Recently I’ve been having trouble finding the source of some variance in WSE results. I am using Windows 11 (version 22H2).
I created a 1D model in HEC-RAS 6.3.1, and for all cross sections, I initially used a Manning coefficient of 0.03. At that time, I considered that my bank station would be at the extremes of all sections. In this way, the central channel’s Manning coefficient was valid for the entire section.
In this situation, I got a certain WSE result, as shown in the image below.
Image 1: linkpicture.com/q/image_381.png
Later, to refine my model, I corrected my bank stations to the proper places, keeping the Manning coefficient identical for the margins and for the center of the channel, without any change in relation to the previous model other than changing the position of the bank stations.
To my surprise, WSE has changed slightly. The image below shows the same section in this new scenario.
Image 2: linkpicture.com/q/image-1_159.png
Initially, I thought that this could have occurred due to the increase in the horizontal discretization of the velocity field, since I now had three refinement regions (left margin, right margin, and central channel), whereas, in the first model, I had only the central channel (as the bank stations were at the ends of the sections). This variation in velocity discretion can even be perceived in the two images above.
But by changing the number of section divisions to have the same velocity field as the first case, the variation in WSE remained. That is, the discretization did not produce that variation in the WSE.
I know the variation in WSE exemplified in the images above is minimal. But in other sections of the model, the change in WSE was more significant. In particular, the discharge capacity of my entire structure changed by 6.7% after I changed only the position of the bank stations.
Therefore, my question is the following: what produced this variation in the WSE when I just changed the position of the bank stations, keeping precisely the same Manning coefficient for the whole model and all the other parameters?
I sincerely appreciate any form of support you can provide me.
Best regards,