CAD for optimal scaling of the 3D model, to compensate the SLS post-processing errors.

Balc, Nicolae ; Berce, Petru ; Pacurar, Razvan 等

1. INTRODUCTION

It is well known the Selective Laser Sintering (SLS) process and technology, which can be used to produce complex parts in different type of materials. Besides the advantages of the SLS process, as compare to the classical manufacturing technologies, there are some disadvantages, as well.

One of the major disadvantages of all the Rapid Prototyping (RP) technologies is the accuracy to produce metal parts, which is bellow the accuracy provided by some classical technologies, such as grinding.

The use of the RP technologies has become more and more frequent in the manufacturing industry, because they offer the quick solution to produce prototypes of the very complex shapes, without being necessary any special tools. Anyhow, there is a lot of work to be done to increase the accuracy of the metal parts produced by RP.

The latest advances in the recent developed processes, such as selective laser melting and electron beam melting, offer the possibility to quickly produce a complex fully dense metal part, but research is steel necessary to be done, in order to improve the accuracy and the surface roughness of the metal parts made using these technologies.

This paper presents a new CAD method to find the necessary scaling factors, in order to compensate the SLS manufacturing errors.

New software has been developed by the authors, in order to compensate the shrinkage which might occur during different post processing methods of the SLS metal parts.

2. SLS ERRORS AND EXISTING METHODS TO COMPENSATE SOME ERRORS

It is very complex the SLS process and there are different types of errors which might be involved [Yan, et al., 1998] and [Tang et al., 2003], caused by: the scanning system, material contractions, layers scanning, etc.

On top of that, there are random errors and systematic errors in connection to a particular SLS machine. The Sinterstation 2000--SLS machine drives the laser beam using the two mirrors galvanometric system, illustrated in figure 1.

F-Theta lens is used to focus the laser beam onto a horizontal working plane. The software package provided by the SLS machine manufacturer does not have the possibility to compensate the distortions caused by the scanning system.

[FIGURE 1 OMITTED]

There is a calibration procedure for the SLS machine, in order to estimate some compensation factors, just for the SLS process onto the Sinterstation 2000 machine, but not for the post processing. For plastic parts made by SLS, the accuracy improvements were discussed by Raghunath [2007] and Yang [2002]. For the PolyJet TM process the scale factors were optimized [Brajlih et al., 2006].

3. NEW METHOD TO CALCULATE THE X, Y, Z COMPENSATION FACTORS

[FIGURE 2 OMITTED]

New original software package has been developed by the authors, to calculate better scaling factors for the SLS process, which take into account the post processing deformations.

The test part, illustrated in figure 2 was designed in order to test and validate the new software, called FOS (Optimal Scaling Factors). The FOS software calculates the scaling factors by taking into account both the systematic errors of the Sinterstation machine measured onto the green parts and the post processing shrinkage. During post processing the SLS metal parts, 3D deformations occur, depending of the type and temperature of the post processing cycle.

It was used the Finite Element Analysis method, to estimate the deformations during post processing the sintered metal parts. The FEA has been done in two stages. First stage consisted in a transient thermal analysis.

[FIGURE 3 OMITTED]

For the undertaken case study, the warming up time was 12 hours (up to 1070[degrees] C), the maintaining period was 3 hours and the cooling period was 12 hours.

The purpose of this stage was to find out the temperature distribution in the part, at different moments, during the heating cycle illustrated in figure 3.

The second stage consisted in a thermoelastic analysis. Two cases have been studied:

--Thermoplastic expansion during the heating process from room temperature to 1070[degrees]C. In this case, the thermal expansion coefficient has been set to a higher value, due to the fact the sintered material is in a powder state.

--Thermal contraction during the cooling process in the oven, from 1070[degrees]C to the room temperature. In this case, the thermal expansion coefficient has been set to a lower value, due to the fact the part is already in a compact state.

As conclusion, the effects of the infiltration process can be modeled by modifying the value of the thermal expansion coefficient ([alpha]--coefficient). The other values for the parameters used within the FEA are presented in figure 4.

[FIGURE 4 OMITTED]

The thermoelastic analysis has been performed imposing the minimum amount of kinematical restraints. More precisely, just the translation degrees of freedom corresponding to three perpendicular surfaces were locked: the basis and two lateral surfaces of the part. In this way, the thermal deformation could take place along all three axes.

[FIGURE 5 OMITTED]

For the 100 mm high part, the thermoplastic expansion was 1.307 mm and the thermal contraction was 1.158 mm, so the errors on Z axis would have been 0.149 mm.

All deformations on x,y,z axis are presented in table 1.

4. RESULTS AND CONCLUSIONS

The new software developed in C++ was tested and validated onto the test parts and some case studies of producing complex metal parts by SLS. The FEA was used to estimate the 3D deformations during post processing by infiltrating the SLS steel parts with epoxy resins. Both experimental data measured onto green parts and post processing data estimated using FEA were used as input data for the new FOS software, which calculates better scaling factors to alter the dimensions of the 3D virtual model, in order to obtain good dimensions of the SLS part, as close as possible to the theoretical dimensions.

5. REFERENCES

T. Brajlih et al., Optimizing scale factors of the PolyJet TM rapid prototyping procedure by genetic programming. J. Achiev. Mater. Manuf. Eng., May-Jun. 2006, vol. 16, iss. 1/2, str. 101-106

N. Raghunath, Pulak M. Pandey, Improving Accuracy Through Shrinkage Modeling by using Taguchi method in Selective Laser Sintering, International Journal of machine Tools & Manufacture 47 (2007), 985-995.

Y. Tang et al, Accuracy Analysis and Improvement for Direct Laser Sintering, International Journal of machine Tools & Manufacture 43 (2003), 985-995.

M. Yan, et al., Analysis of machine accuracy for rapid prototyping of quality components, Proceedings of SPIE, The International Society for Optical Engineering, V 3517 (1998), 91-101.

H.J. Yang et al., A study on shrinkage compensation of the SLS-process by using Taguchi method, International Journal of machine Tools & Manufacture 42 (2002), 1203-1212.

Tab. 1. Estimated deformations, using the FEA

Deformation X Y Z

Warming up 1,451 1,451 1,307
Cooling down 1,285 1,285 1,158
Residual deformations 0,166 0,166 0,149