Back to the programme printer.gif Print

Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'How does the wind blow behind wind turbines and in wind farms?' taking place on Tuesday, 11 March 2014 at 16:30-18:00. The meet-the-authors will take place in the poster area.

Ahmed Elsayed Uppsala University, Egypt
Ahmed Elsayed (1) F P Bahri Uzunoglu (1)
(1) Uppsala University, Hehia, Egypt

Printer friendly version: printer.gif Print




A very simple case will be addressed in our research. simulation will be performed for a mounted cube over a very smooth terrain. For the simulation, WindSim is used as a CFD tool. , we are going to validate the obstacle feature in WindSim to assess its performance and to find out what the conditions and features that should be implemented in WindSim to be used in urban simulations.


As the modern urbanization is increasing, the energy demand is increasing as is very important to understand the flow behavior maximize the out power from the small scale wind turbines throughout assessing the wind resource .a non-linear models should be used to solve such cases as the linear models have a limitation to deal with such cases.WindSim is a CFD tool that solves RANS (Reynolds averaged navier stokes) equations. This tool was developed by a company in Norway to be used in generating a Wind atlas for a complex terrain and then they developed the programs and added more features for the commercial purposes.For complex terrain, a non-linear models should be used to solve such cases as the linear models have a limitation to deal with such cases. A numerical study that has been conucted by solving RANS equations for a surface mounted cube by byY.Tominaga In 2009, will be followed and used in our study to compare our results, also experimental results for the same case by Castro & Robin in 1976(1), will be employed. The following figure (Figure -1) shows the layout of the domain dimensions and boundaries conditions that are set for the study. As it is obvious from figure1, it is very simple case as there will not be height contours or roughness lines or any complex situation. We performed a validation study for obstacle feature in Windsim to validate the feature. A domain was built with the same conditions showed before. Default setting in the program has been used (wind speed above boundary terrain=10m/s, Coupled method used and Standard turbulence model (SKE). After validating study is done, the same setting used in the experimental and numerical study will be applied. Then results will be compared for several turbulence models.

Main body of abstract

a. Convergence and Validation study
Convergence study has been conducted in order to reach convergence and it means that the results are no longer depending on the number of cells (Grid independency). So, we need to start with a number of cells and after that we need to follow the wind speed for instance in a specified place to find out if the wind speed is still changing or not with increasing the number of cells. Figure 2 shows the evolution of the flow behavior with increasing grid cells until reaching convergence. we started with 300,656 cells and then we increased the number of cells incrementally and in the mean time we were watching the wind speed values at a certain place as discussed before. Figure 3 shows how wind speed is started to be steady after reaching grid in-dependency as with increasing the number of cells the wind speed for at that point will no longer increased or decreased significantly . Figure (4) exhibits the vertical wind speed profile and turbulence intensity at two different points, at the center of the cube and behind it by 5 meters. It is cleared that the cube is affecting the vertical wind speed profile above the cube and behind it , also it is very turbulent area behind the cube as shown also in figure 5. At the top of the cube the turbulence intensity is not higher as behind the cube and the wind speed is increasing vertically at the same point as we go away from the cube. Also, as mentioned before it is a validation study for the block file feature in WindSim as we do not know the performance of WindSim in such cases and also according to survey there is no much effort in that regard and we tried first to validate the feature in order to simply understand the problems that might faced during the simulation and the default setting was applied to avoid any conflicts in the simulations and later the setting of the employed cases will be set for the simulation
b. Final simulation
After validating the feature in WindSim ,We are applying now the setting of the experimental and numerical studies which we have indicated before. Four different turbulence models will be used( Standared Turbulence model (SKE),RNG model, k-e with launder and Yakhot and the Realizable k-e model) , Reynolds number will be maintained same with other two studies (RE= 1.1X10^4) and the same inlet vertical speed profile should be employed. The primary simulations are very demanding as we are trying to reach convergence first. Simulations now are being running to achieve the convergence in the mean time, we are communicating with WindSim Team support in order to advise us regarding to our case issues , also sometimes, we contact Pheonix support team for advising. Both team now are working with us to find out the Y+ value in our case,because we should make sure that we are in the right range of Y+ in order to change the grid configuration to suit our case.Thus we are working now to visualize Y+ values by using WindSim or may be another plotting program will be used to visualize the values.


We have a primary results regarding to our simulations. Most of them conducted with the validation and convergence study as we just finish them and the recent step is running simulation with same setting of the employed researches. Although the problem seems very simple but it is very complex shape for the flow to move around it . we did very extensive simulations to reach convergence with the default setting in WindSim and Reynolds number was very high. Simulations demanded several iterations to reach convergence as seen in figure6. We started with the same number that is used in the study we are following almost 300,656 cells and reached 2 million cells. To construct an obstacle in WindSim ,it needs a user with good knowledge with WindSim. So you can build the required grid correctly and same as possible to the one employed in the study that we are following.
For the final simulation in which we apply the same setting in both cases experimental and numerical ones, convergence was difficult with the same Reynolds number and we took that the cube height 10 meter so the wind speed should be 0.023858 m/s. So, simulations now are running with small scale ( like to take smaller value for the cube height ) .We are trying now with WindSim developers to continue in that research and also phoenix team. Recently, we are trying to make sure that the grid is being used is having the favorable Y + value. Y+ value in our case should be around 30 -100.

Learning objectives
After finalizing the case that we recently have, there are two researches that will be conducted in WindSim and depending on our recommendation on implementing Windsim. The first case is conducted to several cubes arranged in different patterns to complicate the situation more and more to study the wind resources for such cases. The second case is with real data in a place in Sweden. We have a real data for a real canyon of buildings

1- Y.Tominaga, T.S., 2009. Numerical Simulation of dispersion around an isolated cubic building. ALSEVIER, (43), pp.3200-10
2- I.Eliasson, B.O., 2006. Wind fields and turbulence statistics in an urban street canyon. Atmosphric environment , (40), pp.1-16.