Laser polishing of parts built up by selective laser sintering

Laser polishing of parts built up by selective laser sintering

ARTICLE IN PRESS International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050 www.elsevier.com/locate/ijmactool Laser polishing of parts...

2MB Sizes 0 Downloads 8 Views

ARTICLE IN PRESS

International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050 www.elsevier.com/locate/ijmactool

Laser polishing of parts built up by selective laser sintering A. Lamikiz, J.A. Sa´nchez, L.N. Lo´pez de Lacalle, J.L. Arana Department of Mechanical Engineering, University of the Basque Country, ETSII, c/Alameda de Urquijo s/n, 48013 Bilbao, Spain Received 27 June 2006; received in revised form 22 January 2007; accepted 23 January 2007 Available online 20 February 2007

Abstract In this work, a surface finish method for parts built-up by selective laser sintering (SLS) is presented. One of the main drawbacks of the SLS technique is the high surface roughness of resulting parts. Therefore, parts have to be polished to be valid for operation conditions. Polishing processes are usually based on manual abrasive techniques. However, in the present paper, a surface polishing method based on laser irradiation is presented. The laser beam melts a microscopic layer on the surface, which re-solidifies under shielding gas protective conditions, resulting in a smoother surface. Laser-polishing tests for lines, planar surfaces and inclined planes have been performed, with satisfactory results in all the cases. The experimental tests were carried out on sintered test parts with an initial roughness of 7.5–7.8 mm Ra. The tested material is a commercial alloy denominated LaserForm ST-100r, composed by sintered stainless steel and infiltrated bronze that it is used mainly for the constitution of injection moulds. Experimental results present final surface roughness below 1.49 mm Ra, which represent an 80.1% reduction of the mean roughness. Finally, a complete analysis of test probes and its metallurgical composition is presented. Considering that the material presents a non-homogeneous structure, the polished surfaces present slightly higher hardness values and are more homogeneous than the initial ones. Thus, polished surfaces do not present any heat affected zone or cracks, which could cause failure during the part operation. r 2007 Elsevier Ltd. All rights reserved. Keywords: Surface finishing; Laser polishing; Selective laser sintering

1. Introduction The number of laser-based systems and processes is continuously growing for many industrial applications in communications, medicine, industry and other strategic areas [1]. One of the emerging techniques is the laser surface treatment processes. Since new types of lasers have been developed and the requirements for surface treatment are tighter, some high added-value part manufacturers consider to apply the laser as a local heat source for the surface treatment of metal parts, regardless of the high costs. One of the most recent processes is the laser surface texturing, which is based on the vaporization of microscopic layers of material. The objective is to modify the surface topography getting a new desired feature such as Corresponding author. Tel.: +34 94 601 4221; fax: +34 94 601 4215.

E-mail address: [email protected] (A. Lamikiz). 0890-6955/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijmachtools.2007.01.013

tribological, functional, aspect, etc. Some works have been focused on improvement of friction conditions in piston rings [2], texturing methods for ceramic implants which improves the wettability of these elements [3] or laser honing of cylinders for high-performance car engines. In the field of surface finishing, laser polishing has been used more than 10 years ago for the polishing of optical glass lens. In [4] experimental results demonstrate that fused quartz surfaces can be polished from 2.0 to 0.05 mm Ry with a 25 W CO2 laser. In this case, the polishing mechanism is based on the fusion of a micro layer of material due to the action of the laser beam. This work proves that there is an acceptable energy density range for the surface polishing between 800 and 1100 J cm2. Under this range, surface roughness variations are not appreciated. Analyzing other existing works, laser polishing has been mainly used on diamond surfaces [5–7] and optical lenses of fused silica or quartz [8]. However, there are only few works studying metallic surface laser polishing, and no

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

experimental results can be found. Furthermore, the quoted references indicate that there is a lack of a systematic and logical research for metal polishing. A deep study of the process and some results are presented in [9]. In this work, laser polishing of selective laser sintering (SLS) parts is applied and reductions up to three times of mean roughness are achieved. The process was carried out by scanning the surface with a CO2 and Nd-YAG laser beam at high speed inside a vacuum chamber. This chamber was previously set to a pressure of 105 bar and then maintained with an argon flow. On the other hand, rapid manufacturing (RM) have become an alternative group of processes that may be the best solution to repair high-added value parts or even build up small functional parts. This trend can be observed not only in the number of research projects or scientific papers, but also in exhibitions and manufacturing industry reports [10]. One of the most extended applications of RM processes is the manufacturing of small-medium injection moulds or small inserts for larger moulds. One of the most interesting features of these techniques is the possibility to make complex internal cooling conducts following the part surface, which may lead in a reduction of the cooling time of plastic injected parts more than 30%. On the last few years, different RM techniques have been developed, mainly for metal deposition applications. These techniques such as: LENS [11], SLS [12,13] or selective laser melting, build up fully functional metal parts from raw powder material. The main advantages are the manufacturing-time reduction and the possibility of building parts with very complex geometry. Among the existing RM methods, SLSis one of the most extended in industry according to the Wholers Associates last report [14]. This process allows the fabrication of complex parts in different materials (polymers, ceramic and some metallic alloys). However, the resultant surface quality is very poor. In general, one of the main disadvantages not only for SLS, but also for all the metal additive processes, is the poor quality of the surface [15]. Mean surface roughness values for RM techniques used to be between 5 and 7 mm Ra while average values for injection moulds has to be less than 1 mm Ra [15]. The same manual abrasive techniques used for machined moulds are used to achieve the final finishing requirements

2041

in parts made by SLS, which can be time consuming and expensive job. Some investigations are addressed to eliminate them. Thus, in [16] a large-area electron beam irradiation is used to polish and treat small mould surfaces. The part is inserted in a vacuum chamber and a sequence of 60 mm diameter electron beam are pulsed, resulting on a high-quality surface with enhanced surface roughness and corrosion resistance. The main drawback of this technique is that the process has to be carried out on a special machine. This work presents a laser-polishing method for metallic surfaces. It is based on the fast melting and re-solidification of a surface microscopic layer by means of a laser raster strategy. Material of the surface peaks is vaporized and/or melted filling the valleys, resulting in a smoother surface. The final goal is to polish a sintered part on the same SLS machine. Once each layer has been consolidated, the same laser that has built the part can polish the exposed part surfaces, adding a final step on the SLS process.

2. Laser-polishing fundamentals Laser-polishing process is based on the melting of a microscopic layer and a fast re-solidifying of the melted material. The affected layer has to be deep enough to melt the roughness peaks, but it must not be deeper than the valleys. Therefore, the energy of the laser beam must be carefully controlled to melt just a microscopic layer. In Fig. 1, a scheme of the laser surface polishing process is presented. The process depends mainly on three factors: the surface material, its initial topography and the energy density of the laser beam. Therefore, once the surface is selected, the process would be successful if the energy density is correctly selected. The laser energy density (energy per surface unit) is much higher than other heat sources such as plasma-arc. Thus, it is possible to melt a microscopic area without affecting the surrounding areas, which is a key factor in laser-polishing process. The energy density depends on the laser power, the laser beam diameter and the time the beam is irradiated on the surface. When a pulsed laser beam is irradiated, this time is directly the pulse duration, but for a continuous wave laser, energy density is inversely proportional to the feed rate. In this last case, it is possible to

Fig. 1. Laser-polishing process scheme.

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

2042

calculate the energy density using ED ¼

6000P , DV f

(1)

where P is the laser power (W), D is the beam diameter (mm) and Vf is the beam feed rate (mm min1). The result is the effective energy density (J cm2). In surface treatment processes, a usual technique is to maximize the beam diameter increasing the focal offset distance (Fig. 1), and finally, fit the power and feed rate in order to obtain the desired energy density. In this way, a larger area can be processed increasing the productivity of the process. A deeper analysis of the process showed in [9], demonstrate the existence of two different regimes denominated surface shallow melting (SSM) and surface over melt (SOM), depending on the depth of the affected layer. Both, the laser energy density and the original surface topography determine the appearance of each one. In the SSM regime, the peaks are melted and valleys are filled resulting on a smoother surface, as it is theoretically expected. However, if the energy density is increased, a deeper layer is melted. If the melted layer depth is below the surface valleys a new regime denominated SOM starts and the obtained surface profile is a smaller frequency and bigger amplitude peak-valley sequence, resulting in a surface roughness increasing. Experimental tests show that it is an optimum value of the density of energy, where a minimum roughness is achieved. If the energy density is increased from this optimum value, the higher the density of energy, the higher roughness is obtained. The critical value of energy density is the limit between SSM and SOM regimes. Nevertheless, the experiments show that both regimes are combined and final surface roughness depends on one dominant regime. If laser parameters are not accurately controlled, it is very difficult to predict which one will be the dominant regime. The most difficult parameter to control is the focal offset distance, measured from the laser focal point to the part surface, because it affects directly to the laser beam diameter. In the experimental set-up used in this work, an accurate control of the focal offset distance has solved this problem and no additional control has been used due to the simple tested geometries. However, for complex geometries, constant focal offset can be very difficult to control. This problem can be solved by using a system similar to the applied in industrial five-axis laser cutting systems. These machines install an additional axis and an inductive position sensor in the laser head; the additional axis is controlled by a closed-loop system that corrects the focal position of the laser instantaneously. 3. Laser-polishing experimental tests and result discussion In order to study the potential of the laser-polishing process, a set of experimental tests was carried out. In these tests a commercial alloy denominated LaserForm ST-100r

was selected. This material is composed by approximately 60% of sintered AISI 420 stainless steel and 40% of infiltrated bronze; its mechanical and thermal properties are optimum for injection moulding applications. The test parts where first built up by SLS resulting in a porous AISI 420 stainless steel part. Then, the resulting part was infiltrated with bronze in order to reduce porosity and increase the mechanical properties. The laser-polishing experiments have been carried out in a 2500 W CO2 slab laser. The optic head was installed in a three-axis gantry machine with 10,000 mm min1 maximum feed rate. The laser optics was originally designed for laser cutting operations; therefore, it has been possible to use the cutting assistance gas nozzle as a protecting gas flow to avoid metal oxidation of the part surface during polishing. The laser presents a Gaussian power density profile. Thus, the maximum energy density is focused on the center of the spot. The optimum beam profile for laser polishing processes is a top-hat profile that presents a largeconstant energy peak. However, most of the RM machines use a CO2 or a Nd:YAG laser with Gaussian profiles, in order to maximize the focusing of the laser beam and improve the resolution of the part details. Therefore, the experiments have been carried out taking into account the beam profile of most RM machines. The main process parameters are the laser beam feed rate, the laser power and the laser beam diameter. However, since the beam diameter is an indirect parameter, consequence of the focal offset distance, the controlled parameter in all tests has been the focal offset distance. The laser beam diameter was measured for each focal offset distance. The objective of the tests is to find a parameter combination to obtain the lower mean roughness (Ra) with the maximum feed rate and beam diameter. Therefore, the target is to increase the process productivity to the maximum value. Measured parameters have been the final topography of the polished surface, the mean roughness and the reduction index calculated as the relationship between the final and the initial roughness. These values have been obtained with a 3D profile and roughness measurement system. The experimental tests have been divided into three groups: simple planar line polishing tests, planar surface polishing tests and 3D-polishing tests. In Fig. 2 the tested parts geometry are shown. 3.1. Simple line polishing tests First, to check the viability of the process, a series of simple line tests were carried out in planar surfaces. The test parts were manufactured by SLS with a layer thickness of 0.15 mm, resulting on an initial mean roughness of 7.5 mm Ra, which is a common value for sintered surfaces. Tested area was a LaserForm ST-100r 25  90 mm planar surface. The experimental tests parameters are presented in Table 1. For all tests, argon was injected coaxially as shielding gas. The experimental results show that the optimum parameters

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

2043

Fig. 2. LaserForm ST-100r test parts built-up by SLS.

Table 1 Laser-polishing test parameters Test no.

Power (W)

Feed (mm min1)

Focal offset (mm)

Laser beam diameter (mm)

Energy density (J cm2)

Roughness reduction (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

600 900 600 1200 1200 900 900 600 900 600 1200 900 1200 900 900 900 600 900 900 1200 900 1200 900 600 600 900 600 1200 600 900 1200 1200

2000 1400 800 800 2000 800 1400 1400 800 800 800 1400 1400 800 1400 1400 1400 2000 1400 1400 2000 1400 1400 800 2000 2000 2000 2000 1400 1400 800 2000

30 20 30 40 30 30 30 20 40 40 30 30 30 20 30 30 30 30 30 40 40 20 40 20 40 20 20 40 40 30 20 20

0.93 0.54 0.93 1.3 0.93 0.93 0.93 0.54 1.3 1.3 0.93 0.93 0.93 0.54 0.93 0.93 0.93 0.93 0.93 1.3 1.3 0.54 1.3 0.54 1.3 0.54 0.54 1.3 1.3 0.93 0.54 0.54

1935.48 7142.86 4838.71 6923.08 3870.97 7258.06 4147.47 4761.90 5192.31 3461.54 9677.42 4147.47 5529.95 12,500.00 4147.47 4147.47 2764.98 2903.23 4147.47 3956.04 2076.92 9523.81 2967.03 8333.33 1384.62 5000.00 3333.33 2769.23 1978.02 4147.47 16,666.67 6666.67

74.8 0.9 (no coherent) 62.8 59.6 82.7 73.3 78.8 78.6 59.5 52.0 71.3 80.2 72.2 74.5 70.0 86.9 81.1 80.9 80.3 78.8 75.6 73.1 73.2 79.5 72.7 77.4 77.4 80.5 80.1 79.5 57.6 59.5

were in the range of 4000–4500 J cm2 energy density. Some dispersion on the results can be observed, which is logical taking into account that the initial surface topography is one of the main parameters (Fig. 3). Once the tests were performed, initial and final topologies were measured. The roughness measurements show mean roughness values below 1.49 mm Ra, which means a 19.9% of the initial surface mean roughness, i.e. 80.1% reduction. Fig. 4 shows the original topography of the original test part surface and one of the laser-polishing

tests. The first appreciable result is that the resulting surface is much smoother than the initial one. The evaluation of the topography of Fig. 4 has to be made taking into account the different scales of both figures. On the other hand, as shown in the figure, the form profile of the test part is conserved. It denotes that the laser-polishing process does not affect to form deviations of the parts but only to surface roughness values. Although resulting surfaces are much smoother than initial one, results are not good enough to consider for

ARTICLE IN PRESS 2044

A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

injection moulding applications. Some previous laserpolishing tests on DIN 1.2344 tool steel show final roughness below 0.8 mm Ra. It can be concluded that the results for SLS do not achieve these values due to the particular material composition. As it has been pointed above , LaserForm ST-100r is a mixture of stainless steel and bronze. The different melting points of these two materials can involve the melting of the bronze whereas the stainless steel is not affected. Therefore, if stainless steel matrix is melted, a deeper bronze layer is melted. Moreover, the best results were for more than 4000 J cm2 energy density, much higher than the used for the DIN 1.2344 tool steel tests. In order to evaluate the influence of all the parameters in the surface roughness, a design of experiments (DoE) was completed. The results of the DoE show the existence of an optimum roughness reduction for a given energy density.

The considered input parameters have been the laser beam power, the feed and focal offset distance (see Table 1). Therefore, a three level and three factor DoE has been considered and the experiment parameters and results have been introduced on the Design-Expertr 5 software. In this way, an ANOVA result on a quadratic model, which shows the best fit with the experimental data. The resultant response surface for the adjusted model is shown in Fig. 5, where a maximum roughness reduction can be observed for an optimum feed rate, focal distance and laser power combination. In the response surface, the optimum laserpolishing parameters can be calculated for different criteria. Then, optimum parameters are shown in Table 2, considering maximum roughness reduction and maximum feed rate criteria. In order to validate the DoE estimations, a new laserpolishing test set was compared with DoE predictions. In

Fig. 3. Laser-polishing experiments results.

Fig. 5. Response surface for LaserForm ST-100r tests for a focal offset distance of 30 mm.

Fig. 4. Left: Original topography of the SLS test part. Right: final topography for a laser-polishing test for LaserForm ST-100r.

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

Table 3, measured and estimated values for roughness reduction are presented. Experimental results shown to be in reasonable agreement with predicted results. It is important to take into account that experimental tests shown in Table 3 were not used to feed the DoE. In Fig. 6 the validation is presented graphically. 3.2. Planar surface polishing tests Once the optimum parameters for linear polishing tests were found, horizontal surface polishing tests were performed. These tests were carried out by overlapping line tests on the surface. A new parameter, denominated overlap index (Oi), is introduced in this case. The overlap index measure the relative distance that a line overlaps with the last one. This parameter depends on two factors: the stepover distance and the width of the lines, which is equal to the beam diameter. A 100% overlap index indicates that the laser beam have polished the same previous one. If overlap index is 0% indicates that there is no any overlap between two consequent passes. In Fig. 7 a scheme of surface laser-polishing strategy and the calculation of overlap index are shown. Table 2 Optimum parameters from the design of experiments with LaserForm ST100r Laser power (W) Laser feed rate Vf (mm min1) Distance (mm) Resulting energy density (J cm2) Roughness reduction (%)

600 1482.94 20 4.580 86.8

2045

Overlap between laser paths have to be enough to cover the gap between two lines, but too high values can lead on an excessive heat input on the same area, which can affect the final part. Thus, different overlap indexes have been tested, obtaining optimum results in the range from 15% to 30%. Taking into account that the beam diameter is about 0.8–2 mm, depending on the focal offset distance, the stepover value ranges between 0.7 and 1.5 mm. Experimental tests were carried out by executing a CNC program fully automatically. The test part was a horizontal plane of a LaserForm ST-100r test part. The planar polishing tests have been carried out with the optimum energy density calculated in linear polishing tests. The distance was increased to 40 mm to cover a larger area in each line. In this way, laser power and speed were fixed at 1100 W and 1100 mm min1 to maintain the optimum energy density. The tests included different overlapping indexes concluding that the best results were obtained for 25% with a reduction of 68.2% of roughness. The topographies before and after the polishing test are shown in Fig. 8, where roughness reduction can be also observed (noticing the different measurement scales). The measured mean roughness in the polished area was below 2.5 mm Ra and most of the areas present values below 1 mm. On the other hand, the measured topography presents an isotropic roughness, with similar roughness values in all directions, as it can be observed on the resulting topography, shown in Fig. 8. It is a well-known effect that the isotropic roughness improves the fatigue life of the parts since it avoids the nucleation of cracks.

Table 3 Tests parameters and measured and DoE estimated roughness reduction Test no.

Power (W)

Feed (mm min1)

Focal offset (mm)

Laser beam diameter (mm)

Energy density (J cm2)

Measured % reduction

DOE % estimated reduction

Difference (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

1200 1200 1200 1200 1200 800 1000 800 600 850 850 900 950 750 700 650 600 600 550

1400 1400 1800 1600 2000 1000 1000 800 800 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000

27 27 27 27 27 27 27 27 35 35 35 35 35 35 35 35 35 35 35

0.85 0.85 0.85 0.85 0.85 0.85 0.85 0.85 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2

6050.42 6050.42 4705.88 5294.12 4235.29 5647.06 7058.82 7058.82 3750.00 4250.00 4250.00 4500.00 4750.00 3750.00 3500.00 3250.00 3000.00 3000.00 2750.00

79.06 75.85 81.62 66.45 82.05 74.57 76.71 79.49 86.11 83.33 86.54 89.74 89.32 76.71 79.70 77.14 83.33 85.04 83.33

75.25 75.25 76.13 76.38 74.48 77.49 74.77 72.21 71.71 77.11 77.11 76.53 75.75 77.66 77.64 77.41 76.97 76.97 76.34

3.81 0.60 5.50 9.93 7.57 2.92 1.94 7.27 14.40 6.22 9.43 13.21 13.56 0.95 2.07 0.27 6.36 8.07 7.00

ARTICLE IN PRESS 2046

A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

Fig. 6. Measured roughness reduction vs. DoE estimated values.

Fig. 7. Surface laser-polishing strategy.

Fig. 8. Left: Initial surface topography. Right: test result for one laser-polished area.

3.3. Three dimensional line polishing tests Finally, in order to check the capabilities of the process for more complex geometries, a set of 3D laser-polishing

tests were performed. A specific test part was designed to develop these 3D tests. Thus, a test part (see Fig. 9) with three different slopes (151, 301 and 451) was build-up by SLS.

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

Results show similar roughness reductions to the horizontal line tests. Final measured mean roughness is between 1.25 and 2.5 mm Ra. The same parameters obtained in linear tests were tested and the results show that optimum results were achieved with the same energy density. It was also checked that the part shape profile does not change in any case. In Fig. 10, the initial roughness and a laser polishing resulting profile is shown. It can be observed that there is only a minimum variation in the part edges, whereas the roughness profile is much smoother than the initial one. The maximum measured deviation on the edges is 25 mm, while the surface reduction is 74.2%, similar to that obtained in horizontal tests.

3.4. Results discussion Experimental results for laser polishing show that the mean roughness decreases for all the tested cases. Experimental results show that the most important parameter is the energy density. On the other hand, the final surface

Fig. 9. 3D test part built up by SLS in LaserForm ST-100r.

2047

roughness depends on the initial surface roughness, thus, the surface roughness reduction presents some dispersion due to different surface topographies of the tested surfaces. Tests show reductions upto 76% with no form profile deviations. These reductions show that the process is able to improve the surface roughness considerably. However, the minimum measured mean roughness was 1.49 mm Ra. This value cannot be considered as a polished surface and final hand polishing should be done to reduce this value below 1 mm. The results can be explained at the view of the material composition. The necessity to melt two different materials with different melting points, forces to increase the energy density. Thus, the melted layer is deeper and the final roughness reduction is lower than the expected. This hypothesis is coherent with the presence of the SOM regime. As far as the form deviations, experimental results show only significant deviations in the edges of the test parts. This effect can be minimized by changing the strategies of laser-polishing paths or increasing the speed just in the edges. For the rest of the cases it has been observed that surfaces conserved the form deviations of the original part, reason why it is possible to ensure that the macrogeometric deviations are not affected by the process. Despite slope variations or small differences of the distance between the laser and the surface, the process results are similar to planar tests. In addition, for 3D polishing case, the trajectories were programed by a CAM system, so it can be considered that the process could be applied fully automatically in a complex surface. The maximum tested inclination angle has been 451, for higher inclinations the laser beam energy may be lost. In this case, two solutions can be implemented. First, a five-axis system which ensures the perpendicular irradiation of the laser beam. Second possibility is, to implement the laserpolishing technique into the SLS machine; thus, a laser-polishing procedure for each layer could be carried

Fig. 10. 3D polishing results in LaserForm ST-100r.

ARTICLE IN PRESS 2048

A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

out, considering that the part surface can be irradiated when the consolidation of each layer finishes. Finally, although obtained roughness values are higher than 1 mm Ra, the final roughness is much more uniform and much less rough than the initial one. This fact allows reducing the manual polishing time drastically and even, for those surfaces with low surface roughness requirements could be considered as a finished product. 4. Integrity of the laser polished surfaces One of the most important aspects to be controlled is the influence of the process in the polished surface properties. Some polishing techniques induce high residual stress which can lead to cracking and fatigue failure. On the other hand, laser processes involve surface melting and resolidification. Therefore, some heat affected zone (HAZ) will be present and must be analyzed in order to control the material behaviour. Although there are more potential applications, laser polishing of sintered parts has been focused basically for small plastic injection moulds or inserts. In these parts, the most common failure cause is thermal fatigue and abrasion. Thermal fatigue can be reduced with high thermal conductivity materials, which evacuate heat rapidly and maintain mould temperature relatively constant. In this way, bronze thermal conductivity is much higher than tool steel, improving the thermal fatigue behaviour of the mould or mould inserts. On the other hand, harder mould materials, reduce abrasion wear and achieve higher production volumes. In this way, LaserForm ST-100r presents a mixed structure combined with

high-porosity areas, which results in 140–160 HV hardness, while the hardness of a standard tool steel ranges from 320 HV. Thus, Laserform ST-100r is a relatively soft mould material and high porosity reduce considerably the mechanical properties of the material. In order to study the variation of mechanical properties and the influence of laser-polishing process on the parts, a complete metallurgical study has been done. The analysis results show that three different areas can be distinguished in the laser polished zone. First, denominated as Area 1, the initial substrate material, which is a stainless steel bounded particles matrix with bronze infiltration. Despite the bronze infiltration, some porosity can be observed in this area due to the sintering process. This porosity affects significantly to mechanical properties of the material. The second area (Area 2) is located surrounding the polished surface. Its composition is almost 100% AISI 314 stainless steel, with no bronze presence. This area is harder than the previous one because there is only stainless steel and, due to the re-solidification process, the material is much more homogeneous than the sintered one. Finally, the third area or Area 3 is the polished area and is located on the part surface. It is composed of a mixture of stainless steel and bronze, with a higher proportion of bronze. The high proportion of bronze is due to an exudation of this alloy from the stainless steel matrix. The resulting material is much more homogeneous than the substrate material and no porosity is appreciated. On the other hand, the polished layer material is harder than the substrate, although it is not as hard as the stainless steel layer. This effect will improve wear resistance of the surface. In Fig. 11 two different laser-polishing test sections

Fig. 11. LaserForm ST-100r laser polished line sections for two tests.

ARTICLE IN PRESS A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

are shown. In this figure, the three mentioned areas can be distinguished in both sections, although the test conditions (showed in Table 4) are different for each test. In Fig. 11, the Vickers microhardness measurements of the different areas are showed too. The initial hardness of the part is 140–155 HV, depending on the indentation location and the proportion of bronze and stainless steel of the measured point. In Area 2, the hardness increases up to 340 HV, this is a common value of AISI H314 stainless steel hardness value. Finally, in Area 3, the measured hardness is between 175 and 190 HV. Thus, the polished surface present a harder and more homogenous structure

Table 4 Laser-polishing test parameters for LaserForm ST-100r Test conditions

Test A

Test B

Laser power (W) Laser feed rate Vf (mm min1) Focal offset distance (mm) Shielding gas

1100 800 40 Argon

800 1100 30 Argon

2049

that should improve the mechanical behaviour for plastic injection moulding. A most detailed view of the Area 3 (Fig. 12), shows a combination of bronze and stainless steel much more homogeneous than the original part structure (Area 1). The structure is formed by a bronze matrix with stainless steel spheres. The stainless steel, due to the higher molten temperature, solidifies while bronze is still melted. Thus, stainless steel solidifies forming spheres. In order to ensure the composition of each area, an EDX analysis has been used. The analysis for Areas 1–3 are shown in Fig. 12, as well as the associated material for each analysis. In the results of Area 1, copper and tin peaks, corresponding to the infiltrated bronze, can be observed. On the other hand, iron, chromium and titanium peaks, which correspond to stainless steel, are detected too. However, for Area 2, iron, titanium and chromium peaks are much more remarked while copper and tin peaks indicate that there is not almost bronze in this area. The Area 3 analysis show copper, tin, iron and chromium peaks, corresponding with a mixture of bronze and stainless steel.

Fig. 12. Details of Area 3 (stainless steel spheres in a bronze matrix) and EDX analysis of Areas 1–3.

ARTICLE IN PRESS 2050

A. Lamikiz et al. / International Journal of Machine Tools & Manufacture 47 (2007) 2040–2050

5. Conclusions

Acknowledgments

In the work presented here, a laser-polishing process for metallic sintered parts by SLS has been described. The results show clear roughness reductions for a commercial material denominated LaserForm ST-100r which consists on sintered AISI 420 stainless steel infiltrated with bronze. The measured reductions are up to 80% reductions in Ra parameter. That means a final roughness of 1.2–1.3 mm Ra whereas the initial surface roughness was in the range of 7.5 mm Ra. Different surface geometries have been tested, including inclined planes. It has been observed that, regarding to the studied parameters, there is a range of energy density levels in which optimum results are obtained. If the laser power increases or the laser head operates more close to the surface, energy density grows and the process moves away the optimum. Therefore, energy density is the most important parameter of the process. Thus, once evaluated the optimum energy density, the optimal parameters are those that optimize feed and spot diameter, in order to obtain the maximum productivity. The experimental results also conclude that the rougher the original surface, the more effective the process is. Linear polishing tests have been carried out, in order to validate the process and to find the optimum parameters. This study has been carried out by using a DoE in addition to the experimental procedure. Once the optimum parameters have been found, polishing of planar surfaces has been tested by overlapping linear tests. In this case, a new parameter denominated overlapping index is introduced. Experimental results show that optimum overlapping index is about 15–30%. Finally, a 3D test part, shaped by three inclined planes has been designed and tested. Polishing tests have been programmed using a CAM system and the experimental results show that the process can be applied for any geometry, while the energy density is maintained. Therefore, this process can be applied to small-medium size moulds by overlapping polishing lines using a commercial CAM system to program the laser-polishing paths. Metallurgical analyses show that the heat affected zones do not present cracks or porosity. It can be concluded that laser affected areas present a more homogeneous composition than the initial ones. On the other hand, resulting surfaces are harder and more homogeneous than the initial sintered material. Finally, it can be concluded that the laser-polishing process only involves a microscopic layer of the treated area, it means that only surface roughness is affected while the form errors or the original macrogeometry is not affected.

This work has been developed in the X-MAPAL project (reference PTR95-0988) subsidized by the ministry of education and science of Spain. In these projects also take part the technological centres Robotiker and AIJU and the companies Maier S. Coop., Batz S. Coop. and Tibi S.L. Special thanks to Dionisio del Pozo, Jesus Maria Lo´pez and, in general all the Inmotion team of Robotiker Technological Centre.

References [1] H. Golnabi, M.H. Mahdieh, Trend of laser research developments in global level, Optics and Laser Technology 38 (2006) 122–131. [2] G. Ryk, I. Etsion, Testing piston rings with partial laser surface texturing for friction reduction, Wear (2006) in-press. [3] L. Hao, J. Lawrence, Effects of Nd: YAG laser treatment on the wettability characteristics of a zirconia-based bioceramic, Optics and Lasers in Engineering 44 (2006) 803–814. [4] H- Y. Wang, D.L. Bourell, J.J. Beaman, Laser-polishing of silica slotted rods, Materials Science and Technology 19 (2003) 382–387. [5] Erdetnir, et al., Durability and tribological performance of smooth diamond films produced by Ar-C60 microwave plasmas and by laserpolishing, Surface and Coatings Technology 94–95 (1997) 537–542. [6] G.A. Shafeev, S.M. Pimenov, E.N. Loubnin, Laser-assisted selective metallisation of diamonds by electroless Ni and Cu plating, Applied Surface Science 86 (1995) 392–397. [7] S. Hogmark, et al., Direct current bias applied to hot flame diamond deposition produces smooth low friction coatings, Wear 200 (1–2) (1996) 225–232. [8] S. Gloor, et al., UV laser-polishing of thick diamond films for IR windows, Applied Surface Science 138–139 (1999) 135–139. [9] J.A. Ramos-Grez, D.L. Bourell, Reducing surface roughness of metallic freeform-fabricated parts using non-tactile finishing methods, International Journal of Materials and Product Technology 21 (2004) 297–316. [10] T. Grimm, Rapid prototyping, Edited by Society of Manufacturing Engineers (2004) ISBN 0-87263-697-6. [11] W. Liu, J.N. DuPont, Fabrication of functionally graded TiC/Ti composites by laser engineered net shaping, Scripta Materialia 48 (9) (2003) 1337–1342. [12] B. O´Donnchadha, A. Tansey, A note on rapid metal composite tooling by selective laser sintering, Journal of Materials Processing Technology 153–154 (2004) 28–34. [13] D. King, T. Tansey, Rapid tooling: selective laser sintering injection tooling, Journal of Materials Processing Technology 132 (2003) 42–48. [14] T. Wohlers, Wohlers Report 2006 vol. 12 (2006) Wohlers Associates, ISBN 0-9754429-2-9. [15] Lu, J. Fuh, Y.S. Wong, Laser Induced Materials and Process for Rapid Prototyping, Kluwer Academic, Dorderecht, The Netherlands, 2001. [16] Y. Uno, et al., High-efficiency finishing process for metal mold by large-area electron beam irradiation, Precision Engineering 29 (2005) 449–455.