Search

An angle sensor can be used to evaluate profiles without any shape references. We regard it suitable for evaluating a large profile and consider a gyro as an angle sensor for evaluating a profile larger than 100 m with an accuracy of better than 1 mm. A gyro can evaluate profiles without restrictions in span or direction; however, angles detected by a gyro typically fluctuate unacceptably for our purpose. We demonstrate that periodical reversal measurement by flipping a gyro is effective in reducing the effect of the fluctuation. Then, we rotate the gyro for continuously realizing the reversal, where the angles of the gyro’s rotating axis against the earth’s rotating axis can be derived without being affected by the fluctuation, and can be used as an angle sensor. Here, we consider a new method using gyro signals rotating around four orthogonally aligned axes. This can improve the accuracy of the derived angles by eliminating the effects of the gyro’s scale factor as well as the fluctuations.

Profile evaluation by detecting tangential angles of the profile is competent for large objects because it inherently requires no reference, which is difficult to define with sufficient accuracy as the object becomes larger. We considered using a gyro for detecting the angles instead of an inclinometer or an autocollimator, which are conventionally used as angle detectors. A gyro can detect angles without angular reference; therefore, profiles can be evaluated without the limitation of a reference. However, angles detected by a gyro generally have considerable fluctuations to ensure accuracy in the μrad range, which is the same level as a highly precise inclinometer. In this work, we adopted a periodic reversal measurement using a rotating mechanism to eliminate fluctuations. Analysis and experimental results show that the angles of the gyro’s rotating axis against the earth’s rotating axis can be derived from the angular signals of two gyros rotating in counter directions, and that this method is effective for reducing the influences of fluctuations.