Optical mount

Optical mount is a mechanical part of optical systems, determining the alignment, robust, and precision. Any variation would affect optical performance in geometric and these factors are vibration, temperature drift, acoustic fluctuation, shock, chemical corrosion...etc.

Mirror is particularly sensitive to any fluctuation because any geometric variation would amplify the change of optical path. For example, angle tilt of mirror causes reflection angle double.

Aerospace field has very tough requirement - huge shock, wide temperature.....
NASA had a report about the the reliable test of the relationship between temperature and mirror mount.

• (1997) Angular Alignment Testing of Laser Mirror Mounts Under Temperature Cycling

This report compared commercial mount and NASA customer-designed mount to proof that NASA's mount is excellent and doesn't wast taxpayer's money. But their data still give us a sight into the temperature effect.
The test setup is to support stable test environment and focus on the temperature effect. Vacuum chamber eliminate the air flow fluctuation during heating. Laser beam reflected from mirror mount would tell the deviation as temperature changes. The distance is long above 3m to amplify the angular deviation, and PSD would digitize the data.

The comparison table shows below:

• movement - test the temperature effect(angular change as temperature change 1°C)
• hysteresis - test if the mirror position can go back its original as temperature remove

For commercial products, these optics component companies are well-known. NASA's conclusion praised Lee's mount for the lowest hysteresis performance and Newfocus for the lowest movement. (Lee's is part of LiNOS, and Newfocus is part of Newport)

It is very interesting in the comparison table. Temperature drift is very complicated and affected by many factors.  Here I just focus on major factors in optomechanical design:

• size
  Large size mount has small coefficient of temperature tilt. It's very contrary to our physical intuition - 1 degree drift has larger geometrical expansion for larger size. 
• material
  For thermal expansion coefficient: Al > Steel > Ceramic
  Data of coefficient of temperature tile generally match this relationship, but hysteresis doesn't. 
• structure
  Because the mirror mount is a composite part - consisted of screw, spring, and joint. The most common type for mirror is kinematic mount, which is easy to adjust. Simple structure would be good in test. Different material accessories also affect the overall performance.

It is not easy to design an excellent optomechanical mount because there are many factors are needed to consider from design to manufacture. From basic physical intuition, you just focus the isolated phenomena and can't help to verify which one is good.

此篇為我以前在學校時,研究光學系統架設,看到最值得好好思考的一篇文章, NASA報告並沒有給什麼決定性結論,但最初從物理直覺思考,就有錯誤偏見-不繡鋼就是比鋁好, 小的就是比較穩, 有鎖定機制的設計會較佳....etc

或許很多人的光學系統並不會這麼在意這些關鍵性的問題,因為精度上的要求並不高 (系統不大,光路不長, 沒有需要真空, 單純只看影像, 單純平面光路....)加上懶的思考以及科系專業視野狹隘,根本就不會在意這些問題

從NASA報告中

Heat Hysteresis以LiNOS的Lee's最佳(溫度回到原點,位置也差不多回到原點)

單位溫度影響以Newfocus最佳 (增加1度C,位置移動最小)

• Newport(併購了Newfocus)應該是在台灣大家最常耳聞的元件廠商,各種產品一應俱全,其不繡鋼的mount還設計的比別人好看,用不少spring去fix, 但是效果如何,還真的要實驗看看才知道。

• LiNOS的mount是用全鋁製的,體積也大,外觀一點都不現代,價格也高,但是performance就是不錯

最簡單是當你screw轉多少,再轉回多少,看光束是否能回到同一點

光機元件是包含不少元件的,像是screw, 結合, 彈簧....,所以NASA這實驗有很多因素要考量,並不是不繡鋼的熱膨脹係數比較低,就一定比鋁的mount好, 有可能整個mount的均勻性是不佳的(可能是製作上的問題),所以東邊熱脹多一些,西邊少一些,整體光路就跑掉了, 或是附件的元件影響了(像是橡膠材與主體的金屬材彼此就受熱影響不同)

機械這種東西並不簡單,是跟工藝技術有關,可以說是"Art"