Minimum Quantity Cutting Fluid Application for Grinding Weld Flash: Surface Integrity Evaluation

One of the final processes in making a bandsaw blade is to cut the coil to a specific length and weld the ends of the blade to form a continuous loop. Resistance welding is the go-to/typical welding process used in the sawing industry for joining alloyed steels to manufacture band saw blades. It also finds its application in industries like railway and power distribution (electrical—bus bars). In the resistance welding process, the excess metal is extruded to the top of the weld joint, and the excess metal is called “flash”. The flash is conventionally removed using a machining process like grinding. The grinding of the flash is usually carried out under no cutting fluid application (dry). One of the main goals is to remove the weld flash without inducing crack-like discontinuities along with the ground weld interface. The surface integrity of the welded joint is essential for the band saw blade life and to prevent any premature failure at the weld joint due to fatigue loading (a band saw blade mainly undergoes cyclic bending fatigue during its service). The effect of grinding on surface integrity has been well researched [1,2] since it has the propensity to cause surface cracks as it is a thermal-dominated process. In general, the weld joints are post-processed to improve the structural integrity and service life. The post-processing methods include heat treatment (HT), weld geometry machining, thermal relief, shot peening, etc. [3]. Some weld applications require weld geometry or flash removal either by grinding or water jet eroding. Such weld geometry removal processes can induce tensile surface residual stresses [3]. A recent review of the weld geometry removal process has concluded that the grinding process does improve the fatigue life when compared with as weld samples, but the use of the cutting fluid in the grinding of weld geometry was not investigated [4]. Whereas researchers have reported that the targeting minimum quantity of cutting fluid on the face turning improves the surface integrity metrics such as surface residual stresses and surface roughness [5–7]. However, there are very few studies to understand the effects of grinding weld-related flash. It is essential to establish a fundamental understanding of the effects of the cutting fluid application on grinding-induced sub-surface integrity (microhardness, residual stresses, and microstructural changes) for weld flash grinding.

Generally, a coolant is used in grinding to reduce the thermal effect on the ground surface. In literature, many researchers have studied the role of using cutting fluid while grinding [8]. Most of the energy used in the grinding process is converted to heat energy and subsequently transferred most of the energy to the ground workpiece if no cutting fluid (or termed as “dry”) is used [9]. Malkin has also reported the effect of thermal damage caused by the grinding process on the workpiece surface [10]. To reduce thermal damage, most of the industry recommends using some kind of cutting fluid while forming and finishing grinding [11]. If cutting fluids are used the amount of heat energy dissipated to the workpiece reduces significantly [12]. Similarly, Inoue and Aoyama have shown that if the depth of grinding is smaller (∼100 µm), the surface produced by the sustainable cutting fluid solutions is comparable with flood coolant [13]. However, Choi et al. have shown that cold air as a cutting fluid generated inferior surface integrity (higher tensile residual stresses and surface roughness) compared to flood coolant due to cold air’s inability to lubricate through the grinding zone [14]. Recently, researchers have used minimum quantity fluid to reduce the grinding forces and grinding wheel wear while grinding ceramic matrix composites [15]. Researchers have studied using sustainable cutting fluid solutions like air or ester oil as a coolant while grinding with some success [16].

As mentioned earlier, most grinding research with cutting fluids focused on form and finish grinding. Some grinding processes like stock removal (grinding), wheels are designed and used without any cutting fluids. The effects of using sustainable cutting fluid solutions on stock removal grinding wheels have not been investigated. Grinding weld flash is one of the stock removal grinding processes. The effect of dry grinding of weld flash on sub-surface integrity metrics as per ANSI B211.1-1986 [17] has not been reported. The surface integrity of the ground surface at the welded joint, which plays a significant role in the life of the bandsaw blade, has not been studied. In general, the last steps for bandsaw blades are: (1) cutting the band saw blades to the length, (2) butt welding the edges using resistance welding, followed by (3) normalizing the weld area, and (4) grinding the weld flash as shown in Fig. 1. The effects of the weld flash grinding process on the weld service life are critical for the bandsaw blade life and have never been reported. Thus, establishing a fundamental understanding of the effects of grinding temperature (by applying minimum quantity fluid) on surface integrity (microhardness, residual stresses, and microstructural changes) is essential. In this study, the effects of using a minimum cutting fluid application under varying cutting fluid combinations on the grinding of weld flash in alloy steel are studied. The use of compressed air (CA) as a sustainable solution with the minimum cutting fluid application was investigated.

The grinding experiments were carried out on 50 mm wide welded band saw samples of ∼750 mm long. The band saw samples were made of medium carbon alloy steel, and it is one of the widely used (backer) materials for band saw blades. The samples were welded and normalized using an Ideal BAS340S welder. The samples were made from the same manufacturing coil to reduce/eliminate any process variation. The welded and normalized samples were ground using an Ideal SMP120 manual grinding machine at four different cutting fluid conditions, namely dry, CA, minimum quantity lubricant (MQL), and minimum quantity coolant (MQC). The weld flash heights were between 2.97 mm and 3.1 mm per side for the welded sample used in this study. Even though this is a manual operation, the depth of weld flash removal for each pass was kept constant at 0.20 mm. Figure 1 shows the setup of the grinding machine used in this research work.

The grinding of weld flash in the bandsaw industry is generally carried out without any cutting fluid/coolant (dry condition). Hence, the Ideal SMP120 grinding machine is not equipped with any fluid delivery system. UNIST Saw Blade Lube System^™ was installed on the grinding machine to perform the experimental study. Figure 2 shows the nozzle and the location of the nozzle on the grinding machine. The nozzle was directed towards the interface between the grinding wheel and the welded sample.

UNIST Coolube^® 2210EP biodegradable lubricant (recommended lubricant by the manufacturer for steel workpiece) was used as the lubricant for MQL application while Nanotech 6800 semi-synthetic water-soluble coolant mixed with water at the ratio of 1:24 was used for MQC application. The flowrate was constant for both MQL and MQC applications at 360 ml/h. In air condition, compressed air at a pressure of 620 MPa was applied using the same outlet used for minimum quantity application.

Figure 3 shows typical weld flash and post grind images of a band saw blade. A total of four welded samples were prepared for each experiment. Each sample had a weld flash on both sides that had to be ground. A total of eight weld flashes were ground for each experimental condition. The grinding of weld flash is a manual-fed operation. The grinding wheel moves across the weld flash, as shown in Fig. 3. A bend test was performed to evaluate the weld strength. The bend test was carried out on a custom-built destructive bend tester in accordance with a standard testing method—ASTM E290-97a [18].

Surface roughness, residual stresses, and microhardness were measured to study the effects of grinding-induced sub-surface integrity. Surface roughness was measured at three different spots—close to the gullet, middle of the band, and close to the back edge (refer to Fig. 3) using Keyence VHX 700 3D digital 4 K optical microscope at ×1500 magnification. The most common techniques to measure residual stresses are X-ray diffraction and hole drilling methods and both the methods produce similar residual stress profiles [19]. The hole drilling strain gage method was used to measure the residual stresses in the middle of the grind samples for all four conditions. The procedure that was utilized to measure the residual stresses using the hole drilling method is in accordance with a standard testing method—ASTM E837 [20]. Hole drilling method was used on EA-06-062RE-120/LE strain gages which were mounted on the samples to measure residual strains. Later on, H-Drill^® software was used to calculate the surface residual stresses [21].

Microhardness was measured along the sub-surface of the ground sample using the Shimadzu HMV-2 T E microhardness tester. The samples were cut closer to the back edge of the ground samples, and three readings at 25 µm, 50 µm, and 75 µm from the ground surface were measured. The area closer to the back edge has the highest potential for thermal damage due to grinding as there is a direction change in the grinding process. A new grinding wheel was used for each experiment. Images of the used grinding wheels were taken at ×30 magnification using Keyence VHX 700 3D digital 4 K optical microscope to study the effects of cutting fluid conditions. Table 1 shows the complete experimental plan for this research work.

One sample from all four conditions was tested using the custom-built bend tester to evaluate the weld quality. The bend test was conducted as per ASTM E290-97a [18]. Figure 4 shows the post bend test images of all four applications. MQL and MQC had cracks from the gullet due to improper removal of weld flash along the gullet. As shown in Fig. 3, the gullet is the area between the two teeth. The weld flash in the gullet area cannot be removed using the grinding process used in this study. The removal of weld flash from the gullet is a manual operation using a deburring tool, and it is beyond the scope of this research work. All four welds passed the destructive bend test, showing that the weld quality was acceptable.

Figure 5 shows the surface roughness data for all four cutting fluid applications. Surface roughness was measured at three different locations along the ground surface, as shown in Fig. 5. As mentioned earlier, the surface roughness is measured using an optical microscope. The error bar in the graph represents one standard deviation. From the graph, it is evident that dry has the highest surface roughness among all four conditions. The surface roughness data for MQL, MQC, and air were relatively similar, showing the significance of cooling and/or lubricating while stock removal grinding of weld flash. The improved surface roughness values due to the usage of the cutting fluids (air, MQL, and MQC) can be attributed to the reduction in grinding temperatures, thus reducing the thermal damage. The high grinding temperature can cause smearing leading to the loading of the grinding wheel. Researchers have attributed the improved surface roughness when using the cutting fluids to the lower machining temperatures compared to dry machining processes [23–26]. Further, usage of the cutting fluid improves the frictional interface along the chip and tool rake face, thus assisting chip flow and improving the surface roughness [27]. Similar results were seen when using flood coolant versus dry slot grinding [28]. Even though the lubrication and temperature effects of air, MQL, and MQC were different, the surface roughness values are statistically similar. Further investigation is needed to understand the reasons for the aforementioned phenomenon. Overall, the sustainable solutions (air, MQL, and MQC applications) in this work have statistically (based on the error bar) improved the surface roughness values compared to dry grinding.

Residual stresses for all the ground samples were measured precisely at the same radial distance from the center of the workpiece to ensure consistency. Figure 6 shows the residual stress profiles along the sub-surface of the medium carbon alloy steel at different steps of a typical band saw manufacturing process. The incoming band saw material (Raw-HT) to the welding stage has a very high compressive surface residual stress due to the HT and blasting process. The next step in the process is welding and normalizing the material. The sub-surface residual stresses of the normalized material were close to zero (up to ∼100 mm from the top surface). This was as expected since the normalizing process removes the residual stresses. Thus, confirming the residual stress measurement method being used and also the process used in the bandsaw manufacturing process. The next step is grinding the weld flash in dry condition (no cutting fluid). This is the process currently being used in the industry. The sub-surface residual stress profile for the dry condition was tensile in nature, without having any cross-over to compressive residual stress for the measured depth. The residual stress profile of the dry grinding process of an annealed weld shows that the dry grinding process has induced tensile sub-surface residual stresses along with the measured depth. This could be attributed to the high grinding temperatures, as other researchers have reported [14].

Figure 7 shows the residual stress profile along the sub-surface of the ground surface for all four cutting fluid applications. Air produced the highest compressive sub-surface residual stress among the four conditions, while dry was the only condition with tensile sub-surface residual stresses. Tensile residual stresses are attributed to thermal loads, whereas compressive residual stresses are attributed to mechanical loads [29]. Also, a similar trend was observed in other machining processes like facing steels and titanium with targeted minimum quantity fluid application [5,26,30–32]. Grinding in dry condition produced a high temperature, thereby resulting in tensile sub-surface residual stresses, whereas, in the other three conditions, the mechanical effect is more dominant. Using a minimum quantity of fluid in this application improved the sub-surface integrity of the weld, thereby increasing the possibility of improved fatigue life of the weld. A fatigue life study to estimate the weld service life is recommended to confirm the observations made in this study.

Figure 8 shows microhardness along the sub-surface of the ground surface for all four experiments. As the grinding wheels are designed to be thermally insulated, the grinding temperatures are transferred to the ground sub-surface causing thermal softening. Chen et al. [33] analyzed the thermal damage in the grinding process. They reported that thermal damage during the grinding of hardened steel would result in the tempering of steel, thereby leading to thermal softening. The sub-surface microhardness of the ground surface reduces due to thermal softening. Similar phenomena could be occurring in this study as all four cutting conditions had lower sub-surface hardness than the bulk hardness. Dry had the lowest sub-surface microhardness suggesting that there could be severe thermal damage/thermal softening. Air had the highest sub-surface hardness suggesting that the condition had the least thermal damage. With the dry condition, the grinding temperatures are higher than the cutting fluid conditions as mentioned in the aforementioned section, thus leading to lower sub-surface hardness. Contrastingly, compressed air application reduced the grinding temperatures, thus reducing the thermal softening effect. Also, the cutting fluids could improve the frictional properties along the tool-chip interface, thereby reducing the grinding temperatures further. Further metallurgical analysis is recommended to validate the thermal softening phenomenon.

Figure 9 shows new unused and used grinding wheel images for all four cutting fluid applications. Loading is visible on all four grinding wheels (pointed in Fig. 9). From Fig. 9, it is evident that the amount of loading in the dry condition is considerably higher than that of the other three applications. This further confirms that the temperature is relatively higher in dry condition leading to increased loading and causing thermal softening. Loading of the grinding could negatively influence the grinding process, leading to premature tool wear, thermal surface damage, and increased grinding forces. These factors could influence the surface roughness, as shown in Fig. 5. Air, MQL, and MQC also use an air jet, thus reducing this adhesion tendency by increasing the formation of interfacial/oxide films. This needs to be confirmed by a later study involving chemical analysis on the surface. Air, MQL, and MQC applications had relatively similar loading, suggesting a similar temperature range. As mentioned earlier, the cutting application has the possibility of improving the frictional properties and reducing the grinding wheel damage. An interesting observation is that the MQL grinding wheel had oil residue, as shown in Fig. 9. With the grinding wheel images and surface integrity studies, a tool wear study is recommended to extend this paper’s hypothesis that using sustainable cutting solutions like MQL, MQC, or air will reduce the grinding temperature and reduce loading/tool wear.

Figure 10 is the radial plot of all three surface integrity parameters studied (surface roughness, sub-surface hardness, and residual stresses) in this research work. Based on the radial plot, air application had the highest surface compressive residual stresses, highest sub-surface microhardness, and relatively low surface roughness. Air has reduced the grinding temperatures, thereby reducing thermal damage. MQL and MQC are relatively similar, producing compressive sub-surface residual stresses and lower surface roughness values. Based on the observations, MQC and MQL had reduced temperatures compared to dry condition. Overall, the sustainable solution used in this work has improved the sub-surface integrity of the ground resistance weld flash surface.

As mentioned earlier, there are very few studies to understand the effect of grinding weld-related flash. The surface integrity of the welded joint is critical for the bandsaw blade life and to prevent any premature failure at the weld joint due to fatigue loading. In this study, the effects of using a minimum cutting fluid application under varying cutting fluid combinations on the grinding of weld flash in alloy steel are studied. Results from the surface roughness, sub-surface residual stresses, and microhardness have proven that the compressed air produced superior surface integrity, i.e., compressive sub-surface integrity, lower surface roughness, and less thermal damaged sub-surface. The service life of the saw blades with the weld flash ground in different cutting fluid conditions (dry, air, MQL, and MQC) is out of scope for this study. The fatigue life of the weld joint and microstructural analysis are recommended as a future study to confirm that the improved surface integrity metrics improved the service life of the band saw blade weld.

The authors would like to thank Nancy Morse Sonner and Sally Morse Dale for their unwavering support towards research and development. The authors would also like to thank Alan Holbrook, Justin Smith, and Josh Critchfield for their help in conducting the experiments. The authors would like to thank Research and Product Development teams at The M. K. Morse Company for assisting in the paper. Finally, the authors gratefully acknowledge their mentor and guide, Dr. A. K. Balaji (University of Utah), without whom this study would not be possible.

There are no conflicts of interest. This article does not include research in which human participants were involved. Informed consent is not applicable. This article does not include any research in which animal participants were involved.

No data, models, or code were generated or used for this paper.