High performance peel grinding of steel shafts using coarse electroplated CBN grinding wheels

In order to show the differing process behaviour for both grain sizes in soft (left) and hard (right) grinding Fig. 3 shows the measured process normal forces for increasing feeds. According to the state of the art, the number of abrasive grains that engage the workpiece influences the process forces. With increasing numbers of grains throughout the contact area, the process forces are enhanced. This leads to lower process forces for larger abrasive grains as larger grains reduce the number of abrasive grains in the contact area. This is evident in Fig. 3 since lower process forces occur over all investigations with the larger B602 grains compared to the smaller B252 grains. While the process forces are reduced by a growing number of engaged grains, the load on a single grain is increased. Additionally, when machining soft shafts, the exponential incline of the normal forces over the investigated feeds for the B252 tool also indicates clogging over the course of the investigations. Due to the smaller chip space of this tool and the comparatively large chip thickness, chips get wedged in the chip space and cannot be removed by the cleaning nozzle. This beginning material adhesion hinders the chip removal further and leads to increasing amounts of rubbing in the contact area, which leads to higher normal forces and continuously increasing clogging of the tool. The B602 tool offers more chip space and consequently does not show the same force development. The forces increase linearly by about 45 N per 0.3 mm feed increase, indicating no clogging and unchanged sharp grains.

In hard machining the forces progress similarly. The normal forces for the B252 tool increase exponentially due to growing material adhesion and the clogging up of the chip space. However, hardened material forms smaller chips than soft material according to the state of the art. This occurs due to the more brittle material behaviour of hardened steels, which subsequently leads to shorter chips. The clogging of the B252 tool in soft and hard grinding indicates an overload of the available chip space. This may be linked to high process temperatures due to the high MRR and causes softening of the material. Therefore, material adhesion is more likely. Additionally, the progressing clogging of the tool reduces the amount of coolant lubricant in the contact area and thereby the cooling efficiency. Since the chip space of the B252 tool is just about half as large as for the B602 tool, the overload of the chip space and the reduced coolant supply leads to the exponential increase in process forces seen in Fig. 3.

The process forces of the B602 tool rise linearly by about 50 N per 0.3 mm feed increase. This increase is slightly higher by 5 N per 0.3 mm feed compared to soft machining due to the higher grain load when machining hardened steel. A more detailed analysis of the tools wear behaviour is discussed in chapter 3.2.

In order to further compare soft and hard grinding, the process forces for both processes using the B602 grinding wheel are presented in Fig. 4 (left). Here, it becomes evident that the normal forces are higher by an average of 58 N in hard machining, while the tangential and axial forces are within a 5 N range of each other. Due to the high material removal rate in this roughing operation, the process temperature is very high, which can be seen in large amounts of sparks during the process. Jin [8, 16] has shown that the temperature of the contact area in High Efficiency Deep Grinding increases up to the melting point of the machined material. As the material softens up to this point, the initial hardness of the material decreases regarding its influence on the chip formation.

The grinding force ratio (GFR) µ = F_T / F_N presented on the right of Fig. 4, sets both normal and tangential forces into relation. It is an indication of a tools overall performance. A high GFR indicates an efficient material removal process with sharp grains and small amounts of friction between grains and work piece. It becomes obvious that the B602 grains are at their performance limit at f > 1.5 mm in soft machining as the GFR declines with higher feed rates. This indicates that the tool begins to clog or the grains begin to dull. This can also be seen by the slight exponential increase in normal process forces of the soft grinding (Fig. 4, left) whereas the forces in hard grinding increase linearly. Consequently, the presented GFR for hard grinding increases over all feeds without stagnation, indicating the potential for higher feeds. The increasing GFR shows the growing efficiency of the material removal at higher feeds. While the material removal is more efficient in soft grinding over the investigated feeds, the growth rate of the GFR is higher for higher feeds in hard grinding, verifying clogging-up and thereby reducing the cutting efficiency in soft grinding. The ever increasing GFR at high feeds shows the efficiency of the hard grinding in comparison to the soft grinding. Brittle material removal leads to less material adhesion to the grains, preserving sharp grains in higher feeds. The overall level of the GFR is lower in hard machining because the increased hardness of the shaft leads to higher normal forces. As the material´s Rockwell hardness is measured by indentation it is a measure of the normal force needed to bring the abrasive grains of the grinding wheel to their working depth. Therefore, the normal forces in hard grinding is higher than in soft grinding, which, for constant tangential forces, leads to a lower level of the GFR. In conclusion, the coarser B602 grains are well suited for rough grinding operations in soft and hard machining, while the finer B252 grains are unsuited for the presented processes.

Electroplated tools consist of only one layer of abrasive grains within a Nickel matrix. Due to the high grain retention forces of this bond, the grains can protrude about 50% of their diameter from the bond, leaving large chip spaces in between. A single layer of grains also means that these tools cannot be dressed. If the grains dull or break or if the chip space is clogged, the tool can no longer be used. Therefore, it is crucial to investigate the wear behaviour of these tools. Microscopic pictures from a Keyence VHX600 digital microscope in Fig. 5 show the tools initial state at the top and the worn state at the bottom after a removed volume of V = 18.500 mm³ which equals one pass of hard grinding for the electroplated tools with grain sizes B252 and B602 in comparison. The material was removed with a depth of cut of a_e = 0.5 mm and a feed of f = 1.8 mm over a length of l = 200 mm. As described before, the B252 tool shows severe clogging to a point where the grain protrusion is eliminated. This leads to rubbing of the clogged material on the work piece, resulting in high process normal forces and temperatures. This explains the exponential incline in process forces for this tool. For the larger grain size B602 (Fig. 5, right) this effect cannot be observed as there is only little material adhesion. After cleaning the clogging from the tools with sharpening stones, the CBN grains themselves on both tools were not visibly worn.

The clogging of the B252 tool can be traced back to the fact that this tool provides an average chip space of about 126 µm while the B602 tool provides 301 µm that equals the grain protrusion of about 50% of the grain size. Consequently, chips get wedged between grains more easily and cannot be removed from the contact area as effectively when the grain protrusion is lower. Additionally, the smaller chip space leads to a lower coolant supply into the contact zone, which impedes the removal of chips and thus leads to increased process temperatures.

The chip space and its dimensions provide space for forming chips within the tools topography. Therefore, not only the chip space but also the chip dimensions need to be considered for the tools’ wear behavior. If the chips are larger than the provided chip space, they adhere to the tool easily.

The normal chip length distribution is shown in Fig. 6 for the B252 tool on the left and for the B602 tool on the right. For both grain sizes, chips from soft and hard machining were measured under a digital microscope Keyence VHX 5000 and depicted accordingly. Due to the ductility of the soft shafts, the chips do not break as easily and longer chips result from the process. This can be seen for both grain sizes, as the chips are longer in soft machining compared to hard machining. This effect is particularly prominent for the B252 tool. The average chip size halfes from l_chip = 593 µm in soft machining to l_chip = 302 µm in hard machining for the same process parameters. With the B602 tool the average chip size decreases by 25% from l_chip = 405 µm (soft machining) to l_chip = 299 µm (hard machining). With the assumption, that every grain provides one main cutting edge at its center, the cutting edges on electroplated tools are roughly a grain size apart which was verified by measurements of the grain distances on microscopic images. This means that the chips on the B252 tool are longer than the distance between cutting edges and therefore larger than the chip space in between grains. This leads to the clogging of the tool whereas if chips are shorter than the distance between grains, as for the the B602 tool, no clogging occurs.

Another visible effect is that the standard deviation for both tools is lower for hard grinding than for soft grinding. The standard deviation is a measure of how widely a parameter varies. Therefore, a small standard deviation is recognizable in a narrow graph for the normal distribution. In case of the B252 tool the standard deviation s_B252,s = 243 µm halves in hard machining to s_B252,h = 115 µm. For the B602 tool it only decreases by about 43% from s_B602,s = 122 µm to s_B602,h = 86 µm. This shows that the investigated chip lengths vary more widely for the smaller grains. This can be accredited to the fact, that one chip might be cut by more than one grain or will get compressed into the available chip space, while other chips expand in between grains.

Thus, it can be summarized that the wear behaviour of the B252 tool is already at its maximum capacity. The increasing clogging of the chip space reduces the number of sharp grains, the amount of coolant supply and the chip removal out of the contact zone. This leads to less effective material removal, increasing process forces and temperatures. The chip length, which is larger than the distance between grains, further increases these effects by closing the chip space. In order to increase the performance of the B252 tool, the chip formation needs to be improved according to the tools available chip space.

As the B602 tool shows no signs of tool wear yet, the grinding wheels performance is not limited over the presented feeds. Conclusively, this tool can be used at even higher feeds to increase its performance.

In order to analyse the thermal and mechanical load on the work piece, the residual stress (RS) was measured on all ground shafts. All measurements were made once as the measurement accuracy was proven to be ± 20 MPa. The penetration depth on the surface of the shafts was τ_max = 5.5 µm in axial and peripheral direction in accordance to DIN EN 1305. For clarity Fig. 7 only presents the RS in the peripheral direction. The axial RS for all investigations behave similarly on a slightly lower level.

Most investigated feeds induce tensile residual stress, as the graphs in Fig. 7 show. For feeds of f = 0.3 mm and f = 0.7 mm residual compressive stress was detected. These stresses result from the thermomechanical load during the process. While high process forces induce residual compressive stress, high process temperatures lead to tensile residual stress. Figure 7 shows increasing residual stress for rising feeds for both grain sizes during soft and hard machining. This increase per feed is most significant for the B252 tools as the clogging of the tools leads to large amounts of friction on the work piece surface, which results in rising temperatures in the contact area. On the other hand, the machining with the B602 tool leads to higher total values of residual stress for all investigated feeds. As the resulting RS can be understood as a ratio between process forces and contact area temperature, high residual stress is induced by high temperatures or low normal forces if the other variable remains constant. Since the process forces are lower for the coarser grains, the mechanical load on the work piece surface is reduced, which explains the higher level of residual stress.

In conclusion, the measured residual stress is a result from the process forces and temperature. As the hardened material leads to higher normal process forces over all investigations for both tools, the level of residual stress is lower. The clogging of the B252 tool leads to increasing contact area temperatures and therefore to higher (tensile) residual stress, while the linear increase of the residual stress of the B602 tool is credited to constant material removal mechanisms at higher removal rates.

As the presented tools are designed for roughing operations, the surface roughness is not a focus point of this paper. It is for future applications, however, important to present the resulting surface roughness when working with these tools. The surface roughness is proportional to the chip thickness which is in turn a function of the grain size. Coarser grains therefore lead to higher surface roughness. Figure 8 shows the surface roughness Rz for increasing feeds for both tools. On the left, the surface roughness of the soft shafts is shown, while the right diagram presents the roughness of the hardened shafts. It is apparent that the roughness after grinding with the B602 tool is about twice as high as the roughness after the use of B252 tools, measuring Rz = 58 µm for a feed of f = 0.3 mm while the B252 grinding results in a roughness of Rz = 31 µm. The roughness progressively increases with rising feeds for both tools to Rz = 82 µm (B602) and Rz = 41 µm (B252) for soft shafts. For hardened shafts, the roughness increases over a range of 35 µm ≤ Rz < 55 µm (B602) and 20 µm ≤ Rz < 40 µm (B252).

Summarizing, the aforementioned clogging of the B252 tool leads to smoother surfaces resulting from additional rubbing of the adhering material with the workpiece surface. The surface roughness of the hardened shafts is overall significantly lower than that of the soft shafts, as the material separation is more brittle and therefore fewer bulgings occur on both sides of a single grain engagement.