Bearing Condition Detection Using Virtual Impact Force Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the condition of bearings in systems, such as those used in motors or rotating systems, face challenges in accurately assessing the condition of the lubricant and the bearing due to factors like contamination and stress, which can lead to inaccurate identification of bearing conditions.
Innovation Solution
A computer-implemented method and system that utilize operation data from sensing units to determine an operation profile of the bearing, which includes vibration, thermal, and frequency responses. This data is then used to create a virtual bearing model trained on comparable bearings, allowing for the determination of an impact force profile and the prediction of the bearing's condition, including stress distribution and remaining life.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data based analysis is used to determine bearing condition, then the method can be implemented, but the accuracy is insufficient when stress measurement is not possible
Solution Approach 1:
The patent uses vibration response as an intermediary parameter to indirectly assess stress conditions in the bearing. Instead of directly measuring stress (which is difficult), the system measures vibration response that correlates with stress levels, using this intermediate measurement to infer the actual stress state and lubricant condition.
Solution Approach 2:
The patent replaces direct mechanical stress measurement with a computational model that uses vibration data and machine learning algorithms to estimate stress conditions. This substitution allows the system to determine bearing condition without requiring direct stress sensors or complex mechanical measurement systems.
2Reliability
If experimental methods with sensing units are used, then bearing condition can be assessed, but placement of sensing units affects accuracy
Solution Approach 1:
The patent creates a virtual bearing model that replicates the physical bearing's behavior and response characteristics. This virtual model is trained on operational data and can predict bearing condition without requiring precise physical sensor placement, as the computational model accounts for various placement scenarios and optimizes the assessment accordingly.
Solution Approach 2:
The patent transforms the assessment approach by changing from direct physical measurement parameters (requiring specific sensor placements) to computational parameters derived from vibration analysis and machine learning. This parameter transformation makes the system less sensitive to exact sensing unit placement while maintaining or improving accuracy.
Data Source
AI summary
A system, apparatus, and method of determining a condition of at least one bearing in a system are provided. The method includes receiving operation data associated with the system from one or more sensing units associated with the system and determining an operation profile of the at least one bearing from the operation data. The operation profile includes a vibration response, a thermal response, and/or a frequency response associated with the at least one bearing. An impact force profile is determined during operation of the at least one bearing based on the operation profile and a virtual bearing model trained on operation profiles and impact force profiles associated with a group of bearings comparable with the at least one bearing. The condition of the at least one bearing is determined based on the impact force profile.