主动凹坑变形湍流减阻控制方案研究
|
摘要
湍流边界层壁面高摩擦阻力的产生与近壁区流向涡密切相关,利用MEMS智能蒙皮技术控制流向涡从而降低湍流摩擦阻力是湍流减阻控制研究中的前沿课题。凹坑变形由于可以对流场产生局部干扰,并能在连续表面下工作,因此被认为是MEMS智能蒙皮技术中一种有应用前景的微制动器模型。本文利用槽道湍流的直接数值模拟,对基于凹坑致动的湍流减阻主动控制方案进行研究。
本文研究的控制方案主要包括反向控制、次优控制和神经网络控制。为了能够深入研究主动控制的特性,本文以连续壁面吹吸控制方案作为带有凹坑的可变形壁面控制的指导方案。在连续壁面吹吸控制方案下,本文对比了反向控制和次优控制的减阻效果,通过改变次优控制经验系数比较得到了次优控制在该方案下所能达到的最佳减阻率。
为了能尽量降低计算量和减小输入信息量,需要对主动控制在物理空间的表达进行分析。本文研究了次优控制权函数在物理空间的分布特性,通过对基于壁面连续吹吸致动的次优控制的直接数值模拟,选定了以控制点为中心在展向一字形的截断区域作为流动信息的获取范围,并通过对次优控制和神经网络控制的表达式比较,将次优控制这一选定方案类比应用在神经网络的控制上。对于凹坑致动,在连续壁面吹吸控制方案的指导下本文利用直接数值模拟对反向控制、次优控制和神经网络控制进行了分析。对于传感器-凹坑致动实用模型的次优控制,利用展向一字权函数截断区域和平均重构流动信息的方案,本文得到了12%~13%的减阻效果,减阻效率为η= 11.44。通过对湍流统计量和相干结构的分析,证实了这种控制方案能够在一定程度上削弱湍流近壁区的条带结构和涡结构,近壁区的湍动能和雷诺应力均受到抑制,优于同样模型下的反向凹坑控制方案。对于神经网路控制也有类似的结果。
The generation of high skin-friction over the turbulent boundary layer is close related to the near-wall streamwise vorticity. Applying MEMS intelligent skin technology to control streamwise vorticity in order to bring down the turbulent friction is the leading issue of turbulent drag reduction control research. The transfiguration of dimple could locally disturb the flow field and also work under the continuous surface, hence, it is considered as the preferred actuator model for MEMS system. In the present study, investigation of flow field with dimpled transfiguration initiative control is carried out by direct numerical simulation of turbulent channel flow.
Initiative control discussed in present study includes opposite control, suboptimal control and neural network control. In order to make a further investigation to initiative control, present study uses continuous wall blowing and suction model as the inductive scheme to the dimpled transfiguration control model. By the training of continuous wall blowing and suction model, the comparison of drag reduction effect between opposite control and suboptimal control is given. By altering the experienced coefficient of weigh function, the best drag reduction rate is acquired.
It is necessary to give an investigation to the feature of weigh function in physical space, so as to bring down the amount of calculation and input information. The present study researched the distribution of suboptimal control weigh function. According to the numerical simulation of suboptimal control with continuous wall blowing and suction on the boundary, a spanwise rectangle range was finally chosen as the available flow information acquired region. With the comparison, the control scheme which had been gained by the analysis to suboptimal control could be analogized to neural network control. To the dimple actuator model, under the induction by continuous wall blowing and suction model, the present study applied DNS to analyze opposite control, suboptimal control and neural network control. The suboptimal control model which consists of dimple actuators and sensors with the selection of available information acquired region and average rebuilding manner, deserved a drag reduction of 12~13%, meanwhile, the drag reduction efficiency wasη= 11.44. According to the analysis to the statistic and coherent structure, it was proved that the chosen control scheme could deprive the streak and vorticity over the fluid boundary layer. The turbulent kinetic energy and Reynold stress near wall are also brought down by suboptimal control, which has a better effect than opposite control. There is homologous result on neural network control.
本文研究的控制方案主要包括反向控制、次优控制和神经网络控制。为了能够深入研究主动控制的特性,本文以连续壁面吹吸控制方案作为带有凹坑的可变形壁面控制的指导方案。在连续壁面吹吸控制方案下,本文对比了反向控制和次优控制的减阻效果,通过改变次优控制经验系数比较得到了次优控制在该方案下所能达到的最佳减阻率。
为了能尽量降低计算量和减小输入信息量,需要对主动控制在物理空间的表达进行分析。本文研究了次优控制权函数在物理空间的分布特性,通过对基于壁面连续吹吸致动的次优控制的直接数值模拟,选定了以控制点为中心在展向一字形的截断区域作为流动信息的获取范围,并通过对次优控制和神经网络控制的表达式比较,将次优控制这一选定方案类比应用在神经网络的控制上。对于凹坑致动,在连续壁面吹吸控制方案的指导下本文利用直接数值模拟对反向控制、次优控制和神经网络控制进行了分析。对于传感器-凹坑致动实用模型的次优控制,利用展向一字权函数截断区域和平均重构流动信息的方案,本文得到了12%~13%的减阻效果,减阻效率为η= 11.44。通过对湍流统计量和相干结构的分析,证实了这种控制方案能够在一定程度上削弱湍流近壁区的条带结构和涡结构,近壁区的湍动能和雷诺应力均受到抑制,优于同样模型下的反向凹坑控制方案。对于神经网路控制也有类似的结果。
The generation of high skin-friction over the turbulent boundary layer is close related to the near-wall streamwise vorticity. Applying MEMS intelligent skin technology to control streamwise vorticity in order to bring down the turbulent friction is the leading issue of turbulent drag reduction control research. The transfiguration of dimple could locally disturb the flow field and also work under the continuous surface, hence, it is considered as the preferred actuator model for MEMS system. In the present study, investigation of flow field with dimpled transfiguration initiative control is carried out by direct numerical simulation of turbulent channel flow.
Initiative control discussed in present study includes opposite control, suboptimal control and neural network control. In order to make a further investigation to initiative control, present study uses continuous wall blowing and suction model as the inductive scheme to the dimpled transfiguration control model. By the training of continuous wall blowing and suction model, the comparison of drag reduction effect between opposite control and suboptimal control is given. By altering the experienced coefficient of weigh function, the best drag reduction rate is acquired.
It is necessary to give an investigation to the feature of weigh function in physical space, so as to bring down the amount of calculation and input information. The present study researched the distribution of suboptimal control weigh function. According to the numerical simulation of suboptimal control with continuous wall blowing and suction on the boundary, a spanwise rectangle range was finally chosen as the available flow information acquired region. With the comparison, the control scheme which had been gained by the analysis to suboptimal control could be analogized to neural network control. To the dimple actuator model, under the induction by continuous wall blowing and suction model, the present study applied DNS to analyze opposite control, suboptimal control and neural network control. The suboptimal control model which consists of dimple actuators and sensors with the selection of available information acquired region and average rebuilding manner, deserved a drag reduction of 12~13%, meanwhile, the drag reduction efficiency wasη= 11.44. According to the analysis to the statistic and coherent structure, it was proved that the chosen control scheme could deprive the streak and vorticity over the fluid boundary layer. The turbulent kinetic energy and Reynold stress near wall are also brought down by suboptimal control, which has a better effect than opposite control. There is homologous result on neural network control.
引文
[1] Chiang, S., Chi-Ming, H. 2000. Recent advances of MEMS application in flow control. FMRL. 2000
[2] Ligrani, P., M., Oliveira, M., M., 2003. Comparison of Heat Transfer Augmentation Techniques. AIAA Journal Vol. 41, No. 3, March 2003
[3] Ligrani, P., M., Harrison, J., L., Mahmmod, G., I., Hill, M., L., 2001. Flow structure due to dimple depressions on a channel surface. Phys. Fluids, Vol.13, No. 11.
[4] Won, S., Y., Zhang, Q., Ligrani, P., M., 2005. Comparisons of flow structure above dimpled surfaces with different dimple depths in a channel. Phys. Fluids 17, 045105 (2005)
[5] Endo, T., Kasagi, N., 2001. Active control of wall turbulence with wall deformation. JSME Intenational Journal, Series B, Vol.44, No.2
[6] Lee, C., Kim, J., Choi, H., 1997. Suboptimal control of turbulent channel flow for drag reduction. J.Fluid Mech. vol.258, pp. 245-258
[7] Lee C., Kim J., 1997. Application of neural networks to turbulence control for drag reduction. Phys. Fluids, Vol. 9, No. 6, June 1997
[8] Gaster, M., Grosch ,C.E., Jackson ,T.L., 1994. The velocity field created by a shallow bump in a boundary layer. University of Cambridge, Old Dominion University & Institute for Computer Applications in Science and Engineering, NASA Langley Research Centre. 1994-05-07~1994-05-20
[9] Zhang, Y. W., Yeo K.S., Khoo B.C., 2006. DNS of low Reynolds numbers turbulent flows in dimpled channels. National University of Singapore. Temasek Laboratories, National University of Singapore; Department of Mechanical Engineering. 2006-11-27
[10] Lambert, S. Fundamental studies of active dimples.
[11] Choi, H., Moin, P., Kim, J., 1994. Active turbulence control for drag reduction in wall bounded flows. J. Fluid Mech. 262, 75–110.
[12] Abergel, F., Temam, R., 1990 On some control problems in fluid mechanics. Theor. Comput Fluid. Dyn. 1, 303.
[13] Choi, H., Temam, R., Moin, P. & Kim, J. 1993 Feedback control for unsteady flow and its application to the stochastic Burgers equation. J. Fluid Mech. 245, 509.
[14] Bewley, T. & Moin, P. 1994 Optimal control of turbulent channel flows. In Active Control of Vibration and Noise (ed. K. W. Wang, A. H. Von Flotow, R. Shoureshi, E. W. Hendricks & T. W. Farabee). ASME DE-Vol. 75.
[15] Fukagata, K., Kasagi, N., 2004. Suboptimal control for drag reduction via suppression of near-wall Reynolds shear stress. International Journal of Heat and Fluid Flow 25 (2004), 341–350
[16] Lorang L. V., Podvin B., QuéréP. L., 2008. Application of compact neural network for drag reduction in a turbulent channel flow at low Reynolds numbers. Phys. Fluids 20, 045104 (Apr. 18th 2008)
[17] Ge M. W., Xu C. X., Cui G. X., Zhang Z. S., 2008 Flow structure due to static and oscillating dimples in channel flows. The 6th Symposium on Turbulence and Shear Flow Phenomenon, June. 22-24 2009, Seoul, Korea.
[18] Simon, H.,著.叶世伟,史忠植译.神经网络原理(原书第2版)计算机科学丛书出版社,2004年1月
[19] Mito, Y., Kasagi, N., 1998. DNS study of turbulence modification with streamwise-uniform sinusoidal wall-oscillation. International Journal of Heat and Fluid Flow 19 (1998), 470–481
[20]许春晓.槽道湍流的直接数值模拟:[博士学位论文].北京:清华大学,1995
[2] Ligrani, P., M., Oliveira, M., M., 2003. Comparison of Heat Transfer Augmentation Techniques. AIAA Journal Vol. 41, No. 3, March 2003
[3] Ligrani, P., M., Harrison, J., L., Mahmmod, G., I., Hill, M., L., 2001. Flow structure due to dimple depressions on a channel surface. Phys. Fluids, Vol.13, No. 11.
[4] Won, S., Y., Zhang, Q., Ligrani, P., M., 2005. Comparisons of flow structure above dimpled surfaces with different dimple depths in a channel. Phys. Fluids 17, 045105 (2005)
[5] Endo, T., Kasagi, N., 2001. Active control of wall turbulence with wall deformation. JSME Intenational Journal, Series B, Vol.44, No.2
[6] Lee, C., Kim, J., Choi, H., 1997. Suboptimal control of turbulent channel flow for drag reduction. J.Fluid Mech. vol.258, pp. 245-258
[7] Lee C., Kim J., 1997. Application of neural networks to turbulence control for drag reduction. Phys. Fluids, Vol. 9, No. 6, June 1997
[8] Gaster, M., Grosch ,C.E., Jackson ,T.L., 1994. The velocity field created by a shallow bump in a boundary layer. University of Cambridge, Old Dominion University & Institute for Computer Applications in Science and Engineering, NASA Langley Research Centre. 1994-05-07~1994-05-20
[9] Zhang, Y. W., Yeo K.S., Khoo B.C., 2006. DNS of low Reynolds numbers turbulent flows in dimpled channels. National University of Singapore. Temasek Laboratories, National University of Singapore; Department of Mechanical Engineering. 2006-11-27
[10] Lambert, S. Fundamental studies of active dimples.
[11] Choi, H., Moin, P., Kim, J., 1994. Active turbulence control for drag reduction in wall bounded flows. J. Fluid Mech. 262, 75–110.
[12] Abergel, F., Temam, R., 1990 On some control problems in fluid mechanics. Theor. Comput Fluid. Dyn. 1, 303.
[13] Choi, H., Temam, R., Moin, P. & Kim, J. 1993 Feedback control for unsteady flow and its application to the stochastic Burgers equation. J. Fluid Mech. 245, 509.
[14] Bewley, T. & Moin, P. 1994 Optimal control of turbulent channel flows. In Active Control of Vibration and Noise (ed. K. W. Wang, A. H. Von Flotow, R. Shoureshi, E. W. Hendricks & T. W. Farabee). ASME DE-Vol. 75.
[15] Fukagata, K., Kasagi, N., 2004. Suboptimal control for drag reduction via suppression of near-wall Reynolds shear stress. International Journal of Heat and Fluid Flow 25 (2004), 341–350
[16] Lorang L. V., Podvin B., QuéréP. L., 2008. Application of compact neural network for drag reduction in a turbulent channel flow at low Reynolds numbers. Phys. Fluids 20, 045104 (Apr. 18th 2008)
[17] Ge M. W., Xu C. X., Cui G. X., Zhang Z. S., 2008 Flow structure due to static and oscillating dimples in channel flows. The 6th Symposium on Turbulence and Shear Flow Phenomenon, June. 22-24 2009, Seoul, Korea.
[18] Simon, H.,著.叶世伟,史忠植译.神经网络原理(原书第2版)计算机科学丛书出版社,2004年1月
[19] Mito, Y., Kasagi, N., 1998. DNS study of turbulence modification with streamwise-uniform sinusoidal wall-oscillation. International Journal of Heat and Fluid Flow 19 (1998), 470–481
[20]许春晓.槽道湍流的直接数值模拟:[博士学位论文].北京:清华大学,1995