Design of self-supported 3D printed parts for fused deposition modeling

Fabian Lischke, Purdue University

Abstract

One of the primary challenges faced in Additive Manufacturing (AM) is reducing the overall cost and printing time. A critical factor in cost and time reduction is post-processing of 3D printed (3DP) parts, which includes removing support structures. Support is needed to prevent the collapse of the part or certain areas under its own weight during the 3D printing process. Currently, the design of self-supported 3DP parts follows experimental trials. A trial and error process is needed to produce high quality parts by Fused Depositing Modeling (FDM). An example for a chamfer angle, is the common use of 45° angle in the AM process. Surfaces that are more flat show defects than inclined surfaces, and therefore a numerical model is needed. The model can predict the problematic areas at a print, reducing the experimental prints and providing a higher number of usable parts. Physical-based models have not been established due to the generally unknown properties of the material during the AM process. With simulations it is possible to simulate the part at different temperatures with a variety of other parameters that have influence on the behavior of the model. In this research, analytic calculations and physical tests are carried out to determine the material properties of the thermoplastic polymer Acrylonitrile - Butadiene - Styrene (ABS) for FDM at the time of extrusion. This means that the ABS is going to be extruded at 200°C to 245°C and is a viscus material during part construction. Using the results from the physical and analytical models, i.e., Timoshenko’s modified beam theory for micro structures, a numerical material model is established to simulate the filament deformation once it is deposited onto the part. Experiments were also used to find the threshold for different geometric specifications, which could then be applied to the numerical model to improve the accuracy of the simulation. The result of the nonlinear finite element analysis is compared to experiments to show the correlation between the prediction of deflection in simulation and the actual deflection measured in physical experiments. A case study was conducted using an application that optimizes topology of complex geometries. After modeling and simulating the optimized part, areas of defect and errors were determined in the simulation, then verified and measured with actual 3D prints.

Degree

M.S.M.E.

Advisors

Tovar, Purdue University.

Subject Area

Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS