Convective heat transfer in the thermal entrance region of parallel-flow noncircular duct heat exchanger arrays

Convective heat transfer properties of a hydrodynamically fully developed flow, thermally developing flow in a parallel-flow, and noncircular duct heat exchanger passage subject to an insulated boundary condition are analyzed. In fact, due to the complexity of the geometry, this paper investigates in detail heat transfer in a parallel-flow heat exchanger of equilateral-triangular and semicircular ducts. The developing temperature field in each passage in these geometries is obtained seminumerically from solving the energy equation employing the method of lines (MOL). According to this method, the energy equation is reformulated by a system of a first-order differential equation controlling the temperature along each line.Temperature distribution in the thermal entrance region is obtained utilizing sixteen lines or less, in the cross-stream direction of the duct. The grid pattern chosen provides drastic savings in computing time. The representative curves illustrating the isotherms, the variation of the bulk temperature for each passage, and the total Nusselt number with pertinent parameters in the entire thermal entry region are plotted. It is found that the log mean temperature difference (T _LM), the heat exchanger effectiveness, and the number of transfer units (NTU) are 0.247, 0.490, and 1.985 for semicircular ducts, and 0.346, 0.466, and 1.345 for equilateral-triangular ducts.

---

Konvektiver Wärmeübergang im thermischen Einlaufgebiet von Gleichstromwärmetauschern mit nichtkreisförmigen Strömungskanälen

Zusammenfassung Die Untersuchung bezieht sich auf das konvektive Wärmeübertragungsverhalten eines Gleichstromwärmetauschers mit nichtkreisförmigen Strömungskanälen bei hydraulisch ausgebildetet, thermisch einlaufender Strömung unter Aufprägung einer adiabaten Randbedingung. Zwei Fälle komplizierter Geometrie, nämlich Kanäle mit gleichseitig dreieckigen und halbkreisförmigen Querschnitten, werden bezüglich des Wärmeübergangsverhaltens bei Gleichstromführung eingehend analysiert. Das sich entwickelnde Temperaturfeld in jedem Kanal von der eben spezifizierten Querschnittsform wird halbnumerisch durch Lösung der Energiegleichung unter Einsatz der Linienmethode (MOL) erhalten. Dieser Methode entsprechend erfolgt eine Umformung der Energiegleichung in ein System von Differentialgleichungen erster Ordnung, welches die Temperaturverteilung auf jeder Linie bestimmt.Die Temperaturverteilung im Einlaufgebiet wird unter Vorgabe von 16 oder weniger Linien über dem Kanalquerschnitt erhalten, wobei die gewählte Gitteranordnung drastische Einsparung an Rechenzeit ergibt. Repräsentative Kurven für das Isothermalfeld, den Verlauf der Mischtemperatur für jeden Kanal und die Gesamt-Nusseltzahl als Funktion relevanter Parameter im gesamten Einlaufgebiet sind in Diagrammform dargestellt. Es zeigt sich, daß die mittlere logarithmische Temperaturdifferenz (T _LM), der Wärmetauscherwirkungsgrad und die Anzahl der Übertragungseinheiten (NTU) folgende Werte annehmen: 0,247, 0,490 und 1,985 für halbkreisförmige Kanäle sowie 0,346, 0,466 und 1,345 für gleichseitig dreieckige Kanäle.

---

Nomenclature A cross sectional area [m²] - a characteristic length [m] - C _c specific heat of cold fluid [J kg^–1 K^–1] - C _h specific heat of hot fluid [J kg^–1 K^–1] - C _p specific heat [J kg^–1 K^–1] - C _r specific heat ratio,C _r=C _c/C_h - D _h hydraulic diameter of duct [m] - f friction factor - k thermal conductivity of fluid [Wm^–1 K^–1] - L length of duct [m] - m mass flow rate of fluid [kg s^–1] - N factor defined by Eq. (20) - NTU number of transfer units - Nu _{x, T} local Nusselt number, Eq. (19) - P perimeter [m] - p pressure [KN m^–2] - Pe Peclet number,RePr - Pr Prandtl number,/ - Q _T total heat transfer [W], Eq. (13) - Q ideal heat transfer [W], Eq. (14) - Re Reynolds number,D _h/ - T temperature [K] - T _b bulk temperature [K] - T _e entrance temperature [K] - T _w circumferential duct wall temperature [K] - u, U dimensional and dimensionless velocity of fluid,U=u/u - , dimensional and dimensionless mean velocity of fluid - w generalized dependent variable - X dimensionless axial coordinates,X=D _h ² /a ² x* - x, x* dimensional and dimensionless axial coordinate,x*=x/D _hPe - y, Y dimensional and dimensionless transversal coordinates,Y=y/a - z, Z dimensional and dimensionless transversal coordinates,Z=z/a Greek symbols thermal diffusivity of fluid [m² s^–1] - * right triangular angle, Fig. 2 - independent variable - T _LM log mean temperature difference of heat exchanger - effectiveness of heat exchanger - generalized independent variable - dimensionless temperature - _b dimensionless bulk temperature - dynamic viscosity of fluid [kg m^–1 s^–1] - kinematic viscosity of fluid [m² s^–1] - density of fluid [kg m^–3] - heat transfer efficiency, Eq. (14) - generalized dependent variable     

Drag of a sphere in an unsteady flow of viscous fluid at small reynolds numbers

An unsteady flow of viscous incompressible fluid past a sphere is investigated. The values of the inertial and unsteady terms in the Navier-Stokes equations are characterized by translational (R) and vibrational (R_k) Reynolds numbers, which are assumed small. The solution is constructed in the form of an expansion with respect to max(R, R _k ^1/2 ) by the method of matched asymptotic expansions. A correction to the Stokes force, correct to o[max(R, R _k ^1/2 )], is calculated. It is shown that the result depends strongly on the ratio R/R _k ^1/2 and goes over into the well-known equations for the cases R 0, R_k 0.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 1, pp. 11–16, January–February, 1988.     

Transient laminar free convection in horizontal cylinders

A numerical study of laminar natural convection inside uniformly heated, partially or fully filled horizontal cylinders is made. A coordinate transformation which simplifies the discretization of the equations of motion and energy is utilized. The resulting system of partial differential equations with their boundary conditions is solved using central differences for various Prandtl and Grashof numbers for two different grid sizes. The flow in completely filled cylinders for which experimental data are available is predicted. Close agreement between steady-state predictions and experiments is obtained for temperature and velocity profiles as well as for the streamline contours and isotherms. The technique is further demonstrated by solving the transient natural convection flow inside a partially filled horizontal cylinder with an adiabatic free surface and subjected to uniform wall heating.

---

Laminare freie Konvektion in horizontalen Zylindern

Zusammenfassung Es wurde eine numerische Berechnung der laminaren, freien Konvektion in gleichmäßig beheizten, teilweise oder ganz gefüllten, horizontalen Zylindern durchgeführt. Dabei wird eine Koordinatentransformation benützt, welche die Diskretisierung der Bewegungs- und der Energiegleichung vereinfacht. Das so resultierende System von partiellen Differentialgleichungen wird, zusammen mit seinen Randbedingungen, unter Verwendung einer Differenzenmethode für verschiedene Prandtl und Grashof-Zahlen sowie für zwei verschiedene Gittergrößen gelöst. Für den vollständig gefüllten Zylinder, für den experimentelle Daten verfügbar sind, wird die Strömung vorhergesagt. Dabei wird für stationäre Zustände gute Übereinstimmung zwischen Rechnung und Experiment erzielt. Dies gilt sowohl für den Verlauf der Stromlinien als auch für den der Isothermen. Das Verfahren wird weiterhin am Beispiel der Berechnung instationärer, freier Konvektion in einem partiell gefüllten, horizontalen Zylinder demonstriert, wobei eine adiabate, freie Oberfläche und gleichmäßige Beheizung der Wand angenommen sind.

---

Nomenclature g acceleration due to gravity, m/s² - Gr _R ^* modified Grashof number =gqR⁴/kv² - Gr _R Grashof number =gTR³/v² - H heat function vector, dimensionless - k thermal conductivity, W/mK - L(Y) cord length associated with coordinateY, dimensionless - Pr Prandtl number=v/ - q wall heat flux, W/m² - R radius, m - r(X, Y,Z) distance of a boundary point from the reference axis, dimensionless - S vector derived from the flow field solution, dimensionless - T temperature, K - T _w wall temperature, K - T reference temperature, K - t time, s - u, v velocity components inx, y directions, m/s - U, V dimensionless velocity components inX- and Y-direction normalized withU - U reference velocity=gqR²/k or gTR, m/s - V velocity vector, dimensionless - W vorticity vector, dimensionless - W vorticity, dimensionless - x, y, z cartesian coordinates, m - X, Y, Z cartesian coordinates normalized with a reference length, dimensionless Greek letters thermal diffusivity, m²/s - coefficient of thermal expansion, K^–1 - ,,, non-dimensional coordinates in the transformed domain - non-dimensional temperature =(T–T)_k/q^R or T–T/T_w–T - v kinematic viscosity, m²/s - non-dimensional time=v/R² Gr_Rt or v/R² G _R ^* t - angle measured from the bottom of the cylinder, rads - ^* angle measured from the axis on (– ) plane, rads - heat potential, dimensionless - angle of incidence of the heat flux vector, rads - non-dimensional stream function - vector potential, dimensionless - grid size, dimensionless - ² Laplacian operator - gradient vector     

Stability of flows of the Hamel type in channels with permeable walls

The stability of nonparallel flows of a viscous incompressible fluid in an expanding channel with permeable walls is studied. The fluid is supplied to the channel through the walls with a constant velocity v₀ and through the entrance cross section, where a Hamel velocity profile is assigned. The resulting flow in the channel depends on the ratio of flow rates of the mixing streams. This flow was studied through the solution of the Navier—Stokes equations by the finite-difference method. It is shown that for strong enough injection of fluid through the permeable walls and at a distance from the initial cross section of the channel the flow approaches the vortical flow of an ideal fluid studied in [1]. The steady-state solutions obtained were studied for stability in a linear approximation using a modified Orr—Sommerfeld equation in which the nonparallel nature of the flow and of the channel walls were taken into account. Such an approach to the study of the stability of nonparallel flows was used in [2] for self-similar Berman flow in a channel and in [3] for non-self-similar flows obtained through a numerical solution of the Navier—Stokes equations. The critical parameters *, R*, and Cr* at the point of loss of stability are presented as functions of the Reynolds number R₀, characterizing the injection of fluid through the walls, and the parameter , characterizing the type of Hamel flow. A comparison is made with the results of [4] on the stability of Hamel flows with R₀ = 0.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 6, pp. 125–129, November–December, 1977.The author thanks G.I. Petrov for a discussion of the results of the work at a seminar at the Institute of Mechanics of Moscow State University.     

On the laminar forced convection with axial conduction in a circular tube with exponential wall heat flux

The values of the fully developed Nusselt number for laminar forced convection in a circular tube with axial conduction in the fluid and exponential wall heat flux are determined analytically. Moreover, the distinction between the concepts of bulk temperature and mixing-cup temperature, at low values of the Peclet number, is pointed out. Finally it is shown that, if the Nusselt number is defined with respect to the mixing-cup temperature, then the boundary condition of exponentially varying wall heat flux includes as particular cases the boundary conditions of uniform wall temperature and of convection with an external fluid.

---

Über laminare Zwangskonvektion mit Längswärmeleitung in einem Kreisrohr mit exponentiell veränderlichem Wandwärmefluß

Zusammenfassung Es werden die Endwerte der Nusselt-Zahlen für vollausgebildete laminare Zwangskonvektion in einem Kreisrohr mit Längswärmeleitung und exponentiell veränderlichem Wandwärmefluß analytisch ermittelt. Besondere Betonung liegt auf dem Unterschied zwischen den Konzepten für die Mittel- und die Mischtemperatur bei niedrigen Peclet-Zahlen. Schließlich wird gezeigt, daß bei Definition der Nusselt-Zahl bezüglich der Mischtemperatur die Randbedingung exponentiell veränderlichen Randwärmeflusses die Spezialfälle konstanter Wandtemperatur und konvektiven Wärmeaustausches mit einem umgebenden Fluid einschließt.

---

Nomenclature A _n dimensionless coefficients employed in the Appendix - Bi Biot numberBi=h _e r ₀/ - c _n dimensionless coefficients defined in Eq. (17) - c _p specific heat at constant pressure of the fluid within the tube, [J kg^–1 K^–1] - f solution of Eq. (15) - h ₁,h ₂ specific enthalpies employed in Eqs. (2) and (4), [J kg^–1] - h _e convection coefficient with a fluid outside the tube, [W m^–2 K^–1] - rate of mass flow, [kg s^–1] - Nu bulk Nusselt number,2r ₀ q _w /[(T _w –T _b )] - Nu _H fully developed value of the bulk Nusselt number for the boundary condition of uniform wall heat flux - Nu _T fully developed value of the bulk Nusselt number for the boundary condition of uniform wall temperature - Nu ^* mixing Nusselt number,2r ₀ q _w /[(T _w –T _m )] - Nu _C ^* fully developed value of the mixing Nusselt number for the boundary condition of convection with an external fluid - Nu _H ^* fully developed value of the mixing Nusselt number for the boundary condition of uniform wall heat flux - Nu _T ^* fully developed value of the mixing Nusselt number for the boundary condition of uniform wall temperature - Pe Peclet number, 2r ₀/ - q ₀ wall heat flux atx=0, [W m^–2] - q _w wall heat flux, [W m^–2] - r radial coordinate, [m] - r ₀ radius of the tube, [m] - s dimensionless radius,s=r/r ₀ - T temperature, [K] - T ₀ temperature constant employed in Eq. (14), [K] - T reference temperature of the fluid external to the tube, [K] - T _b bulk temperature, [K] - T _m mixing or mixing-cup temperature, [K] - T _w wall temperature, [K] - u velocity component in the axial direction, [m s^–1] - mean value ofu, [m s^–1] - x axial coordinate, [m] Greek symbols thermal diffusivity of the fluid within the tube, [m² s^–1] - exponent in wall heat flux variation, [m^–1] - dimensionless parameter - dimensionless temperature =(T _w –T)/(T _w –T _b ) - ^* dimensionless temperature ^*=(T _w –T)/(T _w –T _m ) - thermal conductivity of the fluid within the tube, [W m^–1 K^–1] - density of the fluid within the tube, [kg m^–3]     

Forced convection heat transfer at the boundary layer of a packed bed

Numerical investigations of the nature of the fluid flow pattern and heat transfer at the boundary layer of a packed bed are reported. A volume averaged Navier-Stokes equation is used to predict the fluid flow and a volume averaged heat balance equation the heat transfer. A variable porosity in the packing is assumed in the region near the wall. Simulations are performed using a modified penalty Galerkin finite element method. The case of fully developed hydrodynamic flow and developing thermal flow is studied. The Nusselt number is found to depend on the Reynolds number, Graetz number and ratio of thermal conductivity of the solid and fluid phases. Comparison is made to some experimental literature values.Nomenclature A constant - [A] Navier-Stokes type matrix - B constant - [B] divergence matrix - C _p constant pressure heat capacity - d characteristic length - D _p particle diameter - D _t tube diameter - {F} solicitation vector - Gz Graetz number, z D _t ^–1 Pr _f Re _p - k permeability term - k _f Thermal conductivity of the fluid phase - k _s Thermal conductivity of the solid phase - [K] coefficient matrix for temperature equation - n normal vector - P pressure - Pr _f Prandtl number for the fluid _f C _p k _f ^-1 - r radial coordinate - R _t tube radius - R residual - R _m residual - Re _p Reynolds number for particle, - t tortuosity factor - T temperature - interstitial velocity - z axial coordinate - effective thermal conductivity - penalty parameter - boundary of solution domain - porosity - viscosity - density - test function - solution domain - test function     

Rotation of axisymmetric bodies in hydromagnetics

Summary This paper deals with the distrubance due to steady rotation of axisymmetric bodies in a viscous incompressible fluid of infinite conductivity in which the uniform ambient flow field is collinear with the uniform magnetic field. The known results of Sowerby¹) for the problem Couple on a rotating spheroid in a slow stream in an incompressible viscous fluid are generalized to apply to all ellipsoids of revolution and a special case of the circular disc has also been investigated in detail. The conditions of flow are assumed to be such as to permit the use of Oseen's approximation in the equations of motion. The coefficient of couple has been found to the first order approximation in terms ofR(1–S) orR(S–1) whereR is the Reynolds number andS=H ₀ ² /4W ² the pressure number which plays an important part. The couple exerted on the body is computed for various values ofS andR.     

Unsteady supersonic flow around blunt bodies with detached shock wave

The numerical method of calculating the supersonic three-dimensional flow about blunt bodies with detached shock wave presented in [1–3] is applied to the case of unsteady flow. The formulation of the unsteady problem is analogous to that of [4], which assumes smallness of the unsteady disturbances.The paper presents some results of a study of the unsteady flow about blunt bodies over a wide range of variation of the Mach number M=1.50– and dimensionless oscillation frequency l/V=0–1.0. A comparison is made with the results obtained from the Newton theory.     

Thermal stability of composite superconducting tape under the effect of a two-dimensional dual-phase-lag heat conduction model

Thermal stability of composite superconducting tape subjected to a thermal disturbance is numerically investigated under the effect of a two-dimensional dual-phase-lag heat conduction model. It is found that the dual-phase-lag model predicts a wider stable region as compared to the predictions of the parabolic and the hyperbolic heat conduction models. The effects of different design, geometrical and operating conditions on superconducting tape thermal stability were also studied.a conductor width, (m) - A conductor cross sectional area of, (m²) - As conductor aspect ratio, (a/b) - b conductor thickness, (m) - Bi Biot number - B dimensionless disturbance Intensity - C heat capacity, (J m^–3 K^–1) - D disturbance energy density, (W m^–3) - f volume fraction of the stabilizer in the conductor - g(T) steady capacity of the Ohmic heat source, (W m^–3) - g_max Ohmic heat generation with the whole current in the stabilizer, (W m^–3) - G_max dimensionless maximum Joule heating - h convective heat transfer coefficient, (W m^–2 K^–1) - J current density, (A m^–2) - k thermal conductivity of conductor, (W m^–1 K^–1) - q conduction heat flux vector, (W m^–2) - Q dimensionless Joule heating - R relaxation times ratio (_T/2_q) - t rime, (s) - T temperature, (K) - T_c critical temperature, (K) - T_c1 current sharing temperature, (K) - T_i initial temperature, (K) - T_o ambient temperature, (K) - x, y co-ordinate defined in Fig. 1, (m) - thermal diffusivity (m² s^–1) - dimensionless time - _i dimensionless duration time - dimensionless y-variable - _o superconductor dimensionless thickness - dimensionless temperature - _c1 dimensionless current sharing temperature - 1 dimensionless maximum temperature - dimensionless disturbance energy - numerical tolerance - x width of conductor subjected to heat disturbances, (m) - y thickness of conductor subjected to heat disturbances, (m) - dimensionless x-variable - _o superconductor dimensionless width - stabilizer electrical resistivity, () - _q relaxation time of heat flux, (s) - _T relaxation time of temperature gradient, (s) - i initial - sc current sharing - max maximum - o ambient     

Unsteady viscous shock layer in the case of supersonic motion through an inhomogeneity

The model of a perfect gas is used in a numerical simulation of unsteady effects in a viscous shock layer near the stagnation streamline near the front part of a rotating blunt body in an inhomogeneous external flow with pressure difference P _t8, temperature difference T _t8, and vorticity difference . The evolution of nonlinear disturbances due to the passage of a heated region and a change of the injection regime is followed. A divergence finite-difference scheme of second order of approximation across the shock layer, realized by vector sweeps with allowance for the boundary conditions on the surface of the body and behind the separated bow shock wave, is used.Translated from Izvestiya Rossiiskoi Akademii Nauk, Mekhanika Zhidkosti i Gaza, No. 1, pp. 138–145, January–February, 1992.     

Calculation of viscous compressible gas flow past a corner

We consider the problem of calculating the parameters for supersonic viscous compressible gas flow past a corner (angle greater than ). The complete system of Navier-Stokes equations for the viscous compressible gas is solved in the small vicinity Q₁. (characteristic dimensionl~1/R) of the corner point. The conditions for smooth matching of the solution of the Navier-Stokes equations and the solution of the ideal gas or boundary layer equations are specified on the boundary of Q₁. All these solutions are a priori unknown, and the conditions for smooth matching reduce to certain differential equations on the boundary of Q₁. Here account is taken of the interaction of the flows near the wall surface and in the so-called outer region [1].We note that no a priori assumptions are made in Q₁ concerning the qualitative behavior of the solution, in contrast with other studies on viscous flow past a corner (for example, [2–4]).The Navier-Stokes system in Q₁ is solved numerically, using the difference scheme suggested in [5]. This scheme permits obtaining the steady-state solution by the asymptotic method for large Reynolds numbers R, and also has an approximation accuracy adequate to account for the effects of low viscosity and thermal conductivity.     

Similitude of hypersonic flows of low-density gas over blunt bodies

Laws of similitude of hypersonic flows of monatomic gases have been obtained earlier from asymptotic analysis of the equations as S and confirmed by experimental data and numerical results [1], For diatomic gases, dimensionless numbers have not been deduced by analyzing the equations but by general arguments based on analogy with monatomic gases; they were used to compare experimental and calculated results in [1–3]. In the present paper, dimensionless numbers are derived on the basis of model kinetic equations for a diatomic gas, and limits of their applicability are established. Numerical calculations confirm the exact and approximate laws of similitude and permit a comparison with experimental results. The influence of the laws of viscosity on the drag for a sphere as a function of the Reynolds number Re₀ determined using the viscosity at the stagnation point is investigated.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 2, pp. 130–135, March–April, 1981.     

Creeping motion of spheres through shear-thinning elastic fluids described by the Carreau viscosity equation

Summary Previous work on the creeping flow of viscoelastic fluids past a sphere is reviewed. Theoretical analyses available in the literature were obtained for weakly elastic fluids and therefore they predict only a small influence of fluid elasticity on the drag. In this paper, an approximate theoretical analysis is given for the creeping flow past a rigid sphere in an unbounded medium. The analysis uses a variational principle to solve the equations of motion and continuity in conjunction with the Carreau constitutive equation. The theoretical results are presented in terms of a correction factor to the Newtonian drag coefficient. The correction factor is a function of the power law flow behaviour indexn, the ratio of limiting viscosities ( ₀ – )/₀ and a dimensionless time which reflects the elastic nature of the fluids. The results are presented in graphical form covering a realistic range of these dimensionless groups.In order to verify the theoretical predictions, the drag coefficient of a number of spheres was measured in a series of shear thinning elastic test fluids. The flow properties of the test fluids were independently measured with a Weissenberg Rheogoniometer. The power law index of the test fluids varied between 1.0 and 0.4. Particle Reynolds number based on ₀ was in the range of 410^–6 to 410^–2. The difference between theoretically predicted values of drag coefficient and the experimentally measured values is less than ±7.5%. In addition, it is found that the Carreau viscosity equation can be used to predict the elastic parameter of primary normal stress difference with moderate to good accuracy for all the polymer solutions used in this work.

---

Zusammenfassung Einleitend wird ein Überblick über die früheren Untersuchungen betreffend die schleichende Strömung um eine Kugel gegeben. Die in der Literatur vorliegenden theoretischen Analysen sind auf schwach viskoelastische Flüssigkeiten beschränkt und sagen deshalb nur einen geringen Einfluß der Elastizität auf den Widerstand voraus. In dieser Veröffentlichung wird dagegen eine genäherte theoretische Analyse für die schleichende Strömung um eine starre Kugel in einem unendlich ausgedehnten Medium gegeben, bei welcher zur Lösung der Bewegungsgleichungen und der Kontinuitätsgleichung in Verbindung mit den rheologischen Stoffgleichungen vonCarreau ein Variationsprinzip verwendet wird. Die theoretischen Ergebnisse werden mittels eines Korrekturfaktors zum newtonschen Widerstandskoeffizienten beschrieben. Dieser Korrekturfaktor ist eine Funktion des Potenz-Gesetz-Exponentenn, des Verhältnisses der Grenzviskositäten ( ₀ – )/₀ und einer dimensionslosen Zeit, welche das elastische Verhalten kennzeichnet. Die Ergebnisse werden in graphischer Form unter Zugrundelegung eines realistischen Wertebereichs dieser dimensionslosen Gruppen dargestellt.Um diese theoretischen Voraussagen zu verifizieren, wurde der Widerstandskoeffizient für eine Anzahl von Kugeln in einer Reihe von Scherentzähung aufweisenden elastischen Probeflüssigkeiten gemessen. Die Fließeigenschaften dieser Flüssigkeiten wurden zusätzlich mit dem Weissenberg-Rheogoniometer bestimmt. Der Potenz-Gesetz-Exponent variierte dabei zwischen 1,0 und 0,4. Die auf den Kugeldurchmesser und die Nullviskosität bezogenen Reynolds-Zahlen lagen zwischen 410^–6 und 410^–2. Der Unterschied zwischen theoretisch vorausgesagten und experimentell bestimmten Widerstandskoeffizienten war kleiner als ±7,5%. Außerdem wurde noch gefunden, daß die Viskositätsgleichung vonCarreau dazu verwendet werden kann, den elastischen Parameter erste Normalspannungs-Differenz für alle in dieser Untersuchung verwendeten Polymerlösungen mit mäßiger bis guter Genauigkeit vorauszusagen.

---

Notation C _d drag coefficient - d diameter of sphere - f external body forces in equation of motion [2] - F _d drag force - g acceleration due to gravity - J integral defined in eq. [3] - n a parameter in the Carreau viscosity eq. [6] - p isotropic pressure term in equation of motion [2] - r,, spherical coordinates - R radius of sphere - Re ₀, Re₁ Reynolds numbers defined in eq. [16] - t time - u ⁱ ,u ^j velocities in equation of motion [2] - u _r ,u r and components of velocity - V terminal velocity of sphere in unbounded medium - V volume, in eq. [3] - X correction factor to the drag force, eq. [14] - y,z dimensionless spherical coordinates, eq. [9] - ratio of two Reynolds numbers given by eq. [16] - shear rate - apparent viscosity - ₀, zero shear rate and infinite shear rate viscosities respectively - a parameter in the Carreau viscosity eq. [6] - the dimensionless time, defined in eq. [11] - second invariant of the rate of deformation tensor - a parameter in the stream function, eq. [8] - stream function - _p,_f densities of sphere and fluid respectively With 7 figures and 1 table     

Stabilization of plasma flow in a transverse magnetic field

Results are given of a theoretical and experimental investigation of the intensive interaction between a plasma flow and a transverse magnetic field. The calculation is made for problems formulated so as to approximate the conditions realized experimentally. The experiment is carried out in a magneto-hydrodynamic (MHD) channel with segmented electrodes (altogether, a total of 10 pairs of electrodes). The electrode length in the direction of the flow is 1 cm, and the interelectrode gap is 0.5 cm. The leading edge of the first electrode pair is at x = 0. The region of interaction (the region of flow) for 10 pairs of electrodes is of length 14.5 cm. An intense shock wave S propagates through argon with an initial temperature T_o = 293 °K and pressure p_o = 10 mm Hg. The front S moves with constant velocity in the region x < 0 and at time t = 0 is at x = 0. The flow parameters behind the incident shock wave are determined from conservation laws at its front in terms of the gas parameters preceding the wave and the wave velocity W_S. The parameters of the flow entering the interaction region are as follows: temperature T ₀ ¹ = 10,000 °K, pressure P ₀ ¹ = 1.5 atm, conduction ₀ ¹ = 3000 ^–1·m^–1, velocity of flow u ₀ ¹ = 3000 m·sec^–1, velocity of sounda ₀ ¹ = 1600 m·sec^–1, degree of ionization = 2%, 0.4. The induction of the transverse magnetic field B = [0, B_y(x), 0] is determined only by the external source. Induced magnetic fields are neglected, since the magnetic Reynolds number Re_m 0.1. It is assumed that the current j = (0, 0, jz) induced in the plasma is removed using the segmented-electrode system of resistance R_e. The internal plasma resistance is R_i = h(A)^–1 (h = 7.2 cm is the channel height; A = 7 cm² is the electrode surface area). From the investigation of the intensive interaction between the plasma flow and the transverse magnetic field in [1–6] it is possible to establish the place x* and time t* of formation of the shock discontinuity formed by the action of ponderomotive forces (the retardation wave R_T), its velocity W_T, and also the changes in its shape in the course of its formation. Two methods are used for the calculation. The characteristic method is used when there are no discontinuities in the flow. When a shock wave R_T is formed, a system of nonsteady one-dimensional equations of magnetohydrodynamics describing the interaction between the ionized gas and the magnetic field is solved numerically using an implicit homogeneous conservative difference scheme for the continuous calculation of shock waves with artificial viscosity [2].Translated from Izvestiya Akademiya Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 5, pp. 112–118, September–October, 1977.     

Laminar boundary layer on blunt bodies with account for vorticity of the external flow

This paper presents the technique for and results from numerical calculations of the hypersonic laminar boundary layer on blunted cones with account for the vorticity of the external flow caused by the curved bow shock wave. It is assumed that the air in the boundary layer is in the equilibrium dissociated state, but the Prandtl number is assumed constant, =0.72. The calculations were made in the range of velocities 3–8 km/sec, cone half-angles _k=0°–20°. With account for the vortical interaction of the boundary layer with the external flow, the distribution of the thermal flux and friction will depend on the freestream Reynolds number (other conditions being the same). In the calculations the Reynolds number R, calculated from the freestream parameters and the radius of the spherical blunting, varies from 2.5·10³ to 5.10⁴. For the smaller Reynolds numbers the boundary layer thickness on the blunting becomes comparable with the shock standoff, and for R<2.5·10³ it is apparent that we must reconsider the calculation scheme. With R>5·10⁴ for cones which are not very long the vortical interaction becomes relatively unimportant. The results of the calculations are processed in accordance with the similarity criteria for hypersonic viscous gas flow past slender blunted cones [1, 2].     

Supersonic flow past blunt bodies involving fast nonequilibrium processes

In the numerical integration of the system of equations of relaxation gasdynamics the solution may become unstable. Instability arises in those cases when the characteristic time for the nonequilibrium process becomes less than the characteristic flow time. To ensure stability it is necessary to reduce the integration step. With approach to equilibrium conditions, when the process rate increases, the step reduction may lead to excessive computational time. Preceding studies have overcome the difficulty in solving the one-dimensional [1–3] and two dimensional [4] problems by various techniques, the basic idea being the use of implicit difference schemes for approximating the relaxation equations.In the present paper analogous considerations are used to develop a scheme for calculating supersonic flow past blunt bodies with fast non-equilibrium processes within the framework of [5]. The basic coordinate system , is used to approximate the equations, just as in [5]. However the relaxation equation is solved along a streamline element. Calculations are presented for the air flow past a sphere with account for the oxygen dissociation reaction. The validity of the binary similarity law for this model is verified. As an example of the applicability of the technique, a calculation is made of the flow of a chemically reacting mixture with heat release about a sphere.     

Investigation of chemically reacting and radiating supersonic flow in channels

The two-dimensional time dependent Navier-Stokes equations are used to investigate supersonic flows undergoing finite rate chemical reaction and radiation interaction for a hydrogen-air system. The explicit multi-stage finite volume technique of Jameson is used to advance the governing equations in time until convergence is achieved. The chemistry source term in the species equation is treated implicitly to alleviate the stiffness associated with fast reactions. The multidimensional radiative transfer equations for a nongray model are provided for general configuration, and then reduced for a planar geometry. Both pseudo-gray and nongray models are used to represent the absorption-emission characteristics of the participating species.The supersonic inviscid and viscous, nonreacting flows are solved by employing the finite volume technique of Jameson and the unsplit finite difference scheme of MacCormack to determine a convenient numerical procedure for the present study. The specific problem considered is of the flow in a channel with a 10° compression-expansion ramp. The calculated results are compared with the results of an upwind scheme and no significant differences are noted. The problem of chemically reacting and radiating flows are solved for the flow of premixed hydrogen-air through a channel with parallel boundaries, and a channel with a compression corner. Results obtained for specific conditions indicate that the radiative interaction can have a significant influence on the entire flowfield.Nomenclature A band absorptance (m^–1) - A _o band width parameter (m^–1) - C _j concentration of thejth species (kg mol/m³) - C _o correlation parameter ((N/m²)^–1m^–1) - C _p constant pressure specific heat (J/kgK) - e Planck's function (J/m²S) - E total internal energy (J/kg) - f _j mass fraction of thejth species - h static enthalpy of mixture (J/kg) - H total enthalpy (J/kg) - I identity matrix - I _v spectral intensity (J/m s) - I _bv spectral Planck function - k thermal conductivity (J/m sK) - K _b backward rate constant - K _f forward rate constant - I unit vector in the direction of - M _j molecular weight of thejth species (kg/kg mol) - P pressure (N/m²) - P _j partial pressure of thejth species (N/m²) - P _e equivalent broadening pressure ratio - Pr Prandtl number - P _w a point on the wall - q _R total radiative heat flux (J/m² s) - spectral radiative heat flux (J/m³ s) - R gas constant (J/KgK) - r _w distance between the pointsP andP _w(m) - S integrated band intensity ((N/m²)^–1/m^–2) - S integrated band intensity ((N/m²)^–1 m^–2) - T temperature (K) - u, v velocity inx andy direction (m/s) - production rate of thejth species (kg/m³ s) - x, y physical coordinate - z dummy variable in they direction Greek symbols ratio of specific heats - t _ch chemistry time step (s) - t _f fluid-dynamic time step (s) - absorption coefficient (m^–1) - ,_v spectral absorption coefficient (m^–1) - _p Planck mean absorption coefficient - second coefficient of viscosity, wavelength (m) - dynamic viscosity (laminar flow) (kg/m s) - , computational coordinates - density (kg/m³) - Stefan-Boltzmann constant (erg/s cm² K³) - shear stress (W/m²) - equivalence ratio - wave number (m^–1) - _c frequency at the band center     

Nonisothermal non-Newtonian chemical reactions in a tubular reactor under conditions of variable viscosity

The nonisothermal non-Newtonian chemical reactions in a tubular reactor are investigated. The non-Newtonian fluid is assumed to be characterized by the Ostwald-de Waele power-law model, which represents the majority of laminar flow of food products and many polymer melts and solutions. The temperature effect on the viscosity is considered and is found to be very significant. The effects of other important dimensionless parameters on the reactor performances are examined.Nomenclature c mass fraction of reactant - c ₀ inlect mass fraction of reactant - C _p heat capacity, J/kg K - C dimensionless concentration of reactant, c/c ₀ - C _b dimensionless bulk concentration of reactant - D molecular diffusivity, m²/s - E activation energy, J/kg - H heat of reaction, J/m³ - k ₁ frequency factor, s^–1 - k _t heat conductivity, J/m K kg - K fluid consistency, kg s ^n–1/m - K ₁ dimensionless frequency factor, k ₁ r ₀ ² c ^m–1 exp(–₁)/D - K ₀ constant in Eq. (6) - m order of chemical reaction - n rheological parameter - p pressure, kg/m s² - r radial coordinate, m - r ₀ radius of reactor, m - R dimensionless radial coordinate, r/r ₀ - R _g gas constant, J/kg K - T temperature, K - T ₀ inlet temperature, K - u velocity, m/s - u _b bulk velocity, m/s - U dimensionless velocity, u/u _b - x axial coordinate, m - X dimensionless axial coordinate, xD/r ₀ ² u _b Greek symbols dimensionless parameter, - dimensionless parameter, ₀T₀ - ₁ dimensionless activation energy, E/R _g T ₀ - ₂ dimensionless heat generation, - dimensionless temperature, (T–T ₀)/T ₀ - _b dimensionless bulk temperature - liquid density, kg/m³     

Investigation of supersonic isobaric submerged turbulent jets

Turbulent supersonic submerged air jets have been investigated on the Mach number interval M_a = 1.5–3.4 and on the interval of ratios of the total enthalpies in the external medium and the jet i₀ = 0.01 – 1. Oxyhydrogen jets with oxidizer ratios = 0.3–5 flowing from a nozzle at Mach numbers M_a = 1 and 2.4 have also been investigated. When < 1 the excess hydrogen in the jet burns up on mixing with the air. Special attention has been paid to obtaining experimental data free of the influence on the level of turbulence in the jet of the initial turbulence in the nozzle forechamber, shock waves occurring in the nozzle or in the jet at the nozzle exit, and the external acoustic field. The jet can be divided into two parts: an initial part and a main part. The initial part extends from the nozzle exit from the section x, in which the dimensionless velocity on the jet axis u_m = u_x/u_d = 0.75. Here, u_x is the velocity on the jet axis at distance x from the nozzle exit, and u_a is the nozzle exit velocity. The main part of the jet extends downstream from the section x. For the dimensionless length of the initial part x_m = x/d_a, where d_a is the diameter of the nozzle outlet section, empirical dependences on M_a and i₀ are obtained. It is shown, that in the main part of the jet the parameters on the flow axis — the dimensionless velocity and temperature — vary in inverse proportion to the distance, measured in units of length x, and do not depend on the flow characteristics which determine the length of the initial part of the jet. The angles of expansion of the viscous turbulent mixing layer in the submerged heated or burning jet increase with decrease in i₀ and M_a and are practically independent of the afterburning process.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza. No. 4, pp. 56–62, July–August, 1988.     

A power-law fluid flow past a porous sphere

Summary A model has been developed for the flow of a non-Newtonian fluid past a porous sphere. The drag force exerted on a porous sphere moving in a power-law fluid is obtained by an approximate solution of equations of motion in the creeping flow regime. It is predicted that the effect of the pseudoplastic anomaly on the drag force is more pronounced at large porosity parameters.

---

Zusammenfassung Es wird ein Modell für die Strömung einer nichtnewtonschen Flüssigkeit längs einer porösen Kugel entwickelt. Die auf die in einer Ostwald-DeWaele-Flüssigkeit bewegte Kugel ausgeübte Reibungskraft wird durch eine Näherungslösung der Bewegungsgleichungen für schleichende Strömung gewonnen. Man findet, daß der Einfluß der Abweichung vom newtonschen Verhalten um so ausgeprägter wird, je größer die Porosität ist.

---

A, B, C, D a, b, c, d coefficients in eqs. [10] and [18] - F _D drag force - K consistency index in power-law model - k ₁ ,k ₂ coefficients defined by eq. [18] - m porosity parameter - n flow index in power-law model - P pressure - P ^* dimensionless pressure defined by eq. [4] - P pressure difference - R radius of porous sphere - r radial distance from the center of the sphere - U velocity of uniform stream - u _i velocity component - u _i ^* dimensionless velocity component defined by eq. [4] - Y drag force correction factor defined by eq. [27] - _ij rate of deformation tensor - _ij ^* dimensionless rate of deformation tensor defined by eq. [4] - , spherical coordinates - dimensionless radial distance defined by eq. [4] - second invariant of rate of deformation tensor - ^* dimensionless second invariant of rate of deformation tensor defined by eq. [4] - _ij stress tensor - _ij ^* dimensionless stress tensor defined by eq. [4] - stream function - ^* dimensionless stream function defined by eq. [4] - i inside the surface of the sphere - o outside the surface of the sphere With 1 figure and 1 table