Discussion: “Turbulent Flow Velocity Between Rotating Co-Axial Disks of Finite Radius” (Louis, J. F., and Salhi, A., 1989, ASME J. Turbomach., 111, pp. 333–340)

Abstract

radial outflow. The model proposed by Lam and Bremhorst is used in this study and the predictions compare very favorably with the experimental results of Pincombe for an outflowing cavity and of Owen for the flow between rotating and stationary disks. Based on this validation of the model, the flow between counterrotating disks is predicted as is the variation of the radial tangential wall shear stress with Rossby number. The model does not include any effect of possible turbulence anisotropy due to rotational effects and it would be of interest to map the flow regimes for which the Lam-Bremhorst model can be used. The performance of different turbulent models in rotating disk flows is certainly of interest and I agree with the authors' statement that it would be of interest to map out the flow regimes for which the Lam-Bremhorst model can be used. ReferencesThe present results for the rotating cavity show similar agreement with experiment to those of Morse (1987), who used a different version of the k-e model. Morse presented a more thorough evaluation against experimental data, giving results at higher rotational Reynolds numbers and for both radial outflow and inflow of fluid. It is a pity that the present results are restricted to Re,, = JO 5 when data are available for Re e up to 10 6 . Can the authors give any indication of the performance of the Lam-Bremhorst model at higher values of Re"?As noted by the authors there is some discrepancy between prediction and measurements near the rotor in Fig. 1, which gives the radial velocity profile for flow between a rotating Rolls-Royce pic, Derby, United Kingdom. and stationary disk. As a further guide to the accuracy of the model in this important near-wall region, it would be useful to compare the predicted and measured disk moment coefficients.The predictions for the flow between counterrotating disks in Figs. 9 to 15 show surprising features. For example, from Figs. 9, 11, and 12 it appears that the formation of the Ekmantype layers occurs abruptly when a critical radius is reached. This contrasts with the progressive entrainment of flow into the disk boundary layers identified by Chew et al. (1984). As numerical solution of the governing equations for these flows is known to present some difficulties the question of the accuracy of the present solutions naturally arises. Did the authors encounter any difficulties in obtaining convergence of the iterative solution and to what degree are the results affected by truncation error?