536.244It is shown that, for" a certain proportion between the rib height (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15) In diffusional combustion of a condensed filel, tile intensity of heat-and mass-transfer processes can be controlled by deliberate variatioxl of the fluctuating characteristics of the reacting boundary layer. Its stability to flameout can be achieved by installing gas-dynamic obstacles. Theoretical studies of a turbulent boundary layer with local separation fiwe computational difficulties caused by the complex character of the flow [1,2]. A large body of empirical information accunmlated in tests without combustion [3] shows that the rates of heat and mass transfer behind an obstacle can significantly increase. However, there are differences in reported estimates Of the dimensions of reattaehment and relaxation flow regions and also in determined intensities of the transfer processes and their description. For a reacting boundary layer, no experinlental data are availal)le.The objective of the present work was to experinlentally study mass transfer during vai)orization of a liquid tiM (ethanol) behind ribs of wtrious heights for various turbulence levels of air streams with and without combustion.Experimental Setup and Measurement Procedure. The tests were conducted in an open-type wind tunnel described in [4]. At tim inIet of the confuser, a wire grid with a t x 1 ram-mesh or a grating with 25 threaded holes M27 x 1.5 (with a 43-ram gap between the hole centers) was installed. The confilser of length x0 = 300 mm (contoured according to Vitoshinskii's formula; contraction ratio 1 : 6.8) joiimd the test section 100 x 100 ram. There was no upper wall of the channel, the side walls were transparent, and the 1)ottom wall was made of perforate(l plates (stainless steel). The length of the first plate was 40 mm, and that of the other six plates wr 80 ram. A rib of height h = 0-15 mm was installed at the test-section entrance, across the entire width of the channel. The turbulence level Tu0 = 8, 18, or 26% was regulated by the number of oI)en holes in the grating (25, 9, or 5, respectively); the turbulence level in tests with tlm grid was Tu0 = 1%. These vahles were fi)un(l by averaging the nleasurement results over 81 points of the entrance section spaced 10 nml apart. The measurements were performed with the help of a DISA hot-wire anemometer (single-wire probe, wire diameter 8 Itm, 55M01 bridge and 55D10 linearizer). The unifbrnlity of the air stream at the channel entrance decreased with increasing degree of turbulence, an(1 tile inaxinmnl deviation from the average (across tile channel) value of tile flow velocity reached +14% for Tu0 = 26%.