The velocity measurement method of particle image analyzer is mainly based on digital image processing technology, which calculates the velocity by capturing and analyzing the motion trajectory of particles in the fluid. The core methods include particle image velocimetry (PIV), particle tracking velocimetry (PTV), and velocity inversion algorithm based on motion blurred images. The following analysis is based on principles, implementation methods, and technical characteristics:
1、 Particle Image Velocimetry (PIV)
Principle:
Spread tracer particles (such as micrometer sized particles) in the fluid, illuminate the measurement area with a pulsed laser light source, and continuously capture two frames of images with a high-speed camera. Using cross-correlation algorithm to calculate the average displacement of particle swarm within the same interpretation window in two frames of images, combined with exposure time interval to obtain velocity field.
Implementation steps:
Tracer particle distribution: Select particles with good followability (particle size<50 μ m) to ensure that their motion reflects fluid velocity.
Laser illumination: using pulsed laser to form sheet light, illuminating the measurement plane.
Image acquisition: A high-speed camera synchronously captures two frames of particle images.
Cross correlation calculation: Perform cross-correlation operation on particle images within the interpretation window to obtain displacement vectors.
Velocity calculation: Divide displacement by time interval to obtain velocity field.
Technical characteristics:
Full field measurement: can synchronously obtain two-dimensional or three-dimensional velocity fields.
Non contact: No interference to the flow field.
Accuracy dependence: particle concentration, interpretation window size, and cross-correlation algorithm accuracy.
Application Scenario:
Suitable for velocity measurement of complex flows such as gas-liquid two-phase flow and combustion flow fields, such as particle velocity measurement in the tail flame of solid rocket engines.
2、 Particle Tracking Velocimetry (PTV)
Principle:
Directly tracking the motion trajectory of individual particles in the flow field, and calculating velocity by identifying the position changes of particles in consecutive frames.
Implementation steps:
Particle recognition: using edge detection or machine learning algorithms to extract particle contours from images.
Trajectory tracking: Connecting the same particles in consecutive frames through centroid matching or probabilistic algorithms.
Velocity calculation: Obtain instantaneous velocity based on particle displacement and time interval.
Technical characteristics:
Single particle accuracy: It can obtain the velocity and acceleration of a single particle.
High computational complexity: requires processing a large amount of particle trajectory data.
Applicability: Suitable for sparse particle flow or scenes that require high spatial resolution.
Application Scenario:
Used for studying micro dynamic behaviors such as particle collision and agglomeration, such as diffusion analysis of nanoparticles in solution.
3、 Velocity inversion algorithm based on motion blurred images
Principle:
By controlling the camera exposure time, fast-moving particles can form a trailing image in the image. Use the geometric relationship between drag length, exposure time, and particle velocity to infer velocity.
Implementation steps:
Exposure time control: Select the appropriate exposure time (such as microsecond level) based on the particle movement speed.
Image acquisition: Obtain particle images containing motion blur.
Drag analysis: Extract drag length through image processing algorithms such as threshold segmentation and edge detection.
Technical characteristics:
Low cost implementation: No need for complex laser systems, suitable for industrial sites.
Accuracy limitation: dependent on exposure time control and measurement accuracy of drag length.
Real time capability: enables online measurement of particle velocity.
Application Scenario:
Used for measuring droplet velocity at the inlet of gas-liquid cyclone separators, online monitoring of local average particle velocity in circulating turbulent fluidized beds, etc.
Selection suggestion:
Need for full field velocity distribution: Priority should be given to PIV technology, combined with stereo PIV or tomographic PIV to achieve three-dimensional measurement.
Focus on single particle behavior: using PTV technology, combined with high-speed cameras and machine learning algorithms to improve tracking accuracy.
Real time monitoring of industrial sites: based on motion blur inversion algorithm, balancing cost and accuracy requirements.