Signal Processing Algorithm
A mathematical algorithm converts time-based sensor data into a distribution of signal strength across various frequencies. Engineers apply fast fourier transform spectral analysis to identify the root causes of periodic variations in paper thickness. The calculation isolates the specific rotations or vibrations responsible for quality issues.
Frequency Domain Identification
Patterns of light and dark bands on a finished sheet often stem from mechanical resonances in the press section or calender stack. By examining the results of fast fourier transform spectral analysis, a technician can match a peak frequency to the known RPM of a specific roll. Sensors mounted on the machine frame capture the raw data required for these calculations.
Analysis typically reveals hidden harmonics that simple amplitude meters miss entirely. Mechanical repairs or speed adjustments then target the source of the defect. This transition from a time-based view to a frequency-based view provides clarity on whether a problem is caused by a felt, a wire, a roll or a bearing.
Barring Defect Remediation
Periodic checks ensure that roll wear remains within limits. Continuous monitoring prevents failures by identifying early signs of imbalance. The data remains valid as long as the machine speed is constant during the sampling period.
