SiC Crystal Growth Clogging Prediction via Pressure Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge in growing long silicon carbide (SiC) single crystals is the accumulation of solid deposits within the manufacturing apparatus, which clogs gas flow passages and disrupts the continuous supply of SiC source gas, making it difficult to predict and prevent clogging.
Innovation Solution
A SiC single crystal manufacturing apparatus equipped with a pressure sensor and a computing device that uses machine learning to predict clogging time based on gas pressure measurements and a learning model created from simulations and experimental data, allowing for the anticipation and management of clogging events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Length of moving object
If gas growth method is used to grow long SiC single crystal, then the crystal length can be increased, but solid deposit accumulates and clogs gas flow passages interrupting gas supply
Solution Approach 1:
The system performs preliminary action by predicting clogging time before actual clogging occurs. The prediction unit uses machine learning models to forecast when solid deposits will clog the gas introduction pipe, allowing operators to take preventive measures such as cleaning or maintenance before the clogging actually happens, thus maintaining gas supply continuity while enabling long crystal growth
Solution Approach 2:
The system implements feedback by continuously monitoring gas pressure and feeding this information to the prediction unit. The pressure sensor provides real-time data about gas flow conditions, which the machine learning model uses to update clogging time predictions. This closed-loop feedback system allows dynamic adjustment and early warning of potential clogging issues
2Reliability
If solid deposit accumulates in gas flow passage, then clogging occurs disrupting growth conditions, but continuous monitoring and prediction systems increase device complexity
Solution Approach 1:
The system replaces complex mechanical monitoring approaches with a computational intelligence-based solution. Instead of using multiple physical sensors and mechanical detection devices, the invention uses machine learning algorithms that process simple pressure data to predict clogging. This substitution of mechanical/physical monitoring systems with computational models reduces overall device complexity while maintaining high reliability
Solution Approach 2:
The machine learning prediction unit acts as an intermediary between the simple pressure sensor data and the complex phenomenon of solid deposit accumulation. Rather than directly monitoring difficult-to-measure parameters like deposit thickness or composition, the system uses pressure as an intermediary parameter that the ML model correlates with clogging risk, simplifying the monitoring requirement
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 enables the accurate prediction of clogging times, reducing the risk of apparatus failure and allowing for the continuous growth of long SiC single crystals by identifying potential clogging states through observable gas pressure parameters.
Implementation Method 1
a gas pressure of the supply gas measured by a pressure sensor
Implementation Method 2
the heating device is configured to heat and decompose the SiC raw material gas
Implementation Method 3
a learning model created by machine learning using data calculated from results of simulations of growing the SiC single crystal and results of experiments of growing the silicon carbide single crystal
Data Source
AI summary
A silicon carbide single crystal manufacturing apparatus includes a pressure sensor and a computing device. The pressure sensor is configured to measure a gas pressure of a supply gas containing a silicon carbide raw material gas and introduced into a crucible through a gas introducing pipe. The computing device is configured to perform a prediction of a clogging time, which is a time until the gas introduction pipe is clogged with a solid deposit, based on the gas pressure measured by the pressure sensor a learning model created by machine learning using data calculated from results of simulations of growing the silicon carbide single crystal and results of experiments of growing the silicon carbide single crystal.