Research Summary Report of C01

Prof. Dr.-Ing. Stefan Kollmannsberger                                                                                                                                                 Project Leader, stefan.kollmannsberger@uni-weimar.de

Bauhaus-Universität Weimar (BUW),  Professur Data Engineering im Bauwesen (DEiB)

Digital models for Additive Manufacturing (AM) must consider many different geometric scales. The scales range from micrometers up to tens of meters for metal- or concrete-based processes and mutually influence each other. Project C01 aims to develop consistent geometric and computational descriptions for the relevant AM products on all these scales.

As-built structures naturally deviate from as-designed structures in geometry, topology, and material properties especially in AM. The consequences of such deviations upon the structural behavior are commonly termed the ‘effect of defect’. To this end, this project will design a methodology to digitally assess the effect of defect.

Summary

Understanding material failure is of significance in various fields, ranging from civil engineering, aerospace engineering mechanical engineering to biomedical engineering, and recently, additively manufactured (AM) 3D printed parts. Understanding the cause of fracture and failure of materials can help engineers make crucial decisions regarding the safety of structures. To this end, simulations of components are used to predict critical failure loads, also known as fracture load.

Project C01 employs advanced numerical techniques, including phase field models with spatially varying material properties, coupled with the finite cell method and multi-level hp-refinement [1]. Combined with efficient staggered solution schemes [2], these tools allow the precise simulation of crack propagation. Especially the finite cell method enables simulations on complex geometries, such as CT-Scans of experimental specimen, while mlhp-refinement enhances resolution in areas of interests like crack tips dynamically in the simulation. To validate the simulation model, experiments are carried out.

So far, the developed method has been applied, among others, to rock samples [3], [4] and human bones [4], [5], as shown in Fig. 1, and has given accurate results when comparing to experimental data.

Current state of research

Naturally, the next step is to expand the methodology to AM parts, specifically concrete. This introduces additional challenges due to the highly irregular and anisotropic material properties of concrete.

One way of obtaining these material properties is by CT scanning of as-built parts and matching the resulting Hounsfield Unit (HU), a greyscale value, voxel-wise to their respective material properties. While this mapping is well-established in medical engineering, it has not yet been expanded to mineral based construction materials.

To address this gap, an experimental series was designed and carried out in collaboration with project A01. Specimen produced using particle bed 3D printing (PB3DP) by Selective Cement Activation (SCA) were manufactured, CT scanned, and subsequently tested to evaluate their material and strength properties, see e.g., Fig. 2. Particular emphasis was placed on determining the Young’s Modulus E, as this parameter is expected to establish a correlation between the HU and the critical energy release rate (G_c) of additively manufactured specimens. The post-processing of the experimental results is currently underway.