f4ephantom
Whoa...CFD got something right...
WARNING: Text below exposes the extreeme nerdiness of the author. If you are content in the belief that I am a sane and stable human being, ignore this picture and do not read any further.
So, this collage kindof sums up a small portion of my wind tunnel tests this past semester. While this wasn't the goal of the project, I can't help but sic the computer on what I'm doing. So, I put OVERFLOW to work.
This is Mach 4 flow over a cone-flare model (the flow is from right to left...European style, I know...but that's how my camera was set up and it would take too long to make it 'right'...). The top image is the computational grid, the second image is the computed prediction of the density of the air, the third picture is a schlieren photo of the model, and the fourth has the CFD picture scaled and superimposed on the schlieren picture.
(the regions in the schlieren picture change color indicate a change in density of the air, so the fact that the lines in the photo nearly line up with the color boundaries on the CFD prediction indicates that the computer got it right)
This is run in a low Reynolds number wind tunnel and the model is small, so laminar boundary layers are probably reasonable. The CFD seems to think so anyway. My jaw hit the floor when I saw the comparison. My (albeit brief) experience with this type of shock-boundary layer interaction has been that the computer cannot predict the separation location right...and it nearly nailed it here.
Hopefully I'll get around to posting some more cool looking schlieren pictures from this project...I'll try to spread out the dorkfest though...I don't want to get anyone sick, just in case it is contaigous...
Whoa...CFD got something right...
WARNING: Text below exposes the extreeme nerdiness of the author. If you are content in the belief that I am a sane and stable human being, ignore this picture and do not read any further.
So, this collage kindof sums up a small portion of my wind tunnel tests this past semester. While this wasn't the goal of the project, I can't help but sic the computer on what I'm doing. So, I put OVERFLOW to work.
This is Mach 4 flow over a cone-flare model (the flow is from right to left...European style, I know...but that's how my camera was set up and it would take too long to make it 'right'...). The top image is the computational grid, the second image is the computed prediction of the density of the air, the third picture is a schlieren photo of the model, and the fourth has the CFD picture scaled and superimposed on the schlieren picture.
(the regions in the schlieren picture change color indicate a change in density of the air, so the fact that the lines in the photo nearly line up with the color boundaries on the CFD prediction indicates that the computer got it right)
This is run in a low Reynolds number wind tunnel and the model is small, so laminar boundary layers are probably reasonable. The CFD seems to think so anyway. My jaw hit the floor when I saw the comparison. My (albeit brief) experience with this type of shock-boundary layer interaction has been that the computer cannot predict the separation location right...and it nearly nailed it here.
Hopefully I'll get around to posting some more cool looking schlieren pictures from this project...I'll try to spread out the dorkfest though...I don't want to get anyone sick, just in case it is contaigous...