Fluid Induced Vibration


A Natural Vibration Mode


Frequency Content Of Fluid Flow


Fluid Induced Startup Stress Transients

Integrated into the design team, Engenya provided design guidance and support for the development of multiple generations of an 11 metric ton mobile chemical battery to be deployed for energy storage from renewable energy generation plants.

Static, dynamic and fluid-structure interaction analyses were conducted to evaluate the design in transport loading situations (click here for more details) as well as in operational conditions. One particular concern was fluid induced vibration during startup or operation that could result in any one of the four PVC piping system branches being damaged.

Two approaches were taken: an indirect approach where structural dynamic characteristics of the fluid-filled piping structure are assessed against the fluid dynamic characteristics to identify potential for vibration being induced; and a direct approach where the fluid and structural solutions are coupled directly and the response is calculated. Both methods are accepted but there is a potential for discrepancies. By running both and aiming for matching answers a higher level of confidence could be achieved.

In the first instance, each of the piping branches was analyzed for its natural modes and frequencies in a fluid-filled state. At the same time, a CFD model of the flow transitioning from shut down to operational conditions was run and the frequency content of the flow surface pressure for the individual segments of the branches extracted. A comparison of mode shapes and frequencies with the fluid frequency content at the relevant locations was then conducted to identify potential risk of vibration.

In the second instance, a coupled model was set up and run. In this case, the loads acting on the structure and vice versa are calculated directly through the solver coupling, and the response is obtained in a single step. The coupling set up can influence the response and care must be taken to ensure artificial effects are not introduced. Comparison with the first approach helps address this. The advantage of the second approach is that local effects are identified directly even if the computational cost is higher.

Lockheed Martin currently markets this under the GridStar trademark.