I am working with existing accelerometers attached to the shell of a simply supported cylinder. Accelerometers are oriented radially at +/- 45 degrees from horizontal. There is a forced vibration on the cylinder in the vertical and horizontal directions and the cylinder is also exposed to torsion. I don't know the input forced vibration, only the acceleration and velocity measurements I get from these two accelerometers. My ultimate goal is to calculate the centerline displacement of the cylinder from these two accelerometers so I can apply those loads on an FEA model. With vertical vibration, the vertical displacement is the accelerometer measurement / cos(45deg). With lateral vibration, the lateral dispacement is the accelerometer measurement / sin(45deg). If I have vertical and lateral vibration and then add torsion how do I calculate the vertical and lateral acceleration or velocity of the cylinder centerline. I think the torsion generates a vertical and lateral displacement at the accelerometer, which adds to the accelerometer measurement from the vertical and lateral vibration . I think I have to subtract the displacements due to torsion from the accelerometer measurement to get the centerline motion. Need help to understand if my approach is correct and how to generate the centerline vibration.