Michelson interferometer has been applied to many problems such as tiny length

measurement, coherence length of a laser, thermal dynamic flow, flatness of plane plate and thickness of thin film. Many interferometer approaches have been derived from the Michelson interferometer. In other words, the Michelson interferometer can be modified for use in various types of measurement studies [1-3]. It has been proved especially useful for detecting vibration patterns. In this case, for example, the usual procedure for checking the vibration isolated ability of a holographic table is to employ a Michelson interferometer on the tested table. The phenomena of the Michelson interferometer is simpler to describe and can be easily observed. Many modified Michelson set-ups, such as a fringe stabilizer applied in the long-time exposure holographic experiment, are used to improve the vibration problems of a system. The present paper differs from the above work in that its purpose is to demonstrate the use of a modified Michelson set-up in checking a rotating vibration isolated system.

An interesting comparison of our experimental set-up is the Michelson-Morley

experiment[4] ( in 1881, to investigate the possible existence of ether drift ). With the apparatus on the rotating concrete table and a ten meter optical path of two beams, no fringe shift occurred as great as 10 % of the predicted value (2/5 l ). We can therefore imagine the difficulty of the mechanical vibration isolation of the Michelson-Morley experiment. Here we also used the Michelson interferometer in a rotating object condition but the total dimension of the experimental set-up was much smaller.

Usually the space of the base of the vibration-isolation system is too small to accommodate all the components of the Michelson interferometer and Helium Neon laser light source together. Although a diode laser mounted on the rotating table may solve this problem, a diode laser (with long coherent length and good quality collimator) which is suitable to be a light source of a Michelson interferometer is expensive. So we use a single mode fiber; as in Fig.1, to decrease

the dimension and increase the flexibility of the measuring system. The laser beam is transmitted through a single mode fiber to the rotating vibration isolated system. The fiber encircles the stage loosely so that the stage rotates without twisting it. The direction of rotation is counterclockwise, and conversely the fiber encircles the stage clockwise. So the object can rotate and the fiber loosely without winding. This set-up has potential for use with more tiny structures.

In a Michelson interferometer the mirror 1 has, by reflection on the semireflecting side of the beam splitter, an image in mirror 1'. Everything behaves as if the interferences were produced by an air wedge between mirror 1' and mirror 2. If the air wedge is quite large (usually larger than 2 mm, which depends on the distance from the point of the light source to the two mirrors) and image mirror 1' is exactly parallel to mirror 2, we can observe the rings of the interferometric fringes are perfectly circular. If the image mirror 1' is inclined to mirror 2 and the air wedge is very thin, the rings become hyperbolic.

We can see that the Michelson interferometer with hyperbolic fringes will be powerful in providing various vibration information. According to the image of the wireless CCD camera, we can obtain the angular inclination movements of pitching, rolling, and yawing of the mounting. An air bearing worktable of a roundness measuring machine is used to check the accuracy and sensitivity of our measuring system.

There are many devices for roundness measurement but these seldom mention how to check the vibration isolated quality of the roundness measurement machine. In the workpiece rotation roundness measuring machine, a small worktable rotates about the reference axis, the spindle with the stylus is stationary, and the workpiece is mounted on the worktable rotating at a speed of several revolutions per minute. Roundness is calculated from measurement data as long as the electronic stylus displacement falls within the full scale of the meter for each magnification. Error is involved in the measurement result if the vibration-isolation system of the worktable of the roundness measuring machine is ineffective. If the mirrors of the Michelson set-up are not rigidly mounted on the worktable, then motion of worktable will shake the mirrors and this would cause the interference pattern to change. So a modified Michelson interferometer will help to understand the vibration-isolation conditions of the rotating worktable. With this novel set-up we obtained the excitation information of the system. To test the results, measuring data upon the same rotating isolated table with an electronic sensor was compared.

index