摘要:

The pressure‐driven flow of a rarefied monatomic gas through a two‐dimensional slit is simulated using the direct simulation Monte Carlo technique. Of particular interest is the change in flow field structure as pressure ratio and Knudsen number are varied. Comparisons are made to quantify the limits of validity of free‐molecular theory and approximate, nearly free‐molecular iterative methods. Also addressed is the sensitivity of the numerical solutions to grid structure and boundary conditions. The free‐molecular theory is found to predict quantitative flow field properties (e.g., centerline velocities or downstream flux profiles) reasonably well for large finite Knudsen number with the error dependent on the pressure ratio. The nearly free‐molecular corrections are shown to have limited range of applicability. A previously derived parameter is found to correlate total mass flux well as a function of pressure ratio and Knudsen number over a large portion of the transitional regime.  

ISSN:0899-8213    

DOI:10.1063/1.858778

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

By Monte Carlo simulation the Knudsen layer in front of a surface from which atoms evaporate is studied. Evaporation into a vacuum is simulated by means of an evaporation–condensation geometry. Hard sphere interaction cross sections are employed. With the help of the present simulation data, the Knudsen layer is defined as that region adjacent to the evaporating surface, where the temperature of the flow parallel and perpendicular to the flow direction deviate by at least a given resolution &dgr;. Taking &dgr;=1%, it is found that the Knudsen layer is established after 800 mean‐free flight times; it has an extension of 20 mean‐free paths. It takes 60 monolayers to desorb before a Knudsen layer is formed. The data are generally in good agreement with predictions of analytical theory, where available. The differences observed in the case of evaporation from a binary target are discussed.

ISSN:0899-8213    

DOI:10.1063/1.858779

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

The theory of nearly incompressible (NI) fluid dynamics developed previously for hydrodynamics is extended to magnetohydrodynamics (MHD). On the basis of a singular expansion technique, modified systems of fluid equations are derived for which the effects of compressibility are admitted only weakly in terms of the different possible incompressible solutions (thus ‘‘nearly incompressible MHD’’). NI MHD represents the interface between the compressible and incompressible magnetofluid descriptions in the subsonic regime. The theory developed here does not hold in the presence of very large thermal, gravitational, or field gradients. It is found that there exist three distinct NI descriptions corresponding to each of the three possible plasma beta (&bgr; ≡ the ratio of thermal to magnetic pressure) regimes (&bgr;≪1, &bgr;∼1, &bgr;≫1). In the &bgr;≫1 regime, the compressible MHD description converges in the low Mach number limit to the equations of classical incompressible three‐dimensional (3‐D) MHD. However, for the remaining plasma beta regimes, the imposition of a large dc magnetic field forces the equations of fully compressible 3‐D MHD to converge to the equations of 2‐D incompressible MHD in the low Mach number limit. The ‘‘collapse in dimensionality’’ corresponding to the different plasma beta regimes clarifies the distinction between the 3‐D and 2‐D incompressible MHD descriptions (and also that of 21/2‐D incompressible MHD). The collapse in dimensionality that occurs as a result of a decreased plasma beta can carry over to the weakly compressible corrections. For a &bgr;∼1 plasma, Alfve´n waves propagate parallel to the applied magnetic field (reminiscent of reduced MHD), while for a &bgr;≪1 magnetofluid, quasi‐1‐D long‐wavelength acoustic modes propagate parallel to the applied magnetic field. The detailed theory of weakly compressible corrections to the various incompressible MHD descriptions is presented and the implications for the solar wind emphasized.

ISSN:0899-8213    

DOI:10.1063/1.858780

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

Numerical computations are employed to study the flow field produced by a four‐roll mill. The radius of the cylindersa, the cylinder spacing 2b, and the size 2lof the square container are varied to assess the effects on the kinematics of the flow field. It is found that a ratio ofa/b=0.625 withl/b≥3.0 produces the best approximation to a pure extensional flow. With these parameter values, the extension rate remains constant with an error of less than 1% over an axial regionx/b≤0.5. By contrast, the commonly accepted designa/b=0.772 suggested by Fuller and Leal [J. Polym. Sci. Polym. Phys.19, 557 (1981)] produces a variation in extension rate of 50% over the same region. Streamline patterns and velocity gradient error contours are presented for these two designs.

ISSN:0899-8213    

DOI:10.1063/1.858782

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

The problem of a supercritical uniform stream meeting a stagnant mass of water is considered. The whole oncoming flow is forced to overturn and to fall back and down forever like a waterfall into a bottomless chasm. The interaction between this waterfall and the oncoming flow is neglected. It is shown that the solution occurs only at a unique Froude number equal to 2.994.

ISSN:0899-8213    

DOI:10.1063/1.858783

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

The character of axial flow in the core of a concentrated line vortex is studied using flow visualization. Results show that axial flow can be initiated when the no‐slip boundary condition is imposed on the vortex core over a spatial extent comparable to the core diameter. It is further demonstrated that no‐slip is not a necessary condition to generate the axial flow; a region of spanwise shear (i.e., a slip boundary) is sufficient. Two distinct axial flow velocity profiles are observed; they are either jetlike or spatially undulated corresponding to a maximum axial speed occurring at the core center or outer edge, respectively.

ISSN:0899-8213    

DOI:10.1063/1.858784

出版商:AIP    

年代:1993

数据来源: AIP

摘要:

This Brief Communication presents an alternative approach for determining the Landau constants associated with the implementation of weakly nonlinear theories. Results from this approach are shown to be in agreement with the usual solvability condition for the case of exact resonance. Away from resonance, the results are in agreement with a direct solution of the nonhomogeneous system. This approach also provides accurate results for the case of near resonance, in contrast to the approach currently in widespread use. Following this alternative approach, the Landau constants can be calculated accurately in a straightforward manner without requiring the solvability condition in some arbitrary neighborhood of resonance.

ISSN:0899-8213    

DOI:10.1063/1.858785

出版商:AIP    

年代:1993

数据来源: AIP

ISSN:0899-8213    

DOI:10.1063/1.858786

出版商:AIP    

年代:1993

数据来源: AIP