Experimental investigation and parametric optimization of cryogenic abrasive water jet machining of nitrile rubber using Taguchi analysis

Abstract

Mining/marine bushings, biomedical implants, and many more are among the several applications of lightweight elastomer components. The use of non-traditional machining to manufacture these good-quality components in small or batch-size units is required because the primary manufacturing method requires customised mould and follow-up machining. In this regard, the present work focuses on investigating the performance of machining a nitrile rubber (NR) using suspension-type abrasive water jet (AWJ) under both conventional (room temperature) and cryogenic (liquid nitrogen (LN₂)) conditions at optimal values of process parameters: water jet pressure (WJP, bar), transverse rate (V_f, mm/min), and stand-off distance (SOD, mm). The experimental runs are designed using Taguchi analysis with respective performance parameters: Kerf taper ratio (KT_R) and Material removal rate (MRR, mm³/min). The results show that V_f (Rank 1) is the highest influencing factor on the machining performance of NR under both conditions. The reasons are improved kinetic energy, less collision of garnet abrasive particles and lesser change in the average width of the kerf. The influence of LN₂ showed that the optimal values of KT_R reduced by 11.97% at 200 bar, 40 mm/min, and 1 mm, and MRR increased by 0.65% at 250 bar, 60 mm/min, and 2.0 mm, respectively.

Keywords:

• Cryogenic
• nitrile rubber
• garnet
• Taguchi
• Kerf taper ratio
• material removal rate
• parametric optimization

View correction statement:

Correction

1. Introduction

The use of high-performance, lightweight elastomers to replace traditional metals in various applications such as oil/gas, mining/marine sectors, and many more are gaining traction. Elastomers, like, Nitrile rubber (NR), are of primary interest in automotive, mining, marine, biomedical implants, elastomer toughened epoxy composites, sealing gaskets, bearings, and aerospace industries. The reason for wide application is their excellent oil and chemical resistance, high pressure and wear resistance, viscoelastic nature, low-temperature flexibility, and lightweight property (Campbell, 2012; Kim et al., 2019; Mandloi et al., 2021; Xu et al., 2018). In contrast to metals, the stress–strain relationship in NR is quite nonlinear and does not have a yield point. Upon load removal, NR can return to its original, undamaged state after a significant deformation. Manufacturing single items or small batch sizes is unprofitable using primary shaping/casting processes like injection or compression moulding because it necessitates costly custom-pressing moulds. For small volumes, replacing traditional production with advanced machining methods reduces sudden breakdown and increases handling flexibility in most mining or marine equipments, such as positive displacement or mud motors. However, the strong elastic material behaviour with low Young’s modulus of elastomers poses the biggest challenge to cost-effective machining mechanisms with high precision and high-quality surfaces.

Two strategies can improve the machining quality of elastomers to resolve the high elasticity problem during machining. The first is the higher cutting speed to obtain a stiffer material. The second is the workpiece cooling under the glass-transition range using a cryogenic environment to achieve a higher Young’s modulus. Cooling the elastomer under a cryogenic climate is far more effective than employing faster cutting speeds since the temperature impacts the elastic modulus more than the strain rate. Elastomers attain a glassy phase in a cryogenic environment due to increased modulus and decreased elongation. It transforms the elastomer from its elastic to tough phase and improves its machinability. The cryogenic environment is usually achieved by producing an atmosphere of solid carbon dioxide or liquid nitrogen (LN₂). The conducted studies (Dhokia et al., 2011; Nayak et al., 2012; Shih, Lewis, et al., 2004; Shih, Luo, et al., 2004) suggested that cryogenic cooling improved the machining of elastomers by generating burr-free components with improved surface quality. The latest research findings (Mallick et al., 2022) said that the molecular mobility of elastomers is constrained in a cryogenic environment because of substantial stress production at extremely low temperature (−195°C) of LN₂. It enables the elastomer to endure high applied stresses but prevents it from deforming easily, which causes a fracture. Apart from the material perspective, cryogenic machining is eco-friendly and economical, has no environmental side effects, and there is less possibility of a heat-affected zone in the machining area (Khan et al., 2021; Kumar et al., 2022). Machining elastomers in a cryogenic environment improves quality of products and performance.

The machining of elastomers using traditional methods (like, Control numerical control (CNC) drilling, turning, milling, etc.) in the cryogenic environment has some problems during the process, like inferior surface finish, dimensional instability, and adiabatic shear band formation (Dhokia et al., 2011; Mallick et al., 2022). These defects are due to the cutting forces involved in traditional machining methods and workpiece fixing, necessitating a relatively strong clamping force. Over the past few decades, researchers have addressed the drawbacks of traditional cryogenic machining by making significant advancements and modifications in non-traditional machining techniques. Examples include Low-power CO₂ laser cutting (Banerjee & Bhowmick, 2014), Ultra-high pressure water jet machining (Hu et al., 2014), Cryogenic-assisted air jet machining (Gradeen et al., 2014; Lou et al., 2019; Zhang et al., 2021, 2022b, 2022a), and Abrasive water jet (AWJ) machining (Kowsari et al., 2017; Maurya et al., 2022). The need for sustainable machining methods is increasing rapidly (Ozbek et al., 2021). Among the above-discussed non-traditional machining methods, abrasive water jet (AWJ) machining is an eco-friendly and environmentally secure procedure. It is a sustainable machining method in the era of green manufacturing. It is often used in manufacturing and production industries to machine several engineering materials, especially difficult-to-machine materials. Heat does not affect the process, making it less likely to impact the material’s properties. The quality of the machined component is adaptable, free from adiabatic shear band formation and can machine thick material (Khan et al., 2021; Sharma et al., 2022). The AWJ machining has two dominant jet generation variants, namely, injection-type jet and suspension-type jet. The primary benefit of a suspension-type jet is that, compared to an injection-type, it operates at a much lower operating pressure (Molitoris et al., 2016). In suspension-type, a premixing of abrasive particles with water is required before the water emerges from the orifice to form a jet. It uses a liquid suspension known as slurry, which is subsequently moved to the cutting nozzle. Many studies are available on the experimental investigation of suspension-type AWJ machining of engineering materials (Liu et al., 2022; Qiang et al., 2019; Yokomae et al., 2023). Few works on parametric analysis of suspension-type AWJ machining processes of different thermoplastics are available. Tamannaee et al. (2015) worked on suspension-type AWJ machining of Talc-filled Thermoplastic Olefin (3 mm thickness) by considering jet impingement angle as a dominant process parameter. Kowsari et al. (2017) found that the principal dominating process parameters during suspension-type AWJ machining of Polymethylmethacrylate (3 mm thickness) were the number of machining passes, traverse rate, jet impingement angle, and impact velocity. Out of the available literature, none of the studies worked on the suspension-type AWJ machining of elastomeric materials. Hence, it opens up a new field for researchers to explore the significant process parameters involved in the AWJ machining of elastomers.

After analysing the significant parameters of any engineering process, parametric optimization is crucial in minimising or maximising the intended outcome. “Parametric optimization” is a systematic and efficient means of generating and equating the machining/process parameters to achieve optimal machining performance. The optimization method should be highly prioritised when high-quality products are required in several traditional and non-traditional methods (Akgün et al., 2021; Nas & Kara, 2022). The recent literature on parametric optimization of AWJ machining of several engineering materials is available (Joel & Jeyapoovan, 2021; Karatas et al., 2020; Rajesh et al., 2021; Senthilkumar et al., 2020; Thangaraj et al., 2021). Out of available research studies on parametric optimization, a few authors have worked on process optimization during the AWJ machining of polymeric and metal-matrix composites, such as Karatas et al. (2020) and Thangaraj et al. (2021), respectively, using Taguchi’s parametric optimization method. The literature says that many studies focus on optimizing AWJ machining of brittle and ductile materials. Still, investigations on viscoelastic materials have not been adequately carried out. Using Taguchi’s method, it is necessary to work on the parametric optimization of viscoelastic materials (such as NR) because it helps design experimental test sequences and obtain the optimal process parameters. Maurya et al. (2022) and Maurya et al. (in press) have worked on the experimental investigation of the suspension-type AWJ machining of 10- and 15-mm-thick NR, respectively, used in the positive displacement motors in the marine and mining sectors, under a room-temperature condition. The authors found that the jet pressure, traverse rate of the nozzle, and distance of the nozzle from the workpiece are essential process parameters affecting the machining performance. The authors recommended detailed experimental investigation and optimization of the AWJ machining of elastomers under cryogenic environment to investigate the improvement in dimensional stability and material removal mechanism.

The primary objective of the current work is to measure the AWJ machining performance in terms of Kerf taper ratio (KT_R) and Material removal rate (MRR) for variations in Water jet pressure (WJP), Traverse rate (V_f), and Stand-off distance (SOD) under conventional (room temperature) as well as cryogenic (LN₂) conditions as per the suggested experimental test sequence by the Taguchi’s L₂₇ Orthogonal array (OA). The second objective is to identify the influence of critical process parameters on the machining performance parameters as per the analysis of variance (ANOVA) table obtained from the Taguchi analysis both for conventional and cryogenic conditions. The third objective is to identify the influence of cryogenic conditions on the suspension-type AWJ machining of NR and document its impact on the KT_R and material removal mechanism of NR, showing the novelty of the current research work. The fourth objective is determining the optimal process parameter conditions obtained from the experimental runs and Taguchi analysis. Figure 1 displays the organisation of the current research paper.

Figure 1. Organization of the paper.

2. Experimental materials and methods

The section details the material, experimental setup, process parameters, and methods involved. The flow chart of the methodology used in the current study is summarised in Figure 2.

Figure 2. Methodology of the current work.

2.1. Materials

The workpiece used for the machining process is neat NR. NR is an elastomer whose properties depend on two monomers: acrylonitrile and butadiene. The density of NR (0.98 g/cm³) is comparable to that of other lightweight materials (0.80 to 4.5 g/cm³) (Campbell, 2012); thus, it is considered as a lightweight material. The NR workpieces are moulded in the shape of a cuboid of dimensions 50 mm × 80 mm × 15 mm for conventional condition as per National Organic Chemical Industries (NOCIL) guidelines and ASTM D2000-18 (Standard, 2018). The NR workpiece is moulded and is used in as-bought condition. The 80 mm length of the workpiece is chosen to obtain a cutting area to accommodate a maximum of six cuts on one workpiece and minimise the material cost. For machining under cryogenic condition, the workpiece of dimensions 50 mm × 40 mm × 15 mm is used. A single cut is obtained on each of the workpieces under cryogenic condition. The NR contained acrylonitrile of 34%, which falls in the feasible range of medium nitrile (Borkowski & Borkowski, 2009). FTIR spectra of the NR sample are observed using the Fourier-Transform Infrared (FTIR) spectrometer (Make: Shihradzu, Model: QATR-S single-reflection) in attenuated total reflectance mode as per ASTM D3677-10 (Standard, 2011) to confirm the percentage of acrylonitrile content (Teti et al., 2019). The spectra are generated at 4 cm⁻¹ resolution, 4000–400 cm⁻¹ wavelength scanning range, and 40 scans. The obtained FTIR spectrum is shown in Figure 3. The peaks at 2117.65 cm⁻¹, 2083.42 cm⁻¹, and 959.71 cm⁻¹ wavenumbers are characterised by stretching of a strong and medium intensity connection bond of −C≡N (Ramesha et al., 2019).

Figure 3. Fourier-Transform Infrared (FTIR) spectra of NR.

2.2. Experimental setup and selection of process parameters

The machining is carried out with a custom-made two-axis CNC suspension-type AWJ machine. The authors of the current work developed a cryogenic setup (refer to Figure 4a). The suspension-type AWJ machine setup was developed by previous researchers Devineni et al. (2003) (refer to Figure 4b), having a maximum pressure limit of 1,060 bar (15,000 psi). The machining is performed under conventional and cryogenic conditions, as shown in Figures 5a-5c. The cryogenic continuous flow technique is used in the current work (Figure 4a) to provide the cryogenic environment during the machining because it has been widely used by previous researchers, as summarised in the latest review article (Maurya et al., 2021). Cryogenic coolants are made to flow continuously over the surface of the elastomer during or before machining. In this direction, a continuous liquid nitrogen (LN₂) flow is provided on the top machining surface of the NR workpiece to obtain the cryogenic condition. The workpiece surface temperature is measured using an Infrared thermometer (make: Fluke, model: Fluke-59 Max NA). The temperature is measured three times at the top surface of the NR during and immediately after the machining (refer to Figure 5d) to get the temperature range of cryogenic machining. It is found that the temperature range during cryogenic machining is between −12°C and −35 °C.

Figure 4. (a) Schematic of cryogenic setup and (b) photograph of the custom-made suspension-type AWJ machine setup, along with the cryogenic setup.

Figure 5. Photographs of (a) machined NR under conventional condition, (b) machined NR under cryogenic condition, and (c) temperature measurement using an infrared thermometer gun during and immediately after the machining under cryogenic conditions.

The fixed process parameters used in the current work are shown in Table 1. According to the most recent research (Zhang et al., 2022a), when SOD is 1 mm, a jet impingement angle of 90° can reduce the kerf width to 164.9% of the nozzle diameter. The primary goal of choosing two machining passes was to produce a workpiece with a superior and waviness-free kerf profile. A low-cost abrasive frequently used in suspension-type AWJ machining, garnet with an 80-mesh size, was chosen for the machining (Louis et al., 2007). A further benefit of employing garnet abrasives over other commercially available abrasives is that they result in less AWJ nozzle wear (Haghbin et al., 2015). As noted by earlier researchers (Syazwani et al., 2016), tungsten carbide has the highest abrasion toughness compared to other materials. As a result, a specially designed 1 mm orifice diameter tungsten carbide nozzle (shown in Figure 5a) was designed and developed to drive the suspension mixture onto the surface of NR. The previous researcher’s recommendation to use AWJs for high-quality and efficient machining led to the choice of a 1 mm orifice diameter for the current work (Jegaraj & Babu, 2005). As per the earlier researchers’ studies (Ayed et al., 2017; Khanna et al., 2021; Pusavec et al., 2014), the selection of cryogenic spray conditions plays a vital role during machining. A stainless-steel hose nozzle is used to spray the LN₂ onto the surface of NR during machining under cryogenic condition, as shown in Figure 5b. Several hits and trials were conducted to set the cryogenic spray nozzle and get the proper flow of LN₂. In this direction, the cryocan gauge pressure (Figure 4a) and LN₂ jet spray angle were set at 0.6 bar and 80°, respectively, to avoid freezing the suspension mixture near the machining zone and facilitate the unhindered machining during the cryogenic condition. As per the previous researchers (Deepak & Devineni, 2017), it is desirable to consider the mass percentage of Zycoprint polymer (ω_p) and garnet (ω_a) to be 0.8% and 3%, respectively, to achieve a stable suspension mixture free from the sedimented garnet particles. Trial tests were also conducted to confirm the percentages. The major process parameters selected by Kowsari et al. (2017) and Tamannaee et al. (2015) during AWJ machining of polymethyl methacrylate and talc-filled co-polymer, respectively, were the number of machining passes, V_f, jet impingement angle, WJP, SOD, and impact velocity. Preliminary test experiments (Maurya et al., 2022) were conducted on major process parameters, namely, WJP, V_f, and SOD, to analyze their impact on the suspension-type AWJ machining of NR. It was found from the preliminary experiments that low WJP (150 bar), high V_f (60 mm/min), and large SOD (2 mm) within the prescribed limit caused a severe taper in the slot machined between the top and bottom surfaces of the NR workpieces. Conversely, low WJP (150 bar), low V_f (40 mm/min), and low SOD (1 mm) caused less material to be removed from the top and bottom surfaces of NR. The V_f affected MRR the most, whereas WJP and SOD showed less significance. The current study selected the variable process parameters and their levels in this direction, as listed in Table 2.

Table 1. Fixed process parameters during the suspension-type AWJ machining of NR

Parameters	Values
Type of operation	Slot cutting
Length of cut	20 mm
Jet impingement angle	90°
Number of passes	2
Abrasive	Garnet
Abrasive particle diameter	150 μm
Abrasive material hardness	6.5 to 7.5 Mohs hardness
Orifice	Tungsten carbide
Nozzle orifice diameter	1 mm
Focusing nozzle	Stainless steel
LN2 jet pressure	0.6 bar
LN2 spray angle	80° (with nozzle axis)
LN2 flow rate	1.5 LPM
Mass percentage of Zycoprint polymer (ωp)	0.8%
Mass percentage of Garnet (ωa)	3%

Table 2. Variable process parameters during the suspension-type AWJ machining of NR

Symbol	Process parameters	Units	Levels
			1	2	3
WJP	Water jet pressure	Bar	150	200	250
Vf	Traverse rate	mm/min	40	50	60
SOD	Stand-off distance	Mm	1	1.5	2

3. Methods

The section includes details about the Taguchi analysis, evaluation of performance parameters, and ANOVA.

3.1. Taguchi analysis and evaluation of performance parameters

The statistical Design of Experiments (DOE) tool, Taguchi analysis, is applied to the variable process parameters and their levels to get the L₂₇ OA of experiments with the least number of trials and low variance. As per the suggested number of experiments, the performance parameters (KT_R and MRR) are computed for each machined slot under conventional and cryogenic conditions. The KT_R is computed using equation 1.

(1) $\mathit{K}{T}_{\mathrm{R}}=\mathit{K}{W}_{\mathrm{T}}/\mathit{K}{W}_{\mathrm{B}}$(1)

where, KW_T and KW_B are the mean Top and Bottom Kerf widths, respectively. The KW_T and KW_B are measured using a Toolmaker’s Microscope (Model: TM-505B, make: Mitutoyo, least count: 0.005 mm and magnification of 10×) (refer to Figure 6a) and Optical Microscope (Model: DP 22, make: Olympus, magnification of 50×) (refer Figure 6b). In each machined slot, five different readings from the top and bottom kerf are taken using the Toolmakers and Optical Microscope, and the average KW_T and KW_B are computed to reduce the error.

Figure 6. Kerf width measurement using (a) Toolmaker microscope and (b) Optical microscope.

The MRR is computed using equation 2 in terms of the volume of the material removed by the suspension-type AWJ machining in unit time.

(2) $\mathit{MRR}\left(\mathit{m}{m}^3/\mathit{min}\right)=\mathit{t}\times \mathit{w}\times V_f$(2)

where, $\mathit{t}$ = thickness of the workpiece (15 mm), $\mathit{w}$ = average width of the kerf (mm) = (KW_T +KW_B)/2 and V_f = traverse speed of AWJ (mm/min).

The performance parameters and their signal-to-noise (S/N) ratios under conventional and cryogenic conditions, as per recommended Taguchi’s L₂₇ experimental runs, are summarised in Table 3. As stated in Equations 3(3) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{smaller}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n{y}_i^2\right]$(3) and 4(4) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{larger}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n\frac{1}{y_i^2}\right]$(4) , the logarithm of the performance parameter function is used to estimate the S/N ratios for KT_R and MRR, respectively (S. Kumar et al., 2020).

Table 3. Performance parameters and S/N ratios of KT_R and MRR under conventional and cryogenic conditions

Sr. No.	WJP (bar)	Vf (mm/min)	SOD (mm)	KTR		MRR (mm^3/min)		S/N ratios (KTR) (dB)		S/N ratios (MRR) (dB)
				Con.	Cryo.	Con.	Cryo.	Con.	Cryo.	Con.	Cryo.
1	150	40	1.0	1.79	1.44	795.12	801.09	−5.06	−3.17	58.01	58.08
2	150	40	1.5	1.71	1.48	856.80	884.16	−4.66	−3.41	58.66	58.94
3	150	40	2.0	1.69	1.52	918.35	947.82	−4.56	−3.64	59.27	59.54
4	150	50	1.0	2.20	1.77	995.72	934.95	−6.85	−4.96	59.97	59.42
5	150	50	1.5	2.21	1.80	1018.47	1025.55	−6.89	−5.11	60.16	60.22
6	150	50	2.0	2.28	1.82	1041.05	1091.85	−7.16	−5.21	60.35	60.77
7	150	60	1.0	2.13	2.23	1346.15	1132.43	−6.57	−6.97	62.59	61.09
8	150	60	1.5	2.28	2.24	1308.21	1225.17	−7.16	−7.01	62.34	61.77
9	150	60	2.0	2.54	2.26	1270.07	1288.80	−8.10	−7.09	62.08	62.21
10	200	40	1.0	1.57	1.25	926.09	969.90	−3.92	−1.94	59.34	59.74
11	200	40	1.5	1.47	1.29	973.92	1035.51	−3.35	−2.22	59.78	60.31
12	200	40	2.0	1.42	1.30	1021.61	1081.68	−3.05	−2.28	60.19	60.69
13	200	50	1.0	1.92	1.39	1141.22	1140.57	−5.67	−2.87	61.15	61.15
14	200	50	1.5	1.87	1.42	1146.65	1209.30	−5.44	−3.05	61.19	61.66
15	200	50	2.0	1.86	1.44	1151.90	1253.78	−5.40	−3.17	61.23	61.97
16	200	60	1.0	1.95	1.57	1498.89	1372.55	−5.81	−3.92	63.52	62.76
17	200	60	1.5	2.00	1.60	1440.17	1439.10	−6.03	−4.09	63.17	63.17
18	200	60	2.0	2.11	1.62	1381.24	1476.54	−6.49	−4.20	62.81	63.39
19	250	40	1.0	1.72	1.37	1018.87	1099.11	−4.72	−2.74	60.17	60.83
20	250	40	1.5	1.56	1.38	1052.85	1147.20	−3.87	−2.80	60.45	61.20
21	250	40	2.0	1.46	1.38	1086.68	1175.97	−3.29	−2.80	60.73	61.41
22	250	50	1.0	2.15	1.44	1238.99	1296.57	−6.65	−3.17	61.87	62.26
23	250	50	1.5	2.03	1.45	1227.10	1343.48	−6.15	−3.23	61.78	62.57
24	250	50	2.0	1.95	1.45	1215.03	1366.13	−5.81	−3.23	61.70	62.71
25	250	60	1.0	2.24	1.53	1594.36	1553.27	−7.01	−3.70	64.06	63.83
26	250	60	1.5	2.24	1.55	1514.85	1593.63	−7.01	−3.81	63.61	64.05
27	250	60	2.0	2.29	1.55	1435.14	1604.84	−7.20	−3.81	63.14	64.11

*Con. = conventional and Cryo. = cryogenic

(3) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{smaller}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n{y}_i^2\right]$(3)

(4) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{larger}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n\frac{1}{y_i^2}\right]$(4)

where, n and y correspond to the number of machined slots and objective function. When the “smaller is better” criterion is significant, or when the response function is to be minimised, Equation 3(3) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{smaller}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n{y}_i^2\right]$(3) offers a substantial solution, whereas when “larger is better” or when the response function is to be maximised, Equation 4(4) ${\left[\mathrm{S}/\mathrm{N}\right]}_{\mathit{larger}\mathit{is}\mathit{better}}=-10\ast \log \left[\frac{1}{n}{\sum }_{i=1}^n\frac{1}{y_i^2}\right]$(4) provides a significant solution. In order to better reflect the performance feature, the Taguchi analysis adjusts the chosen objective functions in proportion to the S/N ratios. The factorial effect curves for the S/N ratios of KT_R and MRR under conventional and cryogenic conditions are plotted to determine the optimal levels of the process parameters corresponding to the highest S/N ratios. Higher S/N ratio values indicate control factor settings that reduce the impact of noise components.

3.2. ANOVA

ANOVA, a valuable method to quantify the effect of each process parameter on the performance parameters, is used to identify the significant process parameters. A statistical ANOVA table includes terms such as percentage contribution and P-value. The P-value is defined at a confidence interval of 95% for all the performance parameters. The ANOVA analysis consists of both linear and interaction terms. The results with a P-value ≤0.05 are considered statistically significant. The analysis is carried out using MINITAB 22 software.

4. Results and discussions

The details of S/N ratio analysis, ANOVA, the influence of cryogenic condition during suspension-type AWJ machining of NR and optimal process parameters obtained from experimental and Taguchi analysis results are discussed in upcoming sub-sections.

4.1. S/N ratios analysis

The S/N ratio table is analysed to establish the process parameters’ optimal levels (settings) and the accompanying performance characteristics. Table 3 displays the performance parameters and the corresponding S/N ratios. The highest S/N ratio values from the experimental table are chosen, as explained in the previous section, irrespective of whether the “smaller is better” or “larger is better” to determine the associated optimal process parameters and levels used in the experiments. The optimal values of KT_R obtained from experiments under conventional and cryogenic conditions are 1.42 and 1.25, respectively, with the corresponding S/N ratios of −3.05 and −1.94, respectively. Similarly, the optimal values of MRR under conventional and cryogenic conditions are 1594.36 mm³/min and 1604.84 mm³/min, respectively, with the corresponding S/N ratios of 64.06 and 64.11, respectively.

4.2. ANOVA

The ANOVA table obtained from the Taguchi analysis determines the significant process parameters and their percentage contribution to the performance parameters under conventional and cryogenic conditions. The ANOVA results obtained under conventional and cryogenic conditions are consolidated in Table 4. A thorough investigation of the results obtained from the ANOVA table under both conditions is discussed in sub-sections 3.2.1 and 3.2.2, respectively.

Table 4. ANOVA results for KT_R and MRR under conventional and cryogenic conditions

Source	KTR				MRR
	Con.		Cryo.		Con.		Cryo.
	% contribution	P-value	% contribution	P-value	% contribution	P-value	% contribution	P-value
WJP	15.78	<1×10^−3	47.37	<1×10^−3	16.92	<1×10^−3	37.92	<1×10^−3
Vf	76.50	<1×10^−3	41.67	<1×10^−3	79.46	<1×10^−3	56.00	<1×10^−3
SOD	0.34	<1×10^−3	0.44	<1×10^−3	0.06	<1×10^−3	5.08	<1×10^−3
WJP × Vf	0.74	<1×10^−3	10.34	<1×10^−3	0.74	<1×10^−3	0.01	<1×10^−3
WJP × SOD	1.96	<1×10^−3	0.10	0.04	0.33	<1×10^−3	0.88	<1×10^−3
Vf × SOD	4.61	<1×10^−3	0.03	0.44	2.87	<1×10^−3	0.11	<1×10^−3
Error	0.07		0.05		0.02		0.00
Total	100.00		100.00		100.00		100.00

*Con. = conventional and Cryo. = cryogenic

4.2.1. Under conventional condition (at room temperature)

The percentage contributions of factors WJP, V_f, and SOD to the average KT_R under conventional machining condition are 15.78%, 76.50%, and 0.34%, respectively. The percentage contribution of SOD is vital when considering the interaction terms (WJP × SOD, 1.96%) and (V_f × SOD, 4.61%), even though the percentage contribution of the individual is less. All terms, including individual (WJP, V_f, and SOD) and interactions (WJP × V_f, WJP × SOD, and V_f × SOD), are significant. At the same time, the contribution percentages of factors WJP, V_f, and SOD to the average MRR under conventional condition are 16.92%, 79.46%, and 0.06%, respectively. The percentage contribution of the interaction term V_f × SOD (2.88%) is more than WJP × V_f and WJP × SOD (0.74% and 0.33%). All terms, including individual (WJP, V_f, and SOD) and interactions (WJP × V_f, WJP × SOD, and V_f × SOD), are significant. These findings demonstrated that V_f significantly impacted the average KT_R and MRR under conventional machining conditions more than WJP and SOD. The mean S/N ratios for KT_R and MRR under conventional condition are given in Table 5. Based on the ranks of the parameters in Table 5, it is clear that V_f has the strongest influence on KT_R and MRR under conventional condition, whereas SOD has the least influence. The ANOVA results align with the ranks obtained from the mean of S/N ratios of KT_R and MRR. The factorial plots for the mean S/N ratios for KT_R and MRR under conventional conditions are depicted in Figures 7a-7b, respectively. It can be observed from the plots that high S/N ratios mitigate the impact of noise factors. The factorial plots show that the KT_R at conventional condition is minimised at the WJP = 200 bar, V_f = 40 mm/min, and SOD = 1.5 mm, which correspond to the highest S/N ratios. Conversely, the MRR at conventional condition is maximised at WJP = 250 bar, V_f = 60 mm/min, and SOD = 2.0 mm. This signifies that the ANOVA results align with the ranks obtained from the mean of S/N ratios of KT_R and MRR under conventional machining conditions.

Figure 7. Factorial effects plots for the mean of S/N ratios of (a) KT_R and (b) MRR under the conventional condition.

Table 5. Response table for the mean of S/N ratios for KT_R (smaller is better) and MRR (larger is better) under conventional and cryogenic conditions

Level	KTR						MRR
	Conventional			Cryogenic			Conventional			Cryogenic
	WJP	Vf	SOD	WJP	Vf	SOD	WJP	Vf	SOD	WJP	Vf	SOD
1	−6.33	−4.05	−5.80	−5.17	−2.77	−3.71	60.38	59.62	61.18	60.22	60.08	61.01
2	−5.01	−6.22	−5.61	−3.08	−3.77	−3.85	61.37	61.04	61.23	61.64	61.04	61.54
3	−5.74	−6.82	−5.67	−3.25	−4.95	−3.93	61.94	63.03	61.27	62.55	62.93	61.86
Delta	1.32	2.77	0.19	2.09	2.18	0.22	1.56	3.41	0.09	2.33	2.85	0.85
Rank	2	1	3	2	1	3	2	1	3	2	1	3

*Bold ranks indicate the most significant process parameters.

The mathematical interpretation for increased MRR at increased V_f refers to the fact that the MRR is directly proportional to the V_f and average width of the kerf ($\mathit{w}$), as t is constant (refer to Equation 2(2) $\mathit{MRR}\left(\mathit{m}{m}^3/\mathit{min}\right)=\mathit{t}\times \mathit{w}\times V_f$(2) ). At low V_f, the average kerf width is high, and at higher V_f, the average kerf width is low, but the magnitude of change in w is less when compared to the magnitude of increase in V_f. Therefore, the extent of the rise in MRR with increased V_f is more than WJP and SOD, as shown in Figure 7b. As per the physical interpretation, as V_f increases, the collision of fresh garnet abrasive particles with the eroded particles decreases causing improved MRR. Due to this, the kinetic energy of the garnet abrasive particle improved and resulted in parallel Top and Bottom Kerf widths and reduced KT_R. The cutting mechanism is improved at increased V_f, resulting in more material being removed from the target material in a shorter time.

4.2.2. Under cryogenic condition (in LN₂ environment)

On the other hand, under cryogenic condition, the percentage contributions of factors WJP, V_f, and SOD to the average KT_R are 47.37%, 41.67%, and 0.44%, respectively, as shown in Table 4. WJP and V_f are thus the factors that had the most significant impact on the average KT_R while using LN₂ during machining. The terms WJP, V_f and SOD, WJP × V_f, and WJP × SOD are significant, and V_f × SOD is insignificant (P-value >0.05). At the same time, the percentage contribution of factors WJP, V_f, and SOD to the average MRR under cryogenic condition are 37.92%, 56.00%, and 5.08%, respectively. These results showed that the most influential factor on the average MRR under cryogenic condition is V_f, then WJP and SOD. The percentage contribution of the interaction term WJP × SOD (0.88%) is more than WJP × V_f and V_f × SOD (0.01% and 0.11%). All terms, including individual (WJP, V_f, and SOD) and interactions (WJP × V_f, WJP × SOD, and V_f × SOD), are significant.

Based on the ranks of the process parameters in Table 5, it is clear that similar to conventional condition, V_f has the most substantial influence on KT_R and MRR under cryogenic condition. The mathematical and physical interpretation behind the significant influence of V_f on the KT_R and MRR under cryogenic condition is the same as that explained for machining NR under conventional condition (in section 3.2.1). The factorial plots for the mean S/N ratios of KT_R and MRR under cryogenic condition are depicted in Figures 8a-8b, respectively. The KT_R at cryogenic condition is minimised at WJP = 200 bar, V_f = 40 mm/min, and SOD = 1.0 mm and MRR is maximised at WJP = 250 bar, V_f = 60 mm/min, and SOD = 2.0 mm. Under the cryogenic condition, the NR changes its behaviour from elastic to a tough phase, which results in the possibility of ductile wear mode erosion. Furthermore, the increased V_f decreased the case of a mutual collision of abrasive particles in AWJ. Hence, the abrasive particles eroded the tough surface of NR with more kinetic energy, resulting in reduced KT_R and enhanced material removal.

Figure 8. Factorial effects plots for the mean of S/N ratios of (a) KT_R and (b) MRR under cryogenic condition.

4.3. Influence of cryogenic condition during machining of NR

During the suspension-type AWJ machining of NR under conventional condition, the decrease in both the KW_T and KW_B is caused by the increase in V_f, which decreased the exposure duration and resulted in increased KT_R; Wang and Wong (1999) obtained a similar trend. The researchers also said that the particle velocity varies at every jet cross-section from zero at the nozzle wall to a maximum in the jet core. The particle velocity distribution lines up with the jet’s energy or strength distribution. Therefore, the effective diameter of the jet, which in turn depends on the jet strength in the cutting zone and the target material (NR), determines the width of the kerf under both conventional and cryogenic conditions. The jet’s effective width or diameter grows as WJP rises, resulting in low KT_R under both conventional and cryogenic conditions. Under the cryogenic condition, the molecular chains of the NR become entirely unidirectional and stiffer. It leads to the energy-elastic behaviour of the NR with a high elastic modulus, which provides significant scope to attain ductile cutting. The material removal in NR during suspension-type AWJ machining occurred in the fully developed plastic zone by the cutting and ploughing actions, which are ductile cutting mechanisms, as depicted by researchers Nouraei et al. (2013). The angular and sharp-edged garnet abrasive particles are the primary cause of the cutting action, whereas the spherical garnet abrasive particles are responsible for the ploughing actions. On the other hand, in the deformation wear mode, the Garnet abrasive particle caused the plastic deformation of the NR workpiece. The occurrence of an increase in WJP resulted in the transfer of more kinetic energy to the abrasive particles, which strike over the cooled NR surface. It results in less taper between the top and bottom kerf surfaces under cryogenic condition compared to conventional condition. In essence, the shearing action produced by the abrasive particles in AWJ machining essentially accomplishes the cutting action. As V_f rises, the time span an abrasive-laden stream spends in contact with the target surface decreases. According to the literature (Ishfaq et al., 2019), V_f is typically the most significant contributor to MRR; therefore, this is expected.

4.4. Optimal process parameters obtained from experimental and Taguchi analysis

The levels and values of the process parameters corresponding to the optimal performance parameters obtained from experimental runs, depicted in Table 3, are consolidated in Table 6. After comparing the values of the optimal performance parameters under cryogenic and conventional conditions, it is found that KT_R is reduced by 11.97% under cryogenic condition, and MRR improved by 0.65% (refer to Table 7). Hence, it can be concluded that the suspension-type AWJ machining of NR under cryogenic condition decreases the KT_R and increases the MRR resulting in a better-quality cut and improved machining performance. The projected optimal process parameters obtained from the Taguchi analysis are displayed in Table 8 for the range of values considered for the current study. Under the conventional condition, the WJP at 200 bar, V_f at 40 mm/min, and SOD at 1.5 mm are capable of producing the lowest KT_R. In contrast, the WJP at 200 bar, V_f at 40 mm/min, and SOD at 1.0 mm can produce the lowest KT_R under cryogenic condition. While comparing the MRR under conventional and cryogenic conditions, it is found that experimental condition at 250 bar WJP, 60 mm/min V_f, and 2.0 mm SOD is desirable for the highest material removal.

Table 6. Consolidated optimal parameters of the suspension-type AWJ machining of NR, obtained from the experimental runs of Table 3

Machining condition	Performance parameters	Levels	WJP	Vf	SOD
Conventional	KTR = 1.42	2 1 3	200	40	2.0
	MRR = 1594.36 mm^3/min	3 3 1	250	60	1.0
Cryogenic	KTR = 1.25	2 1 1	200	40	1.0
	MRR = 1604.84 mm^3/min	3 3 3	250	60	2.0

Table 7. Percentage reduction and improvement in the optimal experimental result of KT_R and MRR in cryogenic over conventional suspension-type AWJ machining of NR

Performance parameters	Condition		Percentage reduction in optimal experimental KTR	Percentage improvement in optimal experimental MRR
	Conventional	Cryogenic
KTR	1.42	1.25	11.97	0.65
MRR (mm^3/min)	1594.38	1604.84

Table 8. Predicted optimal levels of the suspension-type AWJ machining of NR obtained from Taguchi analysis

Machining condition	Performance parameters	Levels	WJP	Vf	SOD
Conventional	KTR	2 1 2	200	40	1.5
	MRR (mm^3/min)	3 3 3	250	60	2.0
Cryogenic	KTR	2 1 1	200	40	1.0
	MRR (mm^3/min)	3 3 3	250	60	2.0

5. Conclusions

Experimental investigations of suspension-type AWJ machining of highly elastic, oil/chemical resistance and lightweight NR workpieces are carried out by considering different process parameters such as WJP, V_f, and SOD under conventional and cryogenic conditions to identify the machining performance in terms of KT_R and MRR. The following conclusions are elicited from the conducted study.

• The optimal KT_R and MRR obtained from experimental runs under conventional and cryogenic conditions are 1.42 and 1.25, and 1594.36 mm³/min and 1604.84 mm³/min, respectively.

• According to the ANOVA results, the percentage contribution of V_f is more on the KT_R (76.50%) and MRR (79.46%) than WJP and SOD during the suspension-type AWJ machining of NR under conventional condition. On the other hand, WJP is the most influencing parameter on KT_R and V_f on MRR, with percentage contributions of 47.37% and 56%, respectively, during the suspension-type AWJ machining of NR under cryogenic condition.

• The V_f is the highest influencing factor (having Rank 1) on the KT_R and MRR during suspension-type AWJ machining of NR under conventional and cryogenic conditions. The key reasons are improved kinetic energy due to less collision of garnet abrasive particles and the lesser change in the average width of the kerf compared to the magnitude of the increase in V_f, resulting in reduced KT_R and improved MRR.

• The cryogenic condition positively affects the suspension-type AWJ machining of NR. There is a significant percentage reduction in KT_R (11.97%) as compared to MRR (0.65%) while using the LN₂ during the suspension-type AWJ machining of NR. Under the cryogenic condition, the NR attains a high elastic modulus, providing significant scope for ductile cutting. It reduces taper between the top and bottom kerf surfaces and improves the material removal mechanism.

• The optimal process parameters conditions obtained from the Taguchi analysis to achieve the lowest KT_R are different for conventional and cryogenic conditions (i.e., WJP at Level 2, V_f at Level 1, and SOD at Level 2 for under conventional condition and WJP at Level 2, V_f at Level 1, and SOD at Level 1 for cryogenic condition). On the other hand, for the highest MRR, optimal process parameters conditions are the same for conventional and cryogenic conditions (i.e., WJP at Level 3, V_f at Level 3, and SOD at Level 3).

• The optimal machining performance (i.e., lowest KT_R and highest MRR) is obtained at 200 bar, 40 mm/min, and 1 mm, and 250 bar, 60 mm/min, and 2.0 mm, respectively, during the cryogenic condition from both experimental and Taguchi analysis results.

The influence of cryogenic condition on the machining of NR using suspension-type AWJ is beneficial in terms of improved dimensional stability and material removal rate. Further investigation is recommended by considering a wide range of machining parameters and many machined slots to enhance the performance during the machining of NR workpieces. Also, future work should find out the optimal process parameters under conventional and cryogenic conditions during the suspension-type AWJ machining of NR by considering both performance parameters simultaneously.

Acknowledgments

The paper’s authors want to thank the Manipal Academy of Higher Education, Manipal, for providing technical and financial support to carry out the experimental work. The authors also wish to thank S. K. Rubber industries, Kottara Chowki, Mangalore, Karnataka, India, for providing the nitrile rubber workpieces. Pradeep Kumar Shetty is grateful to MAHE, Manipal, for giving the Dr TMA Pai scholarship.

Disclosure statement

The authors declare that there are no relevant financial or non-financial competing interests to report.

Additional information

Funding

The Manipal Academy of Higher Education, Manipal, Karnataka, India, provided an intramural fund (MAHE/DREG/PhD/IMF/2019).