2022
|
Sadeghpour, E.; Nonn, A. Data-driven models for structure-property prediction in additively manufactured steels Werkstoffsimulation Journal Article In: Computational Materials Science, vol. 215, pp. 111782, 2022, ISSN: 0927-0256. @article{SADEGHPOUR2022111782,
title = {Data-driven models for structure-property prediction in additively manufactured steels},
author = {E. Sadeghpour and A. Nonn},
url = {https://www.sciencedirect.com/science/article/pii/S0927025622004931},
doi = {https://doi.org/10.1016/j.commatsci.2022.111782},
issn = {0927-0256},
year = {2022},
date = {2022-09-15},
urldate = {2022-01-01},
journal = {Computational Materials Science},
volume = {215},
pages = {111782},
abstract = {Data-driven models are developed to predict the mechanical properties of polycrystalline materials. The case study is the prediction of the yield strength of a 3D-printed 316L steel from morphological and crystallographic features. Three different artificial intelligence models including feed-forward (FNN), convolution (CNN), and graph (GNN) neural networks are employed to train the data-driven models and are compared in terms of performance and computational requirements. The dataset required for training is generated by performing crystal plasticity finite element simulations. The FNN model has the smallest input size and takes in some statistical parameters describing the material microstructure, but its accuracy is relatively low. The CNN approach inputs voxel-based realizations of the microstructure and is able to give accurate estimations; however, its training process is time-consuming and computationally expensive. In the GNN approach, the polycrystalline material is represented by a graph whose nodes and lines represent the grains and adjacency between grains. It is observed that GNN yields a better performance compared to the other two approaches and has the capability of handling complex tasks.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Data-driven models are developed to predict the mechanical properties of polycrystalline materials. The case study is the prediction of the yield strength of a 3D-printed 316L steel from morphological and crystallographic features. Three different artificial intelligence models including feed-forward (FNN), convolution (CNN), and graph (GNN) neural networks are employed to train the data-driven models and are compared in terms of performance and computational requirements. The dataset required for training is generated by performing crystal plasticity finite element simulations. The FNN model has the smallest input size and takes in some statistical parameters describing the material microstructure, but its accuracy is relatively low. The CNN approach inputs voxel-based realizations of the microstructure and is able to give accurate estimations; however, its training process is time-consuming and computationally expensive. In the GNN approach, the polycrystalline material is represented by a graph whose nodes and lines represent the grains and adjacency between grains. It is observed that GNN yields a better performance compared to the other two approaches and has the capability of handling complex tasks. |
Wiesent, L.; Spear, A.; Nonn, A. Computational analysis of the effects of geometric irregularities on the interaction of an additively manufactured 316L stainless steel stent and a coronary artery Werkstoffsimulation Journal Article In: Journal of the Mechanical Behavior of Biomedical Materials, vol. Volume 125, 2022. @article{Wiesent2022,
title = {Computational analysis of the effects of geometric irregularities on the interaction of an additively manufactured 316L stainless steel stent and a coronary artery},
author = {L. Wiesent and A. Spear and A. Nonn },
url = {https://www.sciencedirect.com/science/article/pii/S1751616121005117?dgcid=author
},
doi = {https://doi.org/10.1016/j.jmbbm.2021.104878},
year = {2022},
date = {2022-01-01},
journal = {Journal of the Mechanical Behavior of Biomedical Materials},
volume = {Volume 125},
abstract = {Customized additively manufactured (laser powder bed fused (L-PBF)) stents could improve the treatment of complex lesions by enhancing stent-artery conformity. However, geometric irregularities inherent for L-PBF stents are expected to influence not only their mechanical behavior but also their interaction with the artery. In this study, the influence of geometrical irregularities on stent-artery interaction is evaluated within a numerical framework. Thus, computed arterial stresses induced by a reconstructed L-PBF stent model are compared to those induced by the intended stent model (also representing a stent geometry obtained from conventional manufacturing processes) and a modified CAD stent model that accounts for the increased strut thickness inherent for L-PBF stents. It was found that, similar to conventionally manufactured stents, arterial stresses are initially related to the basic stent design/topology, with the highest stresses occurring at the indentations of the stent struts. Compared to the stent CAD model, the L-PBF stent induces distinctly higher and more maximum volume stresses within the plaque and the arterial wall. In return, the modified CAD model overestimates the arterial stresses induced by the L-PBF stent due to its homogeneously increased strut thickness and thus its homogeneously increased geometric stiffness compared with the L-PBF stent. Therefore, the L-PBF-induced geometric irregularities must be explicitly considered when evaluating the L-PBF stent-induced stresses because the intended stent CAD model underestimates the arterial stresses, whereas the modified CAD model overestimates them. The arterial stresses induced by the L-PBF stent were still within the range of values reported for conventional stents in literature, suggesting that the use of L-PBF stents is conceivable in principle. However, because geometric irregularities, such as protruding features from the stent surface, could potentially damage the artery or lead to premature stent failure, further improvement of L-PBF stents is essential.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Customized additively manufactured (laser powder bed fused (L-PBF)) stents could improve the treatment of complex lesions by enhancing stent-artery conformity. However, geometric irregularities inherent for L-PBF stents are expected to influence not only their mechanical behavior but also their interaction with the artery. In this study, the influence of geometrical irregularities on stent-artery interaction is evaluated within a numerical framework. Thus, computed arterial stresses induced by a reconstructed L-PBF stent model are compared to those induced by the intended stent model (also representing a stent geometry obtained from conventional manufacturing processes) and a modified CAD stent model that accounts for the increased strut thickness inherent for L-PBF stents. It was found that, similar to conventionally manufactured stents, arterial stresses are initially related to the basic stent design/topology, with the highest stresses occurring at the indentations of the stent struts. Compared to the stent CAD model, the L-PBF stent induces distinctly higher and more maximum volume stresses within the plaque and the arterial wall. In return, the modified CAD model overestimates the arterial stresses induced by the L-PBF stent due to its homogeneously increased strut thickness and thus its homogeneously increased geometric stiffness compared with the L-PBF stent. Therefore, the L-PBF-induced geometric irregularities must be explicitly considered when evaluating the L-PBF stent-induced stresses because the intended stent CAD model underestimates the arterial stresses, whereas the modified CAD model overestimates them. The arterial stresses induced by the L-PBF stent were still within the range of values reported for conventional stents in literature, suggesting that the use of L-PBF stents is conceivable in principle. However, because geometric irregularities, such as protruding features from the stent surface, could potentially damage the artery or lead to premature stent failure, further improvement of L-PBF stents is essential. |
Trautmannsberger, R.; Marx, P.; Keim, V.; Paredes, M.; Nonn, A. Do simplified pressure decay and backfill models represent the loading scenario during the running ductile fracture scenario in gas transmitting onshore pipelines? Werkstoffsimulation Proceedings Article In: Hertelé, Stijn; Cosham, Andrew (Ed.): Technology for Future and Ageing Pipelines (TFAP 2022), Conference proceedings, pp. 9, Ghent, Belgium, 2022, ISBN: 9780646990613. @inproceedings{Trautmannsberger2022,
title = {Do simplified pressure decay and backfill models represent the loading scenario during the running ductile fracture scenario in gas transmitting onshore pipelines?},
author = {R. Trautmannsberger and P. Marx and V. Keim and M. Paredes and A. Nonn},
editor = {Stijn Hertelé and Andrew Cosham},
url = {http://hdl.handle.net/1854/LU-8751432},
isbn = {9780646990613},
year = {2022},
date = {2022-01-01},
urldate = {2022-01-01},
booktitle = {Technology for Future and Ageing Pipelines (TFAP 2022), Conference proceedings},
pages = {9},
address = {Ghent, Belgium},
abstract = {During the running ductile fracture (RDF) in onshore pipelines, interaction takes place between the three physical components pipe, transported mixture and the surrounding backfill. To minimize the accidental consequences, the ductile crack arrest needs to be ensured for service conditions as a major part of the fracture control stage in the current pipeline design standards. The inaccurate description of these physical components and their interactions revealed the shortcomings of the design methods when applied to modern, high-toughness pipeline steels and two-phase mixture compositions. The coupled fluid-structure-interaction (FSI) model has been employed to describe the crack driving forces in the form of the inner pressure profiles during the mixture decompression. Due to the enormous computational effort of the FSI models, this paper deals with the question whether simplified approaches are justified to represent the load case in the RDF scenario. Therefore, contact pressure profiles along the inner and outer pipe wall were extracted from experimentally verified FSI-RDF simulations to study the high loading scenario during the RDF. In the second step, the numerical data was used to determine a simplified loading model that captures the mixture decompression and soil backfill. The developed model was able to represent the temporal and spatial dependence in the loading scenario during the RDF. Although the comparison with the FSI simulations showed reasonable agreement, the temporal dependence of the crack driving pressure from the decompressing fluid and the counteracting backfill forces clearly emphasized the need for the coupled FSI consideration.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
During the running ductile fracture (RDF) in onshore pipelines, interaction takes place between the three physical components pipe, transported mixture and the surrounding backfill. To minimize the accidental consequences, the ductile crack arrest needs to be ensured for service conditions as a major part of the fracture control stage in the current pipeline design standards. The inaccurate description of these physical components and their interactions revealed the shortcomings of the design methods when applied to modern, high-toughness pipeline steels and two-phase mixture compositions. The coupled fluid-structure-interaction (FSI) model has been employed to describe the crack driving forces in the form of the inner pressure profiles during the mixture decompression. Due to the enormous computational effort of the FSI models, this paper deals with the question whether simplified approaches are justified to represent the load case in the RDF scenario. Therefore, contact pressure profiles along the inner and outer pipe wall were extracted from experimentally verified FSI-RDF simulations to study the high loading scenario during the RDF. In the second step, the numerical data was used to determine a simplified loading model that captures the mixture decompression and soil backfill. The developed model was able to represent the temporal and spatial dependence in the loading scenario during the RDF. Although the comparison with the FSI simulations showed reasonable agreement, the temporal dependence of the crack driving pressure from the decompressing fluid and the counteracting backfill forces clearly emphasized the need for the coupled FSI consideration. |
2021
|
Rajaraman, D.; Keim, V.; Pondicherry, K.; Nonn, A.; Hertelé, S.; Fauconnier, D. Stress state characterization of ductile materials during scratch abrasion Werkstoffsimulation Journal Article In: Wear, pp. 203712, 2021, ISSN: 0043-1648. @article{RAJARAMAN2021203712,
title = {Stress state characterization of ductile materials during scratch abrasion},
author = {D. Rajaraman and V. Keim and K. Pondicherry and A. Nonn and S. Hertelé and D. Fauconnier},
url = {https://www.sciencedirect.com/science/article/pii/S0043164821001010},
doi = {https://doi.org/10.1016/j.wear.2021.203712},
issn = {0043-1648},
year = {2021},
date = {2021-01-01},
journal = {Wear},
pages = {203712},
abstract = {Abrasive wear limits the lifetime of many machine components. Most empirical models relate the abrasive wear resistance to material hardness. In reality, however, other material properties are also influencing as scratch abrasion damage follows from a highly complex stress trajectory upon scratching. Numerical (finite element) simulation of scratch abrasion requires the use of a material damage model, which translates this stress trajectory into material degradation and removal. Most damage models include the first two stress invariants. However, fully incorporating the complex stress trajectories that occur during scratch abrasion may require damage models with dependence of the third deviatoric parameter (Lode angle). This paper serves as an a-priori study to evaluate the stress states that may occur during scratch abrasion. Three mechanisms (ploughing, wedging, cutting) are considered. Hereto, the results of an extensive parametric study using elastic-plastic finite element simulations of a scratch indentation process are discussed. Complex, non-proportional variations in stress state values are observed to occur during scratch abrasion. Distinct stress state trajectories are identified for the three abovementioned mechanisms. These variations are critically discussed to motivate a selection of suitable damage models for rigorous finite element analysis of the wear processes associated with scratch abrasion.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Abrasive wear limits the lifetime of many machine components. Most empirical models relate the abrasive wear resistance to material hardness. In reality, however, other material properties are also influencing as scratch abrasion damage follows from a highly complex stress trajectory upon scratching. Numerical (finite element) simulation of scratch abrasion requires the use of a material damage model, which translates this stress trajectory into material degradation and removal. Most damage models include the first two stress invariants. However, fully incorporating the complex stress trajectories that occur during scratch abrasion may require damage models with dependence of the third deviatoric parameter (Lode angle). This paper serves as an a-priori study to evaluate the stress states that may occur during scratch abrasion. Three mechanisms (ploughing, wedging, cutting) are considered. Hereto, the results of an extensive parametric study using elastic-plastic finite element simulations of a scratch indentation process are discussed. Complex, non-proportional variations in stress state values are observed to occur during scratch abrasion. Distinct stress state trajectories are identified for the three abovementioned mechanisms. These variations are critically discussed to motivate a selection of suitable damage models for rigorous finite element analysis of the wear processes associated with scratch abrasion. |
Xue, L.; Keim, V.; Paredes, M.; Nonn, A.; Wierzbicki, T. Anisotropic effects on crack propagation in pressurized line pipes under running ductile fracture scenarios Werkstoffsimulation Journal Article In: Engineering Fracture Mechanics, vol. 249, pp. 107748, 2021, ISSN: 0013-7944. @article{XUE2021107748,
title = {Anisotropic effects on crack propagation in pressurized line pipes under running ductile fracture scenarios},
author = {L. Xue and V. Keim and M. Paredes and A. Nonn and T. Wierzbicki},
url = {https://www.sciencedirect.com/science/article/pii/S001379442100196X},
doi = {https://doi.org/10.1016/j.engfracmech.2021.107748},
issn = {0013-7944},
year = {2021},
date = {2021-01-01},
journal = {Engineering Fracture Mechanics},
volume = {249},
pages = {107748},
abstract = {The current analyses present results of running ductile fracture propagation in high strength X100 line pipe steels under the influence of anisotropy. Mechanical anisotropy is commonly available in pipe products as a result of the manufacturing process, especially, those subjected to hot/cold-worked deformation. The outcomes of the present analyses show that its effect on the behavior of running ductile fracture in cracked pipes undergoing depressurization is meaningful. For instance, the Crack-Tip Opening Angle (CTOA) not only exhibits a strong dependence to the pipe’s diameter size, but also to the material’s anisotropy nature when compared to a hypothetical isotropic material. Moreover, laboratory scale tests such as those performed on Battelle Drop Weight Tear (BDWT) samples provide useful information about initiation of ductile crack propagation when the anisotropy features are taken into account in the material description.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The current analyses present results of running ductile fracture propagation in high strength X100 line pipe steels under the influence of anisotropy. Mechanical anisotropy is commonly available in pipe products as a result of the manufacturing process, especially, those subjected to hot/cold-worked deformation. The outcomes of the present analyses show that its effect on the behavior of running ductile fracture in cracked pipes undergoing depressurization is meaningful. For instance, the Crack-Tip Opening Angle (CTOA) not only exhibits a strong dependence to the pipe’s diameter size, but also to the material’s anisotropy nature when compared to a hypothetical isotropic material. Moreover, laboratory scale tests such as those performed on Battelle Drop Weight Tear (BDWT) samples provide useful information about initiation of ductile crack propagation when the anisotropy features are taken into account in the material description. |
