[Muzz]

I've been attempting simulations of unsteady flow in an urban environment. Using a k-e turbulence model and segregated flow model, I've been able to get good results for a steady solution.

However, when I try to use implicit unsteady to generate unsteady flow within my environment, the solution tends to the steady state form over large periods of time, with no transient properties. If I try to initialise the simulation with a steady state solution, there are still no changes with time. I'm still using the k-e model, which I realise by definition is time-averaged but shouldn't I still be seeing some kind of change?

[me3840]

Not necessarily; if the problem doesn't have enough transience in the physics then an unsteady solution will become steady given time. However, you could also have a coarse enough grid or timestep to filter out the unsteadyness artificially.

[Muzz]

Originally Posted by me3840

Not necessarily; if the problem doesn't have enough transience in the physics then an unsteady solution will become steady given time. However, you could also have a coarse enough grid or timestep to filter out the unsteadyness artificially.

Thanks for your response. I've been out of the office so apologies for the delay... I switched to DES with the Spalart-Allmaras model and achieved success, however, I use that term loosely. Unsteadiness was developed throughout the domain and it appears reasonable. However, there doesn't seem to be a lot of available data for validating unsteady urban flows so I'm still in the dark a little bit.