07.02.2013 Views

Optimization and Computational Fluid Dynamics - Department of ...

Optimization and Computational Fluid Dynamics - Department of ...

Optimization and Computational Fluid Dynamics - Department of ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

148 René A. Van den Braembussche<br />

Fig. 6.1 2D view <strong>of</strong> the 3D flow at the exit <strong>of</strong> a turbine stage<br />

lyze the fluid flows in the same way Finite Element Analysis (FEA) is used<br />

for stress predictions. They provide detailed information about the 3D flow<br />

around existing blade shapes <strong>and</strong> constitute an attractive alternative for detailed<br />

flow measurements. Complex flow phenomena can now be studied in<br />

what is called “Numerical Laboratories”. Although this has resulted in a<br />

drastic decrease <strong>of</strong> the number <strong>of</strong> prototype testing, there are still two problems<br />

that prevent a more efficient use <strong>of</strong> CFD in the turbomachinery design<br />

process.<br />

The first one results from the difficulty to analyze 3D flows on 2D screens or<br />

drawings. 2D vector plots are only a poor representation <strong>of</strong> the reality. They<br />

can be very misleading as they may suggest that the flow is penetrating the<br />

solid walls (Fig. 6.1). Synthetic environments, also called virtual reality, are<br />

very promising in this respect. These techniques are not only applicable to<br />

mere computer games but will become part <strong>of</strong> everyday reality for engineers<br />

in the next decade [16]. Designers will walk inside blade rows <strong>and</strong> diffusers to<br />

inspect the complex 3D flow structures by tracing the streamlines <strong>and</strong> to find<br />

out what geometrical changes may be needed to improve the performance.<br />

The second problem relates to the abundance <strong>of</strong> information provided by<br />

the NS calculation. The output <strong>of</strong> an NS solver contains all the information<br />

needed to improve the performance. However, it does not provide any information<br />

on what modifications are needed to reach that goal. Three velocity<br />

components, the pressure <strong>and</strong> the temperature in typically 10,000 points (2D<br />

flows) or in more than 1,000,000 points (3D flows) are more than what the<br />

human brain is able to grasp <strong>and</strong> fully exploit in new designs. Most <strong>of</strong> the<br />

available information remains unused as the designer will <strong>of</strong>ten calculate a<br />

global parameter to find out if one geometry performs better than another.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!