| 摘要: | 本研究以外腔二極體雷射做為干涉儀的光源，利用其波長可調與易於操控的特性，來實現絕對距離干涉術並探討其應用。 此論文研究的絕對距離干涉術有兩種：波長掃描干涉術與可變式合成波長干涉術。 所有實驗的基本架構為波長可調外差干涉儀，光源是外腔二極體雷射，波長可從772 nm連續變化至792 nm。此干涉儀以外差技術量測干涉相位、用波長儀決定雷射波長、和採取間接量測方式獲得空氣折射率。波長可調外差干涉儀搭配不同的光路，有不同的應用。 波長掃描干涉術紀錄因波長掃描引起的相位變化，即可推算出絕對光程差。在能力測試實驗中，光路安排為Michelson干涉光路。此能力測試包括測距與尋找零光程差位置。測距範圍50 mm以內，量測誤差小於1 mm，而解析度優於200 nm。重複交互使用波長掃描干涉術與單波長干涉術，最終零光程差位置的準確度可達到nm等級。在大階高量測應用中，改採差動干涉光路來降低環境擾動，及使用準確度較高的波長儀及相位計。當待測階高小於50 mm時，量測偏差值低於100 nm。 實施波長掃描干涉術時，波長必須是連續掃描，但跳模會縮短外腔二極體雷射的波長可用範圍，進而降低量測準確度。可變式合成波長干涉術是本研究針對波長掃描干涉術缺點，所提出的新方法，此干涉術以外腔二極體雷射產生一系列由大到小的合成波長，依序對待測光程差進行量測，由這些合成波長及其對應的小數條紋數即可推算出待測光程差。因為不要求波長連續變化，所以可充分利用雷射增益曲線，而且不需要條紋計數器。 驗證可變式合成波長干涉術的實驗有兩個，第一個實驗是重複先前的大階高量測，在量測範圍小於25 mm時，準確度約80 nm。第二個實驗是不透明平行板的厚度量測，為了免除扭合問題，使用了改良型雙端面干涉光路，10 mm塊規的量測偏差約480 nm。 The main objective of this dissertation is to study the absolute distance interferometry with a tunable external cavity diode laser (ECDL), and its applications. Two different ADIs are investigated in this study. They are wavelength scanning interferometry (WSI) and variable synthetic wavelength interfeometery (VSWI). The basic setup of all experiments is a wavelength-tunable heterodyne interferometer (WTHI). The light source is an ECDL, whose wavelength can be continuously tuned from 772 nm to 792 nm. The WTHI measures the interference phase by heterodyne technique, determines the laser wavelength with wavelength meters, and obtains the refractive index of air by indirective measurement method. For various applications, the WTHI is with different optical layouts. WSI determines an optical path difference (OPD) by directly counting the interference fringes as the wavelength is scanned through a known change in wavelength. In the experiment of testing the capabilities of WSI, the optical configuration is a Michelson interferometer with retrorefletors. The tests include distance measurement, resolution verification, and the identification of central fringe. The error and resolution are nearly 1mm and 200 nm, respectively, when the distance to be measured is less than 50 mm. By alternately employing WSI and single wavelength interferometry, the position accuracy of zero OPD is about several nanometers. In the application of measuring large step heights, the Michelson interferometer is replaced with a differential interferometer, which can reduce the influences of environmental disturbances. Besides, wavelength meter and phase measuring instruments with higher accuracy are used. Three gauge blocks of different lengths, 5 mm, 10 mm and 50 mm, are individually wrung on a steel plate to simulate large step heights. Comparing the results measured by the proposed interferometer with those by the gauge block interferometer reveals that the accuracy is around 100 nm. To implement the WSI with high accuracy, the laser wavelength must be continuously scanned over a wide range. The appearance of mode hops shortens the useful range of an ECDL although the gain bandwidth of the diode is very wide. This study describes a new method, VSWI, which also has no fringe order ambiguity problem but does not require that the laser be continuously tuned. An unknown OPD is sequentially measured at a series of descending synthetic wavelengths. Every synthetic wavelength is a combination of a varied wavelength and the initial wavelength of the ECDL. The OPD is determined following a succession of optical path difference calculations, in terms of the synthetic wavelengths and measured synthetic fractional fringes. The uncertainty in the measurement is gradually reduced as the measuring synthetic wavelength is progressively reduced. The capability of VSWI is confirmed in two experiments. One is to repeat the experiment of measuring large step heights. The results reveal that the uncertainty in the measurement is approximately 80 nm when the measured height is up to 25 mm. The other one is the thickness measurement of opaque plane-parallel parts. The optical configuration is a modified double-ended interferometer. Because of the ring-like measuring arm, the test part dose not need to be wrung on a platen. The results indicate that the accuracy is about 0.5 mm for a 10 mm gauge block. |
