1 Introduction

Fine machining by grinding is of decisive importance in the manufacture or repair of components, as this represents one of the final machining steps and thus sets the surface and subsurface properties. The surface and subsurface properties resulting from grinding determine, the application behavior of the components. In the case of forming tools, for example, the flow characteristics of the workpiece during forming are determined by the surface and subsurface properties [1]. High degrees of forming and the work hardening that occurs lead to locally different mechanical loads during forming, which result in high wear [2]. Through the final grinding process, the subsurface can be functionalized by selectively inducing compressive residual stresses in such a way as to counteract the locally occurring load stresses and thus increase the service life [3].

In order to be able to selectively adjust residual stresses during grinding, the separation mechanisms and, above all, the grinding power applied must be taken into account so that no thermal damage occurs. In order to prevent thermal damage to the components, cooling lubricant, mostly grinding oil, is used in large quantities during grinding [4,5,6]. However, the use of cooling lubricant is associated with both economic and ecological costs. Therefore, one obvious option is to dispense with cooling lubricant.

Dry grinding processes are considered problematic, as it can lead to a decrease of the workpiece quality. Basic functions such as lubrication, cleaning of the grinding tool and cooling are also omitted. Thus, there is a risk of thermal workpiece damage due to grinding burn and considerable tool wear. For this purpose, Tönshoff et al. investigated the process behavior of different cutting materials with regard to their suitability for dry grinding. High workpiece speeds of vft = 2 m/s were used in order to reduce heat exposure times. In internal cylindrical grinding, the potential of vitrified bonded cubic boron nitride (cBN) grinding tools to machine workpieces without negatively affecting the surface and subsurface was demonstrated. However, the wear associated with dry machining was found to be critical. As a consequent conclusion, an optimization of the grinding wheel properties was found to be necessary [7]. Tawakoli et al. address the challenge of dry grinding by reducing the contact area to optimize friction and chip formation. By individually adjusting kinematic parameters, such as the shape of the dresser, defined structures are induced into the grinding wheel. The investigations show that the reduction of the contact area leads to comparably low cutting forces and to less heat generation. Reduced surface finishes due to the small contact area were identified as a disadvantage of this method [8]. In order to realize a dry grinding process, the heat input into the workpiece has to be reduced in order to avoid component damage. In current research, different approaches are being pursued. This paper shows the possibilities of additive manufactured (AM) grinding tools to enable a dry grinding process.

In contrast to the conventional manufacturing route, additive manufacturing of grinding tools offers the possibility of producing any grinding tool geometries close to final contour. On the one hand, this saves resources in the manufacture of grinding tools and, on the other, shortens the process times for preparing them for use, e.g. in profiling. This type of manufacturing also allows the production of geometries that were previously impossible or only possible to a limited extent using conventional manufacturing processes, and whose size can also be scaled. In the investigations presented here, toric grinding pins with a diameter of T = 30 mm were used. This shape and size of the grinding tools offers several advantages. Due to the small diameter for grinding tools, these tools can be integrated in conventional milling machines with standard tool holders and stored in the tool magazine. Furthermore, the toric geometry of the tools results in a small contact area between tool and workpiece, which reduces the heat input into the workpiece and the possibility to machine freeform surfaces in addition to plane surfaces [9, 10].

The combination of resource-saving production of the grinding tools and the possibility of integration in 5-axis machining centers with no need for cooling lubricant offers the possibility of manufacturing complex components economically and ecologically.

2 Materials and Experimental Setup

2.1 Materials

Powder metallurgical high-speed steel AISI M3:2 (ASP®2023) was used for the experimental investigations. The high-speed steel was heat treated at an austenitizing temperature of 1.400 °C by vacuum hardening followed by tempering three times at 560 °C for 1 h each time. After heat treatment, the high-speed steel has a hardness of 62 ± 2 HRC. The chemical composition of the high-speed steel is given in Table 1. The specimens used have a rectangular geometry with dimensions of 41 × 35 × 15 mm3.

Table 1 Chemical composition of AISI M3:2
Full size table

2.2 Grinding Process

2.2.1 Toric Grinding Pins

Vitrified bonded cBN toric grinding pins with a diameter of T = 30 mm and a torus radius of rT = 5 mm were used for the investigations (Fig. 1, left). Due to the fact that there is no cooling during dry grinding and the resulting higher thermal load on the cutting grain, cBN was selected as the cutting material, since diamond tends to graphitize at high temperatures and the iron in the material might decrease this thermal transformation point even further. Two different types of tools were used. On the one hand, conventionally manufactured tools were used, and on the other hand, toric grinding pins were manufactured by an innovative additive manufacturing process by the company BDW-BINKA Diamantwerkzeug GmbH, Germany. The grit size of both tool types was dG = 91 µm at a concentration of C = 125.

Fig. 1
figure 1

Representation of a toric grinding pin(left) and positions of residual stress measurement on the specimen (right)

Full size image

2.2.2 Grinding with Toric Grinding Pins

The grinding investigations were carried out on a CNC 5-axis milling machine tool RMF 600 DS (Röders Tec, Germany). Two different grinding strategies, lateral and frontal, are possible when grinding with toric grinding pins. The lateral grinding strategy in synchronous operation was selected here. The lateral grinding strategy is characterized by the fact that the cutting and feed directions are parallel to each other [11]. The process parameters used are summarized in Table 2.

Table 2 Parameters of the grinding process
Full size table

2.3 Surface Measurements

The topographies of all tool surfaces and workpiece surfaces were measured in a field of 3 × 3 mm2. The measurement was performed with a DUO Vario (Confovis, Germany). The surface measurements were evaluated with the software “µsoft analysis premium 8 ®” (NanoFocus AG, Germany).

For the description of tool wear, the initial condition of the tools was determined first. Subsequently, a volume of VM = 43.05 mm3 was machined, as well as a reference volume of VRef = 6.05 mm3 for the analysis of the surface and subsurface properties. After this defined cutting volume, the surfaces of the tools were measured at three different positions (0°, 120° and 240°) on the circumference of the tool. The wear of the tools is described by clogging and the G-ratio. The G-ratio describes the ratio between the machined volume (1) and the wear volume, whereby the wear volume is described by the radius wear (2):

$$G=\frac{{V}_{M}}{{V}_{T,w}}$$
(1)
$${V}_{T,w}=\frac{90 \pi }{360}*\left({{r}_{T,v}}^{2}-{{r}_{T,n}}^{2}\right)*\pi {r}_{T,n}$$
(2)

with: rT,v is the torus radius before grindingrT,n is the torus radius after grinding, VM is the machined volume, VT,w is the wear volume of the tool

2.4 Residual Stress Measurements

Residual stress measurements of the ground workpieces (Fig. 1, right) were performed by X-ray diffraction using the sin2ψ-method in ω-mode [12]. The measurements were performed on a Seifert 3003 TT dual-circuit X-ray diffractometer (GE Inspection Technologies, Germany) with a Cr-anode in combination with a V-filter and a collimator with a diameter of dc = 2 mm. The applied accelerating voltage was Ua = 30 kV with an anode current of Ia = 35 mA. The information depth was τ = 5.5 µm. All elastic constants and Bragg reflections used are listed in Table 3.

Table 3 Specific parameters used for residual stress measurements
Full size table

3 Results and Discussion

For the technological investigations on dry grinding, a total volume of VM = 592.2 mm3 was machined. With the selected process parameters, this corresponds to a machining time of approx. 6.5 h. After a volume of VM = 43.05 mm3, a reference surface was ground in each case, resulting in 12 reference surfaces. On these 12 reference surfaces, the surface and subsurface properties were analyzed afterwards. Furthermore, after each reference surface generated on the workpiece, the tool surfaces were measured at three different positions on the circumference confocally. The experiments were repeated with another AM toric grinding pin with same specifications. The break-off criterion was clogging on the grinding tool surface, which would lead to damage of the workpiece surface.

3.1 Tool Wear

The characterization of the toric grinding tools before and after the grinding tests, as well as after discrete cutting volumes VM, is based on a qualitative and quantitative evaluation of the surfaces of the grinding tools in the contact area and of the torus radii. The wear behavior is further described by the G-ratio. The calculation is based on formulas (1) and (2).

Before starting the investigations, SEM images (cf. Fig. 2) and topography images (cf. Fig. 3, above) of the tool surfaces of conventionally manufactured and AM grinding tools were recorded. On the left in Fig. 2 is the conventionally manufactured grinding tool, on the right the AM grinding tool. It can be seen that the AM grinding tool has a higher porosity. The conventionally manufactured grinding tool can therefore be described as more compact/denser. The tools show no cracks, clogging or other contamination before the start of the investigations.

Fig. 2
figure 2

SEM image of a conventional sintered and AM toric grinding pin

Full size image
Fig. 3
figure 3

Surfaces of the toric grinding pins after a machined volume of VM = 197.4 mm3

Full size image

The tool surfaces of the two toric grinding pins in the initial state and after a machined volume of VM = 197.4 mm3 are compared by topography images in Fig. 3. The conventionally manufactured tool is shown on the left and the AM-tool on the right. From the qualitative illustration, it can be seen that the conventionally manufactured tool already shows pronounced clogging after a machined volume of VM = 197.4 mm3. The clogging in the contact area can be seen on the entire circumference of the tool. As it can be seen in the SEM images in Fig. 2, the conventionally manufactured grinding tool is sintered more densely and has smaller pore spaces. From this it can be concluded that the pores are clogged after a very short time. Since the grinding process is carried out dry, the lack of cooling lubricant cannot free the pores from chips. As a result, the removed chips weld with the tool surface. The resulting increased effective contact area leads to an increased temperature as a result of increased friction. Due to the massive clogging and welding, grinding burn occurred. Thus, the previously defined break-off criterion was reached. In contrast, the AM grinding tool shows only small amounts of clogging. Due to the porous structure of the AM grinding tools, a correspondingly higher volume of chips can be absorbed.

The AM grinding tools were further used. The states of the tool surfaces after VM = 394.8 mm3 and 592.2 mm3 are shown in Fig. 4. It can be seen that the clogging of the tool surface first increases and then decreases again. After a machined volume of VM = 394.8 mm3, the highest degree of clogging was determined. An increase in clogging with subsequent decrease indicates a self-sharpening effect. This effect is known from vitrified bonded grinding tools dressed by crushing [13]. In contrast to crushed grinding tools, where the dressing process breaks the bond, it is assumed that AM grinding tools have a coarser structure with lower bond hardness as a result of the manufacturing process. The characteristic values Spk and Svk taken from the Abbott-curve (see Fig. 4 below).

Fig. 4
figure 4

Surfaces of the AM toric grinding pins after a machined volume of VM = 394.8 mm3 und VM = 592.0 mm3

Full size image

From the Spk value, which can be used as a measure of the grain overhang, it can be seen that the curve is constant over the entire machined volume. This indicates that there is no significant grain flattening, but rather grain splitting. The Svk value, which can be used as a parameter for clogging, shows an approximately constant course with fluctuations around a mean value of Svk = 97 ± 10 µm. This confirms the assumption that the tool surface is freed from clogging by self-sharpening.

The self-sharpening of the AM grinding tools means that clogs are continuously removed, but this is accompanied by higher radius wear. In the initial state, the conventionally manufactured grinding tool has a torus radius of rT,C = 5.027 mm and the AM grinding tool has a torus radius of rT,AM = 5.029 mm. After a machined volume of VM = 197.4 mm3, the conventional grinding tool has a torus radius of rT,C = 4.96 mm. The AM grinding tool has a torus radius of rT,AM = 4.98 mm after a machined volume VM = 592.0 mm3. According to formula (1) and (2), this corresponds to a G-ratio of 12.34 for the conventional grinding tool and a G-ratio of 101.7 for the AM grinding tool.

3.2 Workpiece Surface

In fine-finishing machining, the resulting surface topography is of key significance. The properties of the generated surfaces significantly depend on the tool specifications and the tool condition. The resulting surfaces of the two differently manufactured grinding pin types after a machined volume of VM = 197.4 mm3 are shown in Fig. 5. Both surfaces show a wave-like structure. This results from the selected grinding strategy, in which the tool is displaced line by line by the path distance. The surface produced with the conventional grinding tool has a rougher structure and a mean arithmetic height SaC = 0.822 µm. In contrast, the surface created with the AM grinding tool has a much finer structure, which is also reflected in SaAM = 0.598 µm. The deviation between the two generated surfaces is thus approx. 37%.

Fig. 5
figure 5

Workpiece surfaces after a machined volume of VM = 197.4 mm3

Full size image

Due to the high wear and failure of conventionally manufactured grinding tools, the surfaces of AM grinding tools are evaluated below. The resulting surfaces after a machined volume of VM = 394.8 mm3 and VM = 592.0 mm3 are shown in Fig. 6. The surfaces show roughness characteristics in a range of Sa = 0.4–1.1 µm after a grinding phase, where the mean arithmetic height is at a value of Sa = 1.47 µm. The mean value is Sam = 0.71 ± 0.22 µm. The scatter range of approx. 30% can be attributed to continuous self-sharpening and low bond hardness and is also reflected in the Spk and Svk values shown in Fig. 4. Furthermore, these provide a reason why no grinding burn occurs on the workpiece surfaces with the AM grinding pins.

Fig. 6
figure 6

Workpiece surfaces of the AM toric grinding pins after a machined volume of VM = 394.8 mm3 und VM = 592.0 mm3

Full size image

3.3 Subsurface Properties

In order to investigate the influence of the tool specifications and the tool wear on the subsurface state, sin2ψ residual stress measurements were performed on the reference surfaces. The results as a function of the machined volume are shown in Fig. 7. It can be seen that all grinding tools used induce compressive residual stresses in the subsurface. This can be explained by the plastic deformation due to material removal. Furthermore, a difference between the residual stresses in parallel and transverse direction can be seen. In cutting direction, part of the plastically deformed material is removed again as chips, which means that the induced compressive residual stresses are lower. Transverse to the cutting direction, the material is not removed. The material which has been pressed out of the cutting path and plastically deformed remains completely intact [14]. The induced compressive residual stresses decrease with increasing machined volume when grinding with conventionally manufactured grinding tools. At a machined volume of VM = 197.4 mm3, the compressive residual stresses reach a minimum with σpar = – 140 MPa parallel and σtrans = – 223 MPa transversal to the cutting direction. The constant decrease of the induced compressive residual stresses can be explained by the continuous clogging of the pores. As shown in Fig. 3, the conventional grinding tool is almost completely clogged in the contact area. Since no free grain layer is available and the relative contact area is increased, the mechanical load on the workpiece is reduced while the thermal load increases. Conventional grinding tools, as already described in the previous section, generated grinding burn, which led to the break-off criterion being reached. In contrast, the induced compressive residual stresses of the AM grinding tools are already higher after a machined volume of VM = 49.35 mm3 by ∆σ = 240 MPa for the parallel and by ∆σ = 165 MPa for the transversal compressive residual stresses. Figure 7 also shows a grinding-in phase up to a machined volume of approx. VM = 148.05 mm3. From this volume on, the residual stress values fluctuate around a mean value of σpar = – 880 MPa or σtrans = – 1310 MPa. The fluctuation around these mean values can be attributed to the different degree of clogging of the grinding tool and corresponds to the opposite course of the Svk values. As already described in Sect. 3.1, the grinding tools are continuously clogged up to a certain degree until self-sharpening sets in and grain splitting occurs. This results in the formation of new sharp grain tips or grain breakout, which reduces the effective contact area. Until the self-sharpening effect occurs, the thermal load increases as a result of increased friction, which leads to a decrease in the compressive residual stresses. The maximum induced compressive residual stresses are parallel to the cutting direction at σpar = – 952.9 ± 12.3 MPa and transversal to the cutting direction at σtrans = – 1498.3 ± 76.8 MPa.

Fig. 7
figure 7

Residual stress as a function of machined volume

Full size image

4 Summary

The prevention of high thermal workpiece stress and resulting workpiece damage represents the major challenge to realize a dry grinding process. Previous studies on dry grinding have shown that both the grinding process and the grinding tool specifications must be adapted to deal with these challenges. Therefore, the aim of the investigations presented in this paper was to demonstrate the general suitability of additive manufactured grinding tools for dry grinding. For this objective, conventionally manufactured vitrified grinding pins and additive manufactured vitrified grinding pins were used for grinding a hardened high-speed steel (AISI M3:2). The evaluation of the performance was done on the basis of tool surfaces, the G-ratio, the workpiece surface produced and the residual stresses induced in the workpiece. The results of the investigations are as follows:

Tool wear The conventionally manufactured grinding tool shows pronounced clogging and welding, leading to grinding burn and failure. In contrast, the additive manufactured (AM) grinding tool exhibits lower clogging due to its porous structure, allowing for higher chip absorption. The AM tools show self-sharpening, with clogging being continuously removed, resulting in higher radius wear. The G-ratio for the conventional tool is 12.34, while for the AM tool it is 101.7.

Workpiece surface The surfaces generated by the conventional grinding tool are rougher compared to those produced by the AM tool. The mean arithmetic height for the conventional tool is 0.822 µm, while for the AM tool it is 0.598 µm, representing a difference of approximately 37%.

Subsurface properties Both the conventional and AM grinding tools induce compressive residual stresses in the subsurface. The compressive residual stresses decrease with increasing machined volume for the conventional tool due to continuous clogging of the pores. In contrast, the AM grinding tools exhibit higher induced compressive residual stresses, with fluctuations around mean values of σpar = – 880 MPa or σtrans = – 1310 MPa. The fluctuation is attributed to the varying degree of clogging and corresponds to the changes in the tool's surface characteristics.

Overall, the results indicate that the AM grinding tools have advantages in terms of reduced clogging, self-sharpening, and higher compressive residual stresses. However, the AM tools also experience higher radius wear. Further investigations are needed to optimize the bond hardness and to investigate the influence of process parameters and tool specifications on surface and subsurface properties.