– The water flow is of interest to know when for new construction or expansion of hydropower plants. For example, when building fishways the flow needs to be in such a way that the fish will find the fish ladder past the hydroelectric dam, says Anders Andersson at the research subject Fluid Mechanics at Luleå University of Technology.
Attraction water facilitates upstream migration
Anders Andersson has, as part of the EU project "From Coast to Mountain," looked at various solutions to improve upstream migration. The goal is to create a sustainable development of fisheries in Vindelälven and Ume River. Because many fish get stuck on the way to their spawning grounds, Anders, and SLU in Umeå, made simulations with attraction water and compared this with measurements of the abundance of fish.
– It's about creating an attraction water to help the fish past hydroelectric dams and other obstructions. The fish are attracted by high water velocities and therefore this needs to be a strong flow where the fish will migrate. We could see a connection between the turbulence in the water and the number of fish, however, the turbulence may neither be too high or too low, says Anders Andersson.
Simulations of water flow over rough surfaces
Besides free surface flow, Anders Andersson investigated flow over rough surfaces, such as walls in a blasted tunnel.
– Rough surfaces have a large impact on the flow, and ideally the resistance should be low. It is important to be able to describe the flow in a proper way to get realistic results from the calculations.
The challenge is to create an effective model for simulations. Too many measure points on the rough walls, so-called resolution, makes it more difficult to model and more expensive to simulate. Too few data points, however, gives an incorrect simulation.
– We discovered there is a big difference in how the flow field looks like if you go below a certain resolution. Therefore it is important to have enough data points, says Anders Andersson and continues:
– In the future, it would have been nice with a model that takes the geometric roughness into consideration, i.e., dissolve the large scale of the roughness of the walls, but numerically takes care of the small scales of roughness. That would lead to a more efficient simulation.