Screw Tightening Anomaly Detection from Torque and Speed Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automatic screw tightening apparatuses fail to accurately determine if a screw is properly tightened, especially when there are minor flaws in the screw thread or if the screw is not in the correct position, leading to incomplete tightening and potential loosening post-shipment due to debris caught on the bearing surface.
Innovation Solution
An automatic screw tightening method that measures time-varying torque and rotational speed data between the screw tightening start and completion points, extracts features from this data, and uses a Taguchi-based overall evaluation measure to determine the fit or unfit state of the screw, comparing the unified numerical index against a threshold to assess the tightened state.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a torque limiter is used to manage constant screw tightening torque, then the screw tightening torque can be controlled to reach a preset value, but the screw tightening may be terminated halfway through if a screw thread includes an anomaly such as a flaw or if the screw is not in a correct position or posture with respect to a threaded hole
Solution Approach 1:
The patent applies preliminary action by measuring motor load current and cumulative rotation amount during the screw tightening process before the torque limiter operates. These measurements are taken in advance to detect anomalies early, allowing the system to identify flawed screws or incorrect positioning before the torque reaches the preset value and terminates the tightening process.
Solution Approach 2:
The patent implements feedback by continuously monitoring motor load current and cumulative rotation amount, comparing these values against predetermined thresholds, and providing real-time feedback to determine whether the screw tightening is proceeding normally. This feedback mechanism enables the system to detect anomalies and distinguish between normal torque variations and actual defects.
2Productivity
If the screw tightening torque reaches the preset value with debris caught on the bearing surface, then the screw tightening completes, but the bearing surface and the screw are not in close contact with each other, and appropriate frictional force cannot be ensured
Solution Approach 1:
The patent applies preliminary action by measuring and evaluating motor load current and cumulative rotation amount before the screw tightening completes. This allows the system to detect the presence of debris or abnormal conditions early in the process, preventing defective tightenings from completing while maintaining efficient operation for normal screws.
Solution Approach 2:
The patent implements feedback by continuously monitoring tightening parameters and comparing them against expected values. When debris is detected through abnormal current or rotation patterns, the system provides feedback to identify the defective screw, allowing differentiation between normal completions and defective ones while maintaining productivity.
3Device complexity
If the motor load current at the torque-up operation time point is used for determination, then the tightened state can be assessed, but determination of an anomaly of a screw before the torque-up operation time point or debris caught between a male thread and a female thread is impossible
Solution Approach 1:
The patent applies preliminary action by measuring motor load current and cumulative rotation amount throughout the entire screw tightening process, not just at the torque-up operation time point. These measurements are taken in advance during the tightening operation, enabling detection of anomalies before they would normally be detected, including debris between threads or screw flaws.
Solution Approach 2:
The patent implements feedback by continuously monitoring motor load current and cumulative rotation amount and comparing these values against predetermined thresholds throughout the tightening process. This continuous feedback enables real-time detection of anomalies, providing much earlier and more accurate anomaly detection than single-point measurement at torque-up.
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
Enables real-time, non-destructive determination of a badly tightened screw, preventing unfit components from being shipped and ensuring proper frictional force on the bearing surface, thus reducing the risk of screw loosening.
Implementation Method 1
a motor that produces rotary motion
Implementation Method 2
a torque limiter that couples the motor and the driver bit for transmitting the rotary motion and uncouples the motor and the driver bit when a preset screw tightening torque is reached
Implementation Method 3
appropriate frictional force cannot be ensured on the bearing surface, and screw loosening may occur
Data Source
AI summary
An automatic screw tightening method includes measuring a datum on time-varying screw tightening torque from a motor and a datum on time-varying motor rotational speed between a screw tightening start time point for a male thread with respect to a female threaded hole and a screw tightening completion time point; extracting a plurality of features from measurement-based data on the time-varying screw tightening torque and the time-varying motor rotational speed; and determining, with use of the features, whether a tightened state of the male thread is fit or unfit. In the determination, determining a unified numerical index from the features is included, and the numerical index is compared with a predetermined threshold in the determination of whether the tightened state is fit or unfit. The numerical index is a Taguchi (T) method-based overall evaluation measure from the features.