TY - JOUR A1 - Powierza, Bartosz A1 - Gollwitzer, C. A1 - Wolgast, D. A1 - Staude, A. A1 - Bruno, Giovanni T1 - Fully experiment-based evaluation of few digital volume correlation techniques N2 - Digital Volume Correlation (DVC) is a powerful set of techniques used to compute the local shifts of 3D images obtained, for instance, in tomographic experiments. It is utilized to analyze the geometric changes of the investigated object as well as to correct the corresponding image misalignments for further analysis. It can therefore be used to evaluate the local density changes of the same regions of the inspected specimens, which might be shifted between measurements. In recent years, various approaches and corresponding pieces of software were introduced. Accuracies for the computed shift vectors of up to about 1‰of a single voxel size have been reported. These results, however, were based either on synthetic datasets or on an unrealistic setup. In this work, we propose two simple methods to evaluate the accuracy of DVC-techniques using more realistic input data and apply them to several DVC programs. We test these methods on three materials (tuff, sandstone, and concrete) that show different contrast and structural features. KW - DVC KW - Finite-element analysis KW - Image processing KW - Stress strain relations KW - Computed tomography PY - 2019 U6 - https://doi.org/10.1063/1.5099572 SN - 1089-7623 VL - 90 IS - 11 SP - 115105 PB - AIP Publishing AN - OPUS4-49671 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - JOUR A1 - Buljak, V. A1 - Oesch, Tyler A1 - Bruno, Giovanni T1 - Simulating fiber-reinforced concrete mechanical performance using CT-based fiber orientation data N2 - The main hindrance to realistic models of fiber-reinforced concrete (FRC) is the local materials property variation, which does not yet reliably allow simulations at the structural level. The idea presented in this paper makes use of an existing constitutive model, but resolves the problem of localized material variation through X-ray computed tomography (CT)-based pre-processing. First, a three-point bending test of a notched beam is considered, where pre-test fiber orientations are measured using CT. A numerical model is then built with the zone subjected to progressive damage, modeled using an orthotropic damage model. To each of the finite elements within this zone, a local coordinate system is assigned, with its longitudinal direction defined by local fiber orientations. Second, the parameters of the constitutive damage model are determined through inverse analysis using load-displacement data obtained from the test. These parameters are considered to clearly explain the material behavior for any arbitrary external action and fiber orientation, for the same geometrical properties and volumetric ratio of fibers. Third, the effectiveness of the resulting model is demonstrated using a second, “control” experiment. The results of the “control” experiment analyzed in this research compare well with the model results. The ultimate strength was predicted with an error of about 6%, while the work-of-load was predicted within 4%. It demonstrates the potential of this method for accurately predicting the mechanical performance of FRC components. KW - Fiber-reinforced concrete KW - X-ray computed tomography (CT) KW - Anisotropic fiber orientation KW - Inverse analysis PY - 2019 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:b43-474728 SN - 1996-1944 VL - 12 IS - 5 SP - 717, 1 EP - 16 PB - MDPI AN - OPUS4-47472 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -