Smoothing the pipeline noisy point positions by corotational beam spline
Proceedings Publication Date
Presenter
Prof. Igor Orynyak
Presenter
Author
Igor Orynyak, Roman Mazuryk, Dmytro Koltsov, Yurii Kuznetsov, Jurius Lubinskis
Part of the proceedings of
Abstract

The inline inertial measurement unit (IMU) produces X-Y-Z positions of pipeline points and provides information about the local deformation of pipelines. Treatment of these results for extraction of the true deformation is complicated by two factors. The first is related to systematic measurement errors due to, for example, vibration of IMU inside the pipeline; and the second one - is with embedded curvature of the axis due to availability of the factory and field pipe bends.

This paper describes the application of a smoothing technique based on corotational beam spline, CBS, and contains several ideas related to the specific pipeline. First, at the beginning of the pipe bend and its end, the spline curvature is taken as a discontinuous function and its jumps in the specified plane are equal to the bend curvature. This drastically improves the predicted deformation, its variation, and continuity even beyond the bend section. Second, the points of measurement are considered as the compliant springs – the lesser their compliance, the smoother the calculated pipeline axis, and the smaller are calculated deformation. To find the necessary “springs’” compliances we perform the iterative procedure when the mean value of the gap between the pipeline axis and spring position attains the prescribed value – mean error of measurement.

Several examples and challenges of real calculation are given. First, the “clueing” of overlapping pipeline sections, say of 2-3 km in length is considered. In this case, the spring compliances should be taken the same, then different neighboring sections meet together very continuously. In case, when we compare the same sections inspected several times, the number of the points of measurement and their placement might be different. In this case, the graphs of curvatures are smoothed differently. So, the compliances are corrected by accounting for the distances between the points.

To view the video or download the paper please register here for free

You already have access? Sign in now.