Optical and Electrical Considerations for Developing Pulsed High-Power LED for Volumetric Particle Tracking Velocimetry

Volumetric Particle Tracking Velocimetry (PTV) in volumes larger than 10³ cm³ with 15 μm Air Filled Soap Bubbles (AFSB) requires a powerful pulsed laser source of few tenth of mJ pulse energy, depending on the optical arrangement of the experiment. Initial experiments in our lab with 15 μm AFSB and 1 mJ pulse energy from a 532 nm laser source showed that in order to image turbulence flow phenomena in volumes about 4 × 10³ cm³ at 1 m working distance and an f-number of 8, one would need a laser source with about 315 mJ pulse energy and 1260 W average power at 4 kHz repetition rate. Such high-power 532 nm lasers are not simple to build and are expensive and complicated to develop. By reducing the working distance and increasing the diameter of the camera aperture, the required laser power was reduced by a factor 18 to 17.5 mJ pulse energy and 70W average power, which is still a very high power laser. As a possible alternative, we report here on the development of a collimated beam, high-power, pulsed 460 nm LED for volumetric PTV measurements of turbulence flow in a large volume seeded with AFSB particles. We report of the optical and electrical consideration that are required in development of such an LED source. First, we combined a practical model for the output power of an LED with a model that estimates the signal level on a CMOS detector in a volumetric PTV experiment with AFSB particles. It predicted that our LED should be driven by 277 A in order to generate 1 mJ of optical pulse energy with 20 μs pulse duration. In order to avoid power drooping at high currentsdensity in the LED, we chose an architecture of LEDs array where each LED unit of the array operates at 20 μs pulse duration, 4 kHz repetition rate and with peak current of 200 A. Each such LED will deliver 0.72 mJ pulse energy and 24 such LED units will supply the required optical density. Second, we report of the optical and electrical features of the collimated , high-power, pulsed 460nm LED unit we built. We tested the LED in a PTV experiment where the LED operated with 0.72 mJ pulse energy, 1 kHz repetition rate and 20 A peak current. The measured signal current from a CMOS-pixel due to Mie scattering AFSB particles was in good agreement with the predicated signal level.