Rapid prototyping technologies, such as Selective Laser Sintering (SLS) involve the production of accurate parts directly from CAD models with little human intervention. With the growing trend toward the use of lightweight materials such as aluminium rapid prototyping technologies have developed in recent years so that the direct production of aluminium components from computer models is now possible.
The infiltration of green SLS aluminium components enables the production of large dimensionally accurate parts. The thermal process first involves heating the green SLS aluminium-nylon parts under a vacuum. After 2 1/2 hours, the atmosphere is changed to nitrogen at atmospheric pressure while the part is heated to 540oC. During this time the nylon melts and is burnt out of the SLS part. Once the part bas ruched 540oC it is held at this for 12 hours, during which time the aluminium nitride skeleton is formed. The atmosphere is then changed to argon and the part is heated to 700oC. It is held at this temperature for about 2 hrs so that the block of 6061 alloy infiltrant melts and infiltrates the part. After this time the part is furnace cooled to approximately 500oC where the atmosphere is changed back to nitrogen and the part is left to furnace cool to room temperature.
The strength of unsupported selective laser sintered aluminium parts during thermal processing is at present unknown. This is important as there is currently no information available in regard to the maximum span that can be self-supported and thus the geometrical limitations of parts. The principle aim of this thesis was to determine the strength of unsupported selective laser sintered aluminium components during thermal processing.
The results of a full thermal cycle, in which the parts were allowed to completely infiltrate determined that green SLS parts with unsupported spans of 60m or less were able to withstand the thermal cycle and infiltrate fully. An SLS aluminium part with an unsupported span of 65mm or more cannot be fully infiltrated and will collapse and fail during the thermal cycle. The spraying of green SLS parts during preparation with a boron nitride sealer does not influence the failure of SLS parts during thermal processing.
The failure of the SLS aluminium parts corresponds to the melting of the nylon. Failure of these SLS parts occurred in the temperature range from 171oC- 177oC. The melting point of nylon is 135oC which is slightly lower than the failure temperature of the SLS component.
Investigation into the effect that different geometries have co the ability of SLS pam to withstand the thermal cycle used an unsupported span of 65mm. The cross-Sectional geometries investigated included a solid square, I-Beam, hollow box and arc, while two other SLS parts were tested that arched length-ways. Results from this investigation have indicated that geometry can affect the failure of SLS parts during the thermal cycle. The outcome that unsupported spans of 65mm and varying cross-sectional geometries did not all have the same result has indicated that geometry may influence the failure of SLS parts during the thermal cycle.
The determination of the stress that the parts were required to withstand was performed with a very simplified model in which a number of assumptions were made. It was assumed that the parts could be modelled as simply supported structures and that there was no influence on stress due to thermal constraints. This is an invalid simplification. The SLS aluminium samples are not simply supported structures in that they are fixed at both ends, which in tum will generate thermal stresses. The stresses calculated due to the self-load of the SLS parts are in the range of I kPa to 6 kPa. Due to the limited knowledge of the properties of SLS parts during thermal processing it was not possible to calculate the stress generated by thenna1 constraints as the coefficient of thermal expansion and elastic modulus of the material at the time of failure is unknown.
Further investigation is suggested in which the SLS aluminium parts are modelled as simply supported beams. This will enable correct values to be obtained for the stress the parts withstand due to self-loading.