Learning Forum

[boonthiam13]

Hey peeps, I have come across couple of journal articles in regards to simulation of centrifugal pump where MRF and Sliding Mesh are the two most common methods to perform steady-state and transient analyses, respectively. Interestingly, results from transient analysis (sliding mesh method) closely approximate the actual pump operating performance, and steady-state simulation usually only serves as initial condition for the transient simulation.nI have tried this method and found the same thing, which is by using transient simulation (of which initial conditions are obtained from steady-state simulation), performance of the pump could be closely approximated. On the other hand, results from steady-state simulation deviates a lot from the actual (experimental) results, for example, at a fixed flow rate, pump head obtained from steady-state simulation is more than two times than that of actual one.nMy question:-nWhy a transient simulation can give a close approximation but not the steady-state simulation?nThanks and looking forward to receiving valuable advices nCheers!nBoon n

[DrAmine]

Transient runs takes into account interactions, vane and wake effects, interferences between the vanes /profiles, unsteadiness of flow and strong interaction along the interface between rotor and stator.nnFrozen rotor holds true if and only if you can move along that rotating interface and you the flow is not changing:this is very ideal but an engineering way as the time to solution is lower than running unsteady!.nnSo the main flow is actually is unsteady but a steady approximation can sometimes bring us to the same level of accuracy but at less efforts:Sometimes!n

[boonthiam13]

ArrayThanks DrAmine. Clear to me now!nn

[DrAmine]

Welcome!n

[boonthiam13]

ArrayHi DrAmine, if the transient model is able converge at each time step, shall we still first run the stead-state analysis to serve as initial condition for the transient model? These two approach shall generate the same results?.Thanks!nBoo