TOP > Report & Column > The Forefront of Space Science > 2014 > Computational Science in Fluid Mechanics

| 1 | 2 | 3 | Numerical simulation to next stage Are there any other opportunities to use advantages of numerical simulation other than the researches on numerical algorithms or physical models mentioned above? Finally, I would like to introduce our attempt “to advance numerical simulation to the next stage�Ecurrently under consideration. The behaviors of realistic performance of designed fluid machinery and the underlying flow physics are affected by numerous intrinsic uncertainty factors such as flow conditions, wall-surface conditions, object-shape variation by fabrication error or deformation. Moreover, the simulation is effected by uncertain factors caused by physical models (e.g., turbulence model, chemical reaction model, etc.). For this reason, we believe that, if we were able to pull out of the conventional deterministic numerical simulation conducted under given conditions on given physical models and extend it to enable quantitative evaluation of influences of uncertain factors existing inherently everywhere, we would better understand the realistic fluid phenomena and contribute even more to design of actual complex systems. Based on our recent research, we are proposing the uncertainty-quantification method, which allows us to accurately quantify how the fluid phenomena or performance of fluid machinery respond to various uncertainty factors. Figure 3 shows an case of how the uncertainty involved in the freestream velocity affects the pressure distribution around transonic wing using the uncertainty-quantification method coupled with numerical simulation. The proposed stochastic analysis differs from conventional deterministic numerical simulation (blue line in lower left figure in Fig. 3) in that the uncertainty analysis allows us to quantify the statistical-mean value and upper/lower limits of its reliability and, furthermore, the probability-density distribution comprising the upper/lower limits of reliability in the complex uncertainty systems. Figure 3. Numerical simulation around a transonic wing under the uncertainty involved in the freestream velocity (assuming normal distribution has around 0.6% deviation against the freestream velocity) Evaluating the impact of the uncertainty involved in the freestream velocity on the pressure distribution around the wing using the uncertainty quantification method based on Kriging model. We believe that our new research on the stochastic analysis not only improves the conventional deterministic-numerical simulation and assesses the quantitative impact of various uncertainty factors but also suggests the next possible stage of numerical simulation. Thus, the analysis of the impact of various uncertainty factors on the quantities of interest for realistic analysis and design of complex systems leads us to: an essential part of validation; a rigorous measure of reliability, particularly important for space science; suggestion of sensitivity and priorities; understanding of fluid mechanics determining such sensitivity and priorities; and contribution to robust design and optimization. Further, we consider that our stochastic approach may apply to wider academic fields having uncertainty. We are also continuing our research from this perspective. Conclusion In this article, I introduced a part of our research, even though subjective, on the relationship between space science and fluid mechanics and how computational fluid dynamics research contributes to the space science. The advancement of fluid mechanics and computational mechanics researches is an important element to conduct space science and space engineering. Numerical simulation can be a very powerful tool for understanding of complex and beautiful fluid-dynamics phenomena and their wide range of uses to our lives. I hope that, as a researcher of fluid mechanics and computational mechanics, I can contribute to the further development of science. (Soshi KAWAI) | 1 | 2 | 3 |

• Back to top