Optimizing the laser parameters to attain maximum hardness in nickel based hardfacing surfaces

1 Introduction

Surface modification of traditional materials (steels, aluminum alloys, titanium alloys, etc.) by laser hardfacing (LH) is attracting more and more due to its momentous growth in industrial applications. Laser surface modification is carried out by intruding the laser energy on the substrate with the addition of desired peripheral ingredients (powder). This results in high-performance exteriors (high in wear, oxidation resistance, corrosion and fatigue) on a low-priced low-alloy bulk material with reduced metal cost and rare alloy elements. The laser energy can also be effortlessly tuned to obtain the anticipated surface properties [1], [2], [3]. An extensive demand for materials with improved properties in terms of their hardness and their resistance to wear, corrosion, and oxidation has been the driving force for the development of various surface hardfacing techniques and materials. Recently, LH has been explored for the deposition of less diluted and fusion-bonded thick and thin metallic coatings on a wide variety of metallic substrate materials with a low heat input. Austenitic stainless steels (AISI 316 LN) are widely used in nuclear reactors owing to their excellent corrosion resistance. However, their low hardness leads to poor surface properties in terms of wear and fatigue resistant which confine many of their industrial applications [4], [5].

Abundant studies have shown superior characteristics of hardfacing alloys deposited by laser surfacing compared to other conventional surfacing techniques [5], [6]. The LH technique has the benefit of depositing a precise thickness of the material on a selected area of the substrate. Cobalt- and nickel-based intermetallic alloys have been developed as an excellent corrosion- and wear-resistant material over a wide range of temperatures and environments [7]. Nickel-based alloys are strengthened by the presence of large volume fraction of hard intermetallic laves phase in a much softer solid-solution phase or a eutectic phase mixture [8]. Therefore, tribology is seldom applied in bulk form, but rather are applied as coatings [9], [10]. Cobalt-base alloys have been extensively studied as overlays by laser cladding technique [11] as well as thermal spray technique [12].

Hemmati et al. [13] reported that dilution of the hardfaced deposit is influenced by laser power. Percentage of dilution is 5, 10, 15, 25, 30 and 35%, respectively. It is discovered that for Fe content rises, borides start to alteration in morphology and slowly reduce. Also, significantly lower amounts of Ni-Si-B eutectic phase’s form because of more Fe content. Abolition of the strengthening precipitates let’s to lower the hardness of the deposits (800–500 HV). Ming et al [14] investigated the effect of powder feed rate (PFR) and translation speed (TS) of a laser cladding with three nickel-based hardfacing powders. Deposits produced at higher PFRs registered a higher hardness than deposits made with lower PFRs. Higher PFR produces a higher percentage of hard phases and, the percentage of hard phases decreases linearly with increasing TS. Zhang et al. [15] attempted a hardfacing by Colmonoy 6 powder on 316 L austenitic stainless steel using CO₂ laser process. They investigated the effects of laser power, traveling speed, defocusing distance, PFR on bead height, bead width, penetration depth and dilution. They found that preheating is essential for preventing cracking in the LH procedure and 450°C is the suitable preheating temperature. The friction and wear test results showed that the friction coefficient of specimens with laser cladding is lower than that of specimens without laser cladding, and the wear resistance of specimens has been increased 53 times after laser cladding, which reveals that laser cladding layer plays a major role in wear resistance. The microstructures of laser cladding deposit are composed of Ni-rich eutectic, boride and carbides.

It is well known that the hardfacing parameters plays a major role in determining the deposit quality as the process facts have not been disclosed so far and hence, the selection of LH process parameters to nickel-based alloys is very difficult. Response surface methodology (RSM) is a collection of mathematical and statistical techniques useful for the modeling and analysis of problems in which a response of one interest is influenced by several variables and the objectives are to optimize this response [16]. In this investigation, RSM was used to reduce the number of experiments and optimize the process parameters that yield the higher hardness.

It is understood that the hardfacing process parameters play a major role in deciding the quality of the deposits. Very limited investigations have been carried out to understand the effect of individual LH parameters on mechanical properties and microstructural characteristics. There is no literature available in optimizing the LH parameters to attain maximum hardness on nickel based hardfaced deposit on 316 LN austenitic stainless steel. Hence, in this study, an attempt has been made to optimize the important LH parameters to attain maximum hardness in nickel based hardfaced deposits on 316 LN austenitic stainless steel by RSM.

2 Experimental work

2.2 Feasible working range of LH parameters

Trial runs were carried out using 12 mm-thick 316 LN austenitic stainless steel plate and nickel based alloy to find out the feasible working limits of LH parameters. The optimization process was carried as per the flow chart (Figure 1). Different combinations of parameters were used to carry out the trial experiments. This was done by varying any one of the factors from minimum to maximum while keeping the other parameters at constant (Table 4). The feasible working limits of the individual parameters were identified by macrostructure (cross section of the deposits). A deposit which showed, a smooth appearance without any visible macro level defects such as crack, pores was chosen as the feasible working parameter. The chosen levels of the selected process parameters with their units and notations are presented in Table 5.

Figure 1:

Flow chart for process optimization.

Table 4:

Macrostructure analysis for fixing the working range of laser hardfacing.

S.No	Process parameters	Parameters range	Macrograph	Name of the defect	Reason for defect
1	Power (P)	P>1900 W		Crack, dilution	Higher heat input
		P<1100 W		Pores and escaping of powder	Insufficient heat input
2	Powder feed rate (F)	F>11 g/min		Cracks	Insufficient specific energy input
		F<3 g/min		High depth of penetration and dilution	Higher specific energy input
3	Travel speed (T)	S>500 mm/min		Cracks	Low heat input
		S<300 mm/min		High dilution	Higher heat input
4	Defocusing distance (D)	D>37 mm		Poor bonding	Low energy density per unit
		D<17 mm		Pores	Higher energy density per unit

Table 5:

LH process parameters and their working range.

S. no	Factor	Unit	Notation	Levels
				−2	−1	0	1	2
1	Power	W	P	1100	1300	1500	1700	1900
2	Powder feed rate	g/min	F	3	5	7	9	11
3	Travel speed	mm/min	T	300	350	400	450	500
4	Defocusing distance	mm	D	17	22	27	32	37

2.3 LH experiments and hardness evaluation

Figure 2 shows the multi-track hardfaced deposit configuration with 50% overlapped used in this investigation. The base metal composed of fully elongated austenitic grains. The LH deposits were made as per the conditions dictated by the design matrix (Table 6) at random order so as to avoid the noise creeping output response. The substrate was preheated to 400°C to relieve the internal stresses and also to reduce the cooling rate to avoid the formation of cracks after deposition [15]. The average deposited thickness was about 0.8–2 mm of the stainless steel. An automatic disk LH machine was employed to conduct the experiments. Few of the fabricated deposits are displayed in Figure 3. After hardfacing, the deposit was cut into small samples for the metallography and hardness study. A Vickers microhardness testing machine (Make: SHIMADZU, Japan; Model: HMV−2T) was employed for measuring the hardness across the hardfaced deposit cross section with a load of 0.5 kg and dwell time of 15 s.

Figure 2:

Single layer hardfacing.

Table 6:

Design matrix and experimental results.

Expt. no.	Coded value				Actual value				Hardness of the deposit (HV)
	P	F	T	D	P (W)	F (g/min)	T (mm/min)	D (mm)
1	−1	−1	−1	−1	1300	5	350	22	573
2	1	−1	−1	−1	1700	5	350	22	475
3	−1	1	−1	−1	1300	9	350	22	778
4	1	1	−1	−1	1700	9	350	22	603
5	−1	−1	1	−1	1300	5	450	22	475
6	1	−1	1	−1	1700	5	450	22	574
7	−1	1	1	−1	1300	9	450	22	794
8	1	1	1	−1	1700	9	450	22	703
9	−1	−1	−1	1	1300	5	350	32	743
10	1	−1	−1	1	1700	5	350	32	568
11	−1	1	−1	1	1300	9	350	32	820
12	1	1	−1	1	1700	9	350	32	602
13	−1	−1	1	1	1300	5	450	32	545
14	1	−1	1	1	1700	5	450	32	581
15	−1	1	1	1	1300	9	450	32	680
16	1	1	1	1	1700	9	450	32	648
17	−2	0	0	0	1100	7	400	27	727
18	2	0	0	0	1900	7	400	27	551
19	0	−2	0	0	1500	3	400	27	487
20	0	2	0	0	1500	11	400	27	770
21	0	0	−2	0	1500	7	300	27	799
22	0	0	2	0	1500	7	500	27	721
23	0	0	0	−2	1500	7	400	17	575
24	0	0	0	2	1500	7	400	37	681
25	0	0	0	0	1500	7	400	27	769
26	0	0	0	0	1500	7	400	27	766
27	0	0	0	0	1500	7	400	27	769
28	0	0	0	0	1500	7	400	27	770
29	0	0	0	0	1500	7	400	27	766
30	0	0	0	0	1500	7	400	27	766

Figure 3:

Laser hardfaced deposits.

3 Developing an empirical relationship

Hardness of deposit is a function of the LH parameters such as power (P), powder feed rate (F), travel speed (T), defocusing distance (D), and it can be expressed as,

Second-order polynomial (regression) equation used to denote the response surface Y is given by,

selected polynomial could be expressed as,

where, b₀ is the mean value of response and b₁, b₂, b₃…b₄₄ are linear interactions and square terms of factors. The value of co-efficient was calculated using Design Expert 7 software at 95% confidence level. The significance of the each co-efficient was calculated from t-test and p-values. The value of “Probe>F” is less than 0.05, indicates that model terms are significant. In this case P, F, T, D, PF, PT, FD, TD, P², F² and D² are the significant terms. The final empirical relationship was built using only these co-efficient and the established final empirical relationship of laser hardfaced deposit of Colmonoy-5 alloy is given below.

The adequacy of the above relation is tested by analysis of variance (ANOVA). The ANOVA test results are given in Table 7 at the desired confidence level of 95%. The relationship may be considered to be adequate. If the calculated value of the F ratio of the developed relationship does not exceed the tabulated value of F-ratio for an anticipated level of confidence, the model is found to be adequate. The Fisher’s F-test with a very low probability value demonstrates a very high significance of the regression model. The goodness of fit of the model is fitted by the determination coefficient (R²). The coefficient of determination was calculated to be 0.99 in response which implies that 99% of the experimental values confirm the compatibility with data as predicted by the model. The R² value should always be between 0 and 1. A model is statistically good the R² value should be close to 1.0. Then adjusted R² value reconstructs the expression with the significant terms. The value of adj. R²=0.981 is also high and indicates the high significance of the model.

The pred. R² value is 0.822 which means that the model could explain 82.2% of the variability in prediction. This is in reasonable agreement with the adj. R² of 0.952. The value of the coefficient of variation as low as 2.17, which indicates that the deviation between experimental and predicted values are low. Adequate measures of the signal to noise ratio, a ratio greater than 4 is desirable. During this investigation, the ratio is 34.47, which indicates an adequate signal. This model can be used to navigate the design space. The correlation graph shown in the Figure 4, it shows predicted and actual hardness of laser hardfaced deposit, it could indicate the deviation between the actual and predicted hardness is low. Table 8 presents the actual and predicted value of hardness.

Figure 4:

Correlation graph.

4 Optimizing LH parameters

The RSM was used to optimize the LH parameters considered in this study. RSM is a collection of a mathematical and statistical method that are beneficial for designing experiments, constructing a mathematical model, exploratory for the optimal combination of input parameters and pressing out the value in graphically [18], [19]. Figure 5 depicts perturbation plot for the response of hardness of the deposits. This plot offers an outline view of the response and displays the transformation of hardness, when each LH parameter moves from the reference point, with all other parameters held constant as the reference value. The design of experiment sets the reference point by default in the middle of the design space. From the perturbation graph and response surface graphs, it can be observed that when the hardness increases with increasing PFR, defocusing distance to the certain level and then decreases. It may be endorsed due to the insufficient energy or low heat input causes the escaping of powders and unmelted partials existing in the deposits. Hardness decreases with increasing laser power, travel speed. It may be believed that the high heat input will increase the depth of penetration and dilution of the deposits [15].

Figure 5:

Perturbation graph.

To obtain the influencing nature and optimized condition of the process on hardness (H), the surface and contour plots which are indications of possible independence of factors have been built up for the proposed empirical relation considering two parameters in the halfway tier, and two parameters in the X-axis and Y-axis as shown in Figure 6. The contour plots help us to predict the response at any zone in the design domain [20]. The apex of the response plot shows the maximum achievable hardness. Characterization involves identifying whether the stationary point is a minimum or maximum response or a saddle point to classify this; it is more straightforward to analyze it through a contour plot. Contour plot plays a vital role in the learning the response surface. It is vibrant from that when the hardness increases with increasing PFR, defocusing distance and hardness decrease with increasing laser power, travel speed.

Figure 6:

Response surface graphs and contour plots.

(A) Interaction effect of power and powder feed rate. (B) Interaction effect of power and travel speed. (C) Interaction effect of power and defocusing distance. (D) Interaction effect of powder feed rate and travel speed. (E) Interaction effect of powder feed rate and defocusing distance. (F) Interaction effect of travel speed and defocusing distance.

To know further about the influencing tendency of process parameters on hardness, three-dimensional diagrams are plotted for a particular processing condition. Figure 6A–F as surface and contour plots for each process parameters. It is clear from Figure 6A that the hardness falls and increases with an increase of process parameters such as laser power and PFR. Hardness mainly depends on dilution and microstructure. Laser power mainly used for melting the powder and excess heat melts the substrate. Keep on increasing the laser power, high volume of substrate material melts. The hardness variation in laser hardfaced sample could affected by the deposit dilution. As higher dilution means lower hardness, increasing P, (constant F, T, D) increases dilution and then hardness decreases. Increasing F (constant P, T, D) diminish dilution so hardness increases because more amount of heat utilized for melting the powder and small amount heat melt the substrate material. Increasing the T (constant P, F, D), powder density per square area (g/mm²) is less so dilution rate is somewhat increased and hardness is reduced. Increasing the D (constant P, F, T) diminish dilution so hardness increased. When increasing the defocusing distance, the beam size gets larger so the energy density per unit of clad pass available becomes less, thus the penetration depth and dilution decrease.

The valley of response plot gives the maximum hardness. These response contours can help in the prediction of the response (hardness) in any zone of the experimental domain [21]. By analyzing the response surface and contour plots as shown in Figure 6A–F, the maximum achievable hardness value is found to be 820.48 HV. The corresponding parameters exist maximum hardness value power of 1314 w, PFR of 9 g/min, a travel speed of 366 mm/min, and defocusing distance of 32 mm at the valley of response surface plot and corresponding domain in the contour plot. Higher F ratio value can give the more significant process parameter. From the F ratio value, it can be concluded that the PFR is contributing the major factor to exploit hardness, followed by power, defocusing distance and travel speed for the range considered in this investigation. To check the prediction capabilities of the developed empirical relationship, three more confirmation tests were carried out with the hardfacing process parameters chosen randomly from the feasible working range (Table 9). The actual response was calculated by taking the average of three results. The developed results reveal that the empirical relationship is accurate since the variation is ±5%. Table 10 summarize the experimental values, predicted values and the variation. Table 11 shows the cross-sectional macrograph, top surface hardness and indentation image of low, medium and high hardness laser hardfaced deposits.

Table 11:

Cross-sectional macrograph and top surface hardness indentation of laser hardfaced deposit.

Deposit no	LH parameters	Macrograph	Hardness with indentation	Hardness (HV0.5)
01	P – 1300 W F – 5 g/min T – 350 mm/min D – 22 mm			475
04	P – 1700 W F – 9 g/min T – 350 mm/min D – 22 mm			603
11	P – 1300 W F – 9 g/min T – 350 mm/min D – 32 mm			820

5 Microstructure of LH hardfaced deposits

LH with powder injection was used to produce multi-track deposits with 50% track overlapping. After the LH, visual inspection of the deposits for both initial compositions revealed a good, continuous appearance without signs of surface cracks or lack of adhesion, as can be seen in the micrograph in Figure 7. Cross sections of the specimens revealed a dendritic structure homogeneously distributed throughout the deposit with a continuous interface, of an unrevealed structure, which corresponds to the dilution zone (Figure 7B). Accordingly, the matrix may be supposed to be composed of a solid solution of Ni-γ with a relatively lower iron and silicon content than the dendrite. Laser-deposited layers in general, three different microstructures from the bottom to top: a plane solidification front microstructure with limited thickness, a transition cellular microstructure, and a columnar-dendritic microstructure. In addition to that, the evolution of fourth region which consists of equiaxed grains at the top (Figure 7D). The microstructure at each region depends on the cooling rate and temperature gradient. In the laser-deposited Ni-based layer, the main solidification structure is characterized by columnar crystals at the bottom and equiaxed crystals on the top surface. Due to the high-temperature gradient (G), and lower solidification rate (V), at the solid-liquid interface, epitaxial growth from the substrate occurs in the bottom area, showing a typical directional solidification characteristic. In the top area of the melt pool, columnar crystals disappear and equiaxed crystals appeared due to the local lower G and higher V conditions [22].

Figure 7:

Various microstructures of hardfaced deposits observed by OM.

(A) Low magnification cross-section. (B) Hardfaced deposit-interface. (C) Hardfaced deposit-middle. (D) Hardfaced deposit-top surface. (E) HAZ. (F) Substrate. (G) Precipitates at top of the deposit. (H) Precipitates at near the interface.

The interface appears well defined with a zone that is apparently free of precipitates (Figure 7E). However, at greater magnification the existence of a new arborescent phase, whose composition is the same as the dendrites indicating an epitaxial growth. In the overlapped tracks, higher concentrations of dark precipitates are observed. This may be due to the remelting of the former track during LH. This remelting will produce higher segregation of eutectics at the interface between overlapping tracks. The existence of these eutectics is mainly depending on Cr and C concentrations at the particular region. Laser deposited Colmonoy 5 coatings consists of three general components: Cr-rich precipitates such as CrB, CrC, Ni solid solution dendrites, Ni-B-Si binary and ternary eutectic phases including NiB, NiSi (Figure 7G and H) [23].

Figure 8 shows the line scan EDS it gathers the values of each element (wt.%) as a function of the distance from the substrate, interface and deposit. As can be seen, substrate consist of iron content is around 70–75%. Near the interface iron content drastically reduced it is conformed that dilution is very low, iron content passing from 80% at the interface to 4% in the deposit. The dilution ensures a perfect bond between the metal and the clad layer, and on the other hand the low level of dilution confirms the quality of the coating that’s improve the hardness of the deposit. The silicon content is around 4%, the chromium content around 8% and the nickel content between 80 and 85% in the deposit. EDS spectra shows that the precipitates in the cluster (spot 1) are rich in Cr with fewer additions of C, Si, Ni and Fe as showed in Figure 9. The second spot EDS are rich in Ni with definite counts of Cr, Fe and Si as showed in Figure 9 (spot 2). In addition, the EDS spectra reveal the eutectic interdendritic constituent containing particles of Ni-rich compounds, preferably Ni-borides.

Figure 8:

EDS line scan of the constituent elements across substrate/deposit interface (optimized condition).

Figure 9:

Scanning electron micrograph of the deposit and spot EDS (optimized condition).

6 Conclusions

1. An empirical relationship was developed to predict the hardness of nickel-based layer deposited on 316LN austenitic stainless-steel substrate with 95% confidence level by incorporating important LH parameters.

2. A maximum hardness of 829.44 HV could be achieved in the deposit made using laser power of 1314 W, PFR of 9 g/min, a travel speed of 366 mm/min, and defocusing distance of 32 mm.

3. Of the four LH parameters, the PFR (based on F value) is the major influencing factor to predict the hardness followed by power, travel speed and defocusing distance.