D³ Projects
The first cohort of D³ started on October 1, 2023. A list of the individual projects as well as detailed project descriptions can be found below. To read more, click on the arrow to the right of the project title.
Project group Data
Hiring:
Postdoctoral researcher
Local Mentors:
M. Kästner, M. Kaliske
International Mentors:
W. Sun (data-driven computational mechanics), F. Auricchio (multiscale continuum mechanics)
Description of the project
The targeted virtual design of engineering components consisting of materials with an underlying micro- or mesostructure, e.g., spinodoid metamaterials requires numerical modeling and simulation across scales with high computational efficiency. Thus, the responsibilities of the postdoctoral researcher range from crystal plasticity tools for bridging small scales to a highly efficient data-driven multiscale approach. For developing numerical tools, mesostructures are created based on descriptors developed in D2. The training and architecture of physics-informed scale bridging neural networks are rendered thermodynamically consistent in cooperation with D1. Multiscale simulations and the assessment of surrogate models are conducted in collaboration with S2, M1 and F1. Spinodoid structures designed in S1 are used to predict the macroscopic behavior. Together with D3, the influence of uncertain input data on structure-property linkages is investigated. DP collaborates with S3 for experimental design.
Hiring: PhD Student
1st Principal Investigator:
I. Sbalzarini (data and computer science, artificial intelligence)
2nd Principal Investigator:
M. Kästner (computational mechanics)
International Mentor:
C.L. Müller (data-driven modeling, inverse problems)
Description of the project
This project studies the use of machine learning and computational inference methods for approximating and inverting the structure-property linkages of metamaterials. The central scientific question is how robust designs can be systematically achieved, how constraints from manufacturability and cost can be included, and whether data-driven methods can improve the designs or suggest entirely new ones. The structure descriptors and parameter spaces developed in D2 are used and complemented by the uncertainties estimated in D3, which are considered as additional oracle in the design centering inversion once they become available. An interesting prospect is to integrate uncertainties directly into design centering, extending the latter to a stochastic variant. The interaction with DP is on the simulations. Finally, interaction with S1 takes place on novel designs resulting from, e.g., affine transformations or meta-grain mesostructures.
Hiring: PhD Student
1st Principal Investigator:
M. Salvalaglio (mathematics, theoretical materials science)
2nd Principal Investigator:
I. Sbalzarini (data and computer science, artificial intelligence)
International Mentor:
S. M. Wise (applied mathematics)
Description of the project
Characterizing general morphological and topological features of metamaterials is crucial for establishing structure-property linkages and design centering. This project provides structural metamaterial descriptors at different length scales, with a central focus on mesoscale features and hyperuniformity. The developed descriptors and parametrizations are used for generating structures in S1 and DP as well as for optimizing them (S1) using methods from D1. Collaboration with D3 is required to provide structural uncertainties. Close contact is maintained to S1 for achieving desirable properties by hyperuniformity and novel alloy image data is obtained from M2.
Hiring: PhD Student
1st Principal Investigator: M. Kaliske (computational engineering, uncertainty quantification)
2nd Principal Investigator: M. Salvalaglio (mathematics, theoretical materials science)
International Mentor: K.Terada (multiscale damage modeling)
Description of the project
The material and geometry in additively manufactured products are subject to uncertainty. The objective of this project is to quantify the polymorphic uncertainty of the mesostructure’s morphology and of the corresponding effective properties as well as functionalization parameters, which can be used subsequently to identify the sensitivity of input parameters on the interface properties and the resilience of the metamaterial. The simulation workflow is developed with S1 and DP. A collaboration with D2 is established to develop and define appropriate mesostructure descriptors. The experimental characterization by S3 yields material parameters for the damage modeling as well as a data basis for the uncertainty quantification and validation. The determined uncertain material parameters are considered within the design centering procedure elaborated in D1. For developing the nonlocal damage formulation, an intense cooperation with S1 is intended. The uncertainty quantification for functionalized coatings is carried out in collaboration with F2 using the surrogates developed by F1.
Project group Materials
Hiring: PhD Student
1st Principal Investigator:
G. Cuniberti (materials modeling and simulation)
2nd Principal Investigator:
M. Kästner (computational mechanics)
International Mentor:
S. Curtarolo (materials databases)
Description of the project
The objective of this project is the identification and application-tailored design of alloys which optimally match mechanical and thermodynamic requirements, bearing the potential to outperform known materials and open new perspectives for target advanced applications. Hereby, nanoscale phenomena such as the quantum-mechanical binding between atoms leading to the formation of atomistic lattices dictate bulk material properties including density, mechanical stiffness and thermodynamic behavior. Thus, a detailed understanding of the interplay between atomistic composition and material properties is essential for an inverse data-driven design of materials with desired properties. To accelerate the design of alloys, a combination of active learning-based, ab-initio high-throughput materials screening using advanced computational methods and targeted experi-mental validation is employed. Promising material compositions are fabricated in M2, experimentally characterized in S3, con-sidered as infill candidates in S2 and functionalized in F1 and F2. Machine learning methods of D1 are applied to data from M1, which are compared and complemented with experiments from M2 and S3. The predicted elastic moduli and critical resolved shear stresses serve as an input for microscale continuum modeling in DP.
Hiring: PhD Student
1st Principal Investigator: J. K. Hufenbach (materials science)
2nd Principal Investigator: G. Cuniberti (materials modeling and simulation)
International Mentor:
Min-Ha Lee (materials science)
Description of the project
To exploit the full potential of additively manufactured metamaterials, tailored alloys for the respective process and the target applications are required. This project aims at the development of AlMg-based alloys for the robust processing of resilient metamaterials by LPBF. The alloy design is supported by data-driven methods and based on scale-bridging characterization for a comprehensive analysis of the process-microstructure-property interactions. This allows to overcome challenges regarding the LPBF-processing of Al alloys and to take full advantage of the design freedom. The alloy development is performed in close collaboration with the data-driven materials scouting performed in M1 and microscopy images are provided to D2 for investigating hyperuniformity. Material parameters of the tailored alloys and designs of spinodoid metamaterials to be produced by LPBF are exchanged with S1 and tested under monotonic and cyclic loading in S3. Further-more, the prepared metallic metamaterials act as substrates for the functionalization in F2.
Project group Structure
Hiring: PhD Student
1st Principal Investigator: M. Kästner (computational mechanics)
2nd Principal Investigator: M. Zimmermann (experimental process-structure-property linkages)
International Mentor: D.M. Kochmann (metamaterials, inverse design)
Associated postdoctoral researcher: K. Kalina (data-driven mechanics)
Description of the project
The deformation and failure behavior of non-periodic metamaterials with spinodoid topologies is predicted using numerical modeling, simulation and homogenization. Structure-property linkages will be explored to discover design principles for resilient spinodal structures including metagrains and their interfaces. S1 collaborates with D3 and DP. While a simple inversion can be implemented independently, advanced data-driven design centering methods are taken from D1. The structure parametrization is developed in collaboration with D2. Material parameters are provided by M2, S2 and S3, while M2 is responsible for manufacturing proposed structures. Finally, interaction with S3 is re-quired for validating the numerical simulations and the chosen design.
Hiring: PhD Student
1st Principal Investigator: C. Leyens (materials science)
2nd Principal Investigator: M. Gude (lightweight engineering)
International Mentor: A. Molotnikov (computational AM engineering)
Associated postdoctoral researcher: J. Moritz (additive manufacturing)
Description of the project
Binder Jetting is a sinter-based additive manufacturing technology that provides the opportunity to create porous metamaterial structures that can be infiltrated to improve their properties. Since this is usually done using the natural porous network of partially sintered bodies, the interactions of the base material and the spinodoid mesostructure require further attention. It is to be understood, which opportunities such an approach opens to create resilient metamaterials, e.g., to substitute conventional alloys that depend on critical raw materials and what are the main process-related influencing factors on the mechanical properties. After an initial phase of development, S1 will suggest promising metamaterial structures to be manufactured. Samples are shared with S3 for mechanical characterization. In collaboration with F2, surfaces of produced metamaterials will be analyzed and optimized towards enhanced sur-face functionalization. The infill material is chosen based on the estimated effect in collaboration with DP. Together with M1 and M2, alternative infill materials are investigated.
Hiring: PhD Student
1st Principal Investigator: M. Zimmermann (experimental process-structure-property linkages)
2nd Principal Investigator: J. K. Hufenbach (materials science)
International Mentor: N. Chawla (4D-PSP-material analysis)
Description of the project
Knowledge about the deformation behavior, structural integrity and damage tolerance of novel materials is essential to implement new approaches to expand existing design spaces. In the case of AM processed mechanical metamaterials, this comprises a 3D scale-bridging observation of stress hotspots and an understanding of the resulting microstructurally driven damage evolution. Hence, this project focuses on a spatially resolved experimental characterization to meet the requirements of the data-driven design addressed in D³ on different scales. The core objective is to identify the required level of data granularity and statistical representation of microstructural parameters for a solid prediction of the reliability of metamaterials. Therefore, an ontology for process-structure-property linkages is developed. The envisioned experiments help assess the alloy composition of M1 and M2, the AM processing in M2 and S2, the hybridization in F2 as well as the spinodal structures designed in S1. The characterization serves as input to the simulation in DP and S1, the morphology description in D2 and the uncertainty quantification in D3. Collaboration with D1 helps to develop algorithms to evaluate DIC data on different length scales. In face of the scale-bridging and central role of S3, a cooperation with DP is intended for planning experiments and laying out the architecture and interface of the data base.
Project group Functionalization
Hiring: PhD Student
1st Principal Investigator: S. Gemming (computational physics and materials science)
2nd Principal Investigator: C. Leyens (materials science)
International Mentor: F. Günther (computational physics)
Description of the project
A fundamental understanding and target-oriented design of interface properties is required to integrate spinodoid structures into technical systems and creates opportunities for improved resilience and graded functionalization. The material properties are dictated by the surface topology as well as the alloy composition, which differs from the bulk material at the interface. Therefore, F1 will employ computer-based modeling of bare and functionalized surfaces of spinodoid Al-Mg-Zr-based structures. F1 will rely on structure and composition data for the bulk (S3) and on surface and interface properties (F2) of the Al-based alloys developed in M2. F1 generates labeled data on interfaces for surrogate models developed by DP under advice by D1 and for the data-driven design of materials from M1 and structures from S1. The generated data are exchanged with F2 for functionalization studies and with D3 for an uncertainty quantification.
Hiring: PhD Student
1st Principal Investigator: M. Gude (lightweight engineering)
2nd Principal Investigator: S. Gemming (computational physics and materials science)
International Mentor: A. Boczkowska (materials science)
Description of the project
Integrating spinodoid structures into technical systems often requires their surface to exhibit certain interface properties that enable an interaction with further material systems. The present project aims at achieving these properties by means of an effective hybridization of metal spinodoid mesostructures with functional polymers in graded composition. In this context, the cross-scale design and the technological implementation of graded functionalization for resilient metamaterials are investigated. This involves the analysis of the effect of input parameters on the interface properties as well as the realization of interfaces with differing functionalities within one hybrid structure. The analyses of the microstructure influence on the bonding properties during the graded functionalization of metamaterials are performed in close collaboration with F1, whereby possible pre-treatment studies are carried out in cooperation with S3. For the specific adjustment of the properties during functionalization of the metamaterials additively manufactured by M2 and S2 and their experimental characterization, an uncertainty quantification is carried out in collaboration with D3 using the generated experimental data and the surrogate developed in F1.