Continuous Casting Surface Quality Control Using Defect Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In continuous casting plants, achieving a predetermined surface quality for steel products is challenging due to spatial and temporal separation of process steps, leading to delayed defect detection and inefficient optimization across the entire process chain, with existing methods failing to effectively correlate process parameters with surface defects and guarantee specified quality.

Innovation Solution

A method that uses historically relevant control variables from steelworks, ladle furnace, and rolling mill, combined with artificial intelligence prediction models and expert systems, to identify causal process parameters for surface defects, adjust these parameters to prevent defects, and ensure a specified surface quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If process optimization is performed within a single plant section, then the complexity of cross-section coordination is reduced, but the overall surface quality across the entire process chain deteriorates due to lack of integrated optimization

Engineering Contradiction:
Improveprocess coordination complexityVSAvoidsurface quality
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The patent merges optimization across multiple plant sections (steelworks, ladle furnace, casting machine, rolling mill) into a unified process chain optimization system. By integrating control variables from all sections and using AI/ML to analyze their combined effect on surface defects, the system achieves comprehensive surface quality improvement while managing complexity through centralized coordination.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system implements feedback by recording surface defects at the surface inspection stage and tracing them back to their causal process parameters in upstream plant sections. This feedback loop enables identification of which control variables in which plant sections most influence surface quality, allowing targeted optimization across the entire process chain.

Inventive Principle:
Principle #23Feedback

2Device complexity

If surface defects are detected only at the end of the process chain, then detection equipment is simplified, but the ability to prevent defects deteriorates due to delayed detection

Engineering Contradiction:
Improvedetection systemVSAvoiddefect prevention capability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The system performs preliminary action by using AI and machine learning to predict surface defects before they occur. By analyzing historical control variables from all plant sections and correlating them with historical surface defects, the system identifies causal process parameters and adjusts them in advance to prevent defect formation, rather than merely detecting defects after they occur.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces traditional mechanical detection systems with AI and machine learning-based predictive systems. Instead of relying solely on physical inspection equipment at the end of the process chain, the system uses algorithms to analyze process data and predict surface defects, enabling earlier and more effective prevention.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If AI and ML are used to correlate process parameters with surface defects, then defect prediction capability is improved, but the method complexity deteriorates due to lack of clear causal identification

Engineering Contradiction:
Improvedefect prediction accuracyVSAvoidcausal analysis method
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses an intermediary approach by introducing AI and machine learning algorithms as mediators between raw process data and surface defect outcomes. These algorithms process and correlate control variables from multiple plant sections with surface defect data, automatically identifying causal relationships without requiring complex manual analysis. The AI/ML models serve as intermediaries that translate complex multi-variable relationships into actionable insights about which process parameters cause surface defects.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4410450A1Method for manufacturing products of certain steel grades with a specified surface quality in a continuous casting plant including a surface inspection
Publication Date: 2024.08.07 SMS GROUP GMBH
  • EP4410450A1 patent drawingFigure 1
  • EP4410450A1 patent drawingFigure 2
  • EP4410450A1 patent drawingFigure 3

AI summary

The invention relates to a method (100) "Prime Optimizer" for the production of products such as thick slabs, thin slabs or long products of a specific steel grade with a predetermined surface quality in a continuous casting plant (1) comprising at least the plant components steelworks (2), ladle furnace (3), casting machine (4), rolling mill (5) and surface inspection (6), based on the steps: - prediction (102) of surface defects using at least one prediction model (13) for surface defects based on the historically relevant control variables and the historical surface defects and their respective positions on the manufactured product; - evaluation (103) of the prediction of the prediction model (13) using an evaluation model (14) to determine the causes of the surface defects predicted by the prediction model (13), wherein the causes of the defects are each at least one control variable of the plant components;and - Determining (104) adapted control variables using an expert system (15) to avoid the predicted surface defects in order to achieve the specified surface quality, wherein the expert system (15) adapts the control variables that were determined by the evaluation model (14) as the cause of the defects.;