Microstructure and surface quality of SLM printed miniature helical gear in LSPwC

Abstract

The present work describes the influence of underwater laser shock peening without coating (LSPwC) on selective laser melting manufactured meso-size (outside diameter ≤ 10 mm) helical gears. Five experiments were conducted using energies in the 200 mJ up to 1 J, while the spot size and overlap were kept constant as 1 mm and 90 %, respectively. Responses were measured and compared in terms of surface residual stresses, surface roughness, and microstructure of LSPwC-treated samples. Results show the development of significant compressive residual stresses in the root of the LSPwC processed helical gear, where it changes the state from tensile +45 MPa to compressive −421 MPa. Surface roughness has shown improvement, while volumetric material peak confirms the reduction by over 50%. Microstructure study was performed at the surface and by cross-section using scanning electron microscopy and electron backscatter diffraction analysis. The grain refinement and change in misorientation were observed, confirming plastic deformation.

Introduction

Gear is one of the oldest and most frequently used machine components. Gears can be applied almost at all the kinematics where power and motion are to be transferred [1]. The demand for the miniaturization of machinery and products has made the gear industries produce miniature gears to suit the client's requirements. Miniature gears found applications in various sectors such as aerospace, automotive, scientific instruments, biomedical and surgical instruments, etc. [2]. The production of these components is complex due to their very small size, the unavailability of tooling in traditional manufacturing processes, and the high initial cost involved in advanced manufacturing [3]. The complexity lies even after their manufacturing, as there is no convincing post-processing operation available in the market which effectively improves their surface quality and integrity to enhance their service life [1, 4]. Non-availability of post-processing operations and poor surface quality attainment by available manufacturing processes led to reduced service life and operating performance of the miniature components. The residual stresses caused by micro-manufacturing procedures or surface modification methods in micro and miniature parts or components play a significant role in assessing their service life. Most micro-manufacturing processes such as micro-milling, micro-powder injection technology, micro-moulding, extrusion, micro-stamping, additive manufacturing (3D printing), etc., do not allow the production of the required residual stresses in the substrate material, and they eventually result in lowering the fatigue life of the developed components. The micro-gears fail due to poor fatigue strengths [1, 5]. Improving the surface properties of essential machine elements has become an indispensable manufacturing step for enhancing their mechanical and metallurgical characteristics, such as fatigue life, corrosion resistance, and resistance to wear and erosion. Several conventional and non-conventional processes are available to enhance the mentioned qualities for the macro-sized components. The industry most commonly used one is mechanical shot peening (SP). Still, it has minimal application in micro and miniature components due to its restricted area and tight micro-geometrical tolerances. Over the last few years, laser shock peening (LSP) has emerged as an industrial-quality technology to induce compressive residual stress deeper than SP processes. LSP induces shallow cold work with a production of considerable surface finish, no dimensional alterations, and enables faster and more precise control over the process. LSP implies that mechanical pressure, not heat, is typically imparted to the material [4–6]. Usually, LSP, where an ablative coating is used, is considered a cold-working process. In the case of laser peening without ablative coating (LSPwC), the surface of the workpiece interacts with the high-pressure, high-temperature plasma; consequently, thermal interaction is usually present. In order to limit the free propagation of the plasma, a tamping layer is necessary. This tamping layer (usually a thin overlay of water) increases the maximum achievable pressure of the plasma. Therefore, it increases the productivity of the process. The plasma's pressure must exceed the material's yield strength to achieve plastic deformation and deep compressive residual stresses. This plastic deformation is often connected with refined microstructure and/or crushed grains [7]. The affected depth is a combination of multiple factors where the material properties, laser parameters (mainly spot size, pulse energy, and power density), and peening parameters (overlap) are intertwined. The usual approach is to use as energetic pulse as possible by the laser system, with as large as possible spot size, with the aim of high productivity and low attenuation of shockwaves into depth. The laser peening without coating (LSPwC) technique was initially examined by Mukai et al. [8] with the goal of producing significant shock wave pressures using low pulse energies and small spot sizes. There has been relatively limited research on the effects of LSPwC on additively manufactured stainless steel alloy components. In-depth research on the impact of LSPwC on conventionally manufactured SS304 austenitic steel plates and grain refinements was conducted by Natraj and Swaroop [9]. Sundar et al. [10] studied LSP's effect on Ti6Al4V alloy, SAE 9260 spring steel, and DIN X20Cr13 martensitic stainless steel with different sacrificial coatings such as tape and paint. Sun et al. [11] studied double-side LSP's effects on Laser-welded aluminium alloy joints filling Al+1.0 wt-% CeO₂. They showed improved residual stresses, porosity, and refined microstructure of LSPeened samples. Other studies on LSPwC and LSP have been conducted on a variety of materials, including Ti-6Al-4 V alloy [12,13], Ti 15 V-3Al- 3Cr-3Sn alloy [14], Al-Si10-Mg [15], 40CrMo Steel [7], etc.

LSPwC can potentially improve the mechanical and metallurgical properties of micro and miniature parts due to their deep penetration capabilities, better surface quality aspects, and unaltered geometry of the treated components. But this can only be achieved by selecting optimum values of laser parameters such as energy per pulse, pulse duration, repetition rate, and temporal and spatial beam shape. Considering the mentioned points, in this paper, laser shock peening has been applied to meso-size helical gears (overall diameter under 10 mm) [16] manufactured utilizing the capabilities of selective laser melting. The challenges in providing a protective tape have been deciphered by performing the underwater laser peening without coating (LSPwC). Peening action was targeted in the fillet gap and root zone of the helical gear, and considering the complexity of the very small gap between the two teeth, the robot was programmed in such a way that the exact centre point should reach in each layering with LSP. Five different values of energy (200 mJ to 1 J) were used to perform the experiment. The outcome of the investigations was recorded in terms of surface residual stresses using an X-ray diffractometer, and surface morphology was analyzed with the help of confocal microscopy. To understand the physical mechanism of the process and the effect of energies on helical gear root and fillet, scanning electron microscopy (SEM) images were studied. Electron backscattered diffractometry (EBSD) study was also performed to analyze the change in the grain sizes and misorientation in LSPwC region. All the analyses were performed before and after LSPwC.

Materials and method

Stainless Steel (SS) grade 316L in powder was used to manufacture the meso-size helical gears in selective laser melting (SLM). SS 316L has been selected as print material for its extensive use in the gear industries, especially in miniature gear manufacturing [1, 17]. Trumpf's Truprint 1000 model was used to produce miniature gears of an outside diameter of 9.6 mm with module 0.8 and a number of teeth of 10. A laser focus diameter of 0.055 mm, slice thickness 0.02 mm, and hatch distance 0.1 mm were used as input parameters for SLM for the used particles as suggested by the manufacturer and previous printing experience. The LSPwC procedures were carried out using the Litron laser system while submerging the gears into the water with the help of robot, as the protective taping in such a small area was not feasible. The underwater environment does not allow for the use of the standard 1064 nm due to its low transmission in water. Therefore, the second harmonic generation (SHG) at 532 nm was used. The 15 ns-long pulses have a circular top-hat beam profile. The laser pulse energy ranged from 200 mJ to 1 J with 200 mJ increments. At the same time, the other variables, including the spot size, overlap (with regard to the diameter of the spot), and repetition rate, remained fixed at 1 mm, 90%, and 10 Hz, respectively. In Figure 1(a), an illustration of the LSPwC arrangement is shown schematically. In order to prevent the flank geometry of the helical gear from being changed by the impact of LSPwC, the root and fillet gap is the intended zone for LSPwC treatment of the meso-size helical gear. The LSPwC treatment was targeted at the root of the gear where cracks are most likely to occur resulting in ultimate failure [16,17]. Figure 1(b) displays a representation of the LSPwC target zone. The complexity of performing LSPwC in such a narrow gap without affecting the flank geometry of the gears was achieved by mounting the gears on a robotic arm, ‘Fanuc M-20iA/20 M’. An auto-synchronized robot movement with Litron LPY ST 7875-10 laser operation was used to achieve laser spot size and overlap accuracy on the treated sample. The robot programming was done so that only the meso helical gears’ root and fillet gap were affected during the treatment. Surface residual stress, surface roughness, and microstructure were investigated both on the treated and untreated gears.

Figure 1

Schematic representation of (a) LSPwC arrangement for processing of meso helical gears; (b) LSPwC target zone; (c) representation of forces on the root; and (d) representation of surface unit normal vector and the laser.

RIGAKU make ‘AutoMATE II’ X-ray diffractometer (XRD) was used to quantify the surface residual stresses using the sin²ψ method. A Mahr, Germany make ‘MarSurf CM explorer’ confocal microscope was used to measure the surface roughness of the untreated and LSPwC-treated gears. The considered surface roughness parameters were evaluated using a 4 mm evaluation length and a 0.8 mm cut-off length (i.e. the average surface roughness Ra, arithmetic mean height Sa). The investigations also considered the characteristics of the material ratio curve. Scanning electron microscopy, using ‘Tescan’ FE-SEM, was used to examine the surface morphology. In addition to the fundamental surface morphology, the physics underlying the results was investigated using kernel misorientation mappings (KAM) maps and electron backscattered diffractometer (EBSD) images.

Results and discussion

The confining medium prevents the plasma from expanding away from the surface, generating shock waves. These shock waves are plastically deforming surfaces, and compressive residual stresses are imparted into the subsurface [18–20]. In water-confined mode, which was used in the experiment, plasma pressure is given by the formula [16]: (1) where is the incident laser power density, α is the fraction of the total plasma internal energy used as thermal energy, and Z is a reduced acoustic impedance given by [16,18]: (2) Where, i.e. and are the shock impedances of the water and the target material i.e. SS316L in the present work. Z_water: 0.157 X 10⁶ g cm⁻² s⁻¹ and Z_target: 4.47 X 10⁶ g cm⁻² s⁻¹.

Plasma pressure is an important aspect of evaluating the effect of LSPwC on the material. To undergo plastic deformation, the plasma pressure must be higher than the dynamic elastic limit of the material, which can be computed using the Hugoniot elastic limit (HEL) [18]. Table 1 presents the calculated values of peak pressure from Equation 1 and considered responses in terms of surface residual stresses for all five experiments performed on meso helical gears under LSPwC. It can be seen from Table 1 that the plasma pressure generated in during the LSPwC at different energies was in the range of 1.9–4.6 GPa, which is sufficient to induce plastic deformation in SS316L material as the plasma pressure is higher than the dynamic elastic limit of the material. The detailed theory of stress wave propagation in materials and plastic deformation can be found in [18]. Table 1 depicts the maximum improvement in compressive residual stresses achieved using the power density of 3.4 GW/cm², while the minimum was recorded in experiment 4, i.e. at 6.79 GW/cm². It is also evident from Table 2 that the values of considered parameters of surface roughness have shown considerable improvements for all the input parameters except at 400 mJ, where the average surface roughness value increases from the base values. These observations were further discussed in detail using the SEM images, as presented in Figures 2 and 3, and EBSD images, as illustrated in Figures 4 and 5. Figure 2 shows the surface SEM images of the unpeened (Figure 2(a–c)) and LSPwC-treated samples (Figure 2(d–h)). It can be seen from Figure 2(a–c) that the unpeened surface has several unmelted or partially melted powder particles over the top surface; these particles are very firmly attached to the surface and thus cannot be removed easily without any post-surface finishing application. Such inappropriate fusion and melting powder particles also lead to higher surface roughness. Whereas, from Figure 2(d–h), it is visible that the LSPwC has helped make the surface flatter by pushing the partially melted powder over the surface in the voids and craters. Improvements can be seen in all the experiments. The detailed study of grain structure and grain size distribution is presented in a subsequent discussion with the help of Figures 3–5.

Figure 2

Scanning electron microscopy surface images of the fillet gap of the flat plate samples and meso helical gear (a,b,c: Unpeened) and (d,e,f,g,h: LSPwC treated at different energies).

Figure 3

Surface optical images for (a) Unpeened gear and (b) LSPwC-treated gears at 400 mJ.

Figure 4

Electron backscatter diffractometer images for fillet gap of meso helical gear for (i) Unpeened (a: IPF; b: Grain structure; c: Kernel Misorientation; d: Grain size vs Area Fraction); and (ii) LSPwC treated (a: IPF; b: Grain structure; c: Kernel Misorientation; d: Grain size vs Area Fraction).

Figure 5

Electron backscatter diffractometer images for root of meso helical gear for (i) Unpeened (a: IPF; b: Grain structure; c: Kernel Misorientation; d: Grain size vs Area Fraction); and (ii) LSPwC treated (a: IPF; b: Grain structure; c: Kernel Misorientation; d: Grain size vs Area Fraction).

Table 1

Details of input parameters and experimental findings.

Input parameters			Average surface residual stresses in (MPa)
Energies (mJ)	Power density GW/cm2	Plasma pressure in GPa	Unpeened (average)		LSPeened
200	1.70	2.06	+ 45	10 (+/-)	−265.5	43.2 (+/-)
400	3.40	2.92			- 421.9	65.1 (+/-)
600	5.10	3.58			−346.2	69.1 (+/-)
800	6.79	4.13			−235.2	33.8 (+/-)
1000	8.49	4.62			−248.5	48.1 (+/-)

Table 2

Details of considered parameters of surface roughness.

Input parameters	Avg. surface roughness in (µm)	Avg. surface roughness in (µm)	Arithmetical mean height in (µm)	Arithmetical mean height in (µm)	Mean Depth of roughness in (µm)	Mean Depth of roughness in (µm)	Maximum height in (µm)	Maximum height in (µm)
Energies (mJ)	Unpeened (average)	LSPeened	Unpeened (average)	LSPeened	Unpeened (average)	LSPeened	Unpeened (average)	LSPeened
200	11.6	7.33	64.1	26.6	81.4	70.7	532.2	287.1
400		19.35		32.9		107.4		368.8
600		6.46		22.7		60.6		283.8
800		11.15		31.1		42.6		206.3
1000		8.42		27.7		48.2		223.1

Figure 3 depicts the mirror-polished optical microscope image of the unpeened sample (Figure 3(a)), and LSPwC-treated sample at 400 mJ (Figure 3(b)). It can be inferred from Figure 3(a) that the SLM printed gears consists of several micropores fundamentally due to gas porosity. Since the powders used for printing these gears are used one thus, they may include some pores within them and result in gas porosity during printing. Also, it can be reaffirmed that there are several unmelted or partially melted powder particles near the root and fillet gaps. While comparing Figure 3(b) with Figure 3(a), it can be noticed that the micropores are reduced, which is an essential benefit of LSPwC, and the surface near the root and fillet gap has smoothened without any micro cracks and defects.

Figure 4 presents the electron backscatter diffractometer images for the fillet gap of meso helical gear; it also shows the kernel misorientation mapping and histographical representation of grain sizes (in µm) versus area fraction. Figure 4(i) presents the details for the fillet gap of an unpeened gear, while Figure 4(ii) shows the effects of LSPwC on fillet gap of the gear at 400 mJ. From Figure 4i(a–c), it can be seen that the large grains in the form of flake are visible, high angle grain boundaries (HAGBs) can be seen in Figure 4i(c), and from Figure 4i(d), it is noticeable that the majority of the grains are of size 120 µm or more. While from Figure 4(ii), it is visible that the more prominent grain breaks into several smaller grains and creates sub-grains due to the impact of LSPwC at 400 mJ, the HAGBs, as can be seen in Figure 4i(c), have changed in low angle grain boundaries (LAGBs). Whereas most of the grain sizes have been reduced to under 30 µm. A similar representation as in Figure 4 has been presented in Figure 5 for the root region of the meso helical gear (Figure 5(i): unpeened) and (Figure 5(ii): LSPwC treated at 400 mJ). Due to the nature of the geometry of helical gears, the tangential interaction of the top-hat beam leads to intense plastic deformation and creates a large number of dislocations in the microstructure during the propagation of the laser shock wave. The dislocation density will then rise close to the sub-grain borders because the sub-grain boundaries continue to absorb the accumulated dislocation structure and climb in the grain. A portion of the sub-grain boundaries that have moved and rotated continuously absorb the nearby dislocation structures, increasing the angle of the grain boundaries until they finally become low-angle grain boundaries. Additionally, as the plastic deformation progresses, the HAGBs continue transforming into LAGBs, refining the initial structure.

Surface roughness plays a vital role in the estimation of the surface quality of the gears. Higher surface roughness has detrimental effects on the performance of gears as they tend to have crack initiation that may propagate through the root of the gear and eventually lead to the gear's failure. Thus, a lower surface roughness value is desired for better functional performance of the gears during use. Therefore, important surface roughness parameters such as average roughness (Ra), mean roughness depth (Rz), arithmetical mean height (Sa), and maximum height (Sz) were studied before and after LSPwC treatment, and the results for the same are presented in Table 2 and Figure 6(a,b). It can be inferred from Figure 6(a) that the LSPwC treatment has improved the Ra and Rz for almost all the energies except at 400 mJ as the unmelted particles stacked over the top of the surface were pushed to the voids and thus making the surface flatter as can be seen from Figure 2(d–h). Similar observation can be noticed in the case of arithmetical mean height Sa and maximum height Sz; the LSPwC treatment has shown more than 50% improvement for almost all the energies used. Further material ratio curve/Abbott Firestone curves, as presented in Figure 6(c) (As-built) and Figure 6(d) (LSPwC treated), were used to study the effect of LSPwC on peak material volume (Vmp), core material volume (Vmc), dale void volume (Vvv), and core void volume (Vvc). It is visible by comparing Figure 6(c) with Figure 6(d) that the LSPwC has reduced all the considered volumetric material parameters by over 50%. Improvement in material ratio curve parameters was again due flattening of the surface since peak reduction from the surface of the material led to lower values of considered parameters.

Figure 6

Graphical representation of considered parameters of surface roughness before and after LSPwC treatment (a) average surface roughness and mean roughness depth; (b) arithmetical mean height and maximum height; and (b) material ratio curve for untreated; and (b) material ratio curve of LSPwC-treated gears.

Conclusions

The present work describes the effect of laser shock peening without coating on SLM printed meso-size helical gears. The following conclusions can be drawn based on the experimental findings:

LSPwC imparted compressive residual stresses in the root and fillet gap, the maximum improvement was observed at 400 mJ where the surface residual stresses change their regime from tensile (+45 MPa) to compressive (−421 MPa).

It was observed that the LSPwC treatment helps in improving the pitting defects and no cracks or surface damage was seen after LSPwC. Some delamination of the unmelted particles during SLM printing was observed.

The microstructure and EBSD results confirm the increase in dislocation density leads to grain refinement during LSPwC.

Considerable improvement was seen in parameters of roughness, particularly, in terms of mean roughness depth, arithmetical mean height, and maximum height. The improvement observed in the parameters of the material ratio curve and roughness together tends to have improved the surface quality of the gears.