۱٫ Introduction

۲٫ Methods

۳٫ Results and Discussion

۴٫ Conclusions and Outlook

Acknowledgments

References

In casting, molding, or additive manufacturing processes, there are some typical issues that can change the geometry of a part and cause porosity or other defects. With the aid of X-ray computed tomography (CT), internal discontinuities and geometry deviations can be accurately detected and visualized. However, the question remains in how far a given defect affects mechanical failure. We aim at bridging this gap by structural mechanics simulations based on CT images. In this study, we describe a method to predict the tensile strength and the location of crack initiation from the simulated stress distributions on the basis of local stress concentrations. We validate the method for tensile rods and real-life aeronautic parts which were additively manufactured from an AlSi10Mg aluminum alloy. Thirty-six specimens were manufactured in total, where different porosity patterns were deliberately inserted. The specimens were CT-scanned in high resolution. Structural mechanics simulations were carried out on basis of the CT images. An immersed-boundary finite elements code is used. The generation of a conforming simulation mesh is not required, making the code suitable especially for complex geometries like porous objects. The same test specimens were subjected to destructive physical tensile tests. We show that there is a very good correlation between the predicted and measured tensile strengths, and that the location of the first crack occurrence can be forecasted accurately.

Introduction

Structural parts almost inevitably differ from their ideal geometry due to the manufacturing process. In casting and injection moulding, gas or shrinkage pores are common kinds of defects. In additive manufacturing by sintering or melting, heat variations can cause imperfect bonding of the material granules. Industrial X-ray computed tomography (CT) allows for the non-destructive detection and visualization of many kinds of superficial or internal defects. With today’s CT scanners reaching spatial resolutions down to the micrometer scale, a CT image constitutes an accurate three dimensional digital representation of the scanned part. The spatial distribution and the morphological features of the porosity can thus be assessed in detail [1, 2]. However, the mere presence of porosity does not necessarily render a part unusable. In fact even highly porous parts can still be good for a specific purpose. In order to minimize false rejections and thus production costs it is desirable to assess the “effect of the defect”, i.e., to what degree a geometric defect affects the mechanical function. Work has been done to relate the detected porosity to fatigue behavior in aluminum castings [3] and laser-melted parts [4], sometimes combining the pore geometry with heuristic modeling of the pore physics [5, 6] or with analyzing stress concentrations at individual pores using finite element analysis [7].