Element Transfer Behavior for CaF₂-Na₂O-SiO₂ Agglomerated Flux Subject in Submerged Arc Welding Process

Abstract

From a thermodynamic perspective, the present study has been performed to investigate the effect of SiO₂ level in agglomerated fluxes on the element transfer behavior of essential elements, by applying CaF₂-Na₂O-SiO₂ agglomerated fluxes with varying SiO₂ contents. Element transfer behavior is quantified by the Δ value. The impact of SiO₂ and heat input upon element transfer behavior is interpreted. Additionally, a possible thermodynamic approach to predict high basicity flux O potential and weld metal composition is proposed and evaluated. It is revealed that the consideration of the gas-slag-metal equilibrium is able to place constraints on the transfer behaviors (of O, Si, and Mn) and formation of gases. In submerged arc welding metallurgy, the empirically determined basicity index models proposed by Tuliani have been applied for more than 50 years to predict flux O potential and weld metal oxygen content. However, it is well known by welding practitioners that the flux basicity index model can only predict the changing trend of flux O potential when the flux basicity index is lower than 2.0. The present study has proposed a new thermodynamic method to identify the flux O potential for fluxes with a basicity index higher than 2.0. Additionally, the experimental evidence for the Mitra kinetic model has been provided.

1. Introduction

Submerged arc welding (SAW), during which the arc plasma and weld pool are shielded beneath the flux cover and molten slag, is one of the most widely applied welding techniques [1,2]. Fine-tuning weld metal (WM) compositions is essential to achieve the desired microstructures and mechanical properties, and efforts worldwide have been focused on the control mechanisms of flux on WM composition subject to SAW [3].

SAW fluxes are produced as fused or agglomerated ones [3,4,5]. Fused fluxes are produced by mixing dry powders and melting them at temperatures of 1500 to 1700 °C; then, the molten flux is chilled by cold water [1,3,6]. Agglomerated fluxes are produced by dry mixing powders and bonding them with additions of potassium silicate or sodium silicate solutions; the agglomerated mixture is then pelletized, dried at low temperatures (lower than 750 °C), broken up, and screened [1,3,6]. Recently, agglomerated fluxes have been widely applied due to lower energy consumption during flux manufacture, in comparison to the fused ones [5,7,8,9].

When fluxes are designed, SiO₂, acting as a network-builder, is added indispensably to improve slag detachability [10,11]. Therefore, an understanding of the control mechanism of SiO₂ in element transfer behavior in the SAW process is essential to ensure a sound weldment. Tremendous focus has been placed on this subject. For instance, Chai et al. [12] investigated the improvement of flux O potential due to SiO₂ addition into fluxes. Zhang et al. [11] designed a series of fused CaF₂-SiO₂ binary fluxes, and quantified the element transfer behavior of O, Si, and Mn between the flux and WM under high heat input SAW. Burck et al. [13], on the other hand, evaluated the effect of SiO₂ addition level on element transfer behavior when manganese-silicate fluxes were employed.

O is the most essential element for submerged arc welded welding that must be carefully controlled [14,15]. It is widely accepted that excessive O may cause several problems, such as promoting porosity, reducing toughness, and decreasing hardenability, and WMs with too low O levels show poor impact toughness since there are insufficient inclusions to promote the formation of acicular ferrite (AF, a structure that exhibits a good combination of strength and toughness) [14,15]. Si is one of the essential elements in WMs that should be carefully controlled, since excessive Si levels (higher than 0.6 wt pct) tend to reduce both elongation and toughness of the weldment [9,16,17]. It is well established that the transfer of Si is governed by slag-metal reaction in SAW [12,16,18,19]. Mn is an AF-promoting agent, and its level, ranging from 0.6 to 1.8 wt pct, can increase AF fraction and depress the formation of polygonal and side plate ferrites [20,21].

During SAW, due to complicated chemical reactions between slag (flux), weld pool, and arc plasma, SiO₂ in agglomerated fluxes tends to improve flux O potential and influence WM composition [3]. However, although agglomerated fluxes have been extensively applied recently to lower energy consumption during manufacture, few works concern the effect of SiO₂ on element transfer behavior subject to SAW [9]. Additionally, the impact of SiO₂ on element transfer behavior under various heat inputs is not clarified.

To gain a better understanding with respect to WM compositional control when agglomerated SiO₂-bearing is applied, the present study is performed to investigate the effect of SiO₂ level in agglomerated fluxes on element transfer behavior, using CaF₂-Na₂O-SiO₂ agglomerated fluxes with varying SiO₂ contents under a heat input of 60 and 20 kJ/cm from a thermodynamic perspective. The fluxes are poised to be employed in low alloy grade steel base metals (BM) to minimize the impact of BM on WM composition [4,22,23].

This study focuses on the WM chemistries of O, Si, and Mn, since they are the essential elements pertaining to the mechanical performance of the weldments of low alloy grade steel [9,19]. For the CaF₂-Na₂O-SiO₂ agglomerated flux system, CaF₂ acts as a dilutant to reduce the melting temperature and O potential of the fluxes, and Na₂O is incorporated to improve the arc stability [3,12,24].

2. Materials and Methods

For each agglomerated flux recipe, reagent grade powders were dry mixed. Then, the powders were bonded by the sodium-silicate solution. The molecular formula of the sodium silicate is a glue-like material, which can be expressed as Na₂O·nSiO₂ with H₂O. The value of n for Na₂O·nSiO₂ is between 1.5 and 3.5. The bonded mixtures were pelletized and dried in the muffle furnace at 973 K for 3 h. In this way, the H₂O in sodium silicate was excluded, and the powders were bonded together. At last, the mixtures were broken up and screened to 14 to 100 mesh. X-ray fluorescence (XRF, model S4 Explorer, Bruker, Germany) was employed to determine the compositions of the fluxes. The analytical flux composition is given in

Table 1. Measured compositions of fluxes (weight percent).

Fluxes	CaF2	Na2O	SiO2	BI
F-1	88.77	1.21	10.02	8.98
F-2	78.87	1.24	19.89	4.03
F-3	68.67	1.23	30.10	2.32
F-4	58.62	1.19	40.19	1.49
F-5	48.84	1.22	49.94	1.00

.

As shown in Table 1, the compositional change in Na₂O is neglectable, and the SiO₂ level improves from 10 to 50 wt pct gradually, with CaF₂ making up the remainder. The basicity index (BI) of each flux is determined by Equation (1) (wt pct) [1,3,25].

$$BI=\frac{{\mathrm{CaO}+\mathrm{CaF}}_2{+\mathrm{MgO}+\mathrm{Na}}_2{\mathrm{O}+\mathrm{K}}_2\mathrm{O}+0.5\times (\mathrm{MnO}+\mathrm{FeO})}{{\mathrm{SiO}}_2{+1/2(\mathrm{Al}}_2{\mathrm{O}}_3{+\mathrm{Cr}}_2{\mathrm{O}}_3{+\mathrm{TiO}}_2{+\mathrm{ZrO}}_2)}$$

(1)

A typical low alloy grade steel, Q345A, was selected as the base metal (BM). Bead-on-plate double-electrodes single-pass SAW (Lincoln Electric Power Wave AC/DC 1000 SD, Lincoln Electric, Cleveland, OH, USA) was performed at the heat input of 60 kJ/cm (DC-850 A/32 V for electrode forward, AC-625 A/36 V for electrode backward, 500 mm/min) and 20 kJ/cm (DC-436 A/30 V, 393 mm/min), respectively.

Inductively coupled plasma optical emission spectrometry (ICP-OES, Perkin Elmer, Waltham, MA, USA) was used to determine the compositions of metallic elements (Si and Mn), while the LECO analyzer was used to determine the contents of O. Compositions of the electrode and BM are given in

Table 2. Measured chemical compositions of BM and electrode (weight percent).

	C	Si	Mn	Ti	Cr	O
Q345A	0.112	0.142	1.540	0.015	0.018	0.003
Electrode	0.127	0.049	1.650	0.015	0.015	0.003

.

A specific Δ value is employed to evaluate the element transfer between the flux and WM; a positive Δ value indicates an elemental gain in the WM from the flux, while a negative Δ value means an elemental loss from the WM to the flux. The magnitude of Δ means the amount of element transfer. The Δ values for O, Cr, and Mn are quantified from Equation (2).

$$\mathsf{\Delta }=M_{\mathrm{WM}}-M_{\mathrm{N}}$$

(2)

In Equation (2), the M_WM implies measured WM composition, while M_N implies the nominal composition (the composition considering only the dilution effect of the BM and electrode); the nominal composition is employed to exclude the compositional dilution effect of the BM and electrode. The M_N value is determined from Equation (3), where M_BM represents the measured composition of the BM, M_el represents the measured composition of the electrode, and d represents the dilution value of the BM.

$$M_{\mathrm{N}}=M_{\mathrm{BM}}\times d+M_{\mathrm{el}}\times (1-d)$$

(3)

Nominal compositions, measured WM composition, and Δ values for O, Si, and Mn elements are summarized in

Table 3. Nominal compositions, measured WM compositions, and quantified Δ values (weight percent).

Weld Metals	Heat Input	WM-1	WM-2	WM-3	WM-4	WM-5
Fluxes		F-1	F-2	F-3	F-4	F-5
(O)A	60 kJ/cm	0.015	0.018	0.021	0.024	0.027
(O)N		0.003	0.003	0.003	0.003	0.003
ΔO		0.012	0.015	0.018	0.021	0.024
(Si)A		0.455	0.561	0.656	1.027	1.252
(Si)N		0.088	0.089	0.095	0.097	0.101
ΔSi		0.367	0.472	0.561	0.930	1.151
(Mn)A		1.190	0.983	0.847	0.754	0.611
(Mn)N		1.604	1.603	1.596	1.593	1.588
ΔMn		−0.414	−0.620	−0.749	−0.839	−0.977
(O)A	20 kJ/cm	0.011	0.013	0.015	0.020	0.022
(O)N		0.003	0.003	0.003	0.003	0.003
ΔO		0.008	0.010	0.012	0.017	0.019
(Si)A		0.362	0.411	0.562	0.894	1.033
(Si)N		0.087	0.087	0.093	0.098	0.102
ΔSi		0.275	0.324	0.469	0.796	0.931
(Mn)A		1.313	1.159	0.945	0.847	0.744
(Mn)N		1.605	1.605	1.598	1.592	1.587
ΔMn		−0.292	−0.446	−0.653	−0.745	−0.843

.

It is technologically impossible to measure the compositions of the gases in the arc plasma, and to capture the molten slag for analytical purposes, since the arc plasma and molten slag are shielded under the granular fluxes; additionally, due to the very large temperature and density gradients, it is unlikely that equilibrium could be attained in SAW [1,9,26]. To solve these difficulties, one may assume that the high temperature and high ratio of surface to the volume would counteract the short time for the chemical reactions to be completed; as such, the equilibrium consideration can be employed to place limits on the chemical reactions and element transfer behavior involved in SAW [9,17,27,28].

In this study, gas-slag-metal equilibrium calculations were performed to aid in the discussion of the quantified data. Equilibrium calculations were performed at 2000 °C, since this temperature is well accepted for the SAW equilibrium [16,18,19]. This approach was employed in our previous study to illustrate the transfer of Ti and O between TiO₂-bearing basic-fluoride fluxes and WMs with varying TiO₂ levels [27]. The details of the thermodynamic calculations are given in Appendix A.

3. Results and Discussion

3.1. Transfer of O

The level of ΔO as a function of SiO₂ content in flux is plotted in Figure 1. As shown by Figure 1, under higher heat input of 60 kJ/cm, the ΔO value increases from 154 ppm to 274 ppm when the SiO₂ content increases from 10 to 50 wt pct; under lower heat input of 20 kJ/cm, the ΔO value increase from 110 ppm to 222 ppm when the SiO₂ content increases from 10 to 50 wt pct. For flux of the same SiO₂ level, an improvement in the ΔO value is observed for all fluxes, as shown by the red shaded area in Figure 1.

Wang et al. [9] and Chai et al. [12] concluded that SiO₂ in fluxes tends to decompose into suboxides in the presence of welding arc plasma via Reaction (4), thereby providing higher levels of O₂. Lau et al. [29] concluded that the flux O potential is governed by the level O₂ pressure in the arc plasma.

$${(\mathrm{SiO}}_2)=\mathrm{SiO}(\mathrm{g})+\frac{1}{2}{\mathrm{O}}_2\left(\mathrm{g}\right)$$

(4)

As is mentioned previously, although it is impossible to measure the compositions of the gases in the arc plasma, the equilibrium p_O2 can be employed to aid in the analysis [17,27,30]. As shown in

Table A1. Gas-slag-metal equilibrium gas components.

Component	Unit	F-1	F-2	F-3	F-4	F-5
CaF2	(Volume fraction)	3.55 × 100	3.23 × 100	2.84 × 100	2.35 × 100	1.77 × 100
SiF4		4.10 × 101	5.04 × 101	5.41 × 101	5.63 × 101	5.73 × 101
SiO		9.04 × 10−1	1.33 × 100	1.80 × 100	2.38 × 100	3.13 × 100
Mn		6.51 × 10−1	5.07 × 10−1	4.16 × 10−1	3.57 × 10−1	3.10 × 10−1
O2	(10−9 atm.)	1.80	2.13	2.26	2.32	2.40

, the equilibrium p_O2 generally improves from 1.8 to 2.4 × 10⁻⁹ atm, which is consistent with the assumption raised by Lau et al. [29] Additionally, it was revealed by Lau et al. [29,31] and Zhang et al. [4,11] that the decomposition of SiO₂ occurs to a larger extent (thereby higher level of p_O2) with a higher heat input. Therefore, it is postulated that the improvement of both SiO₂ addition and heat input drives Reaction [4] to the right side and increases the p_O2 in the arc plasma, leading to higher flux O potential.

It is noted the transfer of O is also influenced by Reaction (5) at the slag–metal interface. [11] To clarify the direction of Reaction [5], the quantified ΔSi value, which is discussed in part 2 (Transfer of Si), is as follows.

$${(\mathrm{SiO}}_2)=[\mathrm{Si}]+2[\mathrm{O}]$$

(5)

3.2. Transfer of Si

The quantified ΔSi value subject to various SiO₂ and heat input levels is plotted in Figure 2. The positive ΔSi value in Figure 2 indicates that Reaction (5) is driven to the right side, that is, Si and O are transferred from the flux to the WM at the slag–metal interface. From an observation from Figure 2, under the heat input of 60 kJ/cm, the ΔSi value increases from 0.367 to 1.151 wt pct with the improvement in the SiO₂ level from 10 to 50 pct; under the heat input of 20 kJ/cm, the ΔSi value increases from 0.275 to 0.931 wt pct with the improvement in the SiO₂ level. For flux of the same SiO₂ level, an improvement in ΔSi is observed, as shown by the red shaded area in Figure 2.

As is mentioned previously, the Si level in WMs should be restricted, since redundant Si tends to reduce elongation and toughness of the weldment. Therefore, from the perspective of the flux selection strategy, the SiO₂ content should be restricted, especially under high heat inputs.

3.3. Transfer of Mn

It was concluded that the transfer of Mn between the flux and WM is governed by Reaction (6) at the slag–metal interface [32,33,34].

$$\left(\mathrm{MnO}\right)=\left[\mathrm{Mn}\right]+\left[\mathrm{O}\right]$$

(6)

Since there is no MnO contained in the flux initially, it is speculated that Reaction (6) is driven to the left side due to the improvement in flux O potential via SiO₂ addition, which can be reflected by the negative ΔMn value plotted in Figure 3.

Figure 3 presents the ΔMn value as a function of SiO₂ content in the flux. It can be observed that the ΔMn value presents a decreasing trend when the SiO₂ content increases from 10 to 50 wt pct. This is expected, since the improvement of flux O potential drives Reaction [6] to the right side, resulting in a more negative ΔMn value [11]. Similarly, due to the increase in flux O potential under higher heat input (see Figure 1), a more negative ΔMn value is obtained, as shown by the blue shaded area [11].

3.4. Effect of Heat Input on WM Composition

Although it is well known that the heat input exerts a significant effect on WM composition, the role of SiO₂ in WM compositional control under different heat inputs is not clarified. As is discussed above, despite the departures from equilibrium, one may still employ equilibrium considerations to place constraints on the chemical reactions involved in welding. Due to the existence of the arc plasma, Zhang et al. [15] summarized the model of plasma (gas)-slag-metal (see the blue point Figure 4), based on which thermodynamic gas-slag-metal equilibrium is established [9,17,27,30]. The thermodynamic calculation procedure is given in Appendix A.

Mitra et al. [32,35] developed a model with respect to the WM compositional prediction in SAW. Mitra et al. [32,35] assumed that the SAW system would maintain a molten status for a longer period under a higher heat input; thus, there is more time for element transfer to take place, although no experiment data were provided to justify it. To verify the assumption of Mitra et al. [32,35], the equilibrium and measured compositions are plotted in and shown in Figure 5.

The equilibrium and measured contents of O, Si, and Mn are plotted in Figure 5a–c, respectively. As shown by the red and blue shaded areas in Figure 5, when the heat input improves from 20 to 60 kJ/cm, the measure contents of O, Si, and Mn are closer to the equilibrium ones.

As is mentioned previously, fluxes are the major sources of O for submerged arc welded metals. Therefore, the flux O potential can be evaluated by the O content in the WM. Herein, we employ the consideration of the gas-slag-metal equilibrium to explain the changing trend of O. As shown in Figure 5a, the flux O potential (WM O content) generally increases with higher SiO₂ levels. With higher heat input, an improvement in flux O potential is observed (see the blue shaded area in Figure 5a).

Similarly, the consideration of the gas-slag-metal equilibrium is able to constrain the changing ΔSi and Si content as a function of the flux formula. Since the nominal Si content nearly holds constant, the ΔSi and Si content follow the same changing trend with the SiO₂ content in the flux. For flux of the same SiO₂ level, an improvement in ΔSi and Si content is observed when the heat input improves from 20 to 60 kJ/cm, since both ΔSi and Si contents are closer to the equilibrium values. Similar trends have been observed in previous studies, but only qualitative analyses with respect to the slag-metal reaction have been considered [11].

This result indicates that, although the thermodynamic equilibrium is not achieved in the overall SAW process, the status of chemical reactions is closer to equilibrium under a higher heat input, which may provide experimental evidence for the Mitra prediction model for SAW.

3.5. Prediction of Flux O Potential

It is well known that welding fluxes are the primary sources of O for WMs. [9,24,31] An empirical flux BI has been proposed by Tuliani et al. [3,25] to predict O content in submerged arc welded metals, as is demonstrated in Equation (1), due to an incomplete understanding of the flux thermodynamic properties during the SAW process of high temperature. Based on the BI value, fluxes are classified into the following categories: acidic (BI < 1.0), neutral (1 < BI < 1.2), and basic (BI > 1.2), as is illustrated in Figure 6.

Based on the experimental data, a general relationship between WM O content and flux BI is summarized in Figure 6, which shows a fairly strong correlation between WM O content and flux BI. It can be observed from Figure 6 that the predicted O content in the WM decreases with increasing flux BI and then reaches a constant [25]. In other words, a disadvantage of the flux BI model lies in the fact that it fails to predict WM O content when flux BI is higher than 2.0, since the predicted O content holds constant, as shown by the red shaded area in Figure 6, although the flux with high BI is widely applied nowadays [1,3,30].

It should be noted that flux BI is an empirical model in nature, since there is no thermodynamic fundamental correlation between flux BI and O content of submerged arc welded metals [1,3,30].

Another approach to predict WM composition is to employ the assumed gas-slag-metal equilibrium consideration involved in SAW [9,30]. Herein, the O content in WMs is predicted by applying both flux BI and gas-slag-metal equilibrium models. With regard to the flux BI model, the O content is predicted using the flux BI value given in Table 1; with regard to the gas-slag-metal equilibrium calculation, the O content is predicted following the procedure given in Appendix A. The predicted O content has been plotted in Figure 7.

It can be observed from Figure 7 that, when the SiO₂ level is lower than 30 wt pct, the predicted O holds constant at 200 ppm; when the SiO₂ level is higher than 30 wt pct, the changing trend of O content with increasing SiO₂ level is predictable (see the red dotted line). However, when the gas-slag-metal equilibrium calculation is performed, the changing trend of O content is predictable for the total compositional range of the flux SiO₂ level.

Another important parameter to be predicted is the negative or positive Δ value, since it is pertinent to the flux design and matching of welding materials [4,11,17]. The predicted Δ values for O and Mn are calculated from Equation [7], where Δ_P implies the predicted value and M_WM implies the gas-slag-metal equilibrium WM compositions. The predicted and measured Δ values are collected in

Table 4. Nominal compositions, measured WM compositions, and predicted Δ values (weight percent).

Weld Metal	WM-1	WM-2	WM-3	WM-4	WM-5
Flux	F-1	F-2	F-3	F-4	F-5
ΔpO	0.022	0.024	0.025	0.026	0.026
ΔO60	0.012	0.015	0.018	0.021	0.024
ΔO20	0.008	0.010	0.012	0.017	0.019
ΔpSi	0.444	0.618	0.616	1.093	1.429
ΔSi60	0.367	0.472	0.561	0.930	1.151
ΔSi20	0.275	0.324	0.469	0.796	0.931
ΔpMn	−0.524	−0.753	−0.886	−0.973	−1.038
ΔMn60	−0.414	−0.620	−0.749	−0.839	−0.977
ΔMn20	−0.292	−0.446	−0.653	−0.745	−0.843

.

The data in Table 4 indicate that the application of the gas-slag-metal equilibrium calculation is capable of predicting the positive/negative Δ value, as well as the changing trend of the Δ value with the SiO₂ addition level.

$${\mathsf{\Delta }}_P=M_{\mathrm{E}}-M_{\mathrm{N}}$$

(7)

When CaF₂-SiO₂-bearing fluxes are applied, gases tend to form, which has been confirmed in our previous study, although no thermodynamic evidence is provided [4,11,22,23]. For instance, it is well known that fluoride gas, such as CaF₂ and SiF₄, tend to form due to the volatilization of CaF₂ and the chemical interaction between CaF₂ and SiO₂ (Reaction (8)) [2]. Additionally, due to the impact of the arc plasma, SiO₂ will decompose into SiO and O₂ gas via Reaction (4), and Mn will be lost from the weld pool via evaporation [2,4,11,30]. From an observation from Table A1, the generation of gases discussed above is predictable from the gas-slag-metal thermodynamic equilibrium calculation, which may provide thermodynamic evidence with respect to the phenomenon reported in our previous studies, when CaF₂-SiO₂-bearing fluxes are applied [4,11,23].

$${2(\mathrm{CaF}}_2{)+(\mathrm{SiO}}_2{)=2(\mathrm{CaO})+\mathrm{SiF}}_4(\mathrm{g})$$

(8)

4. Conclusions

In summary, from the perspective of thermodynamics, the transfer behaviors of O, Si, and Mn are quantified by the Δ value and interpreted subject to various SiO₂ and heat input levels, when CaF₂-Na₂O-SiO₂ agglomerated fluxes are applied. The main findings of this study can be summarized as follows.

• The levels of flux O potential, ΔSi and ΔMn values generally increase with higher SiO₂ content levels and heat inputs.
• Although the thermodynamic equilibrium is not achieved in the overall SAW process, the status of the chemical reactions is closer to equilibrium under higher heat inputs, which may provide experimental evidence for the Mitra prediction model subject to SAW.
• The consideration of the gas-slag-metal equilibrium is, albeit cautiously, able to place constraints on the transfer behaviors (of O, Si, and Mn) and formation of gases.
• In comparison to the traditional flux BI model, the gas-slag-metal equilibrium calculation is capable of predicting the changing trend of flux O potential, even when flux BI is higher than 2.0 for CaF₂-SiO₂-bearing fluxes, which may make up for the deficiency that the flux BI model can only predict the changing trend of flux O potential when flux BI is lower than 2.0.