22
S. P. Donegan and M. A. Groeber
• A select overview of various ICME software tools currently in use, including
both commercial and open-source solutions
• A description of the requirements that a successful ICME workflow manager
must meet
• An example implementation of such a workflow manager
• Presentation of a case study that highlights the utility of an ICME software
infrastructure for solving modern materials problems
This chapter does not discuss details regarding data storage or infrastructure
systems, such as Materials Commons [2], The Materials Data Facility (MDF) [3],
or the Materials Project [4], or visualization tools, such as ParaView [5]. Instead, the
focus is on tools used to generate and analyze materials data in an ICME context.
2 ICME Software Tools
We consider the following general categories for ICME tools: simulation, in which
a physics-based model is used to generate information about a material process,
evolution, or behavior, and analytics, where simulation and characterization data
are postprocessed to produce additional information streams. A key distinguishing
feature in this definition of simulation tools is the use of physics-informed models.
Analytics tools may also be used to model materials, but we distinguish these from
simulation tools as being data-driven. Such data-driven approaches typically use
characterization or simulation data to fit surrogate models that approximate the
underlying material physics without the need for explicit parameterization.
3 Simulation Tools
Materials modeling and simulation has a rich history that extends beyond the
genesis of ICME. Within metals processing, DEFORM ® has been commercially
used since the early 1990s to simulate hot forging processes in both 2D and 3D [6].
Further capabilities include simulation of cold forming, machining, heat treatments,
and microstructure evolution [6]. Similarly, ProCAST is a commercially available
simulation package for casting processes, with support for die casting, investment
casting, and continuous casting [7].
Behavior modeling of structural materials typically consists of solving a set of
constitutive equations with supplied boundary conditions using a numerical method.
The most commonly used numerical approach is the finite element method (FEM),
in which a material volume is discretized into distinct elements on which local
solutions are computed. Commercially available FEM packages include Abaqus [8]
and ANSYS [9], both of which are used extensively within the aerospace supply
chain to simulate material response. Several open-source FEM solutions also exist.
Albany is a modular, general FEM solver for partial differential equations built
S. P. Donegan and M. A. Groeber
• A select overview of various ICME software tools currently in use, including
both commercial and open-source solutions
• A description of the requirements that a successful ICME workflow manager
must meet
• An example implementation of such a workflow manager
• Presentation of a case study that highlights the utility of an ICME software
infrastructure for solving modern materials problems
This chapter does not discuss details regarding data storage or infrastructure
systems, such as Materials Commons [2], The Materials Data Facility (MDF) [3],
or the Materials Project [4], or visualization tools, such as ParaView [5]. Instead, the
focus is on tools used to generate and analyze materials data in an ICME context.
2 ICME Software Tools
We consider the following general categories for ICME tools: simulation, in which
a physics-based model is used to generate information about a material process,
evolution, or behavior, and analytics, where simulation and characterization data
are postprocessed to produce additional information streams. A key distinguishing
feature in this definition of simulation tools is the use of physics-informed models.
Analytics tools may also be used to model materials, but we distinguish these from
simulation tools as being data-driven. Such data-driven approaches typically use
characterization or simulation data to fit surrogate models that approximate the
underlying material physics without the need for explicit parameterization.
3 Simulation Tools
Materials modeling and simulation has a rich history that extends beyond the
genesis of ICME. Within metals processing, DEFORM ® has been commercially
used since the early 1990s to simulate hot forging processes in both 2D and 3D [6].
Further capabilities include simulation of cold forming, machining, heat treatments,
and microstructure evolution [6]. Similarly, ProCAST is a commercially available
simulation package for casting processes, with support for die casting, investment
casting, and continuous casting [7].
Behavior modeling of structural materials typically consists of solving a set of
constitutive equations with supplied boundary conditions using a numerical method.
The most commonly used numerical approach is the finite element method (FEM),
in which a material volume is discretized into distinct elements on which local
solutions are computed. Commercially available FEM packages include Abaqus [8]
and ANSYS [9], both of which are used extensively within the aerospace supply
chain to simulate material response. Several open-source FEM solutions also exist.
Albany is a modular, general FEM solver for partial differential equations built