Vibrationsinducerad utmattning i rörsystem

Vibration-induced fatigue in industrial piping is commonly assessed with the Energy Institute (EI) velocity diagram, a tool that is widely regarded as conservative, albeit not precisely quantified. This thesis investigates both the origin of the diagram and the degree of conservatism built into it, while also identifying the key parameters that practitioners should account for when using the method. The work set out to reproduce the Energy Institute diagram using three progressively more detailed approaches: an analytical Euler–Bernoulli beam model, a beam-element FEA model, and a shell-element FEA model. From each model’s solutions the stress-velocity proportionality was extracted together with their respective eigenfrequency. By assigning the fatigue endurance limit as the target stress, representative of the weld fatigue classes F2 and D for the methods nominal stress and hot-spot respectively, a maximum allowable velocity was calculated. For the analytical model and the beam model the maximum allowable velocity is approximately 110 mm/s RMS, independent of frequency. This value also coincides with the upper bound to the extracted velocities from the shell element model. Capturing the effects of flexible joints and heavy valves, the shell element model reveals a spread of data points of up to two orders of magnitude, with the greatest variation in the low-frequency range. The parameters identified to cause this variation were the valve-to-pipe mass ratio and the flexibility of the joint. The results from the study indicate that the inclination of the categorization lines as well as the width of the “Concern” region of the EI diagram are the results of the non-idealistic beam properties, such as concentrated masses on the span and flexible joints. Statistical lower envelopes fitted to the data show that the Energy Institute Concern and Problem lines correspond closely to the twentieth and first percentiles, respectively, thereby covering about eighty and ninety-nine percent of realistic geometries. The study confirms the guideline’s overall validity and proposes extending its applicability via scaling factors for weld quality, cross-section geometry, and very-high-cycle fatigue conditions.