Effects of build orientation on tensile strength for stereolithography-manufactured ASTM D-638 type I specimens

Abstract

A statistical design of experiments (DOE) approach was used to determine if specific build orientation parameters impacted mechanical strength of stereolithography (SL) fabricated parts. A single platform (25.4 × 25.4-cm cross-section) on the 3D Systems Viper si2 SL machine was designed to hold 18 ASTM D-638 type I samples manufactured in different orientations. The DOE tested three factors: axis, layout, and position. Samples were fabricated parallel with the x-axis or y-axis, or 45° to both axes (called axes 1, 2, and 3, respectively). For each axis, samples were fabricated either flat or on an edge relative to the x–y plane (called layouts 1 and 2, respectively). Three samples were manufactured for each layout and axis combination, and the samples were labeled as positions 1, 2, or 3 depending on the distance from the center of the platform with position 1 being the closest to the center. The results from the statistical analyses showed that axis and position had no significant effect on ultimate tensile stress (UTS) or modulus of elasticity in tension (E). However, layout (or whether a sample was built flat or on an edge) was shown to have a statistically significant effect on UTS and E (at a 95% level of confidence). The differences between average UTS and E for each of the samples built flat and on an edge were ~3.53% (43.2 versus 44.8 MPa) and ~4.59% (763.9 versus 800.7 MPa), respectively. Because of the relatively small differences in means for UTS and E, the statistical differences between layout most likely would not have been identified without performing the multifactor analysis of variance. Furthermore, layout was the only factor that tested different orientations of build layers (or layer-to-layer interfaces) with respect to the sample part, and thus, it appears that the orientation of the build layer (layer-to-layer interfaces) with respect to the fabricated part has a significant effect on the resulting mechanical properties. This study represents one of many to follow that is using statistical analyses to identify and classify important fabrication parameters on mechanical properties for layer-manufactured parts. Although SL is the focus of this work, the techniques developed and presented here can be applied to any layered manufacturing technology producing any ASTM-type specimen with any particular material.