TY - JOUR
T1 - FlexMM
T2 - A standard method for material descriptions in FEM
AU - Groen, Manso
AU - Solhjoo, Soheil
AU - Voncken, Ruud
AU - Post, Jan
AU - Vakis, Antonis
PY - 2020/10
Y1 - 2020/10
N2 - This article discusses a number of key issues concerning simulation-based digital twins in the domain of multistage processes. Almost all production processes are multistage in nature, and so most digital twins involve multiple physical phenomena, process steps and different solvers for the simulations. Good interoperability between model solvers and processes are key to achieving a functional digital twin. Passing information between steps can be challenging, complex and time consuming, especially for material data, because the constitutive model interacts with the full modeling environment: material behavior is interdependent with the history of the process, the solver subroutines and the boundary conditions. This work proposes a flexible yet robust standardization approach, called FlexMM, for dealing with material data, constitutive models, measurement data or mathematical models to overcome part of the abovementioned complexity. The implementation of FlexMM consists of a general rule structure in which constitutive behavior is described, as well as its interaction with the subroutines used by the finite element solver. The definition of the constitutive model is stored in a separate file, in which the material behavior can be described in a user selected format, such as look-up tables, standard statistical models, machine learning or analytical expressions. After a calculation step, the new local material properties are mapped to a file to facilitate the next history-dependent step. In this way, the interaction between the different fabrication steps and processes can be incorporated. A material/process case study is presented to demonstrate the flexibility and robustness of FlexMM.
AB - This article discusses a number of key issues concerning simulation-based digital twins in the domain of multistage processes. Almost all production processes are multistage in nature, and so most digital twins involve multiple physical phenomena, process steps and different solvers for the simulations. Good interoperability between model solvers and processes are key to achieving a functional digital twin. Passing information between steps can be challenging, complex and time consuming, especially for material data, because the constitutive model interacts with the full modeling environment: material behavior is interdependent with the history of the process, the solver subroutines and the boundary conditions. This work proposes a flexible yet robust standardization approach, called FlexMM, for dealing with material data, constitutive models, measurement data or mathematical models to overcome part of the abovementioned complexity. The implementation of FlexMM consists of a general rule structure in which constitutive behavior is described, as well as its interaction with the subroutines used by the finite element solver. The definition of the constitutive model is stored in a separate file, in which the material behavior can be described in a user selected format, such as look-up tables, standard statistical models, machine learning or analytical expressions. After a calculation step, the new local material properties are mapped to a file to facilitate the next history-dependent step. In this way, the interaction between the different fabrication steps and processes can be incorporated. A material/process case study is presented to demonstrate the flexibility and robustness of FlexMM.
KW - FEM analysis
KW - Digital twin
KW - Material modeling
KW - Multi stage modeling
KW - User subroutines
KW - Zero-defect manufacturing
KW - Industry 4.0
KW - DIGITAL TWIN
KW - FLOW-STRESS
KW - TIME-STEP
KW - ELEMENT
KW - MESHES
KW - DISCRETE
U2 - 10.1016/j.advengsoft.2020.102876
DO - 10.1016/j.advengsoft.2020.102876
M3 - Article
SN - 0965-9978
VL - 148
JO - Advances in Engineering Software
JF - Advances in Engineering Software
M1 - 102876
ER -