| Περίληψη: | 3D printing is an ever-growing manufacturing process, gaining a lot of popularity in the last years, with potential for application in a multitude of fields, particularly in the medical field on the subject of prosthetics and orthotics. Its ease of accessibility and both time and cost effectiveness make it an attractive alternative to current manufacturing processes. Due to its recency, however, as well as insufficient documentation for the thermoplastic materials it commonly uses (compared to more conventional materials like metals), there is an information gap, regarding how the plethora of pre-process configurations and parameters can potentially affect the mechanical properties of the 3D-printed parts. The goal of this thesis is to fill this gap, as well as provide a guideline for future studies regarding this subject.
To achieve this goal, a number of specimen was printed, following the guidelines from ASTM D638 standard, which would be tested for their tensile properties. The standard procedure is to input the CAD file of the specimen into the slicer program of the 3D printer, and after some configurations, an STL file is exported, which can be read by the 3D printer, and the part is produced. These configurations have to do with the 3D printing procedure, from nozzle temperatures to specific instructions on how the printer will fill the volume of the part. The latter is very important, as it creates an internal geometry, which greatly affects the mechanical behavior and load response of the 3D printed part, and this is what is tested in this thesis. To do that, control specimen were created, with specific values for four printing parameters that were to be tested: Infill Density, Perimeter Thickness, Shell Thickness and Layer Height. The rest of the specimen had the value of one parameter changed, while the rest remained as they were, in order to test the effect of each parameter individually. For the creation of the specimen, four common 3D-printing materials (PLA, PETG, CPE and ABS) were used, to determine the potential effect that the parameters had on each one. The specimen were then tested on a tensile testing apparatus, to determine their tensile properties, and more specifically, Modulus of Elasticity and Tensile Strength, and compared to determine their differences, taking into consideration the printing time and the cost of their manufacturing. This indicated the effectiveness of each parameter and whether applying that change would be worthwhile, which is especially useful information in large scale operations.
The tensile tests showed that perimeter thickness had the greatest effect on the mechanical properties on the specimen, regardless of the material used, and that by increasing it, both modulus of elasticity and tensile strength would improve significantly, with minimal change in printing time and cost. Shell thickness came second on this matter, also having a noticeable effect on the mechanical properties of the specimen. Increasing the infill density did increase the modulus of elasticity for the specimen, but had a small effect on the tensile strength. In addition, these improvements came with a substantial increase in both printing time and cost, making other adjustments, like the ones mentioned above, more attractive. Finally, in all cases, layer height had a small effect on both modulus of elasticity and tensile strength, compared to the other parameters, while increasing printing time drastically, making it a suboptimal choice, and thus not recommended for improving the mechanical properties of a part, unless required by its geometry.
To validate these experimental results, a Finite Element Analysis model of the tensile specimen was created, which was used in a simulation tensile test. At first, a CAD model of the 3D-printed part, which included the internal geometry, was created in Autodesk Fusion 360, and was then imported in ANSYS Mechanical Student ver., the software used for the simulations. Since ANSYS does not have material data for plastics, they were created from scratch for each material used. With the creation for the materials done, a mesh was then constructed for the specimen, which essentially breaks the model into Finite Elements, to be used in the analysis. Finally, boundary conditions and loads that replicated the tensile tests were configured, the simulation setup was ready. Same simulations were run multiple times with increasing number of FE used, to reach mesh convergence, and the converged data was used to extract the information required (tensile strength and modulus of elasticity), to be compared with the experimental results. Unfortunately, it should be kept in mind that factors like FE usage license limitations, as well as material data inaccuracies by manufacturers and difficulty in finding appropriate material models to express the mechanical behavior of plastics in ANSYS have lowered the quality of the simulation model and its results.
The simulated tensile tests’ results were overall close to their experimental counterparts, indicating that the model was accurate, but lacked precision. There were only a few exceptions, regarding the tensile strength of the specimen, where the model was not able to accurately predict the specimen’s tensile behavior. In addition, the simulation model was more accurate predicting the behavior of PETG and CPE than ABS and especially PLA, which showed larger differences between results. In the end, however, this model was able to predict up to a point, the mechanical properties of 3D-printed specimen, thus validating the experimental results.
Narrow time margins prevented the further improvement of the FEA model. Testing the simulation model without license limitations, as well as a more refined material model would be some basic directions for further research on the FEA part. As far as the experiments go, more parameters like printing orientation and infill patterns, as well as other mechanical tests, like fatigue and compression, would expand the scope of this study and are necessary, in order to have a clearer picture on 3D-printing as a whole.
|
|---|
