DIC (Digital Image Correlation) technology can indeed be used very effectively to accurately identify modal parameters in vibration modal analysis and has unique advantages, but it also has some challenges and applicable conditions.
How DIC technology can be used for modal analysis:
Non-contact full-field measurement: DIC technology uses a high-speed camera to track the deformation of an object's surface (typically coated with a random speckle pattern) during vibration. It calculates the displacement and strain field of each pixel or subset (called a sub-region) in the image sequence.
Acquiring vibration response data: When the structure is subjected to known excitations (such as impact hammers, vibrators, environmental excitations), a high-speed camera records the motion of the structure surface at a frame rate much higher than the vibration frequency (satisfying the Nyquist sampling theorem).
DIC software is used for data processing:
Displacement-time history: For a large number of points on the surface of the structure (theoretically every pixel), DIC provides data on the displacement of each point over time in the X, Y (in-plane) and Z (out-of-plane, requiring a solid or 3D DIC).
Time-domain analysis: For a specific point or region, the displacement-time curve can be directly observed.
Frequency domain analysis: Perform a Fast Fourier Transform (FFT) on the displacement-time signal at a selected point to obtain the frequency response function (FRF) or power spectral density (PSD) at that point. The frequency corresponding to the peak value is the natural frequency.
Modal shape extraction: This is one of the biggest advantages of DIC technology. At a specific natural frequency, by combining the vibration amplitudes (obtained from FFT or PSD) and relative phases of all measurement points at that frequency, the displacement mode shape of that mode can be obtained across the entire field. By processing the displacement data, strain mode shapes can also be obtained.
Damping ratio identification: Identifying the damping ratio is relatively complex, but it can be done using the following methods:
Time-domain method: Apply the logarithmic decay method to the free decay response (such as the response after impact excitation).
Frequency domain method: Utilizing the characteristics of the frequency response function (FRF) near the resonance peak, the half-power bandwidth method is applied.
Modal parameter identification algorithm: Input time-domain or frequency-domain data from multiple points across the entire field into a mature modal parameter identification algorithm (such as PolyMAX, ERA, SSI, etc.) to simultaneously fit the frequency, damping ratio, and mode shape.
1. The core challenges of modal parameter identification
The essence of vibration modal analysis is to solve the mass-damping-stiffness matrix , and traditional methods have limitations:
Finite element simulation: Simplified boundary conditions lead to errors (typical frequency deviation > 5%)
Experimental Modal Analysis (EMA): Improper selection of excitation points can easily lead to the loss of local modes.
2. Advantages of DIC technology for parameter identification
Global displacement field input
Directly obtain displacement time history of 50,000+ measuring points on the structural surface
Avoid modal truncation issues caused by sensor placement
High-density data support
The spacing between measuring points can reach 1/10 of the smallest structural feature size (e.g., identifying vibration at a 5mm weld point on an automotive sheet metal component).
Modal confidence score (MAC) increased to above 0.95.
3. Achieve accurate identification in four steps
Step 1: Incentive Optimization
Random excitation: suitable for broadband modal scanning (0-5kHz)
Step excitation: Focusing on low-frequency modes (<100Hz), improving signal-to-noise ratio by 40%.
Step 2: Data Preprocessing
Sampling rate setting: 5 times or more of the target frequency (e.g., ≥50kHz sampling is required for 10kHz vibration).
Hanning window application: suppressing spectral leakage, achieving a frequency resolution of 0.01Hz.
Step 3: Parameter Extraction Algorithm
Frequency domain decomposition method: rapid identification of dense modes (separation of 32 modes of spacecraft solar panels)
Time-domain recursive method: Accurate calculation of damping ratio (error < 3%)
Step 4: Result Verification
Cross-validation: DIC data vs. laser Doppler vibration meter
Finite element correction: updating the model based on measured data (after correcting the spindle stiffness of a machine tool, the modal frequency error decreased from 7.2% to 0.8%).
