QQ客服 官方微博 反馈留言

题名	高功率光学元件波面功率谱密度分布检测
作者	杨相会
学位类别	硕士
答辩日期	2016
授予单位	中国科学院上海光学精密机械研究所
导师	沈卫星
关键词	高功率激光，中频波面误差，功率谱密度，仪器传递函数
其他题名	Measurement of Wave-front Power Spectral Density Distributions of High Power Optical Elements
中文摘要	用于惯性约束核聚变的高功率激光器为了尽可能地提高输出能量，对光学元件波面质量提出了很高的要求。波面误差中，空间周期位于0.12~33 mm之间的为中频波面误差。中频波面误差严重威胁高功率激光系统的安全运行，是光束非线性小尺度自聚焦的重要诱因，并且影响终端光斑质量。功率谱密度（PSD）函数是一种基于傅里叶分析的方法，可以量化各个空间频率区间的误差，被用于中频波面误差的评估。 基于高功率激光器对中频误差控制的需要，本文致力于光学元件波面功率谱密度检测方向的研究。使用干涉仪检测光学元件波面误差，通过傅里叶变换获得PSD函数分布。干涉仪固有的低通滤波特性会造成中高频信息的丢失，如何获得相对真实的PSD分布，是本文主要研究的问题之一。此外，对于干涉仪检测到的波面误差数据，怎样有效地提取出其中的中频波纹信息，也是本文主要论述的问题。具体包括以下几方面的内容： 1. 详细推导了PSD的定义及数值计算方法，分析了一维和二维PSD的区别和联系。通过对不同空域特征的光学元件进行检测，探究波面误差中频波纹特征和PSD分布之间的关系。实验结果表明：PSD分布上的尖峰是光学元件中频波面误差的反映，尖峰的空间频率和高度分别由波纹的周期和深度所决定。 2. 给出了PSD数值计算流程。介绍了利用双线性插值法和外推法对相位数据中的无效点进行插值填充。对波面相位数据添加汉宁窗，以减小边缘数据截断效应产生的吉布斯噪声，并对修正因子进行了推导。详细介绍了利用二维PSD求取一维PSD坍陷的原理和计算方法。在理论和实验上研究了二维和一维PSD和测量孔径尺寸之间的关系。 3. 为了探究不同干涉仪对中频误差的响应规律，获得相对真实的PSD分布，分别利用4D AccuFiz、ZYGO DynaFiz以及ZYGO GPI三台干涉仪检测周期性波纹和划痕样品，对一维实测PSD曲线进行了对比分析。实验结果表明不同干涉仪对中高频信息的响应度相差较大，干涉仪分辨率越高，响应度也越高。利用仪器传递函数（ITF）曲线对实测PSD分布进行了修正，获得了较为真实的PSD分布。但在某些频率范围内，标定的ITF曲线和实际响应出现差异，造成PSD修正值可信度下降。
英文摘要	In order to improve output energy as much as possible, the wave-front quality of the optical elements employed in high power laser for Inertial Confinement Fusion must be very good. The spatial period range of the mid-frequency wave-front error is 0.12~33 mm. The mid-frequency wave-front error threats the safety of high power laser seriously, which will cause nonlinear small-scale self-focus and affect the focus quality in terminals. Power spectral density (PSD) function is based on Fourier analysis, which can give different spatial frequency errors quantitatively, is used to evaluate mid-frequency wave-front errors. Base on the requirement of mid-frequency errors control in high power laser, this paper devotes to the study of measuring wave-front PSD of optical elements. The wave-front errors are measured by interferometers, and the PSD is calculated by doing Fourier Transform to wave-front data. Interferometers will lose some mid-spatial frequency wave-front information for low-pass filtering characteristics. How to obtain correspondingly true PSD distribution is one of the main problems studied in this paper. In addition, how to extract mid-frequency ripple information from measured wave-front error data effectively is another problem analyzed in this paper. The specific content includes the following several aspects: 1. The definition and calculation of PSD is derived in detail. The difference and relation between 1D PSD and 2D PSD is analyzed. The relationship between mid-frequency ripple characteristics and PSD distribution is discussed by measuring different spatial characteristic optical components. The results show that spikes on PSD distribution is the reflection of mid-frequency wave-front error, the spatial frequency and height of the spikes are determined by the period and depth of the ripple respectively. 2. The numerical calculation flow of PSD is given. Double linear interpolation and extrapolate interpolation is introduced to patch non-valid points. To reduce the Gibbs noise from edge discontinuities, the Hanning window is used to wave-front data and the correction factor is obtained by theoretical derivation. The principle and calculation of 1D PSD collapse of 2D PSD is introduced in detail. The relation between PSD and measured aperture size is studied in theory and experiment. 3. To study the responsivity of different interferometers to mid-spatial frequency errors and obtain correspondingly true PSD distribution, 4D AccuFiz, ZYGO DynaFiz and ZYGO GPI interferometers are used to measure periodic ripple and scratch samples, and a comparative analysis of measured 1D PSD curves is done. The results show that there are significant differences between the responsivity of different interferometers to middle and high spatial frequency errors, the higher the interferometer resolution is, the higher the responsivity of information in middle and high frequency becomes. The instrument transfer function (ITF) is used to correct measured PSD curves, and more incredible PSD distribution is obtained. However, the ITF curve may be different from practical responsivity in some frequency band, which will cause reliability decreased.
语种	中文
内容类型	学位论文
源URL	[http://ir.siom.ac.cn/handle/181231/16984]
专题	上海光学精密机械研究所_学位论文
推荐引用方式 GB/T 7714	杨相会. 高功率光学元件波面功率谱密度分布检测[D]. 中国科学院上海光学精密机械研究所. 2016.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。