In this quarter, we have rebuilt a new multiwavelength light source which boasts over 300mW combined output from 10 different wavelengths and loss of less than 15% for each wavelength and it is much more flexible than the previous version too. Compared to the previous version, the efficiency and construction complexity have been dramatically improved for field use. We also have refined the diode pumped laser which now could give out more pulse energy than before, and this will improve the signal to noise ratio. We are also using the water jet we built in the 2nd quarter to calibrate our laser scattering outside the engine exhaust, and water jet calibration data is used to simulate the scattering results in the exhaust line.

In this quarter, we have started the data collection process of the first field deployable multi-wavelength PM measurement system. This system is now operating in real world on PM emissions from a turbine power generator v.s. known PM standard for the system that we designed and tested in the lab. We proved that we could repeatedly collect same scattering signal under same engine load conditions. We further improved the signal to noise ratio of the system, by shortening the exposure time below 1,500 nanosecond and increasing the peak power. Here, we give detailed description on our investigation of the mechanisms in improving precision of pulsed laser timing.

In this final quarter, we have continued to collect more field data. Here, in this report representative data collected in the field with turbine engine are presented. We also made substantial progress in calibration of standard particles using MOUDI. During the 12th quarter of this project, we collected a myriad of field data at our industrial partner's test site. These data verified the system performances. (1) The system could detect light scattering signal for all 9 wavelength lasers under different load conditions--We verified that the ELIS1024 chip could reliably collect light scattering signal from the 9 wavelength lasers, even the weakest wavelength at 355nm, thanks to our effort in improving the signal to noise ratio of the detector. (2) The data collected for each wavelength channel under the same load is consistent and repeatable--Although different wavelength channel has drastically different signal to noise ratio, after certain averages, we are able to repeat the scattering signal under the same engine conditions. (3) The data collected for each channel under different load conditions are qualitatively consistent with prediction--The data collected for each channel under different load conditions change according to the predictions. We are conducting simulation models to simulate the data and …