Mold Temperature Anomaly Detection Using Infrared Thermography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mold temperature control methods, relying on thermocouples for fixed-point temperature measurements, fail to detect uneven heat distribution and heat pool anomalies effectively, leading to defects in cast products due to their inability to monitor the entire mold surface and requiring expert analysis of thermography images.
Innovation Solution
A mold temperature anomaly sign detection apparatus that uses a thermography system to capture images of the mold surface, preprocesses the data, and generates an inference model through deep learning to predict temperature anomalies, enabling automatic detection and control of mold temperature anomalies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If thermocouples are used for temperature measurement, then the measurement system is simple and cost-effective, but only fixed-point temperature can be detected and heat pool anomalies cannot be detected
Solution Approach 1:
The patent replaces the mechanical contact-based thermocouple measurement system with a non-contact infrared thermography system. This substitution enables full-surface temperature mapping of the mold without physical contact, allowing detection of heat pool anomalies across the entire mold surface while maintaining measurement capability.
Solution Approach 2:
The patent transitions from one-dimensional point measurement (thermocouple at a single location) to two-dimensional surface measurement (infrared thermography across the mold surface). This dimensional expansion allows comprehensive temperature monitoring of the entire mold surface, enabling detection of localized heat pools that would be missed by point measurement.
2Measurement precision
If expert analysis of thermography images is used, then heat pool signs can be detected, but the method depends on individuals and cannot be widely applied
Solution Approach 1:
The patent implements an automated anomaly detection system that performs self-analysis of thermography images using pre-established reference data and anomaly criteria. The system automatically compares current thermal patterns against reference data from normal operation, identifies deviations indicating heat pools, and generates alerts without requiring expert intervention, thereby enabling widespread deployment across different factories and products.
Solution Approach 2:
The patent performs preliminary actions by collecting and storing reference thermography data during normal mold operation before anomalies occur. This reference data is used to establish baseline thermal patterns, which are then compared against real-time measurements to detect deviations. This preparatory step enables automated anomaly detection without requiring expert analysis during actual operation.
3Reliability
If mold lubricant spray is adjusted based on expert analysis, then mold temperature control can be improved, but the process requires manual intervention and cannot be automated
Solution Approach 1:
The patent implements a closed-loop feedback system where thermography images are continuously captured, automatically analyzed for heat pool signs, and used to adjust mold lubricant spray parameters in real-time. The system feeds detection results back to the spray control mechanism, enabling automated adjustment of spray direction and amount based on actual thermal conditions without manual intervention.
Solution Approach 2:
The patent replaces manual expert analysis and adjustment of mold lubricant spray with an automated control system that uses infrared thermography and algorithmic processing. This substitution transforms the manually-controlled spray system into an automated system that responds dynamically to real-time thermal measurements, improving both reliability and automation extent.
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 solution allows for the automatic detection and prevention of mold temperature anomalies, reducing defects in cast products by analyzing thermography images and adjusting mold lubricant application, thereby improving the accuracy and scalability of mold temperature control across multiple factories and products.
Implementation Method 1
by introducing thermography capable of measuring the whole surface of a mold
Data Source
AI summary
A mold temperature control system includes: an inference model generating portion configured to, based on a plurality of pieces of thermo image data of a mold acquired at predetermined intervals and pieces of teaching data associated with the plurality of pieces of thermo image data, learn a predetermined number of consecutive pieces of time-series image data extracted from the plurality of pieces of thermo image data as one piece of sample data to generate an inference model for detecting a sign of a temperature anomaly of the mold; a mold temperature anomaly degree inference portion configured to detect occurrence of the sign of the temperature anomaly of the mold the predetermined number ahead, using the inference model, based on the predetermined number of pieces of time-series image data of the mold; and a warning lamp request transmitting portion configured to control lighting-up of a warning lamp.