Sage Journals: Discover world-class research

Abstract

The fluid jet polishing (FJP) is an ultra-precision and non-contact type polishing process for generating high-quality surface finish in automotive dies manufactured from ductile materials like AISI H13 die steel. This paper presents a study on FJP polishing of AISI H13 die steel material using a custom-designed 3D-printed Ti6Al4V (Ti64) nozzle and silicon carbide (SiC) abrasive slurry. The study presents an effect of pressure, stand-off distance and polishing time on the surface finish while using water-based abrasive slurry mixed with water-soluble cutting oil. The Central Composite Design (CCD) method was used to develop the experimental plan. ANOVA analysis was performed to develop a regression model and find the contribution of each parameter on the percentage change in surface roughness. The regression model developed for the FJP process was validated by polishing the plate specimen of AISI H13 having $500 n m$ average initial surface roughness with the optimized parameters; pressure of $11.1 b a r$ , stand-off distance of $32 m m$ and polishing time of $174.6 s e c$ . The optimized FJP parameters were able to generate the surface roughness of $240 n m$ in one of the confirmation tests. The results of percentage change in surface roughness obtained from the confirmation experiments were found to be very close to the predicted results which validated the regression model developed for polishing of AISI H13 using FJP process. In addition, the optimal feed rate of 20 mm/min was used to polish $30 m m \times 30 m m$ plate made of AISI H13, the surface roughness was reduced from 500 nm to 130 nm in 99.2 min showing the reduction in peaks and valleys with surface improvement of 74% which results in enhanced polishing efficiency.

Keywords

Fluid jet polishing AISI H13 SiC abrasive slurry surface roughness

Introduction

The quality of die and mould surfaces plays an important role in the product quality as well as the productivity of the manufacturing industries. The quality, cost and manufacturing lead time of dies and moulds directly affect the manufacturing productivity and hence the economics of the manufacturing sector. The productivity in dies and moulds manufacturing depends on factors such as the ability to machine them to the specified dimensional accuracy and the required surface finish. The precision dies are used to fabricate components having good mechanical properties, i.e., toughness, hardness and sufficient resistance to withstand high-temperature fatigue cycles. AISI H13 die steel material is one of the preferred materials for precision dies.

The dies used in different applications have complex geometries and curved sculptured shapes.¹ It is desirable to design and manufacture die and moulds parts such that they are free from surface defects, cracks and have better service life.² The dies used to produce the components need to have surface roughness in the range of microns and nanometer scale. Thus, surface polishing is one of the critical processes in die manufacturing which consumes around 30%–50% more time as compared to the machining processes.³ The conventional polishing process, i.e., grinding, lapping and manual polishing are inaccurate, noisy and hazardous processes.⁴ At the same time, the modern die materials are quite hard that require the use of automated polishing procedures. Automation of polishing processes can reduce the lead times for dies and moulds manufacturing and enable the manufacturer to achieve better surface quality with enhanced production efficiency.

Most of the automated techniques used for surface polishing of open dies and moulds are still the slowest ones with the polishing time in certain cases constitute a significant percentage of overall die manufacturing time. The advances in die-polishing technologies have helped to minimize such problems to a certain extent and have enabled the achievement of better surface finishing. But these processes still have scope for improvements to make them more efficient in the present context of demand for higher manufacturing productivity and the stringent requirement of product quality.

Many researchers have proposed various non-conventional post-processing techniques such as chemical polishing (CP) and electropolishing (EP), to improve the surface quality.^5,6 Some have even used integrated techniques combining with electrical polishing (EP), mechanical polishing (MP) and ultrasonic methods,⁷ or use of abrasive jet polishing (AJP)⁸ and magnetorheological finishing process (MRF)⁹ to improve the surface quality. Compared with these non-conventional polishing processes, FJP process is better suited for polishing of dies and moulds, because of better material removal rate, ability to generate small tool influence function (TIF), no heat generation during polishing and better tool life.^10–12

The concept of FJP process was first introduced by Fahnle et al.,¹³ wherein a pressurized jet of slurry (mixture of water and abrasive particles) was discharged through a small nozzle at a pressure of $2.0 b a r$ to $20.0 b a r$ and was allowed to impinge upon the substrate surface. The pressurized slurry jet energies the abrasive particles, generally having average particle sizes from $0.1 μ m$ to $10 μ m$ .¹⁴ The abrasive particles in the slurry jet strike against the workpiece surface and remove the surface irregularities by the means of cracking, cutting, shearing and penetration.¹⁵

Several investigations have been made to improve the surface finish quality achieved through FJP process, which has led to various findings in terms of the suitable range of the process parameters so as to enhance the stability of the effective material removal zone (also called tool influence function or TIF) and the material removal rate. Tsai et al.¹⁶ used Taguchi approach to optimize the process parameters for minimizing the surface roughness for SKD61 die material using silicon carbide (SiC) abrasive and use of water-soluble machining oil. It has been observed that the fluid pressure and impact angle have a significant effect on the surface roughness, in addition, to stand off distance.

Messelink et al.¹⁷ and Yu et al.¹⁸ improved the polishing efficiency by the addition of pressurized gas into the fluid jet but have achieved severe surface defects and poor surface finish on the polished surface than normal FJP process. Shiou and Asmare¹⁹ developed an innovative rotary multi-jet polishing tool to enhance the surface roughness of Zerodur optical glass and optimized the process parameters using the design of experiment (DOE) approach. Wang et al.²⁰ developed a multi-jet polishing process to enhance the polishing efficiency of the existing FJP process and achieve a higher material removal rate without degrading the surface integrity. Anbarasu et al.²¹ developed a multi-staged low-pressure abrasive slurry jet polishing technique for aluminium oxide (Al₂O₃) and cerium oxide (CeO₂) abrasive particles. Multiple passes of FJP process were used to compare the effect of variation of abrasive particle sizes along with the variation of transverse speed of nozzle and impact angle. The authors observed that the average particle size of Al₂O₃ abrasive offered a better surface finish when compared to the cerium oxide (CeO₂) abrasive particles of the same size in finishing pass. Cheung et al.²² developed a curvature adaptive multi-jet fluid polishing method (CAMJP) for freeform surfaces which used a technique of controlling the individual nozzle pressure according to the local surface curvature under the jet.

To maximize the surface finish obtained in FJP process, a better understanding about the interaction of jet pressure, stand-off distance and polishing time is required. In an FJP study on polishing of AISI H13 die steel, Wang et al.²³ studied the effect of stand-off distance on the surface generation mechanism and material removal characteristics. It was found from the experimental results that the suitable range of stand-off distance to polish ductile and brittle material were 25–35 mm and 8 mm, respectively, with 1.1 mm nozzle diameter. The concept of sliding-grinding mechanism presented by Tsai et al.¹⁶ shows that the compound additive comprising water and water-based machining oil in FJP process assists in achieving a better surface finish compared to when only water is used as the abrasive carrying media. Moreover, the higher value of stand-off distance in FJP process, if used along with the slurry jet that contains water-soluble machining oil can result in a stable and symmetrical TIF profile on the polished surface.

It can be seen from the literature that the process parameters that influence the quality and the reliability of FJP process include fluid pressure $(P)$ , stand-off distance $(D)$ , polishing time $(T)$ , impact angle $(α)$ , type of abrasive particles, abrasive particle shape and size, type and quantity of machining oil in abrasive slurry, slurry concentration ratio, nozzle shape, nozzle material, hardness and other mechanical properties of the substrate material. Out of the enlisted process parameters, the most critical are the jet pressure, stand-off distance and the polishing time. So, in the present work, an effort has been made to investigate the simultaneous effect of jet pressure, stand-off distances and polishing time with abrasive slurry containing machining oil using a 3D-printed tool in order to improve the surface finish of tool influence function (TIF) zone in polishing AISI H13 die material. In addition, the optimum values of these process parameters need to be exclusively determined for selecting the suitable range of feed rates for polishing medium and large-sized die surfaces of AISI H13 die steel with improved polishing efficiency.

Material removal mechanism in FJP process

The material in FJP process is removed by the impact of the abrasive particles on workpiece surface. These abrasive particles are accelerated by the drag force of carrier medium which can be lubricating oil and/or water. The particle impingement energy and the resultant particle velocity decide the mechanism of material removal and the quality of the surface finish. The kinetic energy of the abrasive particles depends mainly on the stand-off distance, average particle size and jet pressure.¹⁴ The slurry particles impinging on the surface at an impact angle will lead to the different types of material removal mechanisms, i.e., ploughing, cracking, penetration, cracking and cutting. These different material removal mechanisms are shown in Figure 1.

Figure 1.

Material removal mechanism of FJP process.

The velocity of fluid jet is equally distributed inside the nozzle but produces a considerable velocity difference as it leaves the nozzle, compared to the boundary layer formed by the environmental drag. There are basically three regions/domain when fluid jet flows out from the nozzle²⁴:

Potential Core Region: The flow within this region is in irrotational motion where velocity components are highly stream-lined and carry sufficient amount of kinetic energy for cutting action/workpiece component.

Main Region: Within this segment, the jet axial velocity and dynamic pressure reduce gradually, and turbulence characteristics are highly dominating.

Diffused Droplet Region: In this region, slurry jet and environment medium (air) are thoroughly mixed together by which jet loses its cohesive strength/bond and forms water droplets inside the slurry domain.

The potential core region is applied for cutting action, diffused droplet region is applied for dust-laying and aspirating and the jet in the main region is applied for cleaning and surface finish operations.

There will be only cutting and ploughing type of material removal at lower kinetic energy and smaller average particle sizes.²⁵

In FJP study, the stand-off distance plays an important role and influences the vertical as well as horizontal components of the particle velocity in the shear zone which in turn affects the shape of the tool influence function (TIF) zone. The smaller stand-off distance at higher pressure ranges tends to generate a ‘W-shaped’ sectional profile.²⁶ This is because of the reason that the centre point of the polishing tool tends to be the stagnation point, where the pressure is maximum with almost zero velocity. The velocity is maximum at a radial distance equal to the nozzle outlet diameter. Hence, the centre of the fluid jet has the minimum material removal rate due to the low kinetic energy of the jet slurry.

For an intermediate range of stand-off distance, the height of the central peak of ‘W-shaped’ sectional profile suppresses. With further increase in stand-off distance, the shape of the sectional profile transforms to ‘U-shape’ because when fluid jet eject from the nozzle outlet, the cohesive bond between the abrasive particle becomes weaker due to the environmental drag/effect which in-turn increases the random motion of abrasive particles in the slurry jet domain and hence, the overall kinetic energy of abrasive particle increases with increasing stand-off distance converting the ‘W-shape’ to the ‘U-shape’ material removal characteristics. The ‘U-shaped’ section has a relatively better surface finish in its central region compared to the ‘W-shaped’ section.

Experiment details

The experimental set-up of FJP process developed for 2–14 bar working pressure is shown in Figure 2. The FJP set-up consists of a 3-axis CNC vertical milling centre attached with a BT40 coolant inducer tool holder, a geared stirrer with a range of 180–650 rpm, a vertical multistage centrifugal pump of 14.75 bar, a diaphragm type pulsation dampener, ball valves for regulating slurry flow, a pressure regulating valve (PRV), a non-return horizontal and vertical valves, a Y-stainer filters, pressure sensors (marked as PS1), three pressure gauges (marked as PG1, PG2 and PG3). The x, y and z-axis traverses of the set-up are 800 mm, 350 mm and 380 mm, respectively.

Figure 2.

Schematic diagram of experimental test rig used for study of FJP process.

The nozzle had a convergent-throat design with a 1.0 mm exit diameter and 3.0 mm as the nozzle throat length. The nozzle throat length was decided based upon reducing the effect of carrier fluid attachment against the outlet wall and to achieve the streamlined flow of the slurry jet.^24,27 In order to achieve the streamlined flow and better binding of the abrasive particles in the slurry jet for longer stand-off distances, HOCUT 795-H water-soluble high-performance machining oil was used. The presence of machining oil in the slurry jet also improves the sliding of the abrasive particle across the surface, thereby reducing the embedment effect that improves the surface finish as compared to the surface finish obtained using only water-based abrasive slurry.¹⁶

The experimental setup is designed such that the slurry jet remains vertical and strikes normal to the workpiece surface as shown in Supplementary Figure 1. The slurry, after striking, is collected in a stainless steel container mounted on the CNC table. The slurry is then passed through a number of filters and is returned to the main slurry tank for reuse. As per the working pressure range, it is assumed that there will be no cavitation effect in the set-up.

The surface roughness ( $R_{a}$ ) values were measured using SJ-400 Mitutoyo surface roughness profilometer with 0.8 mm cut-off length. The surface roughness was measured at three different locations and their average value was taken. The 3D topography of material removal profile was measured using Zeta-20 optical microscope. The hardness of the material was measured with the help of Mitutoyo micro-hardness tester, MVH-I with an accuracy (least count) of 0.0001 HV. The hardness tests were carried out with a load of 300 gm and a dwell time as 20 sec.

Pressure stability in the experimental test rig

There may be variations in the nozzle pressure in the experimental test rig due to pulsations generated from the slurry pump. The stability of the inlet pressure is an important factor in the FJP process to achieve the desirable surface finish.¹⁰ The variation in the pressure was observed at the inlet of the coolant inducer assembly for the planned range of pressure using a pressure sensor. In the present work, the pressure sensor was supplied by Applied Measurements Ltd, UK with a pressure range of 0–16 bar which was coupled with NI-USB-6002 data acquisition system to generate the data.

Design of experiment using CCD technique

The parametric study to determine the optimized polishing pressure, stand-off distance and polishing time to maximize the percentage change in surface roughness $(% Δ R_{a})$ in polishing AISI H13 die material is conducted using design of experiment (DOE) technique. These are the most critical parameters for transformation of W-shaped sectioned profile (central uncut material) to U-shaped sectional profile (uniform material removal) as reported by Wang et al.²³ The central composite design (CCD) technique of response surface methodology (RSM) is used in the present work with three variables and five levels of each variable. The CCD method gives the central repeatability of process parameters to ensure the accuracy of the developed data/model. The values of the control factors used in the CCD analysis are given in Table 1. The range of the three control parameters was identified from the literature^16,28–30 along with the initial trials which were conducted on the developed experimental set-up. The selection of abrasive material depends mainly on two factors. Firstly, on the material which will be finished taking into account its density and its physical properties. Secondly, the required level of surface roughness. The commonly used materials for the abrasive particle are silicon carbide, cerium oxide and aluminium oxide, out of which silicon carbide is having the best hardness of 9.5 (on Mohs scale). The values of the other fixed parameters used in the study, which are given in Supplementary Table 1, were selected based upon the literature,^14,23 and the initial experimental trials conducted on the set-up.

Table 1.

Control factors and their levels.

S. No.	Factors	Levels
S. No.	Factors	1	2	3	4	5
1.	Pressure, P (bar)	8	9	10	11	12
2.	Stand-off distance, $D$ (mm)	30	31	32	33	34
3.	Polishing time, $T$ (sec)	150	160	170	180	190

The design of experiment plan developed using CCD method for three control parameters and five levels had 20 set of experiments, shown in Table 2.

Table 2.

Experimental combinations and responses.

Exp. No.	$P$ $(b a r)$	$D$ $(m m)$	$T$ $(s e c)$	Initial average surface roughness $(n m)$ *	Experimental results for surface roughness $(n m)$					Predicted $% Δ R_{a}$ Eq. (2)
Exp. No.	$P$ $(b a r)$	$D$ $(m m)$	$T$ $(s e c)$	Initial average surface roughness $(n m)$ *	$X_{1}$	$X_{2}$	$X_{3}$	Average roughness	Experimental $% Δ R_{a}$ Eq. (1)	Predicted $% Δ R_{a}$ Eq. (2)	% Error
1	10	32	170	500 (4)	270	250	260	260	48	47.36	1.33
2	9	33	180	500 (7)	350	340	360	350	30	31.78	−5.95
3	8	32	170	500 (3)	320	310	330	320	36	33.84	6.00
4	10	32	170	500 (8)	250	260	270	260	48	47.36	1.33
5	9	31	180	500 (7)	280	260	270	270	46	46.03	−0.07
6	10	32	170	500 (9)	250	270	260	260	48	47.36	1.33
7	10	34	170	500 (4)	350	330	340	340	32	31.84	0.50
8	10	32	170	500 (9)	260	250	270	260	48	47.36	1.33
9	9	31	160	500 (10)	350	340	330	340	32	33.78	−5.58
10	10	32	170	500 (2)	250	240	260	250	50	47.36	5.27
11	11	31	160	500 (5)	320	300	310	310	38	35.53	6.49
12	10	30	170	500 (4)	310	320	300	310	38	38.34	−0.90
13	10	32	170	500 (7)	300	280	290	290	42	47.36	−12.77
14	9	33	160	500 (8)	360	350	340	350	30	30.53	−1.78
15	10	32	190	500 (10)	290	300	310	300	40	39.84	0.40
16	10	32	150	500 (9)	360	380	370	370	26	26.34	−1.31
17	11	33	160	500 (6)	280	270	290	280	44	43.28	1.63
18	11	33	180	500 (5)	280	270	260	270	46	44.53	3.19
19	12	32	170	500 (7)	280	260	270	270	46	48.34	−5.09
20	11	31	180	500 (9)	250	260	270	260	48	47.78	0.45

*Standard deviations are listed in parentheses.

In the present work, AISI H13 die steel workpiece material, mostly used in automotive, die manufacturing, aerospace, marine, etc., was machined to the dimension of $75 m m \times 70 m m \times 12 m m$ using a milling operation followed by grinding operation. Thereafter, the machined workpiece was cleaned using the acetone solution and dried before conducting experiments using FJP process. Each of the 20 trials was performed using the self-developed FJP setup shown in Figure 2. The polished spots (TIF) for each of the 20 trials are shown in Supplementary Figure 2. The individual trails were performed by varying the stand-off distance and polishing time for a constant jet pressure with a span of $8 m m$ between each consecutive trial. The jet pressure was set manually while the stand-off distance, the linear traverse in XY-plane over the workpiece plate, and the polishing time was automatically controlled on the CNC machine using a subroutine.

The surface roughness ( $R_{a}$ ) values were measured on each polished spot of approximately 2 mm diameter (TIF) using SJ-400 Mitutoyo surface roughness profilometer with a 0.8 mm cut-off length.

The measured $R_{a}$ values and their averages are shown in Table 2. Further, the percentage change in actual surface roughness $(% Δ R_{a})$ after polishing was determined using equation (1):

% Δ R_{a} = \frac{Initial average surface roughness - final average surface roughness}{Initial average surface roughness} \times 100

(1)

Analysis of variance (ANOVA)

In order to evaluate the effect of process parameters in terms of pressure, stand-off distance and polishing time, a set of experiments were designed. Central composite design (CCD) has been used for the experimental design as it allows for the prediction of the second-order behaviour of the responses for a wider range of process parameters.

A quadratic (second order) polynomial model can be expressed with the help of the following regression equation (2):

Y = β_{0} + \sum_{i = 1}^{k} β_{i} X_{i} + \sum_{i = 1}^{k} β_{ii} X_{i}^{2} + \sum_{j > 1} \sum_{ij}^{β} X_{i} X_{j} + ε

(2)

where Y represents the response variable, $β_{0}$ , $β_{i}$ and $β_{ij}$ are the constants, k represents the number of variables, $ε$ represents the random error and X_i, X_i² and X_iX_j are the first order, second order and interaction terms of the process parameters, respectively.³¹ The response data was analysed based on ANOVA.

ANOVA is a basic statistical method to analyse the experimental data to determine the proportion of influence of a factor or a set of factors on total variation. It is the most reliable method for evaluating the quality of a fitted model.³² In the present work, ANOVA is used to analyse the experimental results and identify the factors which have a significant effect on the output variables, i.e., pressure, stand-off distance and polishing time. The ANOVA test allows a comparison of more than two groups at the same time to determine the existence of a relationship between them. Based on ANOVA, non-significant terms were omitted on the basis of F value, while the significant terms were ordinarily acknowledged at 95% level of confidence.

Supplementary Table 2 lists the results of ANOVA and the F-test in order to evaluate the statistical significance of the aforementioned quadratic regression equation. The moderate Fisher's F-test value (F = 22.19) with a very low probability value (P-value) of prob > F (< 0.0001) demonstrates a high statistical significance of the regression model to represent the actual relationship between the obtained experimental data of percentage change in surface roughness ( $% Δ R_{a}$ ) and three variables. The P-values or probability values show the level of significance of each factor. Lower P-values indicate that the factor values have a higher probability of falling within the ranges which impact the outcome of the experiment. Other indicators include its degree of freedom (df) which is defined as k−1 (where k is number of levels), treatment sum of squares (SSTR), treatment mean squares (MSTR) and F statistics value.

Residual error was calculated statistically but has no physical influence on the experiment. It is, however, part of the ANOVA F-statistics test. The df for residual error is calculated as (total df) – (sum of all treatment df). The SSTR or SSE (sum of squared error in case of residual error) was calculated using equation (3):

S S E = (n_{1} - 1) s_{1}^{2} + (n_{2} - 1) s_{2}^{2} + \dots + (n_{j} - 1) s_{j}^{2}

(3)

where ‘j’ denotes each individual factor, ‘s²’ is the variance and ‘n’ is the number of observations in the jth factor.³³ In the present work, ANOVA was carried out using Minitab software.

Furthermore, the regression model was developed using the least square method in Minitab software, and quality of the fit quadratic regression model was assessed by R² (R-squared), R² (predicted) and R² (Adjusted) values.

The R² (R-squared) value indicates the goodness of fit; or how well the regression model fits the observed data. On the other hand, the relation between the regression model predictions for the new responses is indicated by the R² (predicted). The comparison with Adjusted R² is done to validate the model for new responses. The Adjusted R² adjusts the number of terms in the model. Its value increases only when the new terms improve the model fit. The adjusted R² and predicted R² should be within approximately 0.20 (20%) of each other.³⁴

The quadratic regression equation for the developed model is given in equation (4):

\begin{aligned} % Δ R_{a} = - 4940.10227 - 53.01136 P \\ + 213.98864 D + 21.26932 T - 1.56818 P^{2} \\ - 3.06818 D^{2} - 0.035682 T^{2} + 2.75000 P D \\ - 0.27500 D T \end{aligned}

(4)

The probability values for the control parameters in the quadratic analysis are found to be less than 0.05 which shows the model is significant. The value of the predicted

% Δ R_{a}

determined from equation (4) is also shown in Table 2. Supplementary Figure 3 shows the comparison between the predicted

% Δ R_{a}

and the experimental

% Δ R_{a}

values. It is evident from Supplementary Figure 3 that there is not much variation amid the experimental and the predicted values over the entire range of the study. Also, the

R^{2}

value for the regression model is 94.17%,

R^{2}

-Adj (adjusted

R^{2}

) is 89.92%, and the predicted

R^{2}

is 80.04%. The predicted

R^{2}

value for the model is in reasonable agreement with the

R^{2}

-Adj.

The percentage contribution, given by equation (5), depicts the parameters of significance in the regression analysis:

Percentage contribution of parameter = \frac{Sum of square of parameter}{Total sum of square of model} \times 100

(5)

The contribution of each parameter is shown in Supplementary Table 2. It can be seen from Supplementary Table 2 that the most contributing factor in the regression model is

T^{2}

followed by

D^{2}

, P and T. Supplementary Table 2 indicates that the polishing time is the most dominant factor contributing 41.46% to the predicted value of the percentage change in surface roughness

(% Δ R_{a})

compared to stand-off distance and pressure which contribute 23.02% and 22.45%, respectively. It can also be seen from the table that the pure error contribution is only 3.08%.

Results and discussion

The regression model given by equation (4) was used to analyse the interaction of the control parameters, i.e., pressure, stand-off distance, time and their effect on $% Δ R_{a}$ in a single nozzle FJP polishing. The ANOVA analysis indicates that three control parameters strongly influence $% Δ R_{a}$ .

Effect of pressure and stand-off distance on surface roughness

The surface plot and the contour plot given in Figure 3(a) and (b) shows the simultaneous effect of pressure ( $P$ ) and stand-off distance ( $D$ ) on $% Δ R_{a}$ at polishing time $T = 170 s e c$ . It is clear from Figure 3(a) and (b) that as the fluid pressure increases from $8 b a r$ to $12 b a r$ , there is an increase in $% Δ R_{a}$ value for the stand-off distance from $30 m m$ to $31 m m$ . This is attributed to the reason that the kinetic energy of the abrasive particles for this pressure range and the stand-off distance is just sufficient for removing the material at a rate that is able to provide a better surface finish.³⁵ As seen from the contour plot shown in Figure 3(b), the $% Δ R_{a}$ value is found to be in the highest range (more than 46%) corresponding to the pressure and the stand-off values of $11 b a r$ and $32 m m$ , respectively.

Figure 3.

(a) Surface plot and (b) contour plot showing variation of $% Δ R_{a}$ with $P$ and $D$ at $T = 170 \sec$ .

For the value of stand-off distance beyond $32 m m$ , the $% Δ R_{a}$ declines for the entire pressure range, as can be seen in Supplementary Figure 4. The decrease in the $% Δ R_{a}$ value is due to the fact that the simultaneous increase in slurry pressure and stand-off distance, increases the material removal rate (MRR) in the TIF zone.³⁵ The increase in MRR is due to the increase in the kinetic energy of the abrasive particles which strikes the workpiece surface with higher impact energy.³⁵ At the same time, the increase in stand-off distance makes the slurry jet unstable along with the randomization of the abrasive particles, which leads to the poor surface finish.

Effect of stand-off distance and polishing time on surface roughness

Supplementary Figure 5 shows the simultaneous effect of stand-off distance $(D)$ and time $(T)$ on $% Δ R_{a}$ . It is clear from Supplementary Figure 5 that the $% Δ R_{a}$ increases with the increase in stand-off distance up to a certain value between $30 m m$ to $32 m m$ , beyond which the $% Δ R_{a}$ starts to reduce. This trend is almost consistent for all time levels ranging from 150 sec to $170 s e c$ . The trend for $% Δ R_{a}$ shows that the polishing time between $170 s e c$ and $180 s e c$ is able to generate the highest surface finish for the entire range of stand-off distance from $31 m m$ to $32 m m$ . The decrease in $% Δ R_{a}$ at higher stand-off distance is due to over polishing with the simultaneous increase in these two parameters.²³

The interaction of stand-off distance and polishing time is plotted in Figure 4. Figure 4(a) and (b) indicates that the maximum $% Δ R_{a}$ is observed at stand-off distance of $32 m m$ for the time period between $170 s e c$ and $180 s e c$ .

Figure 4.

Variation in percentage change in surface roughness (% ΔR_a) at pressure $P = 10 bar$ with respect to stand-off distance, D and time, T: (a) surface plot and (b) contour plot.

Optimization

The aim of the present work is to find the FJP process parameters to maximize the $% Δ R_{a}$ value during polishing of AISI H13.

The optimization is done using CCD of Response Surface Methodology (RSM) with Minitab (v17) software. RSM optimizer is used to find the optimum input process parameters at which the best possible output responses are available. It helps to identify the factor settings that optimize a single response or a set of responses. It is useful in determining the operating conditions that will result in a desirable response. The single-objective optimization was carried at ‘unity’ of weight and importance. Here, ‘weight’ determines how the desirability is distributed on the interval between the lower/upper bound and the target, whereas the ‘importance’ determines the relative importance of multiple response variables. The ‘unity’ value of both factors represents that all parameters (P, D & T) are equally important in the study.

The CCD approach, used in the present work, has shown a near symmetric response of the $% Δ R_{a}$ for the entire selected range of the three control parameters, i.e., P, D and T. Supplementary Figure 6 shows the optimum value of 50.27% for $% Δ R_{a}$ obtained from the quadratic regression model for pressure, stand-off distance and time of $11.19 b a r$ , $32.06 m m$ and $174.65 s e c$ , respectively. As shown in Supplementary Figure 6, the vertical red lines indicate the best values for the factor affecting the response and horizontal dotted blue line shows the current setting of parameters and values.

The pressure stability in the experimental test rig was achieved to reduce the variations in the nozzle pressure in the experimental test rig due to pulsations generated from the slurry pump.

Supplementary Figure 7 shows the variation of the pressure for 11.1 bar in the test rig with respect to time. The measured inlet pressure at the coolant inducer assembly shows a variation in the range of ± 0.1 bar for the test run of 5 min duration. The minor variations in the mean pressure can be attributed to the minor turbulence due to the abrasive particles along with the minor variations in the rpm of the slurry pump.

Three confirmation tests were conducted at the control parameters to validate the study on the AISI H13 plate having an initial average (measurement taken at ten different locations) surface roughness of $500 \pm 10 n m$ . Supplementary Table 3 shows the experimental $% Δ R_{a}$ values and the value obtained from the regression model given by equation (4). The percentage error found between the experimental and the predicted $% Δ R_{a}$ values is around 3.36% which validates the accuracy of the developed regression model.

Figure 5(a) shows the initial average surface roughness of the AISI H13 plate, and Figure 5(b) shows the average surface roughness of $240 n m$ achieved in one of the confirmation tests. It is clear from the roughness profile of the polished surface that a significant reduction in the peaks and valleys was achieved using the FJP process as compared to the initial surface.

Figure 5.

Surface roughness profiles for (a) initial surface and (b) final surface at $P = 11.1 bar$ , $D = 32 mm$ and $T = 174.6 \sec$ .

Scanning electron microscope (SEM) images were used to analyse the surface characteristics of the polished TIF zones on the AISI H13 plate which were then compared with the initial surface of the specimens. Figure 6(a) and (b) shows the SEM images of the initial and FJP polished surfaces, respectively. Figure 6(a) shows the presence of the grinding marks and surface irregularities on the initial grinded flat surface of the AISI H13 which were the results of the grinding process. These grinding marks and major surface irregularities were minimized after applying FJP process with the optimum process parameters. Figure 6(b) shows a smoother surface finish achieved in the TIF zone. The polished surface also shows that there were no surface irregularities, no particle embedment of SiC abrasive particles and no cracks present on the surface.

Figure 6.

Surface morphology of (a) initial surface and (b) final FJP finished surface at $P = 11.1 bar$ , $D = 32 mm$ and $T = 174.6 \sec$ .

Area polishing using optimized process parameters

Selection of suitable feed rate ranges is important to achieve a good surface finish with minimum lead time. An attempt has been made to investigate the performance of feed rates on surface roughness of rough and fine grinded AISI H13 die steel workpiece with an initial average surface roughness of 1140 ± 15 nm and 500 ± 10 nm, respectively. The optimal values of pressure and stand-off distance were kept constant at 11.1 bar and 32 mm, respectively. The process parameters used for the experimentation are shown in Supplementary Table 4. A total of 32 feed trials were performed on both rough and fine grinded surfaces in a single raster manner as shown in Supplementary Figure 8(a) and (b), respectively. The travel path of nozzle on workpiece plate was kept as 10 mm in horizontal linear direction which was perpendicular to the direction of grinding. The consecutive gaps between trails were 7 mm and 10 mm in horizontal and vertical direction, respectively. The side step was taken as 1 mm with the diameter of TIF as 2 mm.

After experimentation, the arithmetic surface roughness values were measured on each feed spot, which are shown in Supplementary Figure 9.

Supplementary Figure 9 shows a similar trend for surface roughness on both the rough and fine grinded plates. A lower feed rate of 0.5 mm/min consumes more polishing time during experimentation as polishing time is inversely proportional to the feed rate. The polishing time is also depending on various polishing parameters such as polishing area, side steps and length of tool path. The polishing time (T) is proportional to the length of the tool path and inversely proportional to the feed rate:

T = \frac{L}{f}

(6)

where L is the length of the tool path (mm) and f is the feed rate (mm/min).

As the feed rate increases, the surface roughness value increases but there are some intermediate ranges where surface roughness value decreased and reflected a better surface finish. It can be seen from Supplementary Figure 9 that beyond the feed rate of 20 mm/min, the trend of roughness, $R_{a}$ increases because at higher feed rate the number of impacting particles per unit time is less which are insufficient for removing existing peaks and valleys on the substrate surface. However, this phenomenon occurred after 32 mm stand-off distance. As the stand-off distance is measured beyond 32 mm with 11.1 bar fluid pressure, the cohesive bond between the abrasive particles becomes weaker due to which haphazard mixing (unorganised flow) between slurry jet and abrasive particles creates a highly turbulent fluid stream. Hence, abrasive particles try to splash out from the slurry jet due to the action of higher environmental drag/effect.

To further improve the polishing efficiency, the three specimens of AISI H13 die steel of dimension $30 m m \times 30 m m$ were polished in four multiple passes using three optimal feed rates. The initial surface roughness of the plate was 500 ± 10 nm. The polishing time using the feed rates of 0.5 mm/min and 4 mm/min was too high. The increase in polishing time increases the number of impacting particle per unit time resulting in the formation of new grooves on the polished surface giving the higher surface roughness at lower feed rates. The polishing time with a feed rate of 20 mm/min was 99.2 min, which yields a better surface finish of 130 nm with the reduction of peaks and valleys on the substrate surface as seen in Figure 7. The reduction in surface roughness, $% Δ R_{a}$ , at a feed rate of 20 mm/min was 74. The mirror images of the entire surface of AISI H13 die steel plate before and after the FJP finishing with a feed rate of 20 mm/min are shown in Supplementary Figure 10 (a) and (b), respectively.

Figure 7.

Final surface roughness profiles after FJP finished with a feed rate of 20 mm/min on the entire plate surface.

The surface characteristics of polished die steel were analysed and compared with the initial surface using scanning electron microscopy (SEM) images. The SEM images of the initial surface and FJP polished surfaces are shown in Figure 8 (a) and (b), respectively. The grinding marks and surface irregularities were minimised as the feed moves in raster manner which results in a better surface finish on the substrate surface of AISI H13 die steel.

Figure 8.

Surface morphology of (a) initial surface and (b) final FJP finished surface.

Supplementary Figure 11(a) and (b) shows the measured contour and 3D topography of the finished surface profile. The profile shows a removal of peak to valley on substrate surface using 20 mm/min feed rate due to the reason that velocity is increased in the throat area of the nozzle and becomes maximum when slurry jets comes out of the nozzle. However, abrasive particles strike into the workpiece surface with a higher impact velocity. This impact velocity is further decomposed/divided into normal and tangential velocities.

As the abrasive particles impact the surface with an inclination angle (α), their normal component of velocity is responsible for penetration of the workpiece material, which leads to deformation wear. The horizontal component of velocity is responsible for the cutting action of the particles. In the cutting process, two types of abrasion occur: Type I in which the particle loses all its kinetic energy during the collision whereas, in Type II particle loses only a fraction of its kinetic energy and bounces off the surface. Therefore, the main abrasion mechanism in FJP is a ductile cutting of the surface under the lateral motion of the penetrating abrasives.¹⁴

The abrasion mechanism is also dependent on the hardness of the workpiece material which varies over the period of time. The hardness of the AISI H13 die steel was measured with the help of micro-hardness tester. The rough sample was initially prepared to the required dimensions using a milling operation. Then it was grinded to make the sample workpiece flat and to eliminate the uneven marks or peak-to-valley from the substrate surface so that the impression/indentation which is to be marked during the conduct of hardness test was visible on the surface. Thereafter, the workpiece was cleaned using the acetone solution and dried properly before doing the hardness test on Vickers micro-hardness tester, MVH-1. In this method, a conical diamond indenter was used to permanently deform the surface in the form of impression/indentation. The diagonals of this impression in the form of conical diamond shape/deformed material were properly measured to calculate the value of micro-hardness. The values of micro-hardness were measured at ten different locations on the initial surface and FJP polished surface of AISI H13 die steel and the average of the measured values were taken. The average hardness values before and after polishing of $30 m m \times 30 m m$ AISI H13 die steel plate was found to be 392.25 (392.2534) HV and 396.29 (396.2945) HV, respectively. Thus, the results of the micro-hardness test reveal that the hardness value gets marginally increased after the FJP polishing process due to the fact that polished surface is already relatively smooth and further, the impact of the particles causes hardening effect on the substrate surface of AISI H13 die steel plate.

Conclusion

A parametric investigation on fluid jet polishing of AISI H13 die steel using a single nozzle tool and its surface finish improvement has been discussed in this article. The following conclusions can be drawn from this research work:

The polishing time was the most influencing parameter observed in the FJP study, followed by the stand-off distance and jet pressure. The best performance of the FJP process in terms of the percentage reduction in surface roughness, $% Δ R_{a}$ , of 52% was achieved at pressure of $11.1 b a r$ , stand-off distance of $32 m m$ and polishing time of $174.6 s e c$ .

The accuracy of the predicted values of $% Δ R_{a}$ from the regression model was also predicted using the confirmation experiments, which yielded the results for variation in $% Δ R_{a}$ around 3.36%.

The reduction in surface roughness values from 500 nm to 130 nm at a feed rate of 20 mm/min revealed an improvement in surface polishing efficiency.

The optimized range of parameters revealed in the present work will be beneficial for researchers who are working with higher stand-off distance using FJP process on ductile materials such as AISI H13 die steel.

As the feed rate increases, the surface roughness value increases but there are some intermediate ranges where surface roughness value decreased and reflected a better surface finish.

At higher feed rate and larger stand-off distance, non-uniform material removal takes place that leads to over-polishing of the material.

The present work can be extended by increasing the number of nozzle tool along with the variation in the impingement angle.

Supplemental Material

sj-docx-1-pie-10.1177_09544089221145481 - Supplemental material for Investigation on fluid jet polishing of AISI H13 die steel using a single nozzle tool

Supplemental material, sj-docx-1-pie-10.1177_09544089221145481 for Investigation on fluid jet polishing of AISI H13 die steel using a single nozzle tool by Shubham Choudhary, Ravinder Kumar Duvedi and Jaswinder Singh Saini in Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering

Footnotes

Author contributions

Experimental Setup and Material preparation was done by Shubham Choudhary. The data analysis was done by R. K. Duvedi. The first draft of the manuscript was written by J. S. Saini. All authors commented on previous versions of the manuscript and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Shubham Choudhary

Jaswinder Singh Saini

Supplemental material

Supplemental material for this article is available online.

References

Nguyen

. Prediction and optimization of machining energy, surface roughness, and production rate in SKD61 milling. Measurement 2019; 136: 525–544.

Chen

Wang

, et al. A review on remanufacture of dies and moulds. J Clean Prod 2014; 64: 13–23.

Ahn

Shen

Kim

, et al. Development of a sensor information integrated expert system for optimizing die polishing. Rob Comput Integr Manuf 2001; 17: 269–276.

Marquez

Perez

Rıos

, et al. Process modeling for robotic polishing. J Mater Process Technol 2005; 159: 69–82.

Lee

. Machining characteristics of the electropolishing of stainless steel (STS316L). Int J Adv Manuf Technol 2000; 16: 591–599.

Andrade

Xavier

Rocha-Filho

, et al. Electropolishing of AISI-304 stainless steel using an oxidizing solution originally used for electrochemical coloration. Electrochim Acta 2005; 50: 2623–2627.

Wang

Luan

Pang

, et al. Elastic and electrolytic ultraprecision polishing. Metal Finish 1998; 7: 22–23.

Singh

Shan

Kumar

. Experimental studies on mechanism of material removal in abrasive flow machining process. Mater Manuf Process 2008; 23: 714–718.

Jain

Jayswal

Dixit

. Modeling and simulation of surface roughness in magnetic abrasive finishing using non-uniform surface profiles. Mater Manuf Processes 2007; 22: 256–270.

10.

Beaucamp

Namba

Freeman

. Dynamic multiphase modeling and optimization of fluid jet polishing process. CIRP Ann Manuf Technol 2012; 61: 315–318.

11.

Cheung

Wang

Cao

, et al. Development of a multi-jet polishing process for inner surface finishing. Precis Eng 2018; 52: 112–121.

12.

Beaucamp

Katsuura

Kawara

. A novel ultrasonic cavitation assisted fluid jet polishing system. CIRP Ann Manuf Technol 2017; 66: 301–304.

13.

Fähnle

Van Brug

Frankena

. Fluid jet polishing of optical surfaces. Appl Opt 1988; 37: 6771.

14.

Beaucamp

. Micro fluid jet polishing. In: Yan

(ed) Micro and nano fabrication technology. Micro/nano technologies, 2018. Singapore: Springer, 2018, pp.1.

15.

Cao

Cheung

Ren

. Modelling and characterization of surface generation in fluid jet polishing. Precis Eng 2016; 43: 406–417.

16.

Tsai

Yan

Kuan

, et al. An investigation into superficial embedment in mirror-like machining using abrasive jet polishing. Int J Adv Manuf Technol 2009; 43: 500–512.

17.

Messelink

Waeger

Wons

, et al. Prepolishing and finishing of optical surfaces using fluid jet polishing. Proc SPIE Int Soc Opt Eng 2005; 5869: 38–43.

18.

Kuo

Chen

, et al. Study of air-driving fluid jet polishing. Proc SPIE Int Soc Opt Eng 2011; 8126: 812611.

19.

Shiou

Asmare

. Parameters optimization on surface roughness improvement of zerodur optical glass using an innovative rotary abrasive fluid multi-jet polishing process. Precis Eng 2015; 42: 93–100.

20.

Wang

Cheung

, et al. A novel multi-jet polishing process and tool for high-efficiency polishing. Int J Mach Tools Manuf 2017; 115: 60–73.

21.

Anbarasu

Vijayaraghavan

Arunachalam

. Effect of multi stage abrasive slurry jet polishing on surface generation in glass. J Mater Process Technol 2019; 267: 384–392.

22.

Cheung

Wang

, et al. Curvature-adaptive multi-jet polishing of freeform surfaces. CIRP Ann 2018; 67: 357–360.

23.

Wang

Cheung

, et al. An investigation of effect of stand-off distance on the material removal characteristics and surface generation in fluid jet polishing. Nanomanuf Metrol 2020; 3: 112–122.

24.

Zhang

Tao

, et al. Structure optimization and numerical simulation of nozzle for high pressure water jetting. Adv Mater Sci Eng 2015; 732054: 1–8.

25.

Dai

, et al. Optimization and application of influence function in abrasive jet polishing. Appl Opt 2010; 49: 2947–2953.

26.

Choudhary

Duvedi

Saini

. Effect of material removal characteristics with varying incident angle in fluid jet polishing process. Mater Today: Proc 2022; 62: 215–219.

27.

Mazzoleni

. Design of high pressure waterjet nozzles. Research Reports: 1994 NASA (ASEE Summer Faculty Fellowship Program) 1994; 1–7.

28.

Mao

Yang

Tsai

, et al. Experimental investigation of abrasive jet polishing on the free-form machined surfaces of SKD61 mold steel using SiC particles. Mater Manuf Process 2010; 25: 965–973.

29.

Yan

Tsai

Sun

, et al. Abrasive jet polishing on SKD61 mold steel using SiC coated with wax. J Mater Process Technol 2008; 208: 318–329.

30.

Liu

Chen

Huang

. The polishing of molds and dies using a compliance tool holder mechanism. J Mater Process Technol 2005; 166: 230–236.

31.

Sharma

Pandey

. Rapid manufacturing of biodegradable pure iron scaffold using amalgamation of three-dimensional printing and pressureless microwave sintering. Proc Inst Mech Eng, Part C: J Mech Eng Sci 2019; 233: 1876–1895.

32.

Iqbal

Bhatti

, et al. Response surface methodology application in optimization of cadmium adsorption by shoe waste: a good option of waste mitigation by waste. Ecol Eng 2016; 88: 265–275.

33.

Qasim

Nisar

Shah

, et al. Optimization of process parameters for machining of AISI-1045 steel using Taguchi design and ANOVA. Simul Model Pract Theory 2015; 59: 36–51.

34.

Singh

Jha

Pandey

. Parametric analysis of an improved ball end magnetorheological finishing process. Proc Inst Mech Eng, Part B: J Eng Manuf 2012; 226: 1550–1563.

35.

Supriya

Srinivas

. Machinability studies on stainless steel by abrasive water jet – review. Mater Today: Proc 2018; 5: 2871–2876.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.63 MB