1. publication^'-^. Resultsfor a sample validation problem,not reportedbefore,arepresentedbelow. As basic objectivesformulated in 1989/90, have been approached, new and more demanding requirementsareemerging. Thesearediscussedin thenexttwosections.
2. The four rim seal cavity configuration were gridded with 50-60 cells in the axial direction and 60-70cells in the radial direction. Therotor speed was fixed at 2450rpm (tipspeed =69 m/s) Re, =5 x lo 6),the inlet swirl was set to rotor tip speed. A schematicof the genericdisc cavityand rim seal isshownin Figure 4. Thesimulationsfor each rim seal werecarried out with the main gaspath velocities at 120-150 m/s, and at two differentpurge flow parameters q, =1and 8x 10-3. The additional qt = 2 and 4 x 10%were also used forconfiguration 1. The simulationswere done with central-differencing in space and the k-e model of turbulence and a passive scalar transport equation for the tracer gas. The volume-averaged concentration of COz in the cavitywas then used to calculatethe cooling flow parameter 4.
3. The variation of +asafunctionofthepurgeflow rate for thefour different configurationsisshown in Figures 5a - 5d. Experimental and computed values show verygoodagreement (7-18%at qt= 1 x and ir 2% at qt = 8 x for all four configurations and for a range of coolant flow rates.