Your sketch shows two measurements on the pump discharge nozzle. You have to declare consistent phase directions. When you move the accelerometer from the left side of the nozzle to the right side, then the accelerometer phase orientation changes by 180-degrees. My experience indicates that the correct phase is on the left side (0-degrees), and the phase on the right side was not corrected (unless you did) for mounting orientation. Applying the orientation correction to the right side would make the actual phase 0-degrees (180 - 180), and then both sides of the nozzle would be moving in-phase. The entire motion of the pump would then be clockwise rotation about the pump's primary base. For ODS tests, I define positive X-Y-Z directions for the structure using the right-hand rule. This can be important, for example when making axial measurements on motor and pump on opposite sides on the shaft coupling. In that case, one of the measurement points has to have 180-degrees subtracted from the measured phase. You cannot define pipe strain, as shown, in the vertical direction without at least 4 measurements (90-degrees apart) on the nozzle in the vertical direction. Nozzle loads are typically defined as tension/compression, bending, and shear (lateral/twist). You might draw conclusion from your data that there is a resonant structure (pump-pipe), but you cannot conclude there is excessive pipe/nozzle strain.