could also generate an asymmetrical temperature distribution in the mold and increase the risk of breakouts.⁴⁾ Similar conclusions were obtained through water model experiments,^3,7,8) the mercury model experiment,⁹⁾ and the industrial experiment.¹⁰⁾ Moreover, the nozzle clogging was detrimental to the steel cleanliness by promoting the entrainment of mold flux. Zhang et al.⁴⁾ simulated the flow transport and inclusion motion in a steel CC mold under SEN clogging condition, and proposed that more >200 μm inclusions could be entrained into the molten steel with SEN clogging. Zhou et al.¹⁰⁾ measured the clogging-induced asymmetrical and transient flow pattern in a continuous casting slab strand using nail boards, and found that the percentage of >3 mm level fluctuation increased from 4.7% to 10.6% with a severe clogging of the SEN, meanwhile, the amount of entrained slag-type inclusions with severe clogging was approximately 1.6 times than that with slight clogging. It was demonstrated by Li et al. through the physical simulation by mercury model experiment that the EMS could improve the symmetry of the flow field when the SEN clogging rate was under 50%,⁹⁾ however, a strong EMS would deteriorate the negative segregation in the subsurface and positive segregation in the columnar-to-equiaxed transition (CET) zone of blooms according to the work of Wang et al.^11,12) Moreover, the fragmentation of clogs in the SEN is one of the sources of large-sized inclusions in steels, especially in ULC steels and bearing steels.

Fig. 1. Illustration of detrimental effects of nozzle clogging in the mold. (Online version in color.)

It is clearly indicated that the nozzle clogging is seriously harmful to the continuity of casting production and slab quality. Thus, it is necessary to have a clear understanding on its formation and prevention. The causes, effects, and solutions to clogging in CC nozzles have been briefly reviewed by Rackers in 1995.^13,14) However, after more than twenty years of development, the cleanliness of molten steel has been greatly improved, many new technologies have also been developed. Therefore, it is necessary to re-sort out the formation mechanism and control methods of nozzle clogging, so as to better provide ideas for reducing the occurrence of nozzle clogging and further improve the stability of CC production and the quality of CC products. Michelic et al. has recently reviewed the significance of non-metallic inclusions for the clogging phenomenon in the CC of steel.¹⁵⁾ While currently, the formation and prevention of nozzle clogging during the CC of steel will be more comprehensively reviewed.

2. Characterization of Nozzle Clogging

2.1. Morphology and Structure of Nozzle Clogging

The clogs at the CC nozzle are mainly the aggregates of deoxidation products, solidified steel and complex oxides, and may appear thick, thin, loose, dense, white or colored according to the composition and degree of clogging. Figure 2 shows the macro morphology of a SEN with severe clogging after the casting of a Ti stablized interstitial free (IF) steel¹⁶⁾ and a SEN with slight clogging degree after the casting of a non-quenched and tempered steel.¹⁷⁾ The steel grade has a large influence on nozzle clogging. Figure 3 shows some typical micro surface morphologies of clogs consisted of different particles formed in the continuous casting of different steels, including the Al₂O₃ type from an IF steel,¹⁶⁾ the MgO–Al₂O₃ type from a bearing steel,¹⁸⁾ the CaO–MgO–Al₂O₃ type from a bearing steel,¹⁷⁾ the MgO–Al₂O₃–SiO₂–CaO type from a non-aluminum killed 40Cr steel,¹⁹⁾ and the TiN type from a titanium treated stainless steel.²⁰⁾ In general, the thick and loose SEN clogging were mainly consisted of Al₂O₃ type inclusions and usually existed during the casting of Al-killed steels. On the contrary, the relative thin and dense clogging, which was mainly consisted of silicates or aluminosilicates, might be formed in the casting of non-Al killed steels, which means that non-Al killed steels cause less clogging than Al killed steels. It was also found by Vermeulen et al.²¹⁾ that Al killed high Mn (> 3000 ppm) high C (> 500 ppm) steels would have less clogging than Al killed steels with low Mn content (< 3000 ppm). Generally, steels with a high cleanliness will also have less SEN clogging.

Fig. 2. Typical macro morphologies of SENs with different clogging degree, (a) severe clogging of a Ti-IF steel,¹⁶⁾ (b) slight clogging of a non-quenched and tempered steel.¹⁷⁾ (Online version in color.)

Fig. 3. Typical surface morphology of clogs consisted of different particles, (a) Al₂O₃ from an IF steel,¹⁶⁾ (b) MgO–Al₂O₃ from a bearing steel,¹⁸⁾ (c) CaO–MgO–Al₂O₃ from a bearing steel,¹⁷⁾ (d) MgO–Al₂O₃–SiO₂–CaO from a non-aluminum killed 40Cr steel,¹⁹⁾ and (e) TiN from a titanium treated stainless steel.²⁰⁾ (Online version in color.)

Besides of the surface morphology, the cross-sectional layer structure of clogging should usually be further analyzed so as to reveal the formation sequence and mechanism of nozzle clogging. The optical microscope (OM),²²⁾ scanning electron microscope equipped with an energy dispersive spectrometer (SEM-EDS),²³⁾ and cathodo luminescence microscope (CLM)²²⁾ are usually employed for the cross-sectional detection after the grinding and polishing of samples. The clogging structure was typically consisted of three layers,²³⁾ such as the clogging layer, the penetration layer or solidified steel layer, and the decarburization layer from the molten steel towards the nozzle matrix. Due to the difference of steel grade and the variety of formation mechnism, the main composition of each layer is sometimes different and the clogging may exceed three layers. An example of the five-layer clogging structure revealed using SEM and CLM is shown in Fig. 4.¹⁹⁾ The multi-layer of solidified steel was proposed to be caused by the insufficient preheating of nozzle and the temperature fluctuation of molten steel, while the spinel phase in the surface loose layer was estimated to be owing to the phase transformation of clogs and the precipitation of spinel during the cooling of the surface clogging layer.¹⁹⁾

Fig. 4. Five-layer clogging structure of a SEN after the casting of 40Cr steel revealed by (a) SEM with backscattering mode and (b) CLM.¹⁹⁾ (Online version in color.)

2.2. Quantification of Nozzle Clogging

Precise quantification is the basis of the control of nozzle clogging. Many indices have been used to quantify the nozzle clogging during the CC of steel. Several in-situ quantification parameters are used. The most used one during the production process in the steel plant is the increment of the stopper rod position when the flow rate of molten steel into the mold is controlled by stopper. The stopper rod position will sharply increase at a severe clogging, under which the casting speed has to be reduced and the clogged nozzle needs to be replaced with a new one. Other proposed in-situ parameters include the clog index defined as the ratio between the number of increments in stopper rod position and the number of drops,²⁴⁾ the mold level fluctuation, the cumulative stopper rise between fluctuations in mold level, the stopper relative location,²⁵⁾ and so on.

The nozzle clogging could also been characterized using ex-situ quantification parameters, such as the port width and port height of the clogged nozzle. Based on the area reduction of the cross-section or the outlet pore of an offline clogged nozzle, the clogging severity factor (CSF), defined as the ratio of the area blocked by deposits to the initial cross-section area of the nozzle, was proposed to quantify the clogging, as shown in Eq. (1).²⁶⁾ A similar parameter named the block rate calculated by Eq. (2)⁶⁾ based on the dissection of an offline clogged nozzle was developed as a quantitative indicator of nozzle clogging. Kojola et al. also defined a clogging factor to quantify the amount of clogging by comparing the theoretical mass teeming rate with the actual mass teeming rate, shown as Eq. (3).²⁷⁾ A deviation of the measured mass of cast steel from the calculated mass for the ideal Bernoulli flow after 3 min of casting was also used for evaluation of the castability of liquid steel.²⁸⁾

  

$$\begin{array}{l}CSF={\displaystyle \frac{\text{Extent of reduction in throughput}}{\text{Theoretical flowrate}}}\\ \hspace{1em}\hspace{1em}\hspace{0.5em}={\displaystyle \frac{\text{Area blocked by deposits}}{\text{Initial cross-sectional area}}}\end{array}$$(1)

  

$$R_{\text{block}}=\frac{V_{\text{n}}-V_{\text{c}}}{V_{\text{n}}}\times 100\%$$(2)

where, R_block is the block rate of nozzle, %; V_n is the initial volume of the flow region in the SEN under the slag-line without clogs, m³; and V_c is the actual volume of the flow region in the SEN with clogs, m³.

  

$$\eta =\frac{Q_{\text{h}}-Q_{\text{t}}}{Q_{\text{h}}}$$(3)

where, η is the clogging factor; Q_h is the theoretical mass teeming rate, kg/s; Q_t is the actual mass teeming rate at time t, kg/s.

3. Formation Mechanism of Nozzle Clogging

The formation mechanism of nozzle clogging during the continuous casting of molten steel has been widely investigated and is briefly summarized in Table 1. The nozzle clogging could be formed by one or more mechanisms under different conditions, mainly including the clogging caused by the adhesion and aggregation of high-melting-point deoxidation products, the clogging caused by the inclusion precipitation and the formation of solidified steel layer attributed to the temperature drop of molten steel, the clogging caused by the chemical reaction between nozzle refractory and molten steel or the chemical reaction of new inclusion generation induced by reoxidation, etc.^29,30,31,32) The nozzle clogging in industrial production may be the result of the joint action of various mechanisms. The nozzle clogging is also related to factors of steel temperature, steel composition, casting period, casting speed, and argon gas flow rate in the nozzle. But no relationship was found between the nozzle clogging and the steel superheat.

Table 1. Research summary on the formation mechanism of nozzle clogging.

Type	Steel grade	Clog composition	Detailed formation mechanism	Year	Ref.
chemical reaction and temperature drop	Al‑killed steel	Al2O3 and frozen steel	- Precipitation of Al2O3 in the MnAlSi-oxide nozzle matrix - Reinforeced by freezing steel	1968	33)
temperature drop and chemical reaction and physical adhesion*	Al-killed steel	Al2O3 and frozen steel	- Steel freezing within a clog buildup reinforces the clogging - Greater clogging degree by reoxidation from air aspiration	1995	13)
temperature drop and physical adhesion		Al2O3 and FeAl2O4 embedded in the metal	- Solidification of Fe to the nozzle bore surface - Facilitated by Fe-impregnated alumina and mullite aggregates	2001	34)
temperature drop and physical adhesion	40Cr steel	Al2O3, Al2O3–CaO(–SiO2), MgO–Al2O3, frozen steel	- Insufficient preheating of nozzle - Temperature fluctuation of molten steel - Formation of frozen steel layer, precipitation of MgO–Al2O3	2019	19)
physical adhesion	Al killed steel	Al2O3	- Al2O3 deposition at the inlet and outlet of Al2O3–C SEN.	2000	35)
physical adhesion	Al-killed steel	Al2O3 with a small amount of CaO, MgO and TiO2	- 46–61% of clogging originated from the reduction of chromite-based stuffing sand - 39–53% of clogging came from deoxidation particles from ladle	2008	36)
physical adhesion	Ti bearing ULC steel	Al2O3 and Al2O3–TiOx	- Clogging tendency increases by forming small Al-Ti-oxides	2012	37)
physical adhesion	Al-killed steel	Al2O3	- Deposition of Al2O3 on the nozzle wall	2012	1)
physical adhesion	Al- and S-bearing free-cutting steel	MgO·Al2O3, CaS and CaO·2Al2O3	- Deposition of oxides and CaS formed in secondary refining	2015	25)
physical adhesion	Al-killed steel with 0.020–0.035% S	Al2O3, MgO·Al2O3, sulfide, entrapped steel	- Infiltration of Fe into refractory - Removal of protective zirconia surface - Adhesion of fine Al2O3	2015	38)
physical adhesion**		Al2O3 clusters	- Clogging is a stochastic and self-accelerating process	2018	39)
physical adhesion	Al-killed Ti-stabilized stainless steel	frozen steel and (MgO–Al2O3)rich–CaO–TiOx inclusions	- (MgO–Al2O3)rich–CaO(TiOx) accumulated on SEN wall - Formation of porous clog network - Steel solidified in clog network	2019	40)
physical adhesion	Ti-stabilized ULC steel	Ti-bearing Al2O3	- A high Ti/Al promotes clogging occurrence and intensity	2019	41)
physical adhesion*	Al-killed steel	Al2O3 clusters	- A high steel superheat and properly preheated SEN can inhibit steel solidification before clogging	2020	42)
chemical reaction and physical adhesion	Ti bearing ULC steel	frozen steel, clusters of Al2O3 and spinel	- Al2O3–TiOx inclusions promotes the melt freezing inside SEN deposits	2004	29)
physical adhesion and chemical reaction	Ti bearing IF steel	FeO·TiO2, FeO·Al2O3, Al2O3 and Fe	- Inner layer: loose Al2O3 clusters & granular Al2O3 - Middle layer: dendritic Al2O3 - FeO·TiO2 phase: formed by reoxidation	2010	16)
chemical reaction and physical adhesion	RE alloyed 304 stainless steel	rare earth metal oxides and alumina	- Formation of RE oxides and a alumina rich glaze by reaction between glass coating layer and SEN base materials - Promoting inclusion deposition on SEN wall	2011	43)
chemical reaction and physical adhesion	RE treated bearing steel	cerium aluminate compounds	- SEN decarburization - Steel penetration - Clogging growth	2019 2020	44) 45)
physical adhesion and chemical reaction	Ca-treated Al-killed Ti bearing steel	CaO–TiOx, Al2O3–MgO, Al2O3–CaO–SiO2, and frozen steel	- TiOx: from inclusion adhesion - Silicates: from reaction between nozzle material and steel	2021	23)
chemical reaction***	Fe–C–Al melts	Al2O3 network	- Al2O and SiO gases formed by the reduction of C in nozzle - Al2O3 formed at nozzle-steel interface by the reoxidation of suboxide gases	1992	46)
chemical reaction	Ti-stabilized stainless steel	TiOx as curved chain and Al2O3 as large agglomerates.	- TiOx: from the reaction of TiN with oxygen	1993	47)
chemical reaction***	LCAK steel Ti-killed steel	Al2O3	- LCAK steel: 3SiO2(s)+3C(s)+4[Al]= 2Al2O3(s)+3[Si]+3[C] - Ti‑killed steel: 5SiO2(s)+5C(s)+6[Ti]= 2Ti2O5(s)+5[Si]+5[C]	1994	48)
chemical reaction***	Al-killed steel	Al2O3	- FeO and FeAl2O4 formed by reoxidation act as the adhesive between Al2O3 particles	2001	49)
chemical reaction	LCAK steel	Al2O3	- Reoxidation of steel	2003	50)
chemical reaction	RE alloyed stainless steel	REM oxide cluster	- Reaction of Ce with ZrO2 nozzle refractory - Fast precipitation of inclusions on nozzle walls	2015	28)
chemical reaction	ULC steel and Ca-treated HSLA steel	Al2O3 for ULC steel, CaO–Al2O3 for HSLA steel	- Formation of Al2O3 by the reoxidation of steel by top slag	2016	51)
chemical reaction	Ti-alloyed, Al-killed IF steel	Al2O3, MgO·Al2O3, Al2O3–TiOx	- Formation of Al2O3 by the reoxidation of steel by filler sand	2019	52)
chemical reaction	Ti-ULC steel	TiOx–Al2O3 and Fe drops	- Oxidation of steel by CO gas generated by a carbothermic reaction in SEN - Formation of FetO-containing oxide and alumina, then a Fe and Al2O3–TiOx layer	2018 2019 2020	53) 54) 55)
chemical reaction	Ti-alloyed, Al-killed IF steel	Al2O3 with minor spinel content	- Reoxidation and local carbothermic reaction	2021	56)
chemical reaction	RE treated Q690D steel	CeAlO3, Al2O3 and CaO	- Formation of CeAlO3 by reaction of Ce2O3 in steel with Al2O3 in SEN - Formation of clogging layer by CeAlO3 and other inclusions	2021	57)
chemical reaction*		Al2O3 and Al2O3–TiOx	- Chemical reaction with steel melt - Local temperature and pressure drop - Transportation of inclusions	2021	58)
chemical reaction	Fe–Al–Ce alloy	RE aluminate (Ce–Al–O)	- Decarburization of nozzle material	2021	59)

Note: * indicates numerical simulation results, ** indicates results of simulation and pilot experiment, and *** indicates the mechanism reflected by laboratory experiment results

3.1. Clogging Formation by Physical Adhesion of Inclusions

The physical adhesion mechanism of the formation of nozzle clogging mainly refers to the adhesion and deposition of solid or semi-solid inclusions in molten steel on the inner wall of nozzles and their aggregation into clogs during continuous casting. There are generally four steps for the process: the transport of inclusions to the refractory-steel interface, the adhesion of inclusions on the surface of nozzle refractory, the sintering and deposition of inclusions, and the aggregation and growth of deposits, as illustrated in Fig. 5.⁶⁰⁾ The turbulent flow pattern will enhance the transport of inclusions to the nozzle wall.⁶¹⁾ The stability of inclusion adhesion on the inner wall of the nozzle depends on the roughness of the inner wall of the nozzle, the size of inclusion particles and the velocity of molten steel. The nozzle clogging mechanism during the casting of REM treated stainless steels was investigated via a pilot plant study.⁶²⁾ It was proposed that the nozzle clogging was mainly caused by the adhesion of inclusion and inclusion clusters on the nozzle, which was probably because of microeddy flow near the wall. It was also found that the steel containing small single inclusions clogged faster than that containing mainly large inclusion clusters owing to that the small inclusions could stick to the nozzle wall under high steel flow velocity while the larger ones could not.⁶²⁾

Fig. 5. Illustration of the clogging process by physical adhesion of solid inclusions.⁶⁰⁾

The physical adhesion of inclusions on the nozzle wall is also largely related to the interfacial tension. Solid inclusions in the molten steel have relatively high interfacial energy and have the tendency to reduce the interfacial energy. As a result, solid inclusions tend to agglomerate into larger particles and adhere to the refractory surface of SEN. The surface tension of the molten steel creates a void and an attractive force between solid inclusions or between the inclusion and the nozzle wall.^63,64,65) The attractive force has been calculated to be approximately an order of magnitude greater than the drag force and buoyant force on the inclusion.⁶⁶⁾ Then, a sintered bond between the inclusion and refractory wall can be formed to withstand drag and buoyant forces.

The interfacial energy and wettability of inclusions were determined by measuring the contact angle between inclusions and molten steel. The contact angle between different types of inclusions and steel grades are summarized in Table 2. The contact angle between Al₂O₃ inclusions and molten iron at 1600°C was measured as 135–144° by Ogino et al.⁶⁷⁾ Due to the high interfacial energy, Al₂O₃ inclusions are not wetted with molten steel, which will promote the sintering of Al₂O₃ particles.⁶⁸⁾ Thus, Al₂O₃ based inclusions are most likely to cause nozzle clogging based on the physical adhesion mechanism probably due to the high-melting point, large contact angle with molten steel, and irregular shape. According to the modeling research of Barati et al.,³⁹⁾ the clogging is a stochastic and self-accelerating process including the initial deposition of Al₂O₃ particles on the nozzle wall and the evolution and development of the clog structure. But the composition of clogs formed during the casting of Al-killed steel is not always the pure Al₂O₃, sometimes contains other components, for instance, a small amount of CaO, MgO, and TiO₂, which was inferred to be originated from the reduction of chromite-based stuffing sand.³⁶⁾ The clogs could be consisted of MgO·Al₂O₃, CaS and CaO·2Al₂O₃ in Al-killed high sulfur steels,^25,38) or (MgO–Al₂O₃)_rich–CaO–TiO_x inclusions in the Al-killed Ti-stabilized stainless steel.⁴⁰⁾ The nozzle clogging of Ti bearing ultra-low carbon (ULC) steels has been well investigated,^{16,29,37,41,51,52,53,54,55,56)} and the main components of clogs are Al₂O₃, MgO·Al₂O₃, and Al₂O₃–TiO_x, and sometimes with a small amount of Fe droplets. In Ti-bearing ULC steels the Ti can increase the clogging tendency⁶⁹⁾ and intensity by forming small Al₂O₃–TiO_x particles,^37,41) especially when the Ti/Al ratio was larger than 2.5.⁴¹⁾ One of the reasons is that the titanium decreases the work of adhesion of liquid steel on Al₂O₃, in other words, reinforces the attraction between liquid steel and alumina-type nozzle materials.⁷⁰⁾ It was found that the solidification of molten steel cannot occur before clogging at high steel superheat and properly preheated SEN.⁴²⁾ However, when the porous network of clog is formed by the inclusion accumulation on the inner wall of the SEN, the steel within the clog buildup may freeze and reinfore the clog buildup.^13,40)

Table 2. Contact angles between different inclusions and steel grades.

Temperature (K)	Roughness	Oxygen partial pressure	Inclusion	Contact angle (°)	Steel grade	Ref.
1815	5 nm	10−19 atm	Al2O3 (single crystal)	105.1	pure iron	71)
				132	steel
1873	3 μm	10−3 Pa	Al2O3 (sintered)	136	pure iron	72)
1823				114	tire cord steel
				115	die steel
				103	stainless steel
				99.7	high sulfur steel
1873		10−23 atm	Al2O3 (single crystal)	102	IF steel	73)
1873			Al2O3 (single crystal)	98	Ca treated steel	74)
1815	5 nm	10−19 atm	MgO (single crystal)	99.2	pure iron	71)
				90	steel
1873	0.5 nm	10−3 Pa	MgO (single crystal)	113	pure iron	72)
1823				108.2	tire cord steel
				108.6	die steel
				99.4	stainless steel
				95	high sulfur steel
1873			MgO (single crystal)	137	Ca treated steel	74)
1815	3 μm	10−19 atm	Ti2O3 (sintered)	122	pure iron	71)
				121	steel
1873	0.1 μm	10−3 Pa	TiN (coating film)	115	pure iron	75)
1823				80	tire cord steel
				81.7	die steel
				75	stainless steel
				94	high sulfur steel
1873		10−23 atm	TiN	123	IF steel	74)
1823	0.5 nm	10−3 Pa	MgO·Al2O3 (sintered)	102.6	pure iron	72)
				105.7	tire cord steel
				102.5	die steel
				98.7	stainless steel
				92.6	high sulfur steel
1873		10−23 atm	MgO·Al2O3 (sintered)	110	IF steel	74)
1873			MgO	125	pure iron	76)
1823			MgO	128	pure iron	76)
1823			MgO (95%)	122	low carbon steel	77)
1933	1 nm	10−19 atm	MgO (single crystal)	90	pure iron	78)
1815	5 nm	10−22 atm	MgO (single crystal)	99	pure iron	79)
1873			SiO2	115	pure iron	76)
1873			SiO2	128	pure iron	80)
1823		10−18 atm	SiO2 (single crystal)	135	pure iron	81)
1623			CaO	132	pure iron	76)
1823			CaO	115	pure iron	80)
1623	112 nm		CaO	147	Fe-4%C	82)
1823			CaO	118	electrolytic iron	83)
1823			TiO2	84	industrial pure iron	84)
1815	3 μm	10−20 atm	Ti2O3	128	pure iron	71)
1873			Cr2O3	88	pure iron	76)
1873			ZrO2 (94%)	120	pure iron	80)
1873			ZrO2 (94.5%)	123	pure iron	85)
1823			ZrO2	122	pure iron	76)
1823			ZrO2	119	pure iron	76)
1873	1 μm		ZrO2 (95%)	125	electrolytic iron	77)
1823			MgO–C	108–116	steel	86)
1873			(5–20%)MgO–Al2O3	128–137	electrolytic iron	79)
1933	1 nm	10−19 atm	MgO–Al2O3 (single crystal)	105	pure iron	78)
1833			MgO–Al2O3	108	electrolytic iron	87)
1873			3Al2O3-2SiO2	131.5	pure iron	85)
1823		10−8 atm	74Al2O3-26SiO2	119	pure iron	88)
1873			(5–20%)CaO–Al2O3	113–123	electrolytic iron	79)
1873			(2–20%)CaO–ZrO2	114–132	electrolytic iron	79)
1823			Al2O3	150	Fe-16%Cr	89)
1823	333 nm		Al2O3	95	Fe‑19%Cr‑10%Ni	90)
1823			MgO	133.5	Fe‑Al(1.8×10−5Al)	91)
1823			SiC	initial: 136; equ.: 31	Fe-74%Si	92)
1823			SiC	initial: 100; equ.: 38	Fe-24.7%Si	92)
1793			ZrO2 (89%)	108	Fe-0.16%C	76)
1813			ZrN	84.9	pure iron	93)
1823			BN	112	pure iron	76)
1823			CaS	87	pure iron	76)
1873			BeO	136	pure iron	80)
1823			MgAlON	130	pure iron	94)
1823			Al2O3-(10–30%)C	135–140	pure iron	95)

3.2. Clogging Formation by Temperature Drop

The base material of SEN is usually Al₂O₃–C, which has good thermal conductivity. A heat loss through the nozzle wall will happen during the continuous casting of steels, leading to the freezing of molten steel on the inner wall of SEN, or the newly precipitation of inclusions due to the movement of reaction equilibrium. The insufficient pre-heating of nozzles will accelerate the formation of the solidified steel layer on the inner wall of nozzles. It was found that there will be an initial freezing of the steel at the internal wall when the nozzles had a 200°C lower temperature than the steels liquidus temperature.²⁷⁾ The solidified steel layer with rough surface and porous structure will favour the adherion and deposition of solid inclusions and reinforce the clog buildup.^33,34) In addition, an insufficient superheat of molten steel, especially for the first heat of a casting sequence or a new nozzle, and a large temperature fluctuation of molten steel during the continuous casting process will also induce the formation of frozen steel layer and the precipitation of solid inclusion phase,¹⁹⁾ leading to the further occurrence of nozzle clogging. A typical layer structure of such nozzle clogging is shown in Fig. 4.¹⁹⁾ The clogging in the ladle nozzle bore could also be a result of the solidification of steel to the nozzle bore surface, which was facilitated by the presence of metal-impregnated alumina and mullite aggregates.³⁴⁾ On the other hand, when a porous network of clog is formed by the accumulation of inclusions on the inner wall of SEN, the steel in the clog buildup may freeze,^13,29,40) which further reinforces the clogging.

3.3. Clogging Formation by Chemical Reaction

In many industrial cases, the nozzle clogging may be formed or reinforced by chemical reactions. The clogging generated by chemical reactions usually presents as a continuous phase by the growth of reaction products. The chemical reactions involved in the continuous casting is complicated and can be roughly classified into three types: the decarburization reaction in the base material of SEN, the interaction between molten steel and SEN, and the reoxidation of molten steel, as illustrated in Fig. 6. These three types of reactions would sometimes occur simultaneously.

Fig. 6. Illustration of chemical reactions causing the SEN clogging. (Online version in color.)

3.3.1. Decarburization Reaction

The base material of SENs used in industrial productions is mostly Al₂O₃–C type composed of Al₂O₃, graphite, and a small amount of SiO₂ as the binder. The graphite in the inner surface layer will be oxidized to some extent during the preheating of SEN, resulting in a decarburization layer, meanwhile, a carbothermic reaction or a dissolution of carbon could also occur in the SEN during the continous casting, leading to the formation of a decarburization layer and a rough surface of the inner wall of the SEN,^{53,54,55,56,59)} which could be the original location of the clog deposition. The molten steel would penetrate into the decarburization layer and Al₂O₃-type fragments could also be subsequently transferred into the molten steel.^96,97) In addition, the decarburization reaction will usually promote the nozzle clogging by other reactions.

Tian et al.^44,45) proposed a formation mechanism of SEN clogging originated from the decarburization reaction in the continuous casting of a bearing steel containing RE elements, as illuratrated in Fig. 7.⁴⁴⁾ The first step of the clogging was the formation of decarburization layer on the inner wall of SEN by the graphite dissolution; the second step was the penetration of steel into the decarburization layer and the adhesion of inclusions due to the better wettability with the molten steel; the third step was the growth of clogging and the expansion of the laminar flow interface layer on the inner wall of the SEN, which further accelerated the clogging.

Fig. 7. Proposed formation process of the nozzle clogging originated from the decarburization reaction during the continuous casting of a bearing steel,^44,45) (a) layer structure of nozzle clogging, (b) formation of decarburization layer, (c) steel penetration into the decarburization layer, (d) clogging growth and laminar flow interface expansion. (Online version in color.)

Fukuda et al.⁴⁶⁾ proposed a mechanism of Al₂O₃ deposition on the Al₂O₃–C immersion nozzle focusing on the decarburization reaction products of Al₂O and SiO suboxide gases inside the nozzle. After the diffusion to the contact interface with molten steel, the suboxide gas would be reoxidized into Al₂O₃ at the interface to form the Al₂O₃ deposition.

A glass/silicon powder layer mainly composed of Al₂O₃, SiO₂, and Na₂O was coated on the inner surface of SENs to prevent the decarburization during the preheating of SENs used for RE treated stainless steels. However, the alkaline rich glaze would penetrate into the base refractory during preheating and react with graphite to form CO gas, leading to the reoxidation of RE elements to form RE-oxides at the SEN/steel interface. In addition, the penetrated glaze would also react with Al₂O₃ in the SEN refractory to form a high viscous alumina rich glaze, creating a uneven surface SEN internal surface and facilitating the accumulation of the primary inclusions on the SEN inner walls.⁴³⁾

3.3.2. Nozzle-Steel Interaction

It was found through the chemical interactions in Al₂O₃–C/Fe system applying the sessile drop method that Al₂O₃ whiskers could be generated on the surface of the iron droplet or the Al₂O₃–C substrate,^97,98) as shown in Fig. 8,⁹⁸⁾ which might be a reason for the formation of nozzle clogging during the continuous casting of Al killed steels.

Fig. 8. Morphology of alumina whiskers with corresponding SEM determination.⁹⁸⁾ (Online version in color.)

The nozzle clogging of Ti-bearing ULC steels are usually more severe than normal Al-killed steels, which is attributed to Ti-related chemical reactions.^37,41,69) It was revealed that the SEN clogging of a Ti bearing ULC steel was consisted of three layers with different shaped Al₂O₃ particles.¹⁶⁾ Besides, Ti-containing phases original from several sources could always be detected in the clog. A typical SEN/steel interaction type mechanism for the formation and growth of the initial clog deposit at the nozzle/steel interface during the continuous casting of Ti-bearing ULC steels was proposed by Lee et al. by carrying out a series of experiments employing a rotating finger method, as illustration in Fig. 9.^53,54,55) Five steps were proposed for the mechanism under the CaO-free nozzle:⁵³⁾ (i) The CO gas is generated by carbothermic reactions (4) and (5) in SEN; (ii) The CO moves to the refractory/steel interface through refractory pores; (iii) The CO oxidizes the Al, Ti and Fe to form Fe_tO–Al₂O₃–TiO_x complex oxides, as Reaction (6); (iv) The clog consisted of FeO–Al₂O₃–TiO_x and Al₂O₃ forms and attaches to the inner wall of SEN; (v) The Fe_tO in the complex oxide is gradually reduced by the Al in steel or the C in refractory, forming Fe droplets and Al₂O₃–TiO_x clogs. When components of CaO, ZrO₂, and SiO₂ are contained in the SEN refractory, they will rapidly dissolved into the Fe_tO–Al₂O₃–TiO_x phase. After the reduction of Fe_tO by the Al and Ti in steel, the clog deposit will change to a mixture of reduced Fe droplets and CaO–Al₂O₃–TiO_x–ZrO₂–SiO₂ oxide.⁵⁵⁾ The kinetics of CO gas dissolution into stirred liquid steel and its impact on the nozzle clogging was also discussed.⁹⁹⁾ Besides of CO, the SiO produced by Reaction (4) could also oxidize the Al and Ti in steel to form Al₂O₃ and TiO_x.⁴⁸⁾ Calcium (Ca) treatment is sometimes used to reduce the nozzle clogging intensity of Al-killed Ti-bearing steels. However, a clog deposit on the inner wall of SEN could still be formed with a proposed three-step formation mechanism.²³⁾

  

$$\mathrm{S}\mathrm{i}{\mathrm{O}}_2\left(\mathrm{s}\right)+\mathrm{C}\left(\mathrm{s}\right)=\mathrm{S}\mathrm{i}\mathrm{O}\left(\mathrm{g}\right)+\mathrm{C}\mathrm{O}\left(\mathrm{g}\right)$$(4)

  

$$\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3\left(\mathrm{s}\right)+2\mathrm{C}\left(\mathrm{s}\right)=\mathrm{A}{\mathrm{l}}_2\mathrm{O}\left(\mathrm{g}\right)+2\mathrm{C}\mathrm{O}\left(\mathrm{g}\right)$$(5)

  

$$\mathrm{C}\mathrm{O}\left(\mathrm{g}\right)+\left[\mathrm{M}\right]\left(\mathrm{M}=\mathrm{A}\mathrm{l},\mathrm{T}\mathrm{i},\mathrm{F}\mathrm{e}\right)\to \mathrm{F}{\mathrm{e}}_t\mathrm{O}-\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3-\mathrm{T}\mathrm{i}{\mathrm{O}}_x+\left[\mathrm{C}\right]$$(6)

Fig. 9. A formation mechanism of nozzle clogging for the continuous casting of Ti-bearing ULC steels.^53,54) (Online version in color.)

Chemical reactions are also likely to be involved in the formation of nozzle clogging during the continuous casting of rare earth (RE) treated steels. The Ce₂O₃ and CeAlO₃ inclusions could be formed in the Al-killed steel with a cerium (Ce) content of 0.0028%.⁵⁷⁾ During the continuous casting, Ce₂O₃ inclusions in the molten steel will react with the Al₂O₃ in the SEN refractory to form CeAlO₃, then, the aggregation of CeAlO₃ and Al₂O₃ induces the formation of clogging layer on the SEN surface. The Ce contents on the clogging of a ZrO₂ type nozzle for a RE treated stainless steel was investigated by pilot scale experiments.²⁸⁾ It was found that the nozzle clogging severity would drastically increased with increasing the Ce content in steel up to >0.05%. And the soluble Ce content larger than 0.025% was an important reason for fast nozzle clogging due to the reaction between the Ce in steel and the ZrO₂ in nozzle, inducing the fast precipitation of inclusions and clusters on the nozzle walls.²⁸⁾

3.3.3. Reoxidation

The reoxidation of ULC steel will induce the new formation of Al₂O₃,^51,52) Al₂O₃–TiO_x²⁹⁾ or FeO·TiO₂¹⁶⁾ inclusions and promote the extensive melt freezing inside the SEN deposits, reinforcing the clogging. According to the particle size distribution analysis performed by Karnasiewicz et al.,^52,56) particles generated by reoxidation from different oxidation sources would present different sizes but have the same lognormal shape of population density functions. Many factors can lead to the reoxidation of ULC steel, including air absorption, top slag and tundish covering agent,⁵¹⁾ filler sand of ladle,⁵²⁾ CO gas generated inside the nozzle refractory,^53,54,55) etc. A severe reoxidation of molten steel would induce the formation of FeO_x or FeO·Al₂O₃, which could act as the adhesive between deposits and promote the nozzle clogging.^49,100) However, Suzuki et al.¹⁰¹⁾ proposed that the concentration decrease rate of aluminum in molten steel caused by the oxidation by the permeated air through a SEN refractory tube was estimated 0.35 ppm per hour, which did not attribute to alumina build-up in the SEN.

3.4. Prediction of Nozzle Clogging by Numerical Simulation

The formation mechanisms of nozzle clogging mentioned above are mostly determined by the analysis on an offline clogged nozzle used in a laboratory scale experiment, a pilot scale experiment, or an industrial production. However, the method of experimental analysis is difficult to reveal the detailed formation process of clogging, especially hard to evaluate the influence of the flow pattern of molten steel in the SEN on the clogging. Therefore, the numerical simulation method is introduced into the nozzle clogging research. Many models have been established and developed for the prediction of nozzle clogging, as summarized in Table 3.

Table 3. Prediction models for nozzle clogging.

Author	Model	Result	Year	Ref.
Ikäheimonen et al.	Feedforward neural networks with backpropagation training	- How long time the undisturbed casting can continue based on the production data.	2002	102)
Long et al.	A kinetic model considering different forces to calculate the inclusion trajectory	- Easy deposition of inclusions: smaller inclusion size, more T.O of molten steel, smaller casting speed.	2010	103)
Ni et al.	An extended Eulerian model considering Brownian and turbulent diffusion, turbophoresis, and thermophoresis as transportation mechanisms.	- Non-uniform steel flow leads to an uneven distribution of inclusion deposition rates. - Large inclusions show a large deposition rate.	2014	104)
Gutierrez et al.	A model considering Navier-Stokes equations, standard k-ε model, and Lagrangian discrete phase model	- The highest inclusion deposition rate was just below a low static pressure zone.	2016	105)
Mohammadi-Ghaleni et al.	Eulerian-Lagrangian approach combined with the detached eddy simulation turbulent model	- A higher attaching rate of particles in divergent areas of SEN.	2016	106)
Barati et al.	Transient model including an Eulerian-Lagrangian approach for the transport of NMIs; a stochastic near-wall model to trace NMIs in the near-wall region	- Clogging is a transient process interacting with the melt flow and a stochastic and self-accelerating process.	2018	39)
Lee et al.	Arbitrary Lagrangian–Eulerian model for the fluid-structure interaction	- The dominant frequency decreases with increasing the total clog mass in the SEN.	2021	107)
Wang et al.	Long short-term memory time-series model	- Be capable of predicting the overall trend in quantitative indices for SEN clogging	2022	108)

Ikäheimonen et al. applied feedforward networks with backpropagation training to estimate the time available for undisturbed casting based on the production data related to clogging cases.¹⁰²⁾ However, the neural network model can only predict how long time the casting can continue and determine variables affecting the nozzle clogging. Then, more models are developed to predict the formation process of nozzle clogging, which is helpful to reveal the detailed formation mechanism.

Long et al. developed a kinetic model considering different forces on inclusions, and found that inclusions are more easily to be attached to the SEN wall under smaller inclusion size and casting speed, but could not be entrapped by the nozzle when the size is larger than 100 μm.¹⁰³⁾ Ni et al. established an extended Eulerian model considering Brownian and turbulent diffusion, turbophoresis, and thermophoresis as transportation mechanisms, and found that an uneven distribution of inclusion deposition rates at different locations of the SEN was caused by the non-uniform steel flow in the SEN.¹⁰⁴⁾ It was also proposed that large size inclusions showed a large deposition rate due to the strong effect of turbophoresis,¹⁰⁴⁾ which is inconsistent with Long’s result. Gutierrez et al. developed a model to study the effects of inertial, gravitational, buoyant, pressure gradient, and Saffman forces on the deposited inclusion trajectories and found tha the highest inclusion deposition rate was just below a low static pressure zone.¹⁰⁵⁾ Mohammadi-Ghaleni et al. performed a CFD simulation on melt flow patterns and particle-wall interactions inside the SEN to identify the nozzle clogging mechanism by combining the Eulerian-Lagrangian approach and detached eddy simulation turbulent model, indicated a smaller clog deposit thickness in convergent areas due to the larger velocity and turbulence of the flow while a higher attaching rate of particles in divergent areas of the nozzle.¹⁰⁶⁾ Barati et al. established a transient model employing an Eulerian-Lagrangian approach for the transport of NMIs and a stochastic near-wall model to trace NMIs in the near-wall region, found that clogging is a transient process interacting with the melt flow and also a stochastic and self-accelerating process,³⁹⁾ the simulated evolution of clogging in the nozzle is shown in Fig. 10.³⁹⁾ Lee et al. used the Arbitrary Lagrangian–Eulerian method to model the fluid-structure interaction and predict the vibration frequencies due to clogging. It was shown that the dominant frequency relecting the clogging process was primarily influenced by the total mass of clog and not much affected by the clogging locations in the SEN.¹⁰⁷⁾

Fig. 10. Simulated clogging evolution in the nozzle with time.³⁹⁾ (a) View of the clog region from the vertical section and a cross-section A-A of the nozzle; (b) zoomed view of the flow streamlines and a partially clogged section at 200 s, the yellow part is the clog and the color scale along the streamlines presents the magnitude of the flow velocity. (Online version in color.)

4. Prevention of Nozzle Clogging

Due to the detrimental effect of nozzle clogging on the castability of molten steel and the quality of casting products, many investigations have been carried out on the prevention of nozzle clogging during continuous casting, as briefly summarized in Table 4. In addition to ensuring the preheating effect of the nozzle before use, the prevention methods could be roughly divided into 5 types, including the cleanliness improvement of steel, the liquid modification of inclusions, the fluid flow control inside the nozzle, the coating treatment on the inner surface of nozzles, and the external field treatment such as the electric field and the pulsed electric current.

Table 4. Research summary on the prevention of nozzle clogging.

Type	Method	Key points	Year	Ref.
cleanliness improvement	experiment	- Optimization of slag composition - Slag deoxidation	2012	109)
cleanliness improvement	experiment	- Lowering the oxidation potential - Reducing REM inclusions before CC - Preventing the reoxidation of steel during CC	2013	110)
cleanliness improvement	experiment	- Optimization of refining slag - Improving the soft stirring operation to reduce the steel-slag desulfurization	2014	111)
liquid modification of inclusions	experiment	- Al2O3 modification by Ca treatment	2012	1)
liquid modification of inclusions	experiment	- Replacement Al deoxidation by Si–Al deoxidation to form liquid inclusions after Ca treatment	2019	112)
inclusion modification	experiment	- Increasing Al/Ti ratio in ULC steels to lower the oxidation of Ti	2019	54)
inclusion modification	experiment	- Controlling the liquid fraction of inclusions larger than 20%	2020	113)
fluid flow control	simulation	- Employing a parabolic curve shaped SEN bottom to guide the liquid steel flow in the SEN	2006	114)
fluid flow control and coating material	experiment	- Application the annular step SEN with carbon free liner combined with a multi layer porous upper nozzle	2012	115)
fluid flow control	simulation	- Controlling the internal flow of nozzles by designed chambers or deflectors to decrease the inclusion deposition	2020	116)
surface treatment and fluid flow control	experiment and simulation	- Anticlogging nozzles by applying selected purified raw materials, type of graphite, surface treatment, and new structure for internal flow control of SEN	2020	70)
Base material	experiment	- Changing alumina-graphite nozzles to alumina, zirconia, and zirconia-graphite nozzles to reduce the refractory-steel interactions	2006	117)
Coating material	experiment	- Application of an Al2O3 plasma coated Al2O3–carbon nozzle	2002	118)
protective coating material	experiment	- Employing CaO–ZrO2–C nozzles to form liquid phase by the reaction of CaO in the coat with Al2O3	2003	119)
coating treatment	experiment	- CaTiO3 coating of alumina-graphite SENs	2012	120)
coating treatment	experiment	- Using yttria-stabilized zirconia (YSZ) containing CaTiO3 coated nozzles to form liquid phase by the reaction between Al2O3 and CaTiO3.	2017	121)
coating treatment	experiment	- Using an YSZ coating during the casting of Ce alloyed stainless steels	2018	122)
External electric field	experiment	- Connecting the cathode of power (100 A) to the nozzle	2011	123)
Controlling the electric field	experiment	- Grounding of nozzles, presenting a significant inhibitory effect on the formation of clogging products and the adhesion of slag on the SEN surface	2021 2017	124) 125)
External electric treatment	experiment	- Application of a steady DC electric field to the SEN, decelerating the formation of SEN clogging	2020	45)
External electric field	experiment	- Application of a positive electric field on the wall of the SEN	2021	126)
Pulsed electric current	experiment	- Application of an electric current pulse (ECP) between the SEN and the stopper, decreasing the clogging thickness by 50%	2018	127)
Pulsed electric current	experiment	- ECP inhibits the formation of CeAlO3/LaAlO3 from the reaction between SiO2 in refractory and RE in steel	2021	128)

4.1. Improvement of Steel Cleanliness

Most of the solid inclusions deposited on the nozzle surface are endogenous and originated from the deoxidation of molten steel. Thus, the reduction of dissolved oxygen at the converter endpoint can decrease the inclusion amount formed after the deoxidation and is beneficial to the steel cleanliness. Then, promoting the inclusion removal from the molten steel is expected to reduce the clogging tendency of nozzles and extend the time of continuous casting.

Inclusions are mainly removed during the secondary refining. The removal of inclusions usually consists of two steps, including the transportation of inclusions from the molten steel to the steel/slag interface and the absorption and dissolution of inclusions into the slag. To promote the transportation of inclusions to the steel/slag interface, the fluid flow pattern and stirring strength in the ladle should be optimized by adjusting the argon gas flow rate and the distribution of argon blowing position. Taking the ULC steel as an example, the measures to promote the floating removal of inclusions in steel include: improving the circulating flow of molten steel in the RH refining by increasing the sectional area of snorkels and the argon flow rate, appropriately extending the vacuum treatment time, reasonable soft blowing flow rate and time, and appropriate standing time after refining. To promote the dissolution of inclusions into the slag, the composition of refining slag should be optimized. It was proposed that the absorption efficiency of Al₂O₃ based inclusions is positively correlated with the parameter of ∆C/η of slags.^129,130,131) Ren et al. made an in situ observation on the dissolution of Al₂O₃ particles in CaO–Al₂O₃–SiO₂ slags, accordingly, predicted the dissolution rate of of Al₂O₃ inclusions in Al₂O₃–SiO₂–CaO system slags at different temperatures, the case at 1600°C is shown in Fig. 11,¹³²⁾ indicating that the slag with the CaO/Al₂O₃ ratio around 1.8 is most conducive to the dissolution and absorption of Al₂O₃ inclusions.

Fig. 11. Prediction of the dissolution rate of Al₂O₃ inclusions in Al₂O₃–SiO₂–CaO system slags at 1600°C.¹³²⁾ (Online version in color.)

Inclusions could also be removed in tundish. Generally, the longer the molten steel stays in the tundish, the greater the probability of removing inclusions from the molten steel. Therefore, the tundish flow field is often optimized to promote the removal of inclusions, which is also called “tundish metallurgy”. The optimization of tundish flow field can be realized through turbulence suppressors, retaining walls, retaining dams and weirs. There are also studies on using tundish air curtain retaining walls to promote the floating removal of inclusions. Turbulence suppressors, retaining walls, dams and weirs can be used to optimize the flow field in tundish, and gas curtain type retaining walls in tundish have also been used to promote the floating removal of inclusions.

On the other hand, the reoxidation of molten steel during secondary refining and continuous casting should be avoid. For Al-killed steels, the contents of FeO and MnO in refining slag should be reduced to less than 1.5% by slag deoxidization to prevent their oxidation to molten steel during secondary refining. During soft blowing process, the argon flow rate should be accurately controlled to avoid the reoxidation caused by the exposure of molten steel. During the continuous casting process, good protective casting shall be ensured. The measures taken include using alkaline tundish covering agent to reduce the oxidation of the covering agent to the molten steel, and ensuring the appropriate argon sealing flow at nozzles to avoid the reoxidation caused by air absorption.

4.2. Liquid Modification of Inclusions

For many normal Al-killed steels, the nozzle clogging during continuous casting is usually prevented by the modification of solid Al₂O₃ inclusions into liquid inclusions. Although the inclusion composition could be controlled by refining slag,^133,134,135) calcium treatment is the most common method for liquid modification of Al₂O₃ inclusions^{136,137,138,139,140)} based on Reaction (7) by feeding calcium-containing wires or adding calcium-containing ferroalloys such as ferrosilicon.¹⁴¹⁾

  

$$x\left[\text{Ca}\right]+\left(\frac{x}{3}+y\right){\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}=x\text{CaO}\cdot y{\text{Al}}_{\text{2}}{\text{O}}_{\text{3(liquid)}}+\frac{2x}{3}\left[\text{Al}\right]$$(7)

The modification mechanism of solid Al₂O₃ based inclusions by calcium treatment has been widely investigated.^{136,137,139,140,142,143,144)} It was proposed that the CaS acts as a transition product in the transient evolution of inclusions after calcium treatment,^{136,139,140,144)} and the sulfur content in steel will affect the amount of the transition product and the composition of the equilibrium inclusion phase.¹⁴⁵⁾ Calcium aluminate has several components, among which only 12CaO·7Al₂O₃ (C₁₂A₇) and 3CaO·Al₂O₃ (C₃A) are liquid at casting temperature. Therefore, the liquid modification of inclusions requires precise calcium treatment.

Many related researches have been carried out in recent years, and the influence of steel cleanliness, steel composition, temperature, and operation parameters on the liquid modification of inclusions by calcium treatment has been clarified. The minimum and maximum calcium contents required for liquid modification of Al₂O₃ inclusions in steel with different T.O and T.S contents calculated using FactSage are shown in Fig. 12¹⁴⁶⁾ as an example. It was found that the increase of T.S content in steel will reduce the maximum calcium addition and narrow the “liquid window”, in which inclusions are 100% liquid. The increase of T.O content in steel will widen the “liquid window”, but is detrimental to the cleanliness of molten steel.¹⁴⁷⁾ The aluminum content in steel has no obvious influence on the calcium treatment effect. It was proposed that from the perspective of improving calcium yield and promoting inclusion modification, calcium treatment with SiCa and AlCa cored wires is better than that with pure Ca cored wires under the relatively small ladle capacity attributed to the increasing effect of [Si] and [Al] on the solubility of calcium in molten steel.¹⁴⁸⁾ Another factor influencing the modification effect is the calcium yield during the calcium treatment. A neural network model has been developed by current authors to predict the calcium yield during the calcium treatment process in steelmaking.¹⁴⁹⁾ The influence of different factors on calcium yield could be predicted by the model, that the calcium yield increased with the increase of the silicon content in steel and temperature, and decreased with the increase of the calcium content in steel and the feeding rate of calcium wire.

Fig. 12. Calcium content required for liquid modification of Al₂O₃ inclusions in steel with different T.O and T.S contents, (a) minimum calcium content, and (b) maximum calcium content.¹⁴⁶⁾ (Online version in color.)

4.3. Fluid Flow Control Inside Nozzles

On the one hand, the flow of molten steel in the nozzle can transport the inclusions to the nozzle wall, on the other hand, it can also flush the inner wall of the nozzle to prevent the attachment of inclusions, and even make the clogs fall off. The velocity profile of molten steel in a nozzle is divided into three regions from the flow center to the nozzle wall:¹¹⁶⁾ the turbulent bulk zone, the tansition layer, and the laminar boundary layer. The thickness of the laminar boundary layer is inversely proportional to the flow velocity of molten steel in the nozzle. When the solid inclusions move to the laminar layer, they are easy to be captured by the wall to form the clog.¹⁰³⁾ Therefore, if the thickness of the laminar layer is decreased and further the laminar structure is broken, the probability of nozzle clogging will be reduced to a certain extent.

The most common method used in the industrial practice to prevent nozzle clogging by the flow control is the argon injection through the nozzle wall or stopper rod. The gas bubbling can prevent the contact between the melt and refractory wall by forming a film, can also flush inclusions off the nozzle wall as well as promote their flotation by increasing the turbulence. In addition, the increased pressure in the nozzle caused by the argon injection can inhibit the aspiration of air through porosities in the refractory. However, a too high argon flow rate will lead to more severe mold level fluctuations and more mold flux entrainment, thereby reducing the steel quality.

The fluid flow pattern inside the nozzle could be controlled by adjusting the inner structure of nozzle,^{70,114,115,116,150)} using the swirl blade,^151,152) or introducing the electromagnetic swirling flow technology,¹⁵³⁾ as illustrated in Fig. 13. The structure of the inner surface of nozzles could be designed to annular step,¹¹⁵⁾ coronal block (also named mogul SEN¹⁵⁰⁾), and trapezoidal block.⁷⁰⁾ It was reported that Ar gas bubbles coming out of an annular step SEN were smaller than those coming out of a straight bore SEN,¹⁵⁴⁾ which is more effective in preventing alumina clogging since small bubbles have a tendency to descend along the inner surface of the bore. Industrial tests have also confirmed that the annular step SEN with carbon free liner can significantly reduce the occurrence of nozzle clogging.¹¹⁵⁾ Both the swirl blade and the electromagnetic swirling flow technology illustrated in Fig. 13 are originally applied to improve the uniformity and stability of the outflow from nozzle, stabilize the flow field and temperature distribution and reduce the level fluctuation in the mold.^{151,153,155,156,157)} At the same time, when a swirling flow is generated inside the SEN during the continuous casting, both the axial velocity and the tangential velocity of the molten steel near the inner wall of the nozzle are greater, as shown in Fig. 14, which is conducive to reducing the thickness of the laminar boundary layer and the scouring of the molten steel on the inner wall of the nozzle, so as to effectively reduce the occurrence of nozzle clogging.

Fig. 13. Methods for the flow pattern control inside nozzles, (a) annular step SEN,¹¹⁵⁾ (b) mogul SEN,¹⁵⁰⁾ (c) trapezoidal block SEN,⁷⁰⁾ (d) swirl bladed nozzle,¹⁵²⁾ and (e) electromagnetic swirled SEN.¹⁵³⁾

Fig. 14. Flow field of the molten steel in the SEN with and without swirling: (a) without swirling flow, (b) with swirling flow, (c) velocity field in vertical section of the nozzle outlet without swirling flow, (d) velocity field in vertical section of the nozzle outlet with swirling flow, and (e) distribution of tangential velocity in the middle section of the electromagnetic swirling flow (EMSF) device.¹⁵³⁾ (Online version in color.)

4.4. Coating Treatment of Nozzles

Many nozzle cloggings are attributed to the porous structure of the inner wall of the nozzle and the interaction between inclusions and nozzle refractories. Accordingly, the base material of nozzles are tried to be optimized or the coating treatment on the inner surface of nozzles are carried out to reduce the nozzle clogging during the continuous casting. Trueba et al.¹¹⁷⁾ investigated the effect of nozzle base material on the rate of clogging during the continuous casting of Al-killed steels, found that there was no interactions between the steel and alumina, zirconia, and zirconia-graphite nozzles, and slight interactions were observed in magnesia nozzles, while greater amounts of interactions were observed with alumina-graphite nozzles, inducing a much higher clogging rate in alumina-graphite nozzles than the other nozzles.

However, due to the good mechanical properties and thermal shock resistant properties, alumina-graphite composite SENs are still commonly used for the continuous casting of steel. The larger clogging tendency under the alumina-graphite nozzle is also attributed to the reaction of carbon with other components of the nozzle refractory, resulting in the formation of the porous and rough inner surface of nozzles, on which the frozen steel layer is easier to form and inclusions are easier to adhere. In other words, the high clogging tendency is due to the high adhesion force between the nozzle wall and inclusions. To lower the adhesion force, different anti-clogging materials for the coating treatment of SENs were proposed and briefly summarized in Table 5.

Table 5. Proposed anti-clogging coating materials for SENs.

Material	Mechanism	Year	Ref.
BN	- Formation of boron oxide-rich liquid phase - Prevention of steel infiltration due to poor wettability of BN	1991	158)
Lime-containing liner	- Formation of liquid phase	1993	159)
Al2O3 plasma coating	- No reaction with liquid steel nor with inclusions	2002	118)
CaO–ZrO2, CaO–ZrO2–C	- Formation of liquid phase by the reaction between CaO in the coat and Al2O3 inclusions	2003, 2007	119,160, 161)
Al2O3, ZrO2, ZrO2–C	- Much slighter refractory-steel interactions	2006	117)
Glass/silicon powder layer	- Formation of RE oxides and uneven internal surface of SEN - Promotion of the nozzle clogging	2011	43)
Carbon free liner	- Smoother surface to suppress the adhesion of Al2O3 inclusions	2012	115)
CaTiO3	- Formation of liquid phase by the reaction between Al2O3 and CaTiO3	2007, 2012, 2017	120,121, 160,161)
Y2O3-stabilized ZrO2	- Protection against decarburization of SEN	2018	122)
Purified raw materials and surface smoothing treatment	- Avoid the decarburization reactions	2020	70)

To solve the problem, high purity base materials with a surface treatment were firstly used, including the impregnated surface of the interior wall to substitue partially graphite and the selection of appropriate graphite.⁷⁰⁾ The new nozzle material yielded considerably less clogging than the nozzle with the conventional alumina-graphite. However, the clogging still existed. Then, carbon-free refractory materials are used to coat SENs to prevent nozzle clogging. Vermeulen et al.¹¹⁸⁾ evaluated several materials including Al₂O₃, ZrO₂, Al₂O₃-carbon, SiO₂, and MgO to find a suitable meterial for the SEN coating. The Al₂O₃ was found to react neither with molten steel nor with inclusions, and was concluded to be an anti-clogging material, which was also proved by the good result of the test using an Al₂O₃ plasma coated alumina-graphite nozzle. A much slighter nozzle clogging was also obtained in the industrial practice after casting Al-killed steels when using the carbon-free liner SEN compared to a convertional SEN with alumina-graphite material, as shown in Fig. 15.¹¹⁵⁾

Fig. 15. Comparison of alumina clogging condition between the normal alumina-graphite SEN and the carbon-free liner SEN.¹¹⁵⁾ (Online version in color.)

Another way of the coating treatment of nozzles is to form a liquid layer after the reaction between the coating material and the molten steel or inclusions. Most likely, a fraction of liquid phase is sufficient to provide some protection against clogging.¹²⁰⁾ Since most of inclusions causing nozzle clogging are Al₂O₃-based, CaO containing materials are most used as coating materials. Devic et al.¹¹⁹⁾ developed an anti-clogging nozzles by coating a protective CaO–ZrO₂–C layer with the composition in wt-% of 4.4SiO₂-0.48Al₂O₃-0.06Fe₂O₃-20.7CaO-51.2ZrO₂-22C on the inner wall of conventional alumina-graphite nozzles in the continuous casting of steel slabs. During the casting, the CaO in the coating material reacts with Al₂O₃ agglometrates in the clog to form low melting calcium aluminates, leading to the flushing effect, thereby decreasing the deposition and build-up of clogs and delaying the eventual blocking of the nozzle, with a approximately halved average clog thickness. However, Tsujino et al.¹⁶²⁾ found that the CaO–ZrO₂–C refractory was more hardly to be deposited than Al₂O₃–C type only under a low Al content, on the contrary, it would be more easily deposited in the case of some level of Al content and non-metallic inclusions in steel. The laboratory experiments carried out by Takei¹²⁰⁾ indicated that the CaTiO₃ would be an appropriate coating material according to the partial formation of liquid phase after the reaction between CaTiO₃ and alumina, and thus was proposed to have a high potential to diminish clogging problems in the industry. Svensson et al.¹²¹⁾ concluded that plasma sprayed CaTiO₃ coatings can reduce the clogging tendency during the continuous casting Al-killed low carbon steels through the test of coating materials containing 4.8 wt% and 9.1 wt% CaTiO₃ mixed with yttria stabilized zirconia (YSZ) powder in laboratory trials and pilot plant trials. It was showed that an approximately three times higher steel mass could be teemed through the coated nozzles before clogging occurred compared to trials with uncoated standard nozzles. Then, they implemented an plasma coating material consisted of YSZ to prevent the SEN clogging during the continuous casting of Ce-treated stainless steels.¹²²⁾ Both the pilot plant trials and industrial trials showed that the use of an YSZ coating led to a decreased clogging tendency in comparison to when using an uncoated SEN.

4.5. External Electric Treatment during Continuous Casting

Paik et al.^163,164,165) investigated the surface charging characteristics of different oxides in contact with molten metals introducing the differential potential analysis and induced electromotive force method, and found that most oxides in liquid metals carry positive charges on their surfaces. Meanwhile, Yang et al.¹⁶⁶⁾ investigated the charged characteristics of SEN during continuous casting, and found that the inner wall of the nozzle was negative charged owing to the friction of molten steel with nozzle, and the amount of charges was proportional to the steel flow rate. Thus, non-metallic inclusions in molten steel during continuous casting prefer to migrate and adhere to the inner wall of nozzles, as illustrated in Fig. 16. Based on the phenomenon, methods of reducing the potential difference between molten steel and nozzle refractory, such as the external electric treatment, have been developed in recent years to reduce nozzle clogging during the continuous casting process, as summarized in Table 6. The treatment methods can be mainly divided into three types, including grounding of the nozzle, external direct current (DC) electric field, and external electric current pulse (ECP), as shown in Fig. 16.

Fig. 16. Illustration of charge distributions in SEN. (Online version in color.)

Table 6. External electric treatment for anti-clogging of SEN during CC.

Method	Nozzle material	Steel grade	Key points	Year	Ref.
External electric field	Almandine	LCAK steel and ULCAK steel	Connecting the cathode of power (100 A) to the nozzle	2011	123)
Grounding of SEN	73% ZrO2, 7% Al2O3, 12% C, 8% (SiO2+Na2O); 70% Al2O3-30% C	SWRCH22A cold heading steel and LCAK steel	Grounding of the nozzle	2021 2017	124) 125)
External DC electric field	70.69% Al2O3, 10.69% ZrO2, 12.45% SiO2, 5.89% C, 0.28% others	B-bearing Al-killed cold heading steel	Applying a steady DC electric field to the SEN	2020	45)
External electric field	70.69% Al2O3, 10.69% ZrO2, 5.89% C, 12.73% others	Ce–Al alloyed steel	Applying a positive electric field on the SEN wall	2021	126)
Electric current pulse		LCAK steel	Imposing an electric current pulse (ECP) between the SEN and the stopper	2018	127)
Electric current pulse	52% Al2O3, 24% C, 11% SiO2, 8% ZrO2 and 5% others	RE(Ce, La) bearing steel	Imposing an ECP between the SEN and the stopper	2021	128)

Grounding the nozzle appears to be the simplest method to control the electric field characteristics of the interface between the nozzle and the molten steel. After the nozzle is grounded, the surface charge of the nozzle will flow to the ground, making the nozzle-steel interface electrically neutral and inhibiting the clogging formation on the SEN surface.¹²⁴⁾

To achieve a better inhibitory effect on the nozzle clogging, external electric fields, including DC and ECP, are further employed. In industrial practice, an external electric field is generally applied between the stopper rod and the outer wall of SEN through wire connections.^45,123,127) The external electric field could transfer excess negative charges to the molten steel inside the SEN and maintain a low or even a zero electric potential difference between the SEN and molten steel, so as to decrease the interaction between the two phases, leading to a significant inhibitory effect on the nozzle clogging.¹²⁷⁾ The typical inhibitory effect of applying external electrical fields on the nozzle clogging is shown in Fig. 17. Figure 17(a) shows that the clog thickness in the SEN applying an external DC electric field was significantly smaller than that without external electric field treatment. Meanwhile, as shown in Fig. 17(b), the clogging in the SEN with ECP treatment was much slighter than that without treatment, the thickness of the clogs on the inner wall of SEN decreased from approximately 10 mm to about 5 mm.

Fig. 17. Comparison on the clogging between non-treated SENs and SENs treated with (a) DC¹²³⁾ and (b) ECP.¹²⁷⁾ (Online version in color.)

In addition, it was proposed based on a laboratory scale experiment that the external electrical field can directly affect the decarburization, clogging, and reaction behavior between the nozzle and molten steel, and the decarburization reaction rate and clogging degree of the nozzle connecting the anode of the DC power supply was found to be lower than those of the nozzle connecting the power cathode.⁵⁹⁾

Based on the phenomenon that the electric current density gradient attempts to drive high-resistance particles to migrate to regions with low current density in a non-uniform electric current field,¹⁶⁷⁾ a new mechanism of the inhibiting effect of ECP on nozzle clogging was proposed by Yan et al.¹⁶⁸⁾ according to the observation on the interface between the Al₂O₃–C nozzle material and the RE-containing molten steel treated by ECP. As illustrated in Fig. 18,¹⁶⁸⁾ the application of pulsed electric current can cause the formation of a liquid amorphous phase on the surface of nozzle materials, restraining the decarburization and erosion of nozzles. In addition, the non-uniform electric current field tends to drive inclusions away from the surface of nozzles, reducing the migration and adhesion of inclusions to the nozzle surface.

Fig. 18. A proposed mechanism of the inhibiting effect of ECP on nozzle clogging.¹⁶⁸⁾ (Online version in color.)

5. Future Development and Challenges in Nozzle Clogging Control

There have been many researches on the control of nozzle clogging during the casting of Al-killed steels with a aluminum content lower than 0.1%. However, the problem of nozzle clogging in the continuous casting process of high aluminum steels and RE treated steels still needs to be solved. With the demand for lightweight automobiles, high aluminum steels have developed rapidly in recent years. With a significant increase in aluminum content in steel, the physical parameters such as viscosity and surface tension of the molten steel will first change, and then the types of endogenous inclusions in the steel may also change, such as the generation of AlN inclusions. Furthermore, the increase in reactivity of the molten steel may result in changes in the reaction between the molten steel and the refractory material of the nozzle during the casting process. All of these will lead to changes in the mechanism and degree of nozzle clogging during the continuous casting process of high aluminum steels. Meanwhile, due to the ability of RE elements to modify the morphology of sulfides in steel, refine grain size, and improve the comprehensive performance of steels, RE treated steels are also increasingly being developed. However, due to the fact that RE-containing inclusions are mostly solid in molten steel and have a density similar to that of molten steel, they are difficult to remove from the steel. Additionally, RE elements have high reactivity, resulting in a poor castability of RE treated steels. Thus, one of the main problems in the production process of rare earth treated steels is the severe nozzle clogging during the continuous casting process.

Most of the existing methods for controlling nozzle clogging are still contact based, such as optimizing the composition of nozzle refractory and using coating materials. Due to the reaction between the nozzle refractory and the molten steel, these methods still have a certain impact on the cleanliness of steel. With the increasing demand for steel cleanliness, non-contact methods for controlling nozzle clogging are receiving increasing attention. Therefore, the treatment of external magnetic fields or external electric fields is one of the important development directions. In order to further enhance the inhibitory effect of nozzle clogging, the collaborative control of multiple methods is also one of the development trends in the future.

6. Conclusions

(1) The characteristics of nozzle clogging were briefly summarized. The clogs at continuous casting nozzles are mainly the aggregates of deoxidation products, solidified steel, complex oxides, and sometimes sulfides or nitrides with different appearances depending on the steel grade and clogging degree.

(2) The reasons for the formation of nozzle clogging roughly include the physical adhesion of high-melting-point inclusions, the temperature drop resulted in the inclusion precipitation and the solidified steel layer formation, and the chemical reactions occur during the continuous casting. The nozzle clogging in industrial production may be a result of the joint action of various mechanisms.

(3) The prevention methods of nozzle clogging could be roughly divided into 5 types, including the cleanliness improvement of steel, the liquid modification of inclusions, the fluid flow control inside the nozzle, the coating treatment on the inner surface of nozzles, and the external field treatment.

(4) In the future, more attention needs to be paid to the control of nozzle clogging in the continuous casting process of high aluminum steels and rare earth treated steels. Non-contact control method such as the treatment of external magnetic fields or external electric fields is an important development direction. Multiple methods should be coordinated to achieve a better inhibitory effect on the nozzle clogging.

Acknowledgements

The authors are grateful for support from the National Nature Science Foundation of China (Grant No. 52174293, No. U22A20171, No. 51874031), the Fundamental Research Funds for the Central Universities (Grant No. FRF-BD-20-04A), the High Steel Center (HSC) at North China University of Technology, University of Science and Technology Beijing, and Yanshan University, China.

References

• 1)   B.  Harcsik,  P.  Tardy and  G.  Karoly: Metall. Res. Technol., 109 (2012), 177. https://doi.org/10.1051/metal/2012046
• 2)   H.  Bai and  B. G.  Thomas: Metall. Mater. Trans. B, 32 (2001), 707. https://doi.org/10.1007/s11663-001-0125-4
• 3)   S.-M.  Cho,  S.-H.  Kim,  R.  Chaudhary,  B. G.  Thomas,  H.-J.  Shin,  W.-R.  Choi and  S.-K.  Kim: Proceedings of AISTech 2011, Iron and Steel Society, Indianapolis, Indiana, (2011), 1.
• 4)   L.  Zhang,  Y.  Wang and  X.  Zuo: Metall. Mater. Trans. B, 39 (2008), 534. https://doi.org/10.1007/s11663-008-9154-6
• 5)   P. S.  Srinivas,  D. K.  Mishra,  R.  Gupta,  J. M.  Korath and  A. K.  Jana: Can. Metall. Q., 58 (2019), 308. https://doi.org/10.1080/00084433.2018.1560662
• 6)   C.  Hua,  M.  Wang,  D.  Senk,  H.  Wang,  Q.  Zhang,  J.  Zhi and  Y.  Bao: Metals, 11 (2021), 662. https://doi.org/10.3390/met11040662
• 7)   C.  Hua,  M.  Wang and  Y.  Bao: Metall. Mater. Trans. B, 51 (2020), 2871. https://doi.org/10.1007/s11663-020-01996-5
• 8)   P. S.  Srinivas,  A.  Singh,  J. M.  Korath and  A. K.  Jana: Ironmak. Steelmak., 44 (2017), 473. https://doi.org/10.1080/03019233.2016.1215948
• 9)   B.  Li,  H.  Lu,  Z.  Shen,  X.  Sun,  Y.  Zhong,  Z.  Ren and  Z.  Lei: ISIJ Int., 59 (2019), 2264. https://doi.org/10.2355/isijinternational.ISIJINT-2018-866
• 10)   H.  Zhou,  L.  Zhang,  Q.  Zhou,  W.  Chen,  R.  Jiang,  K.  Yin and  W.  Yang: Steel Res. Int., 92 (2021), 2000547. https://doi.org/10.1002/srin.202000547
• 11)   Y.  Wang,  L.  Zhang,  W.  Yang,  S.  Ji and  Y.  Ren: Steel Res. Int., 92 (2021), 2000661. https://doi.org/10.1002/srin.202000661
• 12)   Y.  Wang,  L.  Zhang,  W.  Yang and  Y.  Ren: J. Iron Steel Res. Int., 29 (2022), 237. https://doi.org/10.1007/s42243-021-00577-3
• 13)   K. G.  Rackers and  B. G.  Thomas: Proceedings of 78th Steelmaking Conference, Iron and Steel Society, Nashville, TN, (1995), Vol. 78, 723.
• 14)   K. G.  Rackers: Master Thesis, University of Illinois at Urbana-Champaign, (1995).
• 15)   S. K.  Michelic and  C.  Bernhard: Steel Res. Int., 93 (2022), 2200086. https://doi.org/10.1002/srin.202200086
• 16)   H.  Cui,  Y.  Bao,  M.  Wang and  W.  Wu: Int. J. Miner. Metall. Mater., 17 (2010), 154. https://doi.org/10.1007/s12613-010-0206-y
• 17)   L.  Zhang: Non-metallic Inclusions in Steels: Industrial Practice (in Chinese), Metallurgical Industry Press, Beijing, (2020), 94. ISBN978-7-5024-8344-9
• 18)   W.  Yang,  L.  Zhang,  Y.  Hu,  F.  Fan and  G.  Li: J. Univ. Sci. Technol. Beijing, 36 (2014), Suppl, 182 (in Chinese). https://doi.org/10.13374/j.issn1001-053x.2014.s1.034
• 19)   S.  Ji,  Y.  Luo,  G.  Zhang,  W.  Jia,  Y.  Ren and  L.  Zhang: Iron Steel, 54 (2019), No.8, 124 (in Chinese). https://doi.org/10.13228/j.boyuan.issn0449-749x.20190210
• 20)   R.  Maddalena,  R.  Rastogi,  S.  Bassem and  A. W.  Cramb: Iron Steelmaker, 27 (2000), 71.
• 21)   Y.  Vermeulen,  B.  Coletti,  P.  Wollants,  B.  Blanpain and  F.  Haers: Steel Res., 71 (2000), 391. https://doi.org/10.1002/srin.200001334
• 22)   H.  Yin and  L.  Laus: AISTech 2018 Proceedings - Iron and Steel Technology Conference and Exposition, Iron and Steel Society, Philadelphia, PA, (2018), 1849.
• 23)   L.  Cheng,  L.  Zhang,  Y.  Ren and  W.  Yang: Metall. Mater. Trans. B, 52 (2021), 1186. https://doi.org/10.1007/s11663-021-02110-z
• 24)   A. R.  McKague,  R.  Engel,  M. A.  Suer,  D. J.  Wolf,  M. D.  Garbowsky and  L. F.  Reed: Iron Steelmaker, 25 (1998), 35.
• 25)   Y.  Guo,  Q.  Wang,  G.  Chen and  S.  He: J. Iron Steel Res. Int., 22 (2015), 87. https://doi.org/10.1016/S1006-706X(15)30144-8
• 26)   N. U.  Girase,  S.  Basu and  S. K.  Choudhary: Ironmak. Steelmak., 34 (2007), 506. https://doi.org/10.1179/174328107X168075
• 27)   N.  Kojola,  S.  Ekerot and  P. G.  Jönsson: Ironmak. Steelmak., 38 (2011), 81. https://doi.org/10.1179/030192310X12706364542704
• 28)   E.  Roos,  A.  Karasev and  P. G.  Jönsson: Steel Res. Int., 86 (2015), 1279. https://doi.org/10.1002/srin.201400344
• 29)   S.  Basu,  S. K.  Choudhary and  N. U.  Girase: ISIJ Int., 44 (2004), 1653. https://doi.org/10.2355/isijinternational.44.1653
• 30)   B.  Harcsik and  G.  Karoly: Steel Res. Int., 84 (2013), 129. https://doi.org/10.1002/srin.201200084
• 31)   J.  Ikaheimonen,  K.  Leiviska,  J.  Ruuska and  J.  Matkala: IFAC Proceedings Volumes, 35 (2002), 143. https://doi.org/10.3182/20020721-6-ES-1901.01175
• 32)   W.-G.  Jung,  O. D.  Kwon and  M.-K.  Cho: Korean J. Mater. Res., 19 (2009), 473. https://doi.org/10.3740/MRSK.2009.19.9.473
• 33)   G. C.  Duderstadt,  R. K.  Iyengar and  J. M.  Matesa: JOM, 20 (1968), 89. https://doi.org/10.1007/BF03378699
• 34)   T.  Wala and  A.  Pawelek: Interceram, 50 (2001), 14.
• 35)   S. K.  Choudhary and  A. J.  Khan: Steel Times Int., 24 (2000), No.3, 21.
• 36)   C.  Toulouse,  A.  Pack,  A.  Ender and  S.  Petry: Steel Res. Int., 79 (2008), 149. https://doi.org/10.1002/srin.200806330
• 37)   C.  Bernhard,  G.  Xia,  M.  Egger,  A.  Pissenberger and  S. K.  Michelic: Proceedings of AISTech 2012, Iron and Steel Society, Atlanta, GA, (2012), 2191.
• 38)   F.  Tehovnik,  J.  Burja,  B.  Arh and  M.  Knap: Metalurgija, 54 (2015), 371.
• 39)   H.  Barati,  M.  Wu,  A.  Kharicha and  A.  Ludwig: Powder Technol., 329 (2018), 181. https://doi.org/10.1016/j.powtec.2018.01.053
• 40)   J.  Li,  G.  Cheng,  Q.  Ruan,  J.  Pan and  X.  Chen: Metall. Mater. Trans. B, 50 (2019), 2769. https://doi.org/10.1007/s11663-019-01708-8
• 41)   P.  Dorrer,  S. K.  Michelic,  C.  Bernhard,  A.  Penz and  R.  Rössler: Steel Res. Int., 90 (2019), 1800635. https://doi.org/10.1002/srin.201800635
• 42)   H.  Barati,  M.  Wu,  A.  Kharicha and  A.  Ludwig: Steel Res. Int., 91 (2020), 2000230. https://doi.org/10.1002/srin.202000230
• 43)   A.  Memarpour,  V.  Brabie and  P.  Jönsson: Ironmak. Steelmak., 38 (2011), 229. https://doi.org/10.1179/1743281210Y.0000000005
• 44)   C.  Tian,  J.  Yu,  E.  Jin,  T.  Wen,  D.  Jia,  Z.  Liu,  P.  Fu and  L.  Yuan: J. Alloys Compd., 792 (2019), 1. https://doi.org/10.1016/j.jallcom.2019.04.010
• 45)   C.  Tian,  L.  Yuan,  J.  Yu,  D.  Jia,  X.  Yao and  T.  Wen: Metall. Mater. Trans. B, 51 (2020), 2095. https://doi.org/10.1007/s11663-020-01926-5
• 46)   Y.  Fukuda,  Y.  Ueshima and  S.  Mizoguchi: ISIJ Int., 32 (1992), 164. https://doi.org/10.2355/isijinternational.32.164
• 47)   Y.  Gao and  K.  Sorimachi: ISIJ Int., 33 (1993), 291. https://doi.org/10.2355/isijinternational.33.291
• 48)   K.  Sasai and  Y.  Mizukami: ISIJ Int., 34 (1994), 802. https://doi.org/10.2355/isijinternational.34.802
• 49)   K.  Sasai and  Y.  Mizukami: ISIJ Int., 41 (2001), 1331. https://doi.org/10.2355/isijinternational.41.1331
• 50)   R.  Dekkers,  B.  Blanpain,  P.  Wollants,  F.  Haers,  B.  Gommers and  C.  Vercruyssen: Steel Res. Int., 74 (2003), 351. https://doi.org/10.1002/srin.200300197
• 51)   Z.  Deng,  M.  Zhu,  Y.  Zhou and  D.  Sichen: Metall. Mater. Trans. B, 47 (2016), 2015. https://doi.org/10.1007/s11663-016-0624-y
• 52)   B.  Karnasiewicz and  E.  Zinngrebe: Metall. Mater. Trans. B, 50 (2019), 1704. https://doi.org/10.1007/s11663-019-01619-8
• 53)   J.  Lee,  M.  Kang,  S.  Kim and  Y.  Kang: ISIJ Int., 58 (2018), 1257. https://doi.org/10.2355/isijinternational.ISIJINT-2018-164
• 54)   J.  Lee,  M.  Kang,  S.  Kim,  J.  Kim,  M.  Kim and  Y.  Kang: ISIJ Int., 59 (2019), 749. https://doi.org/10.2355/isijinternational.ISIJINT-2018-672
• 55)   J.  Lee and  Y.  Kang: ISIJ Int., 60 (2020), 426. https://doi.org/10.2355/isijinternational.ISIJINT-2019-384
• 56)   B.  Karnasiewicz,  E.  Zinngrebe and  W.  Tiekink: Metall. Mater. Trans. B, 52 (2021), 2171. https://doi.org/10.1007/s11663-021-02189-4
• 57)   W.  Liang,  J.  Li,  B.  Lu,  J.  Zhi,  S.  Zhang and  Y.  Liu: J. Iron Steel Res. Int., 29 (2021), 34. https://doi.org/10.1007/s42243-021-00637-8
• 58)   H.  Barati,  M.  Wu,  S.  Michelic,  S.  Ilie,  A.  Kharicha,  A.  Ludwig and  Y.  Kang: Metall. Mater. Trans. B, 52 (2021), 4167. https://doi.org/10.1007/s11663-021-02336-x
• 59)   C.  Tian,  L.  Yuan,  Y.  Li,  D.  Jia,  T.  Wen,  G.  Liu,  J.  Yu and  H.  Li: Ceram. Int., 47 (2021), 22646. https://doi.org/10.1016/j.ceramint.2021.04.278
• 60)   U. D.  Salgado,  C.  Weiß,  S. K.  Michelic and  C.  Bernhard: Metall. Mater. Trans. B, 49 (2018), 1632. https://doi.org/10.1007/s11663-018-1271-2
• 61)   F. G.  Wilson,  M. J.  Heesom,  A.  Nicholson and  A. W. D.  Hills.: Ironmak. Steelmak., 14 (1987), 296.
• 62)   N.  Kojola,  S.  Ekerot,  M.  Andersson and  P. G.  Jönsson: Ironmak. Steelmak., 38 (2011), 1. https://doi.org/10.1179/030192310X12690127076398
• 63)   S.  Ogibayashi: Taikabutsu Overseas, 15 (1995), 3.
• 64)   H.  Duan,  Y.  Ren,  B. G.  Thomas and  L.  Zhang: Metall. Mater. Trans. B, 50 (2019), 36. https://doi.org/10.1007/s11663-018-1478-2
• 65)   H.  Duan,  Y.  Ren and  L.  Zhang: Chem. Eng. Sci., 196 (2019), 14. https://doi.org/10.1016/j.ces.2018.12.008
• 66)   K.  Uemura,  M.  Takahashi,  S.  Koyama and  M.  Nitta: ISIJ Int., 32 (1992), 150. https://doi.org/10.2355/isijinternational.32.150
• 67)   K.  Ogino,  K.  Nogi and  O.  Yamase: Trans. Iron Steel Inst. Jpn., 23 (1983), 234. https://doi.org/10.2355/isijinternational1966.23.234
• 68)   M.  Nakamoto,  T.  Tanaka,  M.  Suzuki,  K.  Taguchi,  Y.  Tsukaguchi and  T.  Yamamoto: ISIJ Int., 54 (2014), 1195. https://doi.org/10.2355/isijinternational.54.1195
• 69)   C.  Bernhard,  G.  Xia,  A.  Karasangabo,  M.  Egger and  A.  Pissenberger: Proceedings of 7th European Continuous Casting Conference, (Dusseldorf), (2011), paper-17.
• 70)   M. G.  González-Solórzano,  R. D.  Morales,  E.  Gutiérrez,  J.  Guarneros and  K.  Chattopadhyay: Steel Res. Int., 91 (2020), 2000049. https://doi.org/10.1002/srin.202000049
• 71)   C.  Xuan,  H.  Shibata,  S.  Sukenaga,  P. G.  Jönsson and  K.  Nakajima: ISIJ Int., 55 (2015), 1882. https://doi.org/10.2355/isijinternational.ISIJINT-2014-820
• 72)   K.  Chen,  D.  Wang,  D.  Hou,  T.  Qu,  J.  Tian and  H.  Wang: ISIJ Int., 59 (2019), 1735. https://doi.org/10.2355/isijinternational.ISIJINT-2019-053
• 73)   E.  Gao,  G.  Zou,  W.  Wang,  F.  Ma and  X.  Luo: Metall. Mater. Trans. B, 48 (2016), 1014. https://doi.org/10.1007/s11663-016-0898-0
• 74)   U. D.  Salgado,  P.  Dorrer,  S. K.  Michelic and  C.  Bernhard: J. Mater. Eng. Perform., 27 (2018), 4983. https://doi.org/10.1007/s11665-018-3468-6
• 75)   T.  Qu,  D.  Wang,  H.  Wang,  D.  Hou,  J.  Tian,  S.  Hu and  L.  Su: J. Iron Steel Res. Int., 28 (2021), 1149. https://doi.org/10.1007/s42243-020-00546-2
• 76)   A. W.  Cramb and  I.  Jimbo: Iron Steelmaker, 16 (1989), 43.
• 77)   K.  Nakashima,  K.  Takihira,  T.  Miyazaki and  K.  Mori: J. Am. Ceram. Soc., 76 (1993), 3000. https://doi.org/10.1111/j.1151-2916.1993.tb06601.x
• 78)   H.  Shibata,  Y.  Watanabe,  K.  Nakajima and  S.  Kitamura: ISIJ Int., 49 (2009), 985. https://doi.org/10.2355/isijinternational.49.985
• 79)   N. E. N.  Shinozaki,  K.  Mukai,  Y.  Takahashi and  Y.  Tanaka: Trans. Iron Steel Inst. Jpn., 80 (1994), 14. https://doi.org/10.2355/tetsutohagane1955.80.10_748
• 80)   K.  Ogino,  A.  Adachi and  K.  Nogi: Trans. Iron Steel Inst. Jpn., 59 (1973), 1237.
• 81)   A. K.  Lahiri,  E.  Kapilashrami,  A. W.  Cramb and  S.  Seetharaman: Metall. Mater. Trans. B, 34 (2003), 647. https://doi.org/10.1007/s11663-003-0035-8
• 82)   J.  Lee and  K.  Morita: ISIJ Int., 44 (2004), 235. https://doi.org/10.2355/isijinternational.44.235
• 83)   T.  Yoshikawa,  K.  Motosugi,  T.  Tanaka and  M.  Ueda: Trans. Iron Steel Inst. Jpn., 97 (2011), 361. https://doi.org/10.2355/tetsutohagane.97.361
• 84)   J.  Michael Humenik and  W. D.  Kingery: J. Am. Ceram. Soc., 37 (1954), 18. https://doi.org/10.1111/j.1151-2916.1954.tb13972.x
• 85)   K.  Ogino and  K.  Nogi: Can. Metall. Q., 22 (1982), 19. https://doi.org/10.1179/cmq.1983.22.1.19
• 86)   E. P.  Heikkinen,  T.  Kokkonen,  R.  Mattila and  T.  Fabritius: Steel Res. Int., 81 (2010), 1070. https://doi.org/10.1002/srin.201000105
• 87)   N.  Fukami,  R.  Wakamatsu,  N.  Shinozaki and  K.  Wasai: Mater. Trans., 50 (2009), 2552. https://doi.org/10.2320/matertrans.M2009204
• 88)   V.  Sahajwalla,  S.  Seetharaman and  E.  Kapilashrami: ISIJ Int., 44 (2004), 653. https://doi.org/10.2355/isijinternational.44.653
• 89)   Z.  Li,  M.  Zeze and  K.  Mukai: Mater. Trans., 43 (2002), 1724. https://doi.org/10.2320/MATERTRANS.43.1724
• 90)   J.  Lee,  J.  Park and  M.  Shin: ISIJ Int., 48 (2008), 1665. https://doi.org/10.2355/isijinternational.48.1665
• 91)   T.  Furukawa,  Z.  Zhang,  T.  Hirosumi,  N.  Saito and  K.  Nakashima: ISIJ Int., 62 (2022), 1352. https://doi.org/10.2355/isijinternational.ISIJINT-2022-093
• 92)   P. J. Y.  Rubio,  L.  Hong,  N.  Saha-Chaudhury,  R.  Bush and  V.  Sahajwalla: ISIJ Int., 46 (2006), 1570. https://doi.org/10.2355/isijinternational.46.1570
• 93)   C. J.  Xuan,  Z.  Zhao and  P. G.  Jönsson: J. Eur. Ceram. Soc., 36 (2016), 2435. https://doi.org/10.1016/j.jeurceramsoc.2016.03.006
• 94)   T.  Matsushita,  Z. T.  Zhang,  S.  Seetharaman and  W.  Li: Metall. Mater. Trans. B, 37 (2006), 421. https://doi.org/10.1007/s11663-006-0027-6
• 95)   R.  Khanna,  M.  Ikram-Ul-Haq,  P.  Koshy and  V.  Sahajwalla: ISIJ Int., 50 (2010), 804. https://doi.org/10.2355/isijinternational.50.804
• 96)   R.  Khanna,  S.  Kongkarat,  S.  Seetharaman and  V.  Sahajwalla: ISIJ Int., 52 (2012), 992. https://doi.org/10.2355/isijinternational.52.992
• 97)   M.  Ikram-Ul-Haq,  R.  Khanna,  S.  Kongkarat and  V.  Sahajwalla: ISIJ Int., 52 (2012), 1801. https://doi.org/10.2355/isijinternational.52.1801
• 98)   X.  Wei,  A.  Yehorov,  E.  Storti,  S.  Dudczig,  O.  Fabrichnaya,  C. G.  Aneziris and  O.  Volkova: Adv. Eng. Mater., 24 (2022), 2100718. https://doi.org/10.1002/adem.202100718
• 99)   J.  Lee and  Y.  Kang: ISIJ Int., 60 (2020), 258. https://doi.org/10.2355/isijinternational.ISIJINT-2019-355
• 100)   R.  Huang,  L.  Zhang,  R.  Jiang,  Y.  Guo,  Q.  Zhou and  H.  Duan: Iron Steel, 56 (2021), 43 (in Chinese). https://doi.org/10.13228/j.boyuan.issn0449-749x.20200222
• 101)   M.  Suzuki,  Y.  Yamaoka,  N.  Kubo and  M.  Suzuki: ISIJ Int., 42 (2002), 248. https://doi.org/10.2355/isijinternational.42.248
• 102)   J.  Ikäheimonen,  K.  Leiviskä,  J.  Ruuska and  J.  Matkala: IFAC Proceedings Volumes, 35 (2002), 143. https://doi.org/10.3182/20020721-6-es-1901.01175
• 103)   M.  Long,  X.  Zuo,  L.  Zhang and  D.  Chen: ISIJ Int., 50 (2010), 712. https://doi.org/10.2355/isijinternational.50.712
• 104)   P.  Ni,  L.  Tord,  I.  Jonsson,  M.  Ersson and  P. G.  Jonsson: Metall. Mater. Trans. B, 45 (2014), 2414. https://doi.org/10.1007/s11663-014-0145-5
• 105)   E.  Gutiérrez,  S.  Garcia-Hernandez and  J. D. J.  Barreto: ISIJ Int., 56 (2016), 1394. https://doi.org/10.2355/isijinternational.ISIJINT-2016-076
• 106)   M.  Mohammadi-Ghaleni,  M. A.  Zaeem,  J. D.  Smith and  R.  O’Malley: Metall. Mater. Trans. B, 47 (2016), 3384. https://doi.org/10.1007/s11663-016-0798-3
• 107)   W.  Lee,  J. G.  Kim,  J.  Jung and  K. Y.  Huh: Steel Res. Int., 92 (2021), 2000549. https://doi.org/10.1002/srin.202000549
• 108)   R.  Wang,  H.  Li,  F.  Guerra,  C.  Cathcart and  K.  Chattopadhyay: ISIJ Int., 62 (2022), 2311. https://doi.org/10.2355/isijinternational.ISIJINT-2022-181
• 109)   K.  Ashok,  J.  Venkatesan and  P. G.  Venkatakrishnan: Adv. Mater. Res., 585 (2012), 364. https://doi.org/10.4028/www.scientific.net/AMR.585.364
• 110)   S.  Zhou: Metallurgist, 57 (2013), 510. https://doi.org/10.1007/s11015-013-9761-6
• 111)   Y. T.  Guo,  S. P.  He,  G. J.  Chen,  S.  Zhao,  Z. R.  Li,  T.  Wu and  Q.  Wang: Ironmak. Steelmak., 41 (2014), 738. https://doi.org/10.1179/1743281214Y.0000000188
• 112)   P. B. P.  Leão,  J. L.  Klug,  C. A. R.  Carneiro,  H.  Caldas and  W. V.  Bielefeldt: Steel Res. Int., 90 (2019), 1900151. https://doi.org/10.1002/srin.201900151
• 113)   W.  Li,  Y.  Wang,  W.  Wang,  Y.  Ren and  L.  Zhang: Metals, 10 (2020), 1205. https://doi.org/10.3390/met10091205
• 114)   R.  Sambasivam: Ironmak. Steelmak., 33 (2006), 439. https://doi.org/10.1179/174328106X118198
• 115)   H.  Niitsuma,  M.  Ogata,  I.  Mclennan,  P.  Cassar and  M.  Camplin: Shinagawa Tech. Rep., 55 (2012), 59 (in Janpanese).
• 116)   E.  Gutiérrez,  J. D. J.  Barreto,  S.  Garcia-Hernandez,  R.  Morales and  M. G.  González-Solorzano: Metals, 10 (2020), 1420. https://doi.org/10.3390/met10111420
• 117)   L.  Trueba,  K. D.  Peaslee,  J. D.  Smith and  M.  Karakus: Steel Res. Int., 77 (2006), 37. https://doi.org/10.1002/srin.200606128
• 118)   Y.  Vermeulen,  B.  Coletti,  B.  Blanpain,  P.  Wollants and  J.  Vleugels: ISIJ Int., 42 (2002), 1234. https://doi.org/10.2355/isijinternational.42.1234
• 119)   S.  Devic,  R.  Galjak and  T.  Bradaric: Ironmak. Steelmak., 30 (2003), 57. https://doi.org/10.1179/030192303225009533
• 120)   K.  Takei: Bachelor Thesis, Royal Institute of Technology, (2012).
• 121)   J. K. S.  Svensson,  A.  Memarpour,  S.  Ekerot,  V.  Brabie and  P. G.  Jönsson: Ironmak. Steelmak., 44 (2017), 117. https://doi.org/10.1179/1743281215Y.0000000065
• 122)   J. K. S.  Svensson,  F.  Larsson,  A.  Memarpour,  S.  Ekerot,  V.  Brabie and  P. G.  Jönsson: Ironmak. Steelmak., 45 (2018), 105. https://doi.org/10.1080/03019233.2016.1245916
• 123)   Y.  Tsukaguchi,  T.  Kato,  S.  Watanabe,  S.  Ooga and  T.  Tanaka: Bull. Jpn. Inst. Met., 50 (2011), 27. https://doi.org/10.2320/materia.50.27
• 124)   X.  Yang,  Y.  Zhang,  W.  Zhan,  F.  Mao,  J.  Yu and  Z.  He: J. Aust. Ceram. Soc., 57 (2021), 1255. https://doi.org/10.1007/s41779-021-00624-y
• 125)   J.  Yu,  X.  Yang,  Z.  Liu,  X.  Hou and  Z.  Yin: Ceram. Int., 43 (2017), 13025. https://doi.org/10.1016/j.ceramint.2017.06.133
• 126)   C.  Tian,  L.  Yuan,  J.  Yu,  D.  Jia,  T.  Wen,  G.  Liu and  H.  Li: J. Mater. Res. Technol., 11 (2021), 404. https://doi.org/10.1016/j.jmrt.2021.01.040
• 127)   X.  Yang,  Z.  Liu and  J.  Yu: J. Mater. Process. Technol., 259 (2018), 341. https://doi.org/10.1016/j.jmatprotec.2018.05.008
• 128)   L.  Yan,  L.  Chen,  C.  Liu,  C.  Liu,  S.  Pang,  X.  Zhang and  L.  Zhang: JOM, 73 (2021), 3910. https://doi.org/10.1007/s11837-021-04869-7
• 129)   B. H.  Reis,  W. V.  Bielefeldt and  A. C. F.  Vilela: ISIJ Int., 54 (2014), 1584. https://doi.org/10.2355/isijinternational.54.1584
• 130)   M.  Valdez,  G. S.  Shannon and  S.  Sridhar: ISIJ Int., 46 (2006), 450. https://doi.org/10.2355/isijinternational.46.450
• 131)   R.  Wang,  Y.  Li,  D.  Li,  Y.  Kang,  Y.  Bao and  Z.  Yan: Steel Res. Int., 91 (2020), 1900440. https://doi.org/10.1002/srin.201900440
• 132)   C.  Ren,  L.  Zhang,  J.  Zhang,  S.  Wu,  P.  Zhu and  Y.  Ren: Metall. Mater. Trans. B, 52 (2021), 3288. https://doi.org/10.1007/s11663-021-02256-w
• 133)   Y.  Ren and  L.  Zhang. ISIJ Int., 57 (2016), 68. https://doi.org/10.2355/isijinternational.ISIJINT-2016-509
• 134)   V. C. D.  Rocha,  J. A. M.  Pereira,  A.  Yoshioka,  W. V.  Bielefeldt and  A. C. F.  Vilela: Metall. Mater. Trans. B, 48 (2017), 1423. https://doi.org/10.1007/s11663-017-0935-7
• 135)   J. H.  Shin and  J. H.  Park: ISIJ Int., 58 (2018), 88. https://doi.org/10.2355/isijinternational.ISIJINT-2017-456
• 136)   Y.  Ren,  L.  Zhang and  S.  Li: ISIJ Int., 54 (2014), 2772. https://doi.org/10.2355/isijinternational.54.2772
• 137)   N.  Verma,  P. C.  Pistorius,  R. J.  Fruehan,  M. S.  Potter,  H. G.  Oltmann and  E. B.  Pretorius: Metall. Mater. Trans. B, 43 (2012), 830. https://doi.org/10.1007/s11663-012-9660-4
• 138)   C.  Liu,  Y.  Kacar,  B.  Webler and  P. C.  Pistorius. Metall. Mater. Trans. B, 52 (2021), 2837. https://doi.org/10.1007/s11663-021-02287-3
• 139)   N.  Verma,  P.  Pistorius,  R.  Fruehan,  M.  Potter,  M.  Lind and  S.  Story: Metall. Mater. Trans. B, 42 (2011), 711. https://doi.org/10.1007/s11663-011-9516-3
• 140)   N.  Verma,  P.  Pistorius,  R.  Fruehan,  M.  Potter,  M.  Lind and  S.  Story: Metall. Mater. Trans. B, 42 (2011), 720. https://doi.org/10.1007/s11663-011-9517-2
• 141)   M.  Li,  S.  Li,  Y.  Ren,  W.  Yang and  L.  Zhang: Ironmak. Steelmak., 47 (2020), 6. https://doi.org/10.1080/03019233.2018.1521566
• 142)   G.  Ye,  P.  Jonsson and  T.  Lund: ISIJ Int., 36 (1996), S105. https://doi.org/10.2355/isijinternational.36.Suppl_S105
• 143)   Y.  Ren,  Y.  Zhang and  L.  Zhang: Ironmak. Steelmak., 44 (2017), 497. https://doi.org/10.1080/03019233.2016.1216632
• 144)   L.  Zhang,  Y.  Liu,  Y.  Zhang,  W.  Yang and  W.  Chen: Metall. Mater. Trans. B, 49 (2018), 1841. https://doi.org/10.1007/s11663-018-1289-5
• 145)   Y.  Liu,  L.  Zhang,  Y.  Zhang,  H.  Duan,  Y.  Ren and  W.  Yang: Metall. Mater. Trans. B, 49 (2018), 610. https://doi.org/10.1007/s11663-018-1179-x
• 146)   L.  Zhang,  F.  Li and  W.  Fang: Steelmaking, 32 (2016), No.2, 1 (in Chinese).
• 147)   L.  Holappa,  M.  Hamalainen,  M.  Liukkonen and  M.  Lind: Ironmak. Steelmak., 30 (2003), 111. https://doi.org/10.1179/030192303225001748
• 148)   S.  Wu,  Y.  Zhang,  W.  Yang and  L.  Zhang: Steel Res. Int., 93 (2022), 2200264. https://doi.org/10.1002/srin.202200264
• 149)   W.  Wang,  L.  Zhang,  Y.  Ren,  Y.  Luo,  X.  Sun and  W.  Yang. Metall. Mater. Trans. B, 53 (2022), 1. https://doi.org/10.1007/s11663-021-02363-8
• 150)   M.  Nakamura,  S.  Inoue and  T.  Horiuchi: Shinagawa Tech. Rep., 49 (2006), 59 (in Japanese).
• 151)   J.  Chen,  Z.  Su,  D.  Li,  Q.  Wang and  J.  He: ISIJ Int., 58 (2018), 1242. https://doi.org/10.2355/isijinternational.ISIJINT-2018-095
• 152)   Y.  Tsukaguchi,  O.  Nakamura,  S.  Yokoya and  S.  Takagi: ISIJ Int., 91 (2005), 581. https://doi.org/10.2355/tetsutohagane1955.91.7_581
• 153)   X.  Zhu,  D.  Li,  C.  Wu,  K.  Marukawa and  Q.  Wang: Acta Metall. Sin. (Engl. Lett.), 31 (2018), 1317. https://doi.org/10.1007/s40195-018-0808-3
• 154)   N.  Hiroki,  A.  Takahashi,  Y.  Nanba,  N.  Tsukamoto,  Y.  Kurashina and  K.  Yanagawa: Shinagawa Tech. Rep., 36 (1993), 75 (in Japanese).
• 155)   S.  Yokoya,  S.  Takagi,  S.  Ootani,  M.  Iguchi,  K.  Marukawa and  S.  Hara: ISIJ Int., 41 (2001), 1208. https://doi.org/10.2355/isijinternational.41.1208
• 156)   S.  Yokoya,  S.  Takagi,  M.  Iguchi,  Y.  Asako,  R.  Westoff and  S.  Hara: ISIJ Int., 38 (1998), 827. https://doi.org/10.2355/isijinternational.38.827
• 157)   S.  Yokoya,  Y.  Asako,  S.  Hara and  J.  Szekely: ISIJ Int., 34 (1994), 883. https://doi.org/10.2355/isijinternational.34.883
• 158)   E.  Luhrsen,  A.  Ott,  W.  Parbel,  J.  Piret,  R.  Ruddlestone and  B.  Short: Proceedings of the 1st European Conference on Continuous Casting, Florence, 1 (1991), 1.
• 159)   P. M.  Benson,  Q. K.  Robinson and  H. K.  Park: Proceedings of the 76th Steelmaking Conference, Iron and Steel Society, Dallas, TX, (1993), 533.
• 160)   R. B.  Tuttle,  J. D.  Smith and  K. D.  Peaslee: Metall. Mater. Trans. B, 38 (2007), 101. https://doi.org/10.1007/s11663-006-0001-3
• 161)   R. B.  Tuttle,  J. D.  Smith and  K. D  Peaslee: Metall. Mater. Trans. B, 36 (2005), 885. https://doi.org/10.1007/s11663-005-0090-4
• 162)   R.  Tsujino,  A.  Tanaka,  A.  Imamura,  D.  Takahashi and  S.  Mizoguchi: ISIJ Int., 34 (1994), 853. https://doi.org/10.2355/isijinternational.34.853
• 163)   Y. H.  Paik,  J. H.  Pan and  W. J.  Yoon. J. Colloid Interface Sci., 244 (2001), 444. https://doi.org/10.1006/jcis.2001.7966
• 164)   Y. H.  Paik,  W. J.  Yoon and  H. C.  Shin. J. Colloid Interface Sci., 269 (2004), 354. https://doi.org/10.1016/S0021-9797(03)00629-5
• 165)   Y. H.  Paik,  H. C.  Shin and  J. M.  Lee: Met. Mater., 4 (1998), 995. https://doi.org/10.1007/BF03025967
• 166)   X.  Yang,  J.  Yu,  Z.  Liu, X.i Hou and B. Ma: Ceram. Int., 43 (2017), 2881. https://doi.org/10.1016/j.ceramint.2016.11.012
• 167)   K. N.  Tu,  C. C.  Yeh,  C. Y.  Liu and  C.  Chen: Appl. Phys. Lett., 76 (2000), 988. https://doi.org/10.1063/1.125915
• 168)   L.  Yan,  L.  Chen,  C.  Liu and  X.  Zhang: Metall. Mater. Trans. B, 52 (2021), 1603. https://doi.org/10.1007/s11663-021-02127-4