I have been trying for some time to set up a simulation that resembles that of the study conducted by a member of the Sim Scale team located here. I am relatively new to using CFD and have received really good and helpful advice which i am very grateful for.
I have tried my best to replicate the analysis conducted by @dlynch in the paper previously linked but I am unsure on how i can verify if my results are valid when compared (i understand it will not be exact). I have looked into checking my mesh quality and all seems good to me, when i compare the results for the forces and moments for the base point i have set they do not perfectly match when considering forces and moment in XYZ, in fact some values are not similar and off by magnitudes of 10.
I would really appreciate it if someone could take a look and provide me with useful information on how to improve my work so far.
Project located here
I have wasted many hours of computational time trying to resolve issues my self and have now admitted defeat. Thank you in advance
Your flow region is too small, which causes the walls to be too close to the building. I would leave the space of 3 heights at the top, sides and upwind and 10 downstream.
Your ABL definition is strange, shouldn’t it just vary with the height?
Hi @ggiraldof
Thank you for you help with this project to date. I have updated the flow domain and based the values used off a research paper i found.
In terms of the ABL definition am i correct in saying it must be defined in 3D, We discussed the ABL definition in the past here
Have i misunderstood the previous discussion?
Yes i have read this article previously, I believe when last going through this you had stated that i needed to define the X and Y coordinates in the table? Defining the X as constant at the location of the domain inlet,and defining Y at the lowest negative value, 0 and the upper positive value?
I have made the amendments you suggested and re-meshed the model.
Is the new ABL input acceptable as well as the new domain? (the domain was created based off a research paper so not identical to what you suggested)
Further to this i checked the mesh in Mesh quality and noticed the non orthogonality was 76, which is too large when compared to the recommended values.