Green Compact Shape Prediction With Temperature-Dependent Viscosity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for predicting the shape of a green compact during sintering, particularly in liquid-phase sintering, lack accuracy due to the absence of a well-established physical model, leading to significant deformation and increased costs from trial and error in design adjustments.
Innovation Solution
An information processing method using a finite element method (FEM) with a viscous constitutive equation and machine learning, incorporating a temperature/viscosity curve to predict the shape of a green compact, optimizing parameters through design of experiments and machine learning to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning model is used to predict green compact shape, then prediction capability is provided, but prediction accuracy is insufficient
Solution Approach 1:
The invention changes the parameter representation from simple geometric dimensions to physical properties including viscosity values at different temperatures. By incorporating temperature-dependent viscosity parameters and their rates of change, the prediction model captures the rheological behavior of the green compact during sintering, significantly improving prediction accuracy and reliability
Solution Approach 2:
The invention creates a composite prediction model that combines machine learning algorithms with physical sintering models. The hybrid approach integrates data-driven patterns recognition with physics-based viscosity-temperature relationships, producing more reliable and accurate shape predictions than either approach alone
2Manufacturing precision
If trial and error design adjustments are made, then shape deformation issues can be addressed, but costs increase and time is lost
Solution Approach 1:
The invention performs preliminary shape prediction and deformation analysis before actual fabrication using the trained machine learning model. By predicting the sintered shape from green compact geometry and identifying potential deformation issues in advance, designers can adjust parameters beforehand, eliminating the need for costly trial-and-error iterations during production
Solution Approach 2:
The invention implements a feedback mechanism where prediction results are compared with actual sintering outcomes. The prediction model continuously learns from real-world data, refining its accuracy over time. This closed-loop system enables progressive improvement of shape prediction without requiring repeated physical trials
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
The method enables precise prediction of the final compact shape, reducing deformation and design iterations, thereby minimizing costs and improving fabrication efficiency.
Implementation Method 1
a temperature/viscosity curve in which, when a temperature has reached a predetermined temperature threshold value, a viscosity decreases from a first viscosity to a second viscosity
Data Source
AI summary
An information processing method is performed by an information processing apparatus and includes receiving a shape of a fabrication object after sintering, predicting a shape of the fabrication object before the sintering based on the shape of the fabrication object after the sintering and a prediction condition stored in a storage unit, and outputting information relating to the predicted shape of the fabrication object before the sintering. The prediction condition includes information based on a temperature threshold value that is a temperature at which a viscosity of the fabrication object decreases from a first viscosity to a second viscosity due to the sintering, the first viscosity, and the second viscosity.