Sage Journals: Discover world-class research

Abstract

Fluxes in submerged arc welding of high-strength low-alloy steels are not readily available, flux compositions are not clear and compositions are patented. This study aims at the design, development and optimization of flux for submerged arc welding of high-strength low-alloy steel. Extreme vertices design suggested by McLean and Anderson is used to formulate twenty-one fluxes to study the effect of flux constituents on tensile strength, percentage elongation, impact strength, diffusible hydrogen content and microstructure of the weld metal. Mathematical models for ultimate tensile strength, percentage elongation, impact strength and diffusible hydrogen content for welded specimens versus flux constituents have been developed. From the experiments, it is found that tensile strength and elongation are affected the most with Al₂O₃ content, whereas CaO and CaF₂ contents have significant effect on impact strength. Synergic binary effect of CaO·CaF₂, CaO·MgO and Al₂O₃·CaF₂ mixtures is more than other binary mixtures on the mechanical properties. Developed regression models have been checked for adequacy using t-test for regression coefficients and analysis of variance (F-test) for whole regression equation. Finally, optimum flux composition giving optimum mechanical and microstructural characteristics is suggested.

Keywords

High-strength low-alloy extreme vertices design synergism antisynergistic diffusible hydrogen content and acicular ferrite

Introduction

Among the arc welding processes, submerged arc welding (SAW) is preferred over other methods of welding because of its inherent qualities such as easy control of process variables, high quality, deep penetration, smooth finish, capability to weld thicker sections and prevention of atmospheric contamination of weld pool.^1–3 SAW of thick sections is generally associated with problems such as distortion and residual stresses, leading to cracking and hardening in heat-affected zone (HAZ) as well as in base metal, especially in high-strength steels.^4–6 But the most important problem in welding high-strength low-alloy (HSLA) steels is to prevent brittle fracture of welded joints due to increased strength of HSLA steels (with high carbon equivalent (CE)). Brittle fracture is caused by structural transformations in the welded joint and also by the embrittling effect of dissolved hydrogen⁷ and high value of CE (equation (1)) of the base metal.^8,9 To take full advantage of the developments in HSLA steel,^3,10 welding fluxes yielding welds with increased strength, toughness and minimizing diffusible hydrogen content must be developed¹¹

CE = % C + \frac{% Mn}{6} + \frac{(% Cr + % Mo + % V)}{5} + \frac{(% Ni + % Cu)}{15}

(1)

The main elements of submerged arc welding are SAW apparatus, electrode wire, base metal and flux. In SAW, flux plays an important role in the stability of the arc and mechanical and chemical properties of the final weld deposit, and the quality of the weld may also be affected by the care and handling of the flux.¹²

Many researchers have tried to study the effect of flux constituents and basicity index. Pandey et al.¹³ studied the influence of welding parameters and flux basicity index on the weld chemistry and element transfer and concluded that element transfer solely depends on the flux basicity index. Mechanical properties¹⁴ of weld metal depend on weld metal chemistry or element transfer and microstructure, which further depend on flux, operating variables and weld thermal history. Kanjilal et al.¹⁵ developed a model for acicular ferrite (AF) for C–Mn steel. It was found that MgO is most effective as an individual and CaO in interaction with other elements in controlling the amount of AF content in the weld metal. High-strength, quench and tempered (Q&T) plates having a yield strength of 670 MPa (CE: 0.6) are susceptible to a crack-sensitive microstructure and cold cracking during welding.⁸

It is clear from the literature that weld metal chemistry is mainly affected by transfer of elements to or from the slag, which further depends on welding consumables (mainly flux) and operating variables for the same joint design and edge preparation and weld thermal history. So element transfer to or from the slag solely depends on the overall flux composition and/or flux basicity index for a set of electrode wire and base metal.

Despite extensive research carried by a number of authors, hydrogen-assisted cold cracking remains a prevalent issue in higher strength steel welds.¹⁶ Terashima and Tsuboi¹⁷ reported tremendous reduction in diffusible hydrogen levels of weld metal with an increase in the flux basicity index from 0 to 3. Chew¹⁸ demonstrated a considerable reduction in weld metal hydrogen with an increase in the coating CaCO₃ content.

Possible methods to reduce hydrogen-assisted cracking/diffusible hydrogen content are discussed by Jindal et al.¹¹ Among these, design of flux is one of the methods to minimize diffusible hydrogen content.

Plessis et al.¹⁹ have suggested that the diffusible hydrogen content of weld metal can be reduced by increasing the flux basicity index of shielding metal arc welding (SMAW) electrodes. The results showed that the addition of 22% CaF₂ in the flux reduced the hydrogen content by about 30%, whereas the addition of more than 22% CaF₂ increased the hydrogen content. Since flux plays an important role in the formation of weld metal and its resulting properties,¹⁵ the objective of this study is to get optimum mechanical properties by design, preparation and development of SAW agglomerated flux.

In the present investigations, individual constituents and their binary and ternary mixtures have been studied with the aim to get optimum flux formulated by varying flux ingredients in CaO–MgO–CaF₂–Al₂O₃ flux system using statistical mixture design for submerged arc welding of HSLA steel, which is not available in the literature. Regression equations for mechanical properties and diffusible hydrogen content in terms of individual constituents and their binary interactions have been developed, which may be further used to achieve desired values of mechanical properties. In the optimization of these properties, weight may be specified for respective individual property according to the importance and desirability of that property. With the help of such relations, flux designer may design and optimize flux to get desired and optimized mechanical properties for HSLA steel (API 5L X65).

Design of experiment

For design of experiment, extreme vertices design suggested by Anderson and McLean^20,21 is used for flux formulation. The method suggests that constrained mixture design for a mixture of q components having lower and upper bounds on some or all of the components may be represented mathematically as

0 \leq a_{i} \leq x_{i} \leq b_{i} \leq 100

(2)

and

\sum_{i = 1}^{q} x_{i} = 100

(3)

where $i = 1, 2, 3, \dots, q$ and $a_{i}$ and $b_{i}$ are the constraints imposed by the experimenter on $x_{i}$ .

The primary requirement while deciding the percentage composition of various flux ingredients is that their melting temperature should be lower than that of base metals as the flux should melt first before the base metals and should remain in the molten state even after the solidification of weldment so as to avoid the atmospheric contamination.²² Based on these aspects, the phase diagrams of various systems are analyzed, and the percentage composition of flux ingredients is broadly chosen. The phase diagram for CaO–Al₂O₃–MgO ternary system is shown in Figure 1. Then, the range of these flux ingredients is further narrowed, based on the available literature.

Figure 1.

Phase diagram for CaO–Al₂O₃–MgO ternary system showing different equilibrium temperatures.²³

In the work presented here, four components CaO, Al₂O₃, CaF₂ and MgO are taken as per the following constraints

\begin{array}{l} 10 \leq CaO (x_{1}) \leq 35 \\ 8 \leq {Al}_{2} O_{3} (x_{2}) \leq 40 \\ 10 \leq {CaF}_{2} (x_{3}) \leq 40 \\ 10 \leq MgO (x_{4}) \leq 40 \end{array}

(4)

\sum_{i = 1}^{4} x_{i} = 78

(5)

Twenty-one agglomerated fluxes were prepared by varying the composition of four components, CaO, Al₂O₃, CaF₂ and MgO, and keeping SiO₂, TiO₂, MnO and Bentonite as constants as given in Table 1.

Table 1.

Design matrix for flux formulation.

Type of point	Experiment No.	Varied mixture components				Basicity index
		CaO (% wt.)	Al₂O₃ (% wt.)	CaF₂ (% wt.)	MgO (% wt.)
Vertex	1	35.0	8.0	10.0	25.0	3.7
Vertex	2	10.0	8.0	40.0	20.0	3.7
Vertex	3	10.0	40.0	10.0	18.0	1.1
Vertex	4	20.0	8.0	10.0	40.0	3.7
Vertex	5	10.0	8.0	20.0	40.0	3.7
Vertex	6	10.0	18.0	10.0	40.0	2.5
Vertex	7	20.0	8.0	40.0	10.0	3.7
Vertex	8	35.0	8.0	25.0	10.0	3.7
Vertex	9	18.0	40.0	10.0	10.0	1.1
Vertex	10	35.0	23.0	10.0	10.0	2.1
Vertex	11	10.0	40.0	18.0	10.0	1.1
Vertex	12	10.0	18.0	40.0	10.0	2.5
Centroid	13	10.0	22.0	23.0	23.0	2.2
Centroid	14	21.7	8.0	24.2	24.2	3.7
Centroid	15	21.3	22.8	10.0	23.8	2.1
Centroid	16	21.3	22.8	23.8	10.0	2.1
Centroid	17	35.0	13.0	15.0	15.0	3.0
Centroid	18	12.7	40.0	12.7	12.7	1.1
Centroid	19	13.3	11.3	40.0	13.3	3.2
Centroid	20	13.3	11.3	13.3	40.0	3.2
Centroid	21	18.6	18.9	20.3	20.3	2.4

The mixture design consists of 12 vertices of truncated tetrahedron, 4 centroids of hexagonal faces, 4 centroids of triangular faces and 1 overall body centroid. In the first step, 12 vertices of polyhedron were selected from 4 × 2⁴⁻¹ = 32 combinations (q × 2^q−1) of 4 components satisfying lower and upper bounds and total proportion of the mixture yielding 12 vertices of truncated tetrahedron in three-dimensional (3D) space (Figure 2). Then, of the 12 vertices combinations grouped into three or more having either of the vertices with the same proportion $x_{i}$ , so 8 such combinations were selected forming the centroids of the faces of truncated tetrahedron and 1 as the overall (body) centroid. Hence, a total of 21 experiments were performed; experiment numbers 1–12 represent 12 vertices, experiment numbers 13–20 represent 8 centroids and experiment number 21 represents 1 overall centroid (Table 1).

Figure 2.

Graphical demonstration of flux mixture design in 3D space.

The output response characteristics are given in the form of second-order regression model as

y = \sum_{i}^{q} b_{i} x_{i} + \sum_{i < j} \sum^{q} b_{ij} x_{i} x_{j}

(6)

where y is the output response variable and $b_{i}$ and $b_{ij}$ are the least squares model regression coefficients. The terms $b_{i} x_{i}$ and $b_{ij} x_{i} x_{j}$ in the regression model represent the individual effect and interaction effect of the mixture components respectively. If the output response variable y increases with the regression coefficient $b_{ij}$ positive, then the excess is called the synergism of the binary mixture, whereas if the output response variable y decreases with the regression coefficient $b_{ij}$ positive, then the decrease is called the antisynergism of the binary mixture and vice versa.²¹ Hence, the positive and negative values of the regression coefficient, $b_{ij}$ , define synergism and antisynergism of the binary mixtures, respectively.¹⁵

Experimentation

Preparation of fluxes

For the preparation of fluxes, constituting elements were weighed separately on a digital weighing balance (accuracy 1 mg) according to the weight percentage shown in Table 1 and then mixed thoroughly in a container with sodium silicate binder (20% weight)^24–26 for about 30 min to get homogenous semi-solid mass. Sodium silicate is added for better arc stability and binding the individual ingredients together. Then, the solid mass was dried in air for 24 h and baked in the muffle furnace at 700 °C for nearly 1 h. After cooling, these fluxes were crushed and sieved and then were kept in airtight bags.

Welding of plates

For welding, the edges of HSLA plates were prepared by making a V-groove of 60°. Twenty-one weld samples were made using two HSLA steel (API 5L X65) plates, each of size 250 × 150 mm and thickness 18 mm. Base plates are butt welded on the SAW machine (ADOR Tornado SAW M-800) available at M.M.U., Mullana (Ambala). All the experiments were conducted using a 3.2-mm-diameter electrode wire (EH-14) and optimum welding parameters given in Table 2 with reverse polarity. The optimum welding parameters were selected on the basis of trial runs and performing optimization experiments using response surface method (RSM) technique by varying the parameters, current range 349–651 A, voltage range 24–32 V and travel speed range 17–33 m/h, and keeping the nozzle-to-tip distance constant at 25 mm with formulated flux number 11 (neutral flux) having a basicity index of 1.1. Composition of the flux has significant effect on chemical and mechanical properties of the weldments, but the effect of flux constituents on weld bead geometry has not been reported in earlier literature. The study by Chai and Eagar²⁷ also indicates that weld metal chemistry is primarily dependent on weld metal flux composition and independent of operating parameters. The study also showed that weld bead geometry is dependent on weld parameters only, namely, welding current, voltage, travel speed, electrode diameter and electrode polarity, and independent of flux composition. Weld parameters are optimized on the basis of appearance of weld bead, weld form factor and dilution (Figure 3). Form factor and dilution^1,12,28 are given by

Table 2.

Weld parameters used for welding.

Welding parameters	Current I (A)	Voltage V (V)	Weld speed S (m/h)
	410	30	20

Figure 3.

Weld bead geometry.

Form factor = \frac{Width of bead, W}{Depth of bead, d}

(7)

Dilution (%) = \frac{Area of penetration}{Area of penetration + Area of reinforcement}

(8)

Chemical analysis

Chemical composition of base metal and filler wire (welding electrode) (Table 3) was determined by Atomic Absorption Spectroscope Foundry-Master UV.

Table 3.

Chemical composition of base metal and electrode.

Chemical composition	C	Si	Mn	P	S	Cu	Nb	Ti
Base metal (API 5L X65)	0.26	0.26	1.45	0.03	0.03	0.09	0.07	0.008
Electrode wire (EH-14)	0.15	0.1	1.8	0.03	0.03	0.32	—	—

Mechanical testing and microstructural analysis

Mechanical studies were performed on welded plates. For the analysis, welded plates were cut so as to get two specimens for tensile testing, two specimens for impact strength testing and one specimen for microstructural analysis from transverse weld. Tensile tests (ASTM E8M-00a) on weld coupons were performed on Heico computerized UTM (capacity 60 ton) at TERII, Barna (Kurukshetra); ultimate tensile strength (UTS), ultimate load, peak extension and peak strain were observed. Charpy V-notch impact test (ASTM E23-07ae1) was performed on impact specimens on impact testing machine at room temperature. Microstructural analysis of the welded samples was conducted by Leica optical microscope.

Results

The results of the various mechanical properties, UTS, percentage elongation, impact strength and diffusible hydrogen content are presented in Table 4.

Table 4.

Properties of weld metal samples.

Experiment No.	Ultimate tensile stress (MPa)			Percentage elongation			Impact strength (N m)			Hydrogen content (mL/100 g)
	Sample 1	Sample 2	Mean	Sample 1	Sample 2	Mean	Sample 1	Sample 2	Mean
1	565.7	510.4	538.1	23.8	21.7	22.8	63.8	157.0	110.4	9.1
2	535.6	511.5	523.6	30.1	21.1	25.6	49.1	98.1	73.6	9.2
3	528.4	591.84	560.1	37.8	24.6	31.2	107.9	112.8	110.4	8.2
4	582.7	571.55	577.1	26.8	28.9	27.9	127.5	112.8	120.2	7.8
5	520.2	572.04	546.1	22.5	25.8	24.1	98.1	89.8	94.0	6.2
6	527.1	527.1	527.1	32.0	27.6	29.8	100.0	110.0	105.0	7.9
7	647	630	638.5	41.6	41.6	41.6	101.9	98.1	100.0	5.1
8	614.8	583.5	599.2	33.0	26.7	29.9	88.3	88.3	88.3	4.0
9	422.2	600	511.1	20.1	29.6	24.9	39.2	117.7	78.5	8.7
10	460	416	438.0	19.4	18.4	18.9	39.2	56.4	47.8	9.8
11	600.1	589.64	594.9	25.4	27.0	26.2	132.4	147.2	139.8	5.4
12	573.4	584.6	579.0	28.6	28.0	28.3	107.9	132.4	120.2	6.2
13	540.1	567.9	554.0	25.3	24.6	25.0	98.1	186.4	142.2	6.9
14	635.3	574.7	605.0	33.3	35.9	34.6	132.4	137.3	134.9	6.1
15	520.6	579.5	550.1	35.9	28.8	32.3	148.1	107.9	128.0	7.5
16	610.3	581	595.7	32.4	28.7	30.5	157.0	103.0	130.0	4.1
17	496.4	507.3	501.9	21.1	22.9	22.0	93.2	103.0	98.1	8.9
18	545.9	544.1	545.0	25.3	27.3	26.3	122.3	117.7	120.0	6.5
19	625.7	559.9	592.8	42.3	24.4	33.4	93.2	122.6	107.9	5.2
20	575.4	573.5	574.5	32.1	28.9	30.5	98.1	122.6	110.4	7.1
21	600.5	585.9	593.2	28.5	28.5	28.5	132.4	141.8	137.1	5.5

Development of regression models for mechanical properties

Using the observed values of mechanical properties from experimentation, least squares regression equations have been developed in terms of percentage composition of individual components (CaO, Al₂O₃, CaF₂ and MgO). Second-order quadratic regression models are formed in terms of predictors of individual effect of mixture components (CaO, Al₂O₃, CaF₂ and MgO) and interaction effect of binary mixtures (CaO·Al₂O₃, CaO·CaF₂, CaO·MgO, Al₂O₃·CaF₂, Al₂O₃·MgO and CaF₂·MgO).

UTS = - 4.9531 CaO + 7.6824 A l_{2} O_{3} + 2.7335 Ca F_{2} + 6.0239 MgO + 0.0344 CaO \cdot A l_{2} O_{3} + 0.5013 CaO \cdot Ca F_{2} + 0.3182 CaO \cdot MgO + 0.1326 A l_{2} O_{3} \cdot Ca F_{2} - 0.0218 A l_{2} O_{3} \cdot MgO - 0.0197 Ca F_{2} \cdot MgO

(9)

Percentage elongation = - 1.2589 CaO + 0.2875 A l_{2} O_{3} + 0.4626 Ca F_{2} + 0.2042 MgO + 0.0231 CaO \cdot A l_{2} O_{3} + 0.0521 CaO \cdot Ca F_{2} + 0.0365 CaO \cdot MgO - 0.0134 A l_{2} O_{3} \cdot Ca F_{2} + 0.0140 A l_{2} O_{3} \cdot MgO - 0.0163 Ca F_{2} \cdot MgO

(10)

Impact strength = - 4.6949 CaO - 0.8874 A l_{2} O_{3} - 3.1084 Ca F_{2} - 1.9637 MgO + 0.0484 CaO \cdot A l_{2} O_{3} + 0.2148 CaO \cdot Ca F_{2} + 0.2447 CaO \cdot MgO + 0.2145 A l_{2} O_{3} \cdot Ca F_{2} + 0.1044 A l_{2} O_{3} \cdot MgO + 0.0906 Ca F_{2} \cdot MgO

(11)

H_{2} content = 0.4271 CaO + 0.1596 A l_{2} O_{3} + 0.3822 Ca F_{2} + 0.0326 MgO - 0.0033 CaO \cdot A l_{2} O_{3} - 0.01977 CaO \cdot Ca F_{2} - 0.0031 CaO \cdot MgO - 0.0110 A l_{2} O_{3} \cdot Ca F_{2} + 0.002 A l_{2} O_{3} \cdot MgO - 0.0011 Ca F_{2} \cdot MgO

(12)

Analysis of regression models

Prediction equations (9)–(12) have been analyzed using t-test and analysis of variance (ANOVA) (F-test). Individual regression coefficients ( $b_{i}$ and $b_{j}$ ) have been tested by comparing t₀ value with tabulated t value (at 95% significance level for dof_error),^20,21 as shown in Appendix 1. Whole mixture model has been checked separately for linear and quadratic least squares models as well as for full model tested by calculating F and P values at 95% significance level. R² values for all the prediction models are greater than 0.90, which indicates that the predicted values are in close agreement with the actual/observed values of the mechanical properties of weld metal as seen from Figure 4. Summary of the regression equations indicating the individual and interaction effects of binary mixtures on the mechanical properties of weld metal is given in Table 5.

Figure 4.

Predicted versus actual values of the mechanical properties and H₂ content: (a) ultimate tensile strength, (b) percentage elongation, (c) impact strength and (d) diffusible hydrogen content.

Table 5.

Summary indicating the effect of different mixtures on the mechanical properties and diffusible hydrogen content of weld metal.

Type of effect	Term	UTS	Percentage elongation	Impact strength	Diffusible H₂ content
Individual effect	CaO	Decrease	Decrease	Decrease	Increase
	Al₂O₃	Increase	Increase	Decrease	Increase
	CaF₂	Increase	Increase	Decrease	Increase
	MgO	Increase	—	Decrease	Increase
Binary mixture effect	CaO·Al₂O₃	—	Synergism	Synergism	Antisynergism
	CaO·CaF₂	Synergism	Synergism	Synergism	Antisynergism
	CaO·MgO	Synergism	Synergism	Synergism	Antisynergism
	Al₂O₃·CaF₂	Synergism	Antisynergism	Synergism	Antisynergism
	Al₂O₃·MgO	—	Synergism	Synergism	Synergism
	CaF₂·MgO	—	Antisynergism	Synergism	Antisynergism

UTS: ultimate tensile strength.

Discussion

Effect of flux constituents on UTS

Al₂O₃ is the most effective and MgO is the second most effective component among the individual flux mixtures having an increasing effect on UTS. CaF₂ also increases UTS, whereas CaO decreases weld metal UTS. Binary mixtures such as CaO·CaF₂, CaO·MgO and Al₂O₃·CaF₂ have synergistic (increasing) effect, whereas CaO·Al₂O₃ has no significant synergistic effect. Other mixtures such as Al₂O₃·MgO and CaF₂·MgO have no significant antisynergistic effect on UTS (Table 5).

Al₂O₃ tends to increase Si content of weld metal and deoxidizes weld pool, whereas MgO increases Mn and C content of weld metal and also controls O₂ of weld pool, thereby increasing tensile strength. CaO tends to pick up carbon from weld metal thus lowering UTS. Binary synergistic effect of CaO·CaF₂, CaO·MgO and Al₂O₃·CaF₂ is due to the binary synergistic effect on Mn and C content and antisynergistic effect on O₂ of weld metal.

Effect of flux constituents on impact toughness

All the individual flux ingredients, CaO, CaF₂, MgO and Al₂O₃, tend to decrease impact toughness. CaO is the most effective component to decrease toughness. Binary mixtures of all the components have synergistic (increasing) effect on impact toughness (Table 5).

CaO increases weld metal oxygen content, thus decreasing AF content and thereby decreasing toughness. Binary synergistic effect of CaO·CaF₂ and CaO·Al₂O₃ is due to the increase in AF (as is clear from optical micrographs of weld metal), which may further be due to the binary synergistic (increasing) effect on Mn and C content and antisynergistic (decreasing) effect on O₂ of weld metal, which are also reported in the literature.^15,29,30

Effect of flux constituents on diffusible hydrogen content

From the results (see equation (12) and Table 5), it can be seen that the main effect or individual effect of all the flux constituents is to increase the H₂ content, whereas the interaction effects or binary mixtures (CaO·Al₂O₃, CaO·CaF₂, CaO·MgO, Al₂O₃·CaF₂ and CaF₂·MgO) of the constituents tend to decrease the H₂ content. Hence, the binary mixtures of all constituents except Al₂O₃·MgO have antisynergistic effect on diffusible hydrogen content.

Figure 5 is a plot of observed values of H₂ content versus composition of CaF₂ (%) in the formulated fluxes. This figure indicates the overall effect of CaF₂ in the flux on the H₂ content. Second-degree regression fit plotted in this figure indicates the trend of diffusible hydrogen content with respect to CaF₂, which shows that H₂ content tends to decrease up to 28% of CaF₂; beyond this, diffusible hydrogen content starts increasing which is in agreement with the study by Plessis et al.¹⁹

Figure 5.

Plot of diffusible hydrogen content versus CaF₂.

Minimum diffusible hydrogen content has been observed for flux numbers 8 and 16, which have relatively high percentage composition of CaF₂. Decrease in hydrogen content in weld metal is due to a number of reasons as fluoride in fluorspar reacts with hydrogen to form insoluble HF, CaF₂ also reacts with SiO₂ in the flux to form SiF₄, CaO·SiF₄ functions as a shielding gas and reduces the partial pressure of hydrogen and CaO tends to increase the basicity of the slag, thereby decreasing hydrogen content.

Contour surface plots for various properties

Contour plots indicating predicted values of mechanical properties are shown in Figure 4. Different regions on contour surface³¹ show variation of mechanical property, so every contour curve marked on the surface gives constant value of mechanical property and each dotted point on the plot indicates one of the twenty-one design points. Figure 6 shows contour plots for UTS, percentage elongation impact toughness and diffusible hydrogen content for different proportions of flux components CaO, CaF₂ and Al₂O₃ with constant MgO = 20% content.

Figure 6.

Contour surface plots for mechanical properties and H₂ content: (a) ultimate tensile strength, (b) % elongation, (c) impact strength and (d) diffusible hydrogen content.

Discussion of microstructural analysis

Optical micrographs for weld zone of the weld joints are taken for all coupons. Optical micrograph of the base metal (API 5L X65) contains ferritic–pearlitic microstructure consisting of uniform mixture of ferrite, F (α-iron), and pearlite (α-iron and Ferric carbide—Fe₃C) shown as white and dark flakes, respectively, in different phase volume fractions (Figure 7(a)). Fusion zone of all weld metals contained acicular ferrite (AF), grain-boundary ferrite (GBF), Widmanstatten ferrite (WF) and upper bainite and lower bainite in general. AF formed at grain boundaries rather than blocky ferrite due to larger grain size is called WF.³²

Figure 7.

Optical micrographs (at ×100 magnification): (a) base metal and (b) weld zone for sample no. 16.

It can be seen from the respective micrographs for different samples showing fine-grained AF, GBF, WF for weld sample number 16 having uniform distribution of ferrite and pearlite (Figure 7(b)), AF and GBF are observed in weld sample number 18 and volume of ferrite content appears to be minimum (Figure 8(a)), and bainite structures are observed in weld sample number 21 (Figure 8(b)). Improvement of AF and GBF in sample number 16 is due to increase in Al₂O₃ content in the flux, which improves grain refinement, thereby improving AF content in weld metal.

Figure 8.

Optical micrographs of weld zone (at ×100 magnification): (a) weld zone for sample no. 18 and (b) weld zone for sample no. 21.

Multiresponse optimization of mechanical properties

The optimization problem for maximizing tensile strength and impact toughness and minimizing diffusible hydrogen content is formulated as a multiobjective, multivariable, nonlinear optimization problem. These properties have been optimized simultaneously using composite desirability optimization method proposed by Derringer and Suich.³³ This method makes use of an objective function D(x), called desirability function, which transforms an estimated response into a value called desirability. Composite desirability is the weighted geometric mean of individual desirability for the responses. The factor settings with maximum total desirability are considered to be the optimal parameter combinations.³⁴

Weights are used to fix the importance of a response/objective in the multiobjective optimization. Larger the weight for an objective, greater is the improvement in performance of that objective with a certain degree of degradation in the performance of other objective. The composite desirability is given by³⁵

D = [d_{1}^{w_{1}} \cdot d_{2}^{w_{2}} \cdot \dots \cdot d_{n}^{w_{n}}]^{\frac{1}{(w_{1} + w_{2} + \dots + w_{n})}}

(13)

D = {[Π_{i}^{n} d_{i}^{w_{i}}]}^{\frac{1}{(\sum_{i}^{n} w_{i})}}

(14)

where n is the number of responses, $d_{i}$ is the desirability of particular response and $w_{i}$ are the weights satisfying $0 < w_{i} < 1$ and $(w_{1} + w_{2} + \dots + w_{n}) = 1$ .

The present study aimed to optimize parameters to get maximum tensile strength and impact toughness and minimum diffusible hydrogen content. Four solutions with equal weights, for tensile strength, impact toughness and more weightage for hydrogen content, were found at different levels of desirability as presented in Table 6.

Table 6.

Optimum flux mixtures for optimized tensile strength, impact toughness and hydrogen content.

No.	CaO	Al₂O₃	CaF₂	MgO	Tensile strength	Percentage elongation	Impact toughness	H₂ content	Desirability
1	14.97	24.83	28.19	10.00	605.9	29.2	141.7	4.05	0.959
2	18.15	21.18	28.67	10.00	611.4	31.6	134.2	3.88	0.946
3	14.40	29.66	23.95	10.00	597.1	27.6	142.2	4.30	0.930
4	18.38	15.27	34.35	10.00	622.0	35.0	122.2	4.18	0.901

Model validation

Confirmatory experiments were conducted using randomly selected flux mixtures to validate the regression model and to ensure the repeatability and reliability of the predicted values (Table 7). Six random flux mixtures have been selected; weld samples have been mechanically tested. From the experimental values, it has been observed that error (%) for each response is less than 5% in almost all the samples.

Table 7.

Percentage error of predicted values and experimental values.

Flux constituents				Predicted values			Experimental values			Error (%)
CaO	Al₂O₃	CaF₂	MgO	UTS	Imp.St.	H₂	UTS	Imp.St.	H₂	UTS	Imp.St.	H₂
35.0	8.0	10.0	25.0	531	111.9	9.56	511	107.2	10	3.77	4.20	4.59
20.0	8.0	10.0	40.0	587	122.6	7.33	575	128.9	7.1	2.04	5.14	3.12
20.0	8.0	40.0	10.0	635	96.76	4.65	610	101.1	4.9	3.94	4.49	5.31
10.0	18.0	40.0	10.0	581	121.6	5.95	603	119	6.1	3.79	2.14	2.49
21.3	22.8	23.8	10.0	597	133	4.09	612	139	4.3	2.51	4.51	5.24
18.6	18.9	20.3	20.3	585	142.5	5.72	564	138	6	3.59	3.16	4.94

UTS: ultimate tensile stress; Imp.St.: impact strength; H₂: diffusible hydrogen content.

$Error (%) = \frac{Actual value - Predicted value}{Predicted value} \times 100 .$

Conclusions

Statistically designed flux mixtures have been formulated by mixing eight flux components, varying four components CaO, Al₂O₃, CaF₂ and MgO and keeping SiO₂, TiO₂ and MnO constant, using the extreme vertices design method.

The prediction equations for the weld mechanical properties, in terms of individual flux constituents and their binary mixtures, have been formed for welding of HSLA steel.

The difference between predicted values and experimental values of mechanical properties for weld metal is not significant at 95% confidence level.

Individual effect of CaO is to decrease impact toughness, but binary mixtures of CaO with CaF₂ and MgO have a synergistic effect on weld metal UTS and increase impact toughness.

Flux ingredient Al₂O₃ tends to increase UTS and decrease impact toughness, but Al₂O₃·CaF₂ has a synergistic (increasing) effect on both UTS and impact toughness.

MgO has a major increasing effect on UTS but a decreasing effect on toughness.

CaF₂ increases UTS and decreases impact toughness, but interaction of CaF₂ with other components has a positive synergistic effect on all properties.

Al₂O₃ in the flux improves grain refinement, thereby improving AF content in the weld metal.

Four solutions for flux mixtures are provided, yielding optimized weld metal mechanical properties and diffusible hydrogen content of weld metal of HSLA steel.

All the flux constituents have an increasing effect on diffusible hydrogen content, whereas binary mixtures of CaF₂ tend to decrease diffusible hydrogen content.

Diffusible hydrogen content decreases up to 28% of CaF₂, but beyond this hydrogen content starts increasing.

Footnotes

Appendix 1

Analysis of regression models using ANOVA (F-test)

	Model	DOF	SS	MS	F value	P value	R² (%)
UTS	Regression	9	37,441.4	4160.2	24.24	0.000	95.20
	Linear	3	10,439.3	3586.5	20.90	0.000
	Quadratic	6	27,002.1	4500.3	26.23	0.000
	Residual error	11	1887.5	171.6
	Total	20	39,328.9
Percentage elongation	Regression	9	453.397	50.377	13.15	0.000	91.49
	Linear	3	108.422	63.225	16.50	0.000
	Quadratic	6	344.975	57.496	15.00	0.000
	Residual error	11	42.153	3.832
	Total	20	495.549
Impact strength	Regression	9	11,234.0	1248.23	53.79	0.000	97.78
	Linear	3	1480.1	359.60	15.50	0.000
	Quadratic	6	9754.0	1625.67	70.05	0.000
	Residual error	11	255.3	23.21
	Total	20	11,489.3
Diffusible hydrogen content	Regression	9	51.1415	5.6824	9.06	0.001	88.12
	Linear	3	13.6258	5.6966	9.09	0.003
	Quadratic	6	37.5157	6.2526	9.97	0.001
	Residual error	11	6.8966	0.6270
	Total	20	58.0381

DOF: degrees of freedom; UTS: ultimate tensile strength; SS: sum of squares; MS: mean sum of squares.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Houldcroft

. Submerged arc welding. 2nd ed. England: Abington Publishing, 1989.

Murugan

Gunaraj

. Prediction and control of weld bead geometry and shape relationships in submerged arc welding of pipes. J Mater Process Tech 2005; 168: 478–487.

Parmar

. Welding processes and technology. New Delhi, India: Khanna Publishers, 1992, pp.200–210.

Kou

. Welding metallurgy. 2nd ed. New Jersey: John Wiley & Sons Inc., 2003, p.405.

Parmar

. Welding engineering and technology. New Delhi, India: Khanna Publishers, 1999.

Prasad

Dwivedi

. Some investigations on microstructure and mechanical properties of submerged arc welded HSLA steel joints. Int J Adv Manuf Tech 2008; 36: 475–483.

Pokhodnya

. Welding materials: current state and development tendencies. Weld Int 2003; 17(11): 905–917.

Datta

Mukerjee

Jha

. Weldability characteristics of shielded metal arc welded high strength quenched and tempered plates. J Mater Eng Perform 2002; 11(1): 5–10.

AWS. Welding inspection handbook. 3rd ed. Doral, FL: American Welding Society, 2000.

10.

Ramirez

. Characterization of high-strength steel weld metals: chemical composition, microstructure and nonmetallic inclusions. Weld J 2008; 87: 65s–75s.

11.

Jindal

Chhibber

Mehta

. Issues in welding of HSLA steels. Adv Mat Res 2011; 365: 44–49.

12.

AWS. Welding processes, welding handbook, 2. 8th ed. Doral, FL: American Welding Society, 1997.

13.

Pandey

Bharti

Gupta

. Effect of submerged arc welding parameters and fluxes on element transfer behaviour and weld-metal chemistry. J Mater Process Tech 1994; 40: 195–211.

14.

Kanjilal

Pal

Majumdar

. Combined effect of flux and welding parameters on chemical composition and mechanical properties of submerged arc weld metal. J Mater Process Tech 2006; 171: 223–231.

15.

Kanjilal

Majumdar

Pal

. Prediction of acicular ferrite from flux ingredients in submerged arc weld metal of C-Mn steel. ISIJ Int 2005; 45(6): 876–885.

16.

Davidson

. Advances in hydrogen management: the science-based design of low hydrogen consumables for the future. Aust Weld J 1998; 43: 33–39.

17.

Terashima

Tsuboi

. Hydrogen in submerged arc weld metal produced with agglomerated flux. Weld J Jpn 1976; 45: 28–33.

18.

Chew

. Prediction of weld metal hydrogen levels obtained under test conditions. Weld J 1973; 52(9): 386s–391s.

19.

Plessis

JDU

Toit

MDU

Pistorius

. Control of diffusible weld metal hydrogen through flux chemistry modification. Weld J 2007; 86: 273s–280s.

20.

Anderson

McLean

. Design of experiments: a realistic approach. New York: Marcel Dekker, 1974.

21.

Cornell

. Experiments with mixtures: designs, models, and the analysis of mixture data. 3rd ed. New York: John Wiley & Sons, 2002.

22.

Bhandari

Chhibber

Arora

. Effect of electrode coatings on diffusible hydrogen content, hardness and microstructures of the ferritic heat affected zones in bimetallic welds. Adv Mat Res 2011; 383–390: 4697–4701.

23.

Zhang

Lee

. Use of phase diagrams in studies of refractories corrosion. Int Mater Rev 2000; 45(2): 41–58.

24.

Houldcroft

. Welding process technology. 1st ed. Great Britain: Cambridge University Press, 1977.

25.

Singh

Pandey

. Economics of recycling submerged arc welding slag as a flux. In: Proceedings of the 8th international conference on trends in welding research, Pine Mountain, Georgia, 1–6 June 2008, pp.806–810, Ohio, ASM International.

26.

Kumar

Mohan

Khamba

. Development of agglomerated acidic flux for submerged arc welding. Est J Eng 2010; 16(2): 135–141.

27.

Chai

Eagar

. The effect of SAW parameters on weld metal chemistry. Weld Res Suppl 1980; 59(3): 93s–98s.

28.

Murugan

Gunaraj

. Application of response surface methodology for predicting weld bead quality in submerged arc welding of pipes. J Mater Process Tech 1999; 88: 266–275.

29.

Burck

Indacochea

Olson

. Effects of welding flux additions on 4340 steel weld metal composition. Weld J 1990; 3: 115–122.

30.

Dallam

Liu

Olson

. Flux composition dependence of microstructure and toughness of submerged arc HSLA weldments. Weld J 1985; 64(5): 140–151.

31.

Hummel

. Introduction to phase equilibria in ceramic systems. New York: Marcel Dekker, 1984.

32.

Messler

. Principles of welding. Weinheim Wiley-Vch Verlag Gmbh & Co, 2004.

33.

Derringer

Suich

. Simultaneous optimization of several response variables. J Qual Technol 1980; 12(9): 214–219.

34.

Harington

. The desirability function. Ind Qual Control 1965; 21: 494–498.

35.

Castello

Montgomery

. Modified desirability functions for multi response optimization. J Qual Technol 1996; 28(3): 337–345.