A new PIV method to measure powder flow velocity in laser metal deposition: an Eulerian-based approach

The influence of particles velocity and mass flow dynamics continues to be a topic of interest to the scientific community, especially for laser metal deposition process. The particles behavior between the nozzle and the substrate influence the energy absorption of the particles (along their individual trajectories), the efficiency of raw material consumption, and the quality of the additively manufactured parts. Particle velocity information is essential to develop optimal process windows. A new adaptive particle image velocimetry (PIV) method to in-flight velocity measurement for metallic particles from a three-port coaxial feed nozzle is proposed. The method has three stages. The pre-processing stage removes motion blur in high-speed images streams and highlights particles using a new particle sharpening filter. The second stage calculates the displacement vectors using a new cross-correlation algorithm called Cuckoo detrended cross-correlation. The third stage makes validation of the displacement vectors, reducing the false-positive rate in the calculation of speeds. The robustness of the proposed method is validated by comparing results against noise-contaminated synthetic images and experimental images. Besides, the experimental results and performance are compared against three different PIV software for various feeding rates (< 91g/min). The results show high precision with an uncertainty less than 1m/s. Optimal configurations for PIV measurements are proposed based on different particle densities to guarantee less human intervention in the experimental set-up.