Turbulence research - Progress and problems

Progress in calculation methods and in experiments related to calculated methods for understanding and prediction of turbulent flow is reviewed, along with constructive comments on topical problems which need further study. Navier-Stokes simulations and advanced conditional-sampling experiments are perhaps the most interesting of the new developments, although they are unable to answer all the questions in the near future. Improvement in knowledge of real-life effects appears to be more important than refinement of calculation methods for classical shear layers. Research should be directed toward identifying the phenomena that make real-life turbulence more complicated than classical shear layers and determining the way in which these phenomena affect the terms in the Reynolds-stress transport equations or their boundary conditions.