Study on surface roughness in laser-assisted machining of Si3N4 ceramics under different material removal modes

Ye-zhuang PU, Yu-gang ZHAO, Jian-bing MENG, Hai-yun ZHANG, Guo-yong ZHAO, Qian LIU, Pan-pan SONG

(Institute for Advanced Manufacturing, Shandong University of Technology, Zibo 255049, China)

Abstract: Laser assisted machining (LAM) is effective for machining brittle and hard materials such as engineering ceramics. Because it has larger technological space, is more economical, and can achieve a better surface integrity. In this paper, two sets of single-factor experiments were performed in which the laser power and the cutting depth were changed respectively. The change of the surface roughness under different material removal modes for LAM of Si3N4 was studied using a variety of parameters. The results show that the values of Ra and Rq are smaller in the plastically machining state than that obtained when the material is removed through a brittle fracturing or thermal damage occurs. When there is no thermal damage, the value of Ra and Rq decrease with the increase of the softening degree of the workpiece. In the ductile material removal mode, the distribution of height amplitude curve of the surface tends to be more symmetrical and more normally; the surface contour is more fine.

Key words: Laser assisted machining, Silicon nitride ceramics, Surface roughness

1 Introduction

Engineering ceramics are increasingly used in national defense, aerospace, mechanical, biomedical, informational, nuclear, and new energy fields due to their superior performance that metal and polymer materials do not possess [1-2]. As an important engineering ceramic material, silicon nitride (Si3N4) ceramics have many excellent properties, such as low specific gravity, high hardness, high temperature resistance, corrosion resistance, wear resistance and good biocompatibility. They are used to manufacture bearings of rocket turbopump, high temperature key components of engine and gas turbine, seals and cutting tools, etc. However, the cutting force and the cutting temperature are very high when engineering ceramics are cut under traditional machining method, because of their inherently high strength, high hardness and high brittleness. As a result, micro cracks will be produced on the machined surface, which greatly reduces the service performance of engineering ceramics parts, and the tool life will be reduced. The root cause is that machining is done in the brittle region, not in the plastic region. In order to solve this problem, laser assisted machining (LAM) is proposed. Extensive experiments conducted on various ceramics, such as Si3N4[3-9], aluminum oxide (Al2O3) [10-11], zirconia (ZrO2) [12], and other ceramic composite materials [13-17], have proved that laser-assisted machining (LAM) is an effective method to solve the above problems. Laser is used as the heat source to heat the ceramic workpiece in LAM. The high-power laser beam focused on the surface of the ceramic material can heat the material to a very high temperature in a short time before the material is cut, reduce the strength of the material, and then change the cutting performance of the material. Thus, the transition from brittleness to plasticity is realized. And, only when the workpiece is heated to a suitable softening state, the material can be removed plastically, and then better surface integrity and longer tool life can be achieved. If the heating temperature is too high, the machined surface can be burned by excessive heat. If the temperature is too low, the workpiece will not be softened sufficiently, and the material removal will still be in the way of brittle fracture.

Shin et al. [3-5] studied the removal mechanisms, the thermo-mechanical characteristics, and the surface integrity of Si3N4in LAM. Under the irradiation of a laser, with the increase of workpiece temperature, the viscosity of the intergranular glass phase decreases. When the temperature exceeds the glass transition temperature, Si3N4grains are repositioned, and the fluid glass phase material flows to form a new grain boundary under the action of cutting tools. Thus, the material is plastically removed. Among the operating parameters, laser power has the most important effect on the temperature of the workpiece, followed by the cutting depth. When the material is plastically removed, the machined parts have low surface roughness, moderate compressive residual stress, and no surface/subsurface microcrack or thermal damage. LAM can achieve lower cutting forces and specific cutting energy, longer tool life, and higher material removal rates. Damage-free silicon nitride parts with complex geometric features were machined, which further proved the feasibility and advantages of LAM. Kang and Lee [6] studied the laser-assisted milling process of Si3N4. A new back-and-forth preheating method was proposed, and a constitutive equation was determined. Lee et al. [7] studied the microstructural variations, the machining characteristics, and the oxidation mechanism of Si3N4from increasing the temperature. Wu [8] investigated the laser absorption and heat conduction process in LAM Si3N4, performed laser-assisted turning and laser-assisted milling of Si3N4experiments, and analyzed the surface integrity. Pu et al. [9] analysed comprehensively the chips, tool wear and the machined surface under different material removal modes in LAM of Si3N4. When the material is plastically removed, the chips are continuously band-like. Tool wear occurs on the rake face. The surface is smooth with no cracks or pits. When thermal damage occurs, band-like chips are longer. Tool wear can barely be seen, and the surface has a porous structure. When the material is removed through a brittle fracture, the chips become particle-like. The cutting edge is tipped, and cracks or pits can be seen on the surface.

Surface roughness is one of the technical requirements for evaluating product accuracy and designing parts, and it is also one of the important indexes for measuring machining performance [18]. Surface roughness refers to the microscopic geometric characteristics such as peaks and valleys on the machined surface. Surface roughness has a great influence on the wear resistance, corrosion resistance, sealing, fatigue strength and assembly reliability, etc. Therefore, the research and analysis of surface roughness will be beneficial to the prediction and control of surface quality. The current research has proved that the value of the arithmetical mean deviation Ra produced by LAM is lower than that produced by conventional machining, but the relationship between the surface roughness and the softening degree of the material has not been pointed out. And the surface roughness regarding the LAM-machined Si3N4surfaces is limited to Ra which is not sufficient or adequate. In this paper, two sets of single-factor experiments were performed in which the laser power and the cutting depth were changed respectively. And the change of surface roughness characterized using a variety of parameters under different machining states was studied.

2 LAM experiment

2.1 Experimental system

The experimental system of LAM of Si3N4is shown in Fig.1. The continuous laser with wavelength of 1070 nm emitted by the ytterbium fiber laser (YLR-150/1500-QCW-MM-AC, IPG Photonics Co.) was focused on the surface to be machined by the laser optical focusing system fixed on the apron of the CNC turret lathe (CKD6136i, Dalian Machine Tools Group). In continuous wave mode, the maximum average power can reach 250 W. The diameter of the laser spot on the workpiece surface can be changed by adjusting the relative position between the focusing system and the workpiece. The relative position of the laser spot and the cutting tool remains unchanged, and they can move synchronously. MCLNR2020K12 tool holders with CBN-tipped carbide tool inserts (CNGA120408 BNK30, Halnn Superhard) were used for all of the cutting operations.

Fig.1 Experimental system for laser assisted machining (LAM) of silicon nitride.

2.2 Experimental material

The gas-pressure sintered Si3N4workpiece (Acro New Materials (Dalian) Co., Ltd.) was cylindrical, and was held using a three-jaw chuck. The size is φ10 mm×100 mm. The material properties are listed in Table 1, and these values were measured by the materials manufacturer.

Table 1 Properties of Si3N4 workpieces.(The reference standard is ASTM F 2094.)

2.3 Test matrix and operating parameters

A single-factor experimental matrix was performed in which only the laser power was changed. The experimental parameters used are reported in Table 2.V,f,d,lrepresent the cutting speed, the feed, the depth of the cut, and the length of the cut, respectively. The laser power (P), the preheating time (t), the diameter of the laser faculae (D), the laser-tool axial distance (L), and the circumferential laser-tool angle (φ) were selected based on a simulation and previous studies. Fig.2 is the schematic diagram of the relative position of the laser optics, the cutting point, and the workpiece.

Table 2 Operating conditions for LAM experiments

Fig.2 The schematic diagram of the relative position of the laser optics, the cutting point, and the workpiece

In order to ensure the accuracy of the cutting depth, two layers of ceramic material were removed in each experiment. Between the two processes, the workpieces ensure adequate cooling. After the second cutting, the chips, tool wear and the machined surfaces were analyzed to infer the removal mode of the materials. Each group of parameters was repeated twice.

The surface roughness of the workpiece was examined using a CLSM(VK-X200 Keyence, Japan). The sampling length is 0.8 mm, and the measuring direction is along the tool feed direction. 19 equally spaced measuring lines were measured in two randomly selected areas with the size of 1 mm×1 mm from each workpiece. The average value of the four sets of test data is regraded as the surface roughness value under this set of machining parameters. When the laser power was 0 W, the machined surface was not measured because there were clearly visible pits and edge chipping. Fig.3 shows the measured values of surface roughness parameters with different laser powers.

Fig.3 The measured values of surface roughness parameters with different laser powers.

In [9], it was concluded that when the laser power was between 195 W and 135 W, the material was removed plastically; when the laser power was over 195 W, thermal damage occurred; and when the laser power was lower than 135 W, brittle fracture became the main removal mode of the material.

3 Results and discussion

3.1 Surface roughness parameters

The arithmetical mean deviationRaand the root mean square deviationRqare the most commonly used surface roughness parameters. It can be seen from Fig.3 thatRaandRqhad the same change trend. In plastically machining state, the values ofRaandRqwere less than the values obtained in other state, and they were the least when the laser power is 195 W. When thermal damage occurred, the values ofRaandRqincreasd with the increase of the laser power. When there was no thermal damage, the values ofRaandRqdecreased with the increase of the laser power.

The skewnessRskreflects the degree of asymmetry of the contour height amplitude curve relative to the average line. A skewness value of zero represents the symmetrical height distribution. A negative skewness indicates that the area of the peaks on the machined surface is small. So the peaks can be worn away quickly and a good bearing surface can be formed easily. Fig.3 shows that although theRskvalues were all positive except the one produced when the laser power was 100 W, theRskvalues were more closer to 0 in the plastically machining state. The kurtosisRkureflects the sharpness of the machined surface. When the kurtosis value is 3, the profile height distribution curve conforms to the standard normal distribution and the surface is ideal. When the kurtosis value is over 3, the surface is mostly sharp. When the kurtosis value is less than 3, the surface is relatively gentle.

The mean width of the profile elementsRsmis used to characterize the distance between peaks and valleys. The smaller the value ofRsmis, the finer the surface profile is and the better the sealing performance is.Rsmis usually used in conjunction withRaorRzand is not used alone.

The maximum peak heightRp, maximum valley depthRv, maximum heightRzare isolated peaks or valleys and do not actually represent the entire surface.Rcis the mean height of the profile elements.

3.2 Mechanism

In traditional cutting, the main factors that affect the surface roughness are: cutting parameters and tool geometry parameters, tool material and workpiece material, workpiece diameter and machining length, the extended length of tool handle and workpiece, insert installation errors, cutting forces, cutting temperature, tool wear, vibration and stiffness of the system, etc.

Compared with the traditional cutting, LAM is an intense thermo-mechanical coupling process due to the laser irradiation. In this paper, cutting parameters and tool geometry parameters, tool material and workpiece material, workpiece diameter and machining length, the extended length of tool handle and workpiece are fixed. The insert installation errors, the vibration and stiffness of the system can not be calculated, but can be ignored. With the increase of the laser power, the workpiece temperature increases, the softening degree of the material increases and the cutting force decreases. It can be seen from Fig.4 that as the degree of softening decreases, the tools changed from almost no wear to the rake face wear, and eventually became the cutting edge tipping. Fig.5 shows the SEM images of the machined surface under different laser powers. When the laser power was 225 W, the excessive temperature caused the oxidation of the surface. The porous structure formed by the burst of nitrogen bubbles under the surface resulted in the increase of the value ofRaandRq. With the decrease of the laser power, brittle fracture became the primarily way to remove materials. Cracks or pits appeared on the machined surface, and the value ofRaandRqincreased. But when the laser power was 40 W, the value ofRaandRqdecreased. The reason is that the laser power is too low to soften the material, and the intense friction between the flank face and the machined surface makes the surface smoother. Therefore, the softening degree of the material is the most important factor affecting the surface roughness in LAM.

Fig.4 Micrographs of the CBN inserts

Fig.5 The SEM images of the machined surface under different laser powers[9]

4 Experimental verification

In order to verify the above conclusion, another set of single-factor experiments was performed, which only changed the cutting depth. The laser power was fixed at 195 W, the cutting depths were 0.1 mm and 0.3 mm respectively, and other parameters were the same as the experiments in section 2.3. Therefore, the experiment with the cutting depth of 0.2 mm was the case 3 in table 2. Under the irradiation of the laser, the temperature gradually decreases from the surface of the workpiece to the heart. As the cutting depth increases, the softening degree of the cutting area decreases. It can be seen from Fig.6 that when the cutting depth is 0.1 mm, the porous structure on the surface indicates that thermal damage has occurred at this state. When the cutting depth is 0.3 mm, the surface is smooth and flat, and no cracks or pits can be seen. It can be inferred that the material is plastically removed without thermal damage. The relationship between the value of each surface roughness parameter and the cutting depth shown in Fig.7 shows that the change trend of parameter value with the softening degree of workpiece is consistent with the above conclusion.

Fig.6 The SEM images of the machined surface under different cutting depth

Fig.7 The measured values of surface roughness parameters with the cutting depth

5 Conclusion

(1) In LAM, the softening degree of the material is the most important factor affecting the surface roughness.

(2) The values ofRaandRqare smaller in the plastically machining state than that obtained when the material is removed through a brittle fracturing or thermal damage occurs. When thermal damage occurs, the values ofRaandRqincrease with the increase of the laser power. When there is no thermal damage, the values ofRaandRqdecrease with the increase of the softening degree of the workpiece.

(3) In the plastically machining state, the distribution of height amplitude curve of the surface tends to be more symmetrical and more normally; the surface contour is more fine.