Additive Manufacturing Material Property Prediction via Thermal History Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting material performance in additive manufacturing, such as laser powder-bed fusion, are hindered by the large scale difference between local thermal processes and component dimensions, requiring extensive and time-consuming simulations, making it impractical for industrially relevant parts due to high computational demands and storage requirements.
Innovation Solution
A method that reduces the data set of solidification gradients and velocities to enable interpolation-based prediction of microstructure characteristics and material properties, focusing on thermal history and using microstructural modeling to determine material properties efficiently, thereby reducing computational resources and storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full thermal history simulation is performed for each scan vector to accurately predict material properties, then prediction accuracy is improved, but computational time and resources become prohibitively large
Solution Approach 1:
The component is divided into multiple regions, and the thermal history simulation is performed selectively for representative regions rather than the entire component. This segmentation allows accurate prediction of material properties in critical areas while avoiding unnecessary computational expenditure in less critical regions.
Solution Approach 2:
Instead of performing complete thermal history simulations for all scan vectors throughout the entire component, the method applies partial action by simulating only the necessary portions (representative regions) that capture the essential thermal behavior. This partial simulation approach provides sufficient accuracy for material property prediction while dramatically reducing computational time and resources.
2Measurement precision
If detailed microstructural modeling is performed at the scale of individual grains, then microstructure prediction accuracy is improved, but computational resources and storage requirements become unmanageable
Solution Approach 1:
The microstructural analysis is segmented to focus on representative regions rather than the entire component. By analyzing microstructure development in selected representative areas, the method captures essential microstructural features and their impact on material properties while keeping data volumes manageable.
Solution Approach 2:
The method extracts and analyzes only the critical microstructural features and thermal history data necessary for predicting material properties. By taking out and focusing on the essential microstructural parameters from the complex simulation data, the approach maintains prediction accuracy while significantly reducing the quantity of data that must be processed and stored.
3Manufacturing precision
If complete thermal history analysis is performed for the entire component, then local variation capture is improved, but computational complexity and processing time increase dramatically
Solution Approach 1:
The component is segmented into multiple regions with different thermal characteristics. Thermal history analysis is performed for representative regions that capture the dominant thermal behaviors and local variations. This segmentation allows the method to account for local variations in microstructure and material properties without requiring complete analysis of every region, thereby reducing computational complexity.
Solution Approach 2:
The thermal history data obtained from representative regions is used universally to predict material properties across the entire component. By capturing the essential thermal behaviors in representative regions, the method creates a universal basis for prediction that applies to other regions with similar thermal characteristics, reducing overall computational complexity while maintaining manufacturing precision.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for faster and more accurate prediction of material performance at the part scale, reducing computational time and storage requirements, and improving the design and generation of components with desired mechanical properties, minimizing mismatches between simulation and reality.
Implementation Method 1
powder metal is molten and solidified using a laser beam
Implementation Method 2
heating/cooldown happens at a very local scale (e.g., on the order of microns) and short timespans (e.g., on the order of 106 K/s cooldown)
Implementation Method 3
The consequence is that heating/cooldown happens at a very local scale
Data Source
AI summary
A method for predicting a material property of a component made by additive manufacturing includes: defining an area of interest and generating a mesh in the area of interest; providing a temperature model during the additive manufacturing process; providing a process parameter set; calculating a thermal history in the area of interest based on the process parameter set using the temperature model; and determining a solidification gradient and a solidification front velocity in the area of interest from the thermal history. To obtain reasonable material predictions, the method may further include: reducing the data set of solidification gradients (SGR) and solidification front velocities; determining microstructure characteristics for the reduced data set by microstructural modelling; determining the material property for the reduced data set using a material property model; and interpolating of material property from the solidification gradient and solidification front velocity for the nodes within the area of interest.