. . . : . . : is. ; 7. . . . '. . $ V AVA > : 1 . ... 1 13. . . T OF ORNL P 2724 . $ ܙܘ . T* . . . NO 3 . I . ; . * ? $ . 1 11 ... . . . e . . > . : . 1 MICROCOPY RESOLUTION TEST CHART NATIONAL BUREAU OF STANDARDS -1963 ORNU-P-2124 . HC. 1970 300. mw 65 CONF-6641 DEC 6 1966 MAGNETOHYDRODYNAMIC STABILIZATION EXPERIMENTS WITH JET-DRIVEN VORTEX FLOWS* RELEASED FOR ANNOUNCEMENT J. J. Keyes, Jr. Oak Ridge National Laboratory Oak Ridge, Tennessee IN NUCLEAR SCIENCE ABSTRACTS Confined, vortex-like flows generated by fluid jets impinging tangen- tially on the concave interior surface of a straight tube with inward radial flow of the working fluid are under consideration as a possible means for effecting containment of a very high-temperature gaseous nuclear fuel for propulsion and electrical power generation applications. Since it may be necessary to effect essentially complete separation of fuel molecules from those of the through flowing coolant by a diffusive mechanism, severe hydro- dynamic limitations are imposed - among which is the need for generation of very high tangential velocity with low mass through flow, under as nearly laminar conditions as possible. Unfortunately, as you might have anticipated, the flow is turbulent at values of tangential Reynolds modulus much below the range of practical interest. Turbulence, of course, results in high frictional losses aná limits separation by the mechanism of eddy mixing. Our interest in magnetohydrodynamic stabilization of vortex flow is based on the knowledge that, at the extreme operating temperature envisioned for these devices, the gas will be ionized and, hence, electrically conducting (a plasma). Furthermore, since we are interested in electrical power generation by magneto- hydrodynamic interaction (for example, between the rotating plasma and an axial magnetic field to induce a radial electric fiela), there will be a magnetic field available; so it is natural to inquire as to the effect of this field on flow stability. *Research sponsored by the . s. Atomic. Energy Commission" under contract with the Union Carbide Corporation. . .. rs .. . . - . . . i o . , . .... :.'nin ' Ha . . Vidi ... .. . W onto di W LYLE Let me turn, now, to a brief review of some exploratory experiments in which we have studied the influence of intense axial magnetic fields up to 75 kgauss on jet-driven vortex flow of an electrolytic conductor (concentrated aqueous NH,C1 solution). Because of the time limitation, the emphasis of my presentation will be on results of flow visualization studies, with a minimum of interpretation. Professor Chang, in the next paper, will summarize results of a related analytical investigation and compare these results with some quantitative experimental data. The vortex tube employed in this investigation is illustrated in cross section in Slide 1. It is 10 cm ID x 40 cm long, in which the flow field 113 generated by injection through 0.6-mm slits entering tangentially and extending“. the full length of the tube. We operated with 1, 2, or 4 slits feeding. The slits are sufficiently thin that the flow inside the slits 13 laminar. The fluid spirals radially inward and is withdrawn through an exit opening at the center of one end. The tube wall is stainless steel and the end walls Lucite. In the film to be shown shortly, you will be viewing along the tube axis at dye traces introduced through a feed slit at a point halfway between the end walls.' Incidentally, the flow will appear to rotate clockwise in the film – reversed . from what is shown here. The transition to instability is characterized by a three-dimensional oscillatory notion which appears to originate in the boundary layer along the concave wall, and to initiate instability, characterized by gross radial con- vection, in the interior of the flow field - as you will see in the films. Although we did not observe stationary Tay.l.or-Görtler cells, due perhaps to the disturbing influence of the jets and the strong radial and axial convection, it may nevertheless be possible that the Taylor-Görtler mechanism is responsible for the transition (1.e., the transition may be. triggered by this mechanism). in This is really speculation, however - we don't have a satisfactory explanation for the transition as we observe it. In order to give some quantitative indication of the influence of the magnetic field on stability, consider Slide 2, in which the observed tangentia), Reynolds modi:lus at transition (i.e., the critical Reynolds modulus) based on the tangential velocity near the peri:phery and the tube radius, 18 shown as a function of the Hartmann modulus, a dinensionless magnetic field strength based on the tube radius. The magnetic Reynolds modulus is, of course, very small in these experiments. Values of the radial Reynolds modulus (1.e., through flow Reynolds' modulus) varieå from 20 to about 100. This slide 1llustrates the very significant stabilizing influence we observed. For example, with four slits feeding, the critical Reynolds modulus increased from 1000 to about 10,000 as the Hartinann modulus increased from o (corresponding to zero magnetic field) to about 172 (corresponding to 75 kgauss). The results for 2- and l-slit injection lie below those for 4 slits. Now I would like to show a short film depicting some of our flow visualiza- tion results. LEGAL NOTICE This report was prepared as an account of Government sponsored work. Neither the United Slates, nor the Commission, nor any person acung on behalf of the Commission: A. Makes any warianty or representation, expressed or implied, with respect to the accu- Film racy, completeness, or u83fulness of the information contained in this roport, or that the use. of any information, apparatus, method, or process disclosed in this report may not Irfringe privately owned righto; or B. Assumes any liabil'lies with respect to the use of, or for daniages resulting from the use of any information, apparatus, method, or process disclosed in this report, A8 veed in the above, "person acting on behalf of the Commission" includes any em- ployee or contrator of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Commission, or employee of such coatractor prepares, disseminates, or provides acce88 to, any Information pursuant to his employment or contract with the Commission, or his employment with such contractor. I ( "Rea = 9500 NO 0 + 172 The first part of this sequence with no applied field reveals the unstable RRR flow field at this highly supercritical Reynolds modulus. You can see that dis- turbances which seem to originate in the boundary layer-jet interaction zone are amplified as they move inward to form three-dimensional vortex loops. Bear in mind that this is a two-dimensional view of a three-dimensional flow phenomenon. Targe amplitude eddy structure resulting in gross radial mixing is evident. Notice also that the dye coloration is essentially uniform near the center due to the strong mixing, and it is apparent that this flow field would not be satisfactory for applications requiring molecular separation (point out shadow of exit tube). Film break (long pause). Now the field is being slowly increased. The first evidence of a sta- bilizing influence is the appearance of spiral ring structure in the interior, indication that laminarization is commencing. Observe that flow near the wall is still unstable as shown by the fluctuating dye trace; the amplitude of the fluctuation decreases, however, as the field is .continual.ly increased. Sta- bilization therefore occurs first in the interior and, as the magnetic field increases, the stable region spreads outward to the wall (pause). Finally, ... i at the full field (75 kgauss), we see the stabie, laminar flow in which the dye displays the streamline pattern as the fluid spirals radially inward. This is what we would like for applications involving molecular separation. Transient. Here we see a trace of dye injected through a feed slit under fully stabilized conditions for comparison with the first pictures you say with no magnetic field. Note that the trace remains intact, with no detectable dfs. . persion (point out sharp leading edge). 'Can follow the pattern of the streamlines all the way in to the center. I "Reg = 9500 "Ha 172 +0. Here we slowly decrease the magnetic field back to zero to show the desta- bilization of the flow. If you look closely near the wall, you will begin to see oscillations in the outer dye rings - indicating that destabilization begins near the periphery. Notice the increasing amplitude or these disturbances leading to disintegration of the dye trace as the field is continually decreased; notice also the rapid spreading of the unstable region into the interior (pause). Finally, with no magnetic field, you see again the unstable situation with gross mixing as attested to by the nearly uniform dye coloration in the Intericr. III III / "Rea = 26,000 Ha = 0 In this sequence you will see the highly turbulent flow at Reynolds modulus min ... of 26,000, with no magnetic field. Note the large ampll. Cude oscillations in the dye trace near the wall. The trace is breaking up less than 90 deg downstream from the point of injection, By the time the dye makes one complete revolution, it is so well dispersed as to be barely recognizable. IV ) "Re = 26,000 1 "Ha = 172 Here we see the effect of the field when the Reynolds modulus is well above the critical value corresponding to N = 172. Note, first of all, that, while the periphery is still turbulent, partial stabilization is nevertheless achieved - as is evident by reduction in the amplitude of the fluctuations near the wall. That is, the damping effect of the field reduces the growth rate of the distur- . bances originating in the boundary layer. Perhaps more significant is the appearance of a quasi-laminar flow in the interior, as suggested by the more or less well-defined dye ring structure. END FILM. I might add that we have observed significant stabilization effects of the magnetic field at values of tangential Reynolds modulus an order of magnitude or more above the critical value, suggesting that we may still realize beneficial effects of the field on wall friction and especially on diffusional sepa ration for applications which require operation at Reynolds modulus much greater than that corresponding to complete stabilization with the maximum field which is attainable in practice. + -- . - . SECTION DIVIDER - L EXIT OPENING --- - I SHELL . SUR:?T RING - -0.025 IN. -----.. !1! DI... -( ) 90° WILL SECTION 1 LOUTER JACKET -FEED AMNULUS TX ? IN DIAM. X IG IN LONG SLIT-FED VORTEX TUBE L ALYA ALA 11,000 ORNL-DWIG 56-1195R - 1 10,000 9000 8000 10-cm DIAMETER - 4 SLITSO 20. NÃO, REYNOLDS MODULUS AT TRANSITION TO INSTABILITY 6000 ... 5000 4000 -10-cm DIAMETER - Ĉ SLITS 3000 -10-cm DIAMETER - I SLIT 2000 0 + 2.8-cin DIAMETER - 64 SLITS -28-cm DIAMETER-2 SLITS LII.. ... 40 50 60 70 80 90 100 110 NHỮ, HARTMANN AODULUS 10 20 30 120 130 140 150 160 170 180 SUMMARY OF HYDROMAGNETIC STABILIZATION EXPERIMENTS WITH 28-cm AND 10-cm DIAMETER JET-DRIVEN VORTEX TUBES ..11: .. 15 . -.- . " JS DATE FILMED 1 / 3167) ... "