Recent advancements in magnetic abrasive finishing: a critical review

Figure 1(b) shows two forces acting on the MAPs during finishing, namely normal force (F_n) and tangential force (F_t) on the workpiece. The indentation is caused by normal force and keeping all ferromagnetic particles together, whereas tangential force (F_t) is responsible for shearing off-chip at micro-level due to relative motion between the FMAB and workpiece. Both forces are responsible for removing the asperities in the micron size from the surface until the desired level of surface finish is achieved. Figure 1(a) illustrates the schematic representation of the magnetic field distribution between the north and south poles. The magnetic field gradient at position 'A' exerts a force on a ferromagnetic particle has two components, ${F}_{x}$ and ${F}_{y},$ namely, horizontal and vertical components. ${F}_{x}$ and ${F}_{y}$ can be obtained from equations (1) and (2) [63].

$$\begin{eqnarray}&&{F}_{x}=V{\chi }_{m}{\mu }_{o}H(\frac{\partial H}{\partial x})\end{eqnarray} \tag{ 1 }$$

$$\begin{eqnarray}&&{F}_{y}=V{\chi }_{m}{\mu }_{o}H(\frac{\partial H}{\partial y})\end{eqnarray} \tag{ 2 }$$

Where x is the magnetic force direction, and y is the magnetic equipotential line direction, ${\chi }_{m}$ is the ferromagnetic particle susceptibility, ${\mu }_{o}$ denotes permeability of the vacuum, V denotes volume of ferromagnetic particles, H denotes the magnetic field strength at A, $\frac{\partial H}{\partial x}$ and $\frac{\partial H}{\partial y}\,$ are magnetic field gradients in respective $x$ and y-directions.

Further forces ${F}_{x}$ and ${F}_{y}$ at each particle resolved into the normal and tangential direction in two-dimensional plain (X-Y plane). The normal force is applied perpendicular to the workpiece, pressing the MAPs against the workpiece surface and responsible for depth of penetration. Higher normal force can result in a more aggressive material removal rate, but excessive force may lead to surface damage or increased surface roughness. Therefore, a balanced normal force is crucial to achieve the desired MRR while maintaining surface roughness. However, the tangential force in MAF is responsible for moving the MAPs along the workpiece surface. It influences the cutting action of the abrasive particles and, consequently, the MRR. Controlling the tangential force is crucial for preventing surface defects such as scratches or uneven finishes. Proper tangential force adjustment helps to achieve a uniform and controlled abrasion action, leading to a smoother surface finish. The normal force on the workpiece was calculated using equation (3) [64].

$$\begin{eqnarray}&&{\,F}_{N}={H}_{{mt}}\,{t}_{p}^{2}\,\tan {\theta }_{a}\,{n}_{a\,}\end{eqnarray} \tag{ 3 }$$

Where ${t}_{p}\,$ is the depth of indentation, ${H}_{{mt}}$ is workpiece hardness, ${\theta }_{a}$ is semi-included, angle, ${n}_{a}$ is the number of abrasive particles participating in finishing.

The normal force is influenced by the magnetic field gradient, while the tangential force is determined by the strength of the magnetic field produced by either a permanent magnet or an electromagnet. The MAPs orient themselves in a parallel manner to the direction of the magnetic field lines due to the dipole lines of force produced by a magnet [65]. Voltage and current of an electromagnet are responsible for the stiffness of FMAB [66, 67]. An electromagnet can have a direct current (DC) or alternating (AC) power supply. An electromagnet with an AC power supply can generate an alternating magnetic field. Alternating magnetic fields can re-arrange abrasive particles in every new cycle, due to which fresh abrasive particles participate in material removal. Due to this reason, good surface finish can be obtained compared to a permanent magnetic field [62, 68, 69]. However, the loss of magnetic field strength varies in a cycle, so an electromagnet with a DC power supply is preferable.

Researchers have consistently reported dynamic trends in MRR and surface roughness during finishing processes. In the initial stage of the process, MRR is more and subsequently starts decreasing towards the end of the finishing [70]. The phenomenon is interlinked with the evolving surface roughness and irregularities on the workpiece surface. As the surface undergoes finishing, the apparent contact area increases, leading to a decreasing shearing effect on irregularities and a transition to a steady-state removal process [71].

MAPs are the combination of ferromagnetic and abrasive particles (SiC, Al₂O₃, Cubic boron nitride), which play a crucial role in the surface finishing procedure. The proportion of ferromagnetic and abrasive particles creates suitable pressure and force by FMAB on the workpiece. Ratio of ferromagnetic and abrasive particles used for preparation of MAPs is usually in the range of 60:40 to 80:20 [72]. In MAF, MAPs are used in two forms: (i) unbonded MAPs and (ii) bonded MAPs as shown in figures 3(a) and (b) respectively. Unbonded MAPs are homogeneous mixtures of ferromagnetic and abrasive particles with lubricant. Bonded MAPs are made by either gas atomization, sintering or rapid solidification of mixture of ferromagnetic and abrasive particles to form a matrix [73–77].

Zoom In Zoom Out Reset image size

In the case of unbonded MAPs, whenever the magnetic field is applied, carbonyl iron particles (CIP) form chains between the poles of the magnet [78, 79]. In this case, abrasive particles can move freely inside the magnetic field region. However, in case of bonded MAPs, abrasive particles are restrained from free movement. The selection of ferromagnetic and abrasive particles employed in the process depends upon the characteristics of the material of the workpiece that needs to be finished. It is necessary to select appropriate MAPs based on the requirement of surface roughness. When unbonded MAPs are used in the process, some abrasive particles are embedded in the finished surface of the workpiece. Such phenomena are absent in the case of bonded MAPs [77].

Depending upon the material mechanical and physical properties, various researchers has investigated different types of abrasive particles, including silicon carbide (SiC) [80, 81], chromium oxide (Cr₂O₃) [47], diamond paste [82], aluminum oxide (Al₂O₃) [83–86], Boron carbide [87] and cubic boron nitride (CBN) [88] in different forms.

According to the profile to be finished, different machine tools may be needed to operate a MAF attachment setup, such as a lathe, vertical milling machine, drilling machine etc The workpiece processed with the MAF process can be of two categories: finishing of external surfaces, such as cylindrical or planar surface and finishing of internal surfaces of the cylindrical workpiece (capillary tube, cardiovascular stent tubes etc). The list of different machine tools, abrasives, and materials used in MAF by different researchers is listed in table 2.

2.3.1. Plane MAF setups used in literature

Plane surface workpieces [97, 98] with free-form or complex surfaces use planar tool attachment because FMAB can adjust the shape according to the profile of workpiece surfaces (e.g. free form surface, complex curve surface and plane surface). These MAF attachments are mainly attached to either a vertical milling machine [99, 100] or a lathe machine with a specially designed fixture. Barman and Das [101] designed and developed a MAF setup for a plane workpiece by utilizing permanent magnet to produce a magnetic field between the workpiece and tool, as shown in figure 4. The FMAB formed by a permanent magnet could not press the tool on the workpiece surface because magnetic field lines generated by the permanent magnet form a complete circuit only around itself. A permanent magnet improved the surface finish, but the finishing efficiency of MAF was poor due to the field line not penetrating the workpiece [102]. In order to increase the stiffness of FMAB and finishing efficiency of the MAF process, Kwak placed a magnet on both sides of the workpiece to confine the magnetic flux density and found that there were further improvements in surface roughness than single side permanent magnet due to an increase in stiffness of FMAB [103]. Pandey et al [92] proposed a chemo-assisted double-disk magnetic abrasive finishing on a silicon wafer for further improvement in response variables and reported improvement in surface finish and MRR. Further research shifts from permanent magnet to the electromagnet to study the forming behaviors of FMAB. Magnetic field intensity can be adjusted for an electromagnet by changing the supply voltage and current. Singh et al [49] used an electromagnet for forming the FMAB through DC power supply. The impact of different variables, including working gap, abrasive weight, voltage, and rotational speed, was analyzed, and it was found that voltage is the most significant factor for temperature rise and surface finish improvement.

Zoom In Zoom Out Reset image size

2.3.2. Cylindrical (external and internal surface) MAF setups used in literature

Figure 5(a) illustrates the schematic diagram of the cylindrical MAF process configuration, where a cylindrical workpiece is placed between the magnetic poles [104–106]. Here, tools are mostly stationary and provide rotational motion to the workpiece. Both types of abrasive, either bounded or unbounded, can be used. When the workpiece begins to rotate, prepared MAPs form FMAB when exposed to a magnetic field, and finishing occurs due to the workpiece movement relative to the FMAB. Jain et al [43] investigated surface finish and MRR on the outer surface of cylindrical workpieces with respect to the circumferential speed and working gap. For further improvement, the researcher provided either vibration to the workpiece or combined the MAF process with the non-conventional machining process [44, 107]. The MAF process can also finish the cylindrical internal surface of workpiece by rotating the workpiece or magnet. Yamaguchi & Shinmura were the first to propose finishing an internal pipe surface using the MAF process. Figure 5(b) illustrates the internal cylindrical MAF process. The magnetic jig consists of two magnets that are integrated by a yoke. It is arranged in an N-S-N-S like pattern to make a compact magnetic circuit in the pipe. Because of the rotation of the workpiece, a relative motion occurs between the MAPs inside the workpiece. Finishing takes place around the interior surface of a cylindrical workpiece. Accordingly, we get a smoothed inner surface. Singh & Singh [108] employed a DC power supply and two electromagnetic poles that are oriented perpendicularly to each other to generate a magnetic field. They found that the percentage improvement in surface finish was approx 94% with a surface roughness value of 0.04 μm. Verma et al [109] designed and developed a setup utilizing a permanent magnet for finishing the internal surfaces of pipes, grooves, and vertical surfaces. In their investigation, they focused on the SS304 steel pipe and found that magnetic flux density was the most influential parameter, and they were able to achieve up to 56 nm surface finish at optimum process parameters. Kajal et al [110] performed MAF on the internal surface of a gun barrel and achieved the surface finish up to 150 nm. This reduced friction and heat generation between the bullet and gun barrel. Deng et al [42] used MAF to remove surface defects such as cracks and scratches from the internal surface of Ni-Ti alloy tube, which is utilized in the implementation of cardiovascular stent tubes.

Zoom In Zoom Out Reset image size

Researchers conducted various studies by combining MAF using vibration-assisted/ ultrasonic-assisted processes, as shown in figure 7. Vibration was given to the workpiece or tool in horizontal or vertical direction. In the UAMAF process, surface finish is achieved due to the relative motion of the vibrating tool or vibrating workpiece [115–118]. It was reported that the required surface finish was obtained using UAMAF in lesser time (Refer figure 8(b)) compared to MAF with higher MRR for different workpiece materials such as aluminum, SS304 stainless steel, AISI5200 steel [40, 87, 119]. It was reported that new particles replace worn-out abrasive particles due to vibration. The circumferential marks produced by rotational motion (known as morphology) are eliminated with the tool vibration along the workpiece surface was also reported. Magnetic field becomes dynamic in nature and applies variable forces on MAPs due to vibration. Misra et al [89] reported that the UAMAF process gave a more efficient performance than traditional MAF for a similar set of processing parameters and reported a 68.5% reduction in surface roughness with a loss of only 36.6 mg material. Mulik & Pandey [120] performed UAMAF on high carbon antifriction bearing steel (AISI5200) and found that normal force and cutting torque were more than traditional MAF and concluded that 65% improvement in the surface roughness at optimum parameters than traditional MAF. Anjaneyulu and Venkatesh [121] conducted UAMAF on Hastelloy C-276. They observed a notable 26.8% increase in force and a 26.6% rise in power consumption. Interestingly, this resulted in enhanced process efficiency compared to the traditional MAF. Pak et al [122] used sintered MAPs with glass powder to enhance the process efficiency of UAMAF process. Under optimized conditions, the UAMAF process improved the PCSR by 86.62% for DIN 1.2738 tool steel. They reported a 48.62% improvement in PCSR in comparison to conventional MAF.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In conclusion, UAMAF offers superior surface quality, shorter processing time, higher material removal rate, and precise process control compared to traditional MAF for difficult-to-machine and non-magnetic materials. Ma et al [71] compared MAF and UAMAF and reported 2 to 4 times higher cutting force in UAMAF than traditional MAF, which decreases with increasing machining gap due to energy loss shown in figure 8(a). They achieved a surface roughness reduction of about 59% compared to MAF.

UAMAF method has various advantages over conventional finishing processes. Compared to traditional MAF, the UAMAF process reduced abrasive wastage and energy consumption due to increased process efficiency [123, 124], thereby, producing less impact on the environment.

The chemo-assisted magnetic abrasive finishing (CAMAF) technique was developed to accelerate the MAF process and resolve the shortcomings of chemo-mechanical polishing (CMP) [125] and the MAF process. The hybrid process integrates the advantages of both CMP and MAF processes, suitable for silicon wafers [126], Inconel [114], and other difficult-to-machine materials. In the chemical process, peaks on the surface of the material are dissolved, and a soft layer of the material will be generated (passive layer). Subsequently, MAF easily removes the passive layer. Pandey & Pandey [92] developed a CAMAF process for finishing silicon wafers and analyzed the impact of input parameters, namely working gap, percentage weight of KOH, mesh size, and polishing speed on the surface roughness. Figure 9 illustrates the surface roughness values for different trials. Sihag et al [127] investigated the use of H₂O₂ chemicals in UAMAF and its impact on MAF for tungsten workpieces. They reported that rotational speed had the greatest impact on the percentage change in surface roughness (37.7%), followed by abrasive weight percentage (27.74%), working gap (16.12%), and H₂O₂ concentration (13.75%) as shown in figure 10. Overall, 79.50% improvement in surface roughness was reported at optimum level of input process parameters. Anjaneyulu and Venkatesh [128] performed chemo-UAMAF on Hastelloy C-276 and found that surface roughness was improved by 98.3% from initial R_a = 1.3 μm to R_a = 0.0224 μm with maximum material removal of 22.19 mg.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In conclusion, CAMAF improved the efficiency of the traditional MAF process. However, the chemical used in the process may not suitable for all materials and machines since it may contain hazardous or toxic substances that require proper handling, storage and disposal to avoid environmental contamination [129, 130]. Selecting less toxic, biodegradable or readily recyclable chemicals can reduce the environmental risks. Moreover, continuous research and development efforts are also necessary to optimize the CAMAF process. Further, to solve the environmental issue, a closed-loop control system of chemicals protects the machine components from the chemicals by making handling them easy.

Researchers synchronized MAF with electrochemical machining and developed hybridized electrochemical magnetic abrasive finishing (EMAF) setup as shown in figure 11. This overcame the issue of the decrease in efficiency of MAF in case various difficult-to-machine metals, superalloys and non-magnetic materials such as Al₂O₃ composite, 316L, SUS304 stainless steel etc [10, 107, 131, 132]. During the EMAF process, the electrochemical reaction is responsible for flattening the peaks by anodic dissolution and forming a passive layer on the workpiece. The passive layer formation mechanism in the electrochemical process is shown in figure 12(a), and the whole EMAF process is shown in figure 12(b) [64]. This occurs when the workpiece surface interacts with the electrolyte (${{NaNo}}_{3}$). Ionization reactions take place represented by equations (4) and (5) play a significant role [131].

$$\begin{eqnarray}&&{\,{NaNO}}_{3}\to {{Na}}^{+}+{{NO}}_{3}^{-}\end{eqnarray} \tag{ 4 }$$

$$\begin{eqnarray}&&{\,H}_{2}O\to {H}^{+}+{{OH}}^{-}\end{eqnarray} \tag{ 5 }$$

Cations migrate toward the cathode, and anions move toward the anode, guided by the ionization tendency of cations and the discharge capacity of anions, reactions (6) and (7) take place at the cathode and anode [131]

$$\begin{eqnarray}&&{2H}^{+}+{2e}^{-}\to {H}_{2}\uparrow \end{eqnarray} \tag{ 6 }$$

$$\begin{eqnarray}&&{2{OH}}^{-}\to {H}_{2}O+\frac{1}{2}\,{O}_{2}+{2e}^{-}\end{eqnarray} \tag{ 7 }$$

Hydrogen gas (${H}_{2}$) is generated at the cathode, and oxygen gas (${O}_{2}$) is produced at the anode. The generated oxygen reacts with metal ions, resulting in the formation of a passive layer. This passive layer remains in contact with the workpiece surface under low current densities supplied in this hybridized process. However, during the EMAF process, the passive layer experienced local abrasion due to MAF. This abrasion leads to abrasion-assisted-dissolution, causing the rapid removal of the locally de-passivated metal surface. Subsequently, further passivation occurs, as depicted in figure 12(b) of the EMAF process [64]. There is no physical contact between the workpiece and the tool (electrode) throughout the finishing process. Therefore, the electrode does not disintegrate and enhances the service life [133, 134]. Sun and Zou [135] proposed EMAF on stainless steel SUS304. In the first 4 min, EMAF was performed, and then traditional MAF was performed for the next 26 min to remove all the passive layers formed during EMAF. They found an improvement in surface finish (up to 20 nm) and MRR (9.75 times) than traditional MAF. Judal and Yadava [136] examined how variations in electrolytic current, magnetic flux density, and rotational speed of the workpiece influence material removal and surface roughness during the EMAF process. Fang et al [137] examined the impact of the magnetic field on the movement of ions and chemical reactions. They observed negative ion movement towards the workpiece in a cycloidal fashion when the magnetic field was present, enhancing the surface finish. In conclusion, EMAF combines the capabilities of electrochemical machining and MAF to provide a versatile and effective process for achieving a good surface finish. In conclusion, EMAF improved the process efficiency of the traditional MAF process. However, it has almost similar environmental issues as CAMAF, except the chemical replaced by the electrolyte in combination with low-intensity current is used, which is less severe to an environmental issue than CAMAF.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Surface roughness and MRR are the critical factors for evaluating the finishing processes. Thus, it is important to predict or visualize how the input process parameters along with the working environment, affect the response parameters of the process. MAF aims to obtain a superior surface finish with minimum material removal, like other finishing processes. This will ensure that shape and size inaccuracy in a part can be minimized [118]. The magnetic field strength in MAF is responsible for the forces acting by the FMAB on the specimen to control the MRR. Kumar and Yadava [47] performed mathematical modeling based on Maxwell's equation for predicting the magnetic flux generated and the magnetic forces. Further, they found that magnetic flux density significantly affects the rise in temperature compared to rotational speed. Mishra et al [139] developed a mathematical model to anticipate the MRR for the UAMAF process. They considered both the continuous and temporary removal of materials that occur at the same time during the operation. Unlike steady state, transient MRR includes initial surface roughness. Different researchers developed mathematical models for predicting normal and tangential forces for MAF process. The models developed by Shukla et al [75] and Misra at al. [140] had a high degree of consistency with the actual experimental values of forces. Gao et al [93] developed a model for predicting MRR based on depth of indentation, finishing trajectory, and abrasive particle number.

Empirical models are also in practice for predicting output parameters based on input process parameters. However, the development of an empirical model reported by various researchers requires experimental data and knowledge of the phenomenon rather than theoretical understanding. From experimental data, Chaudhari & Judal [141] developed a regression model for surface roughness and MRR. They used multi-objective optimization techniques to find the optimal parameters of the developed model and verified the experimental result. Mulik et al [141] established a regression model for finishing force and torque in the case of UAMAF. The following parametric equations (8) and (9) were derived for the force and torque, respectively, from the experimental data.

$$\begin{eqnarray}&&F=14.8+0.049\,{X}_{1}-0.814\,{X}_{2}+0.059\,{X}_{3}-0.774\,{X}_{4}+0.196\,{X}_{5}\,\end{eqnarray} \tag{ 8 }$$

$$\begin{eqnarray}&&T=10.50+0.0718\,{X}_{1}-1.22\,{X}_{2}+0.0564\,{X}_{3}-1.04\,{X}_{4}+0.267\,{X}_{5}\,\end{eqnarray} \tag{ 9 }$$

The variables ${X}_{1},$ ${X}_{2},$ ${X}_{3},$ ${X}_{4},$ and ${X}_{5}$ are voltage supply, electromagnet rotational speed, percentage weight of abrasive, working gap, and pulse duration.

Soft computing techniques are also becoming popular for the optimization of MAF. Few researchers already reported some techniques for solving complex problems based on artificial intelligence and natural selection to provide results that are more precise. Ahmad et al [53] used ANN-GA to enhance the process performance of MAF and compared to the result with the conventional Taguchi-ANOVA analysis. The results obtained by ANN-GA exhibited a 7% improvement compared to the Taguchi-ANOVA analysis. Various soft computing techniques utilized by researchers are listed in table 3.

Both numerical and empirical models are robust techniques, but their accuracy for prediction depends on the type of problem and conditions. Numerical models in machining/finishing struggle to anticipate beyond given boundary conditions, dynamic problems, mesh size and do not consider the effect of agglomeration of abrasive particles [151, 152]. Due to computer constraints, they may miss the key factors or simplify essential variables. Empirical models based on observed data could be biased, misinterpret the cause and effect, and become obsolete after overextending the range of selected variables. Both have limitations; therefore, it is essential to use their outputs carefully in finishing decisions while understanding their strengths and drawbacks.

Various numerical studies of the MAF process were carried out to predict the effect of shapes, slot/without slot in magnet, and magnetic field distribution on the surface finishing of sample during the MAF. Srinivas et al [153] performed a numerical study to evaluate the effects of four different magnets, namely, fan-shaped, arc-shaped, ring-shaped and electromagnet magnetic poles on the efficiency of MAF. In their research, authors considered the cylindrical workpiece for the analysis. The results of their article showed that an arc-shaped magnetic pole gave maximum MFD; whereas ring-shaped magnetic pole showed minimum MFD. Jayswal et al [154] investigated the effect of the slot (on the lower end of an electromagnet) on the surface finish of the sample. They found that magnetic field concentration in the slot increased forces around it and got a better surface finish than without slot in less time (Refer figures 13(a) and (b)). Cui et al [155] performed numerical analysis of MAF on curved surfaces. They analyzed the effect of slot angle on magnetic field strength and observed that magnetic field strength increases with a decrease in edge angle. Trapezoidal slot with 75° edge angles on a spherical magnetic pole gives better field strength than a cylindrical. Mishra et al numerically analyzed the temperature and magnetic flux distribution between the workpiece and FMAB interface. They reported that the predicted minimum and maximum temperatures were found to be in the range of 34℃−51℃, which did not affect the workpieces surface finish and surface texture [58]. Singh et al [156] studied the influence of electric current and MFD on temperature for Al-6060 and stainless-steel workpieces using the FEM. They found that the resulting temperature rise (due to finishing) for the stainless-steel workpiece was higher than Al-6060. This is due to low conductivity of stainless steel and magnetic characteristics compared to Al-6060.

Zoom In Zoom Out Reset image size