Pore analysis and mechanical performance of selective laser sintered objects

In this work, systematic studies were carried out on SLS (selective laser sintering) printed samples, with two different geometries, standard test samples dumb-bells (dog bones) and tubes (phi 30 mm and 150 mm long), consisting of two different materials, viz. PA12 (polyamide) with and without the addition of carbon fibres (CFs). These samples were tested according to their respective ISO standards. The standard test samples exhibited relatively small differences with regards to printing directions when PA12 was used alone. Their tensile strengths (sigma(m)) were approx. 75%-80% of the injection-moulded sample. The addition of carbon fibres significantly enhanced the tensile strengths, namely 50% greater for the vertically printed test sample and more than 100% greater for the horizontally printed samples, compared to the respective objects consisting of PA12 alone. The strong difference in printing directions can be attributed to the orientation of the carbon fibres. Mechanical tests on the SLS printed tubes confirmed the trends that were found in the standard test samples. Porosity and pore structure inside the SLS printed tubes were studied by combining optical microscopy and X-ray micro-tomography with image analysis. It was found that porosity was a general phenomenon inside the SLS printed samples. Nevertheless, there were significant differences in porosity, which probably depended on the properties of the materials used, both with and without carbon fibres, thus causing significant differences in light absorption and heat conductivity. The printed samples made of PA12 alone possessed quite a high level of porosity (4.7%), of which the size of the biggest pore was hundreds of microns. The twenty biggest pores with an average size of 75*10(4) mu m(3) accounted for 43% of the total porosity. However, the porosity of the printed samples made from PA12 + CF was only 0.68%, with the biggest pore being only tens of microns. The corresponding average pore size of the 20 biggest pores was 72*10(3) mu m(3), which was one order of magnitude smaller than the printed samples made from PA12 alone. Pores inside the SLS printed samples were probably responsible for a spread in the mechanical properties measured, e.g. tensile strengths, tensile (Young's) modulus, strain at break, etc. The ratios of their standard deviations to their corresponding mean values in the standard test samples could probably be used as an indicator of porosity, i.e. the bigger the ratio, the higher the porosity.