Infrared Real-Time Yarn Monitoring in Circular Knitting Machines

As the textile industry continues its transition toward greater automation and intelligent manufacturing, the ability to monitor yarn status in real-time on circular knitting machines has become increasingly vital. Yarn breakage, misfeeding, and other issues can significantly impact production quality and efficiency. Traditional sensor-based detection methods often fall short in meeting the high-speed, high-accuracy demands of modern circular knitting operations.

In this context, this paper presents a novel yarn monitoring method that leverages infrared sensitization and photoelectric signal processing technology to track yarn movement without physical contact. By integrating hardware circuits with specialized software algorithms, the proposed system enhances the reliability and precision of fault detection in yarn feeding systems. Experimental validation and on-machine testing confirm the effectiveness of this approach, positioning it as a promising innovation for Chinese-made circular welt knitting machines.

Circular knitting machines are core components in the production of seamless textiles, sportswear, and other knitted fabrics. These machines operate at high speeds, often using hundreds of needles, and rely heavily on the uninterrupted and accurate delivery of yarn to maintain consistent fabric quality.

In conventional setups, yarn faults like breakage, slippage, or tension fluctuations are primarily detected using mechanical, piezoelectric, or capacitive sensors. Although these methods offer basic fault detection capabilities, they come with limitations such as poor adaptability to complex environments, mechanical wear, environmental interference, and direct contact with yarns, which can introduce additional tension and reduce yarn strength.

Given these limitations, there is an urgent need for advanced, non-contact, and adaptive solutions that enable continuous monitoring of yarn conditions without interrupting the knitting process. This has spurred interest in optical sensing technologies, particularly infrared-based systems, for real-time, precise fault detection.

These sensors typically respond to mechanical vibrations or strain caused by yarn tension or movement. While sensitive, their direct contact with the yarn can introduce unintended tension, impacting the mechanical properties of the yarn and potentially affecting fabric quality.

Piezoelectric sensors are also prone to wear and signal instability in high-speed or humid environments, reducing long-term reliability.

Capacitive sensors detect yarn faults by measuring changes in dielectric constants, but they are highly sensitive to ambient humidity and temperature. This makes them less suitable for environments with fluctuating conditions.

Electromechanical systems often rely on physical switches or detectors, which may fail to detect subtle yarn movement anomalies or misfeeds.

Some advanced knitting machines incorporate camera-based visual inspection systems. While these systems provide detailed yarn behavior information, they are expensive, complex to implement, and often require significant computational power and lighting control, limiting their practicality in standard production environments.

To overcome the limitations outlined above, we propose an external infrared-sensitized monitoring scheme. This method enables non-contact detection of yarn movement and abnormalities such as breakage and slack feeding. Key benefits include:

Non-invasive monitoring: No physical contact with the yarn, preserving yarn integrity.

High-speed response: Real-time signal processing enables immediate detection of irregularities.

Environmental adaptability: Infrared sensing is less affected by ambient light, humidity, or dust.

The core of the system lies in converting physical yarn motion into modulated photoelectric signals using an infrared emitter-receiver pair. When the yarn passes through the infrared beam, it intermittently blocks the light, producing a modulated signal that corresponds to yarn movement. The resulting signal is then filtered, amplified, and processed through a microcontroller to determine the yarn’s operational status.

Infrared Emitter and Receiver: A pair of diodes positioned across the yarn’s feeding path. The emitter sends a constant infrared beam, while the receiver detects any interruptions caused by yarn movement.

Signal Amplifier: Converts the low-level photodiode current into a usable voltage signal and filters noise to ensure signal integrity.

Microcontroller (MCU): Handles real-time analysis of signal patterns. Key functions include edge detection, time-interval analysis, and anomaly recognition.

Alarm Interface: If yarn breakage or abnormal motion is detected, an output signal is sent to activate an alarm or halt machine operation to prevent defective output.

Initialization Phase: Establishes baseline signal thresholds under normal operating conditions.

Sampling and Data Acquisition: Continuously samples the infrared signal at set intervals using ADC (analog-to-digital converter) features of the MCU.

Pattern Recognition: Analyzes the frequency and amplitude of the signal to determine whether yarn motion is consistent or interrupted.

Fault Classification: Differentiates between yarn breakage, slippage, or complete stoppage using predefined logic trees and machine learning principles in advanced versions.

To evaluate the performance of the proposed system, tests were conducted on a standard circular welt knitting machine under various operational conditions:

Speed Range: 600 to 1200 RPM

Yarn Types: Cotton, polyester, blended yarns

Ambient Conditions: Variable humidity (40% to 75%) and temperature (15°C to 30°C)

Key Findings:

Detection Accuracy: The system achieved a 96.4% success rate in identifying yarn breakage across all yarn types.

Response Time: The average delay between yarn fault occurrence and detection was less than 150 milliseconds.

False Positive Rate: Kept below 3.5%, primarily due to temporary signal disturbances, which can be further reduced by adaptive thresholding.

System Durability: After 500 continuous operation hours, the infrared emitter-receiver pair showed no performance degradation.

These findings affirm the system’s effectiveness in providing real-time, reliable yarn monitoring without adding mechanical burden or compromising product quality.

The comparative analysis clearly illustrates the superior adaptability, accuracy, and efficiency of the infrared-based system, especially for medium- to high-speed knitting environments.

This system’s modular design allows for easy scalability across multiple yarn feeders or machines. As industry trends push toward smart factory solutions and IoT integration, future iterations may include:

Wireless Data Transmission: Allowing remote monitoring via mobile or cloud interfaces.

AI-Powered Diagnostics: Enabling predictive maintenance and pattern recognition for early-stage yarn wear detection.

Integrated Lighting Control: Enhancing infrared accuracy by dynamically adjusting beam intensity based on ambient conditions.

The system can also be integrated with existing MES (Manufacturing Execution Systems) platforms to provide real-time production metrics and fault logs for improved workflow management.

The development of an infrared-based external yarn monitoring system marks a significant step forward in smart textile manufacturing. By addressing the limitations of contact-based and environmentally sensitive detection technologies, this solution offers reliable, real-time monitoring capabilities essential for modern high-speed circular knitting operations.

Through robust hardware integration, intelligent signal processing, and proven on-machine performance, the system enhances fault diagnosis accuracy, reduces downtime, and improves overall production efficiency. As the textile machinery industry continues to evolve, such innovations are crucial in supporting China’s transition to high-end, intelligent manufacturing of circular welt knitting machines.