Polychromators are most frequently used in Thomson scattering (TS) diagnostics to analyze the scattered light spectrum and intensity so that the plasma electron temperature (Te) and density (ne) can be derived. For Te measurements, the spectral response of the polychromator channels and the relative spectral responsivities need to be calibrated. The spectral response is calibrated with a bromine tungsten lamp and a monochromator in a conventional way. A novel method for calibrating the relative spectral responsivities of the polychromators is described in detail. A broadband pulsed Light Emission Diode (LED) is used, which has a spectral irradiance similar to that of the TS spectrum, and the LED can be driven in pulse mode with the pulse width similar to the TS signal pulse width of about 10–20 ns full width at half maximum. This new method allows for the calibration to be done after the polychromator is fully installed, and in situ system calibration can be easily performed, showing the advantages of accuracy, simplicity, efficiency, and flexibility. For ne measurements, absolute sensitivity calibration is done by Rayleigh scattering with argon gas. Formulas for calculating the plasma density from the calibration data and the polychromator signals from the off-laser wavelength channels are presented.

The generation of noise is harmful to worker’s health and the stable operation of the system. Due to the high speed and high head of the high-speed centrifugal pump, the noise performance needs improvement. Therefore, a high-speed centrifugal pump was studied both in experiment and numerical way. The noise performance of the high-speed pump is improved by optimizing the design of the impeller and the guide vane. The blade leaning is also applied as an improvement with forming the X-shape stagger between rotor and stator. Base on the numerical simulation, the characteristic curves and unsteady characteristics of different optimized cases are predicted. The optimal case is selected as the final experimental object to verify the effectiveness. Numerical result indicated that the pressure fluctuation at blade pass frequency (BPF) and the fluctuated radial force is the main component of the original pump. The noise spectrum distribution of the original pump is concentrated near the BPF in the experiment, which means that the rotation effect of the impeller is the main reason for the noise of the pump. By improving the hydraulic design, the pressure fluctuation and the radial force in the pump can be effectively suppressed. But many factors affect the pump noise, the improvement of hydraulic design is limited.