Theoretical analysis of quasi-one-dimensional compressible magnetohydrodynamic channel flow

https://doi.org/10.1016/j.amc.2021.126500Get rights and content

Highlights

  • The dimensionless dynamical system for compressible MHD channel flow is derived.

  • A new magnetic interaction number N is obtained by choosing a length scale that characterizes the variation rate of the cross-sectional area.

  • The theoretical analysis yields five critical values of N that affect the variation characteristics of Mach number and velocity.

  • Eight equivalence classes of phase plane are formed through isomorphism of labeled graphs.

Abstract

A theoretical analysis of quasi-one-dimensional, steady, inviscid, compressible channel flow at a low magnetic Reynolds number with variable cross-section is performed in this paper. A second-order nonlinear dynamical system describing the variation of physical parameters is investigated in the phase plane. The characteristics of all possible channel flow with a constant electromagnetic field are obtained by the phase plane and the isomorphism of labeled graphs. It is revealed that the magnetic interaction number novelly derived from the variation rate of the cross-sectional area could significantly affect the phase trajectory in the phase plane of dimensionless velocity and Mach number. Meanwhile, the phase trajectory is only dependent on this magnetic interaction number. Further analysis of the second-order dynamical system discovers five critical values in the range of the magnetic interaction number. These five critical values can divide the whole range into eight subsets under the isomorphism of labeled graphs, forming eight equivalence classes of the phase plane. The study of such eight equivalence classes reveals that the variable cross-section flow in the phase plane can be viewed as a linear superposition of constant cross-section magnetohydrodynamic flow and isentropic flow, the weight ratio being exactly the magnetic interaction number. Consequently, for the divergent channel, the acceleration region in the supersonic zone is larger than that of the constant cross-section flow, while for the convergent channel, only the acceleration region in the subsonic zone is larger than that of the constant cross-section flow.

Introduction

As researches and explorations in the aerospace field continue to advance, the role of Magnetohydrodynamics (MHD) is becoming more prominent. In the last century, the AJAX program was proposed by Russia, using the MHD energy-bypass technology to broaden the operating Mach number of air-breathing scramjet and reduce the total pressure loss [1], [2], [3]. Currently, some research institutes such as NASA, Air Force Research Laboratory (AFRL), and Central Aerohydrodynamic Institute (TsAGI) developed MHD acceleration wind tunnels and have achieved certain results [4], [5], [6], [7], [8], [9]. Theoretically speaking, the aforementioned air-breathing hypersonic vehicles with MHD energy-bypass technology and the hypersonic MHD acceleration wind tunnel can be abstracted as a compressible channel flow of gas with low conductivity in an electromagnetic field. Consequently, it is of significance to deeply study the characteristics of MHD channel flow.

The quasi-one-dimensional (quasi-1D) MHD channel flow at low magnetic Reynolds number has been studied since the 1950s [10]. Rosa [11] and Chekmarev [12] performed integrations on the governing equations for constant-cross section in quasi-1D, steady, inviscid channel flow. For the case of variable cross-section but maintaining a constant density, Resler and Sears [13] gave analytical expressions for the velocity, temperature, and pressure along the flow direction. In the meantime, Resler and Sears investigated the variation patterns of the velocity and Mach number along the flow direction under the conditions of constant cross-section and constant density. It has been concluded that the relative magnitudes between the flow velocity and three critical velocity values (u1, u2 and u3 in the literature) would change the asymptotic properties of the flow significantly [13], [14]. In addition, the phenomenon of choke occurs when the velocity crosses the speed of sound even in a constant cross-section channel flow, but some special tunnels still enable a continuous transonic process [13], [15]. Tang et al. [16] obtained three modes of operation for a supersonic MHD generator through theoretical analysis of a constant cross-section channel flow. The modes of supersonic flow at the exit of the channel depend on the relations between the applied electric field and the critical electric field Ecr. Unfortunately, the corresponding conclusions and further theoretical analysis for variable cross-section channels are still not available.

All the above analyses were performed to study the variation of gas parameters along the flow direction. However, Culick [17] considered this problem from a pathbreaking perspective: he analyzed the MHD channel flow by studying the phase trajectory consisting of the allowed constant cross-sectional flow parameters in the velocity-Mach number phase plane, whose direction is marked by the second law of thermodynamics. In this way, the critical values of the velocity that affect the extreme and asymptotic characteristics of the flow parameters will be visualized as curves in the velocity-Mach number phase plane. Moreover, Powers et al. [18], and Wu [19] plotted the critical velocity in the dimensionless current density-electric field plane as a limiting circle, which can describe the MHD accelerator/generator and the supersonic/subsonic flow simultaneously. However, the lack of physical quantities related to the gas dynamics makes it difficult to consider the coupling effect of the electromagnetic field with the gas’s thermodynamics in a variable cross-section. Recently, Makinde et al. [20], and Hamrelaine et al. [21] obtained approximate solutions for conductive viscous incompressible channel flow (MHD Jeffrey-Hamel flow) under a uniform magnetic field by the perturbation method. Nevertheless, such approximate solutions are difficult to consider the compressible flow.

From the above review, it can be recognized that, on the one hand, the existing analysis of quasi-1D MHD channel flow has been carried out for constant cross-section, but the results have not been extended to the case of variable cross-section. Since most of the cases required in practical engineering are variable cross-section channel flows, and the coupling effects caused by variable cross-section with the electromagnetic field will produce very complex phenomena, deeper investigations of constant/variable cross-section channel flow in the electromagnetic field are of theoretical and engineering value. On the other hand, for the mathematical methodology adopted in theoretical analysis of MHD channel flow, although the direction of the phase trajectory in the phase plane can be obtained by employing the entropic increase principle ds0, other features such as the asymptotic properties, turning points, and fixed points [22] are related to the adjacency of the acceleration/deceleration regions. These characteristics can be easily expressed using graph theory in mathematics. To the authors’ knowledge, no work has been done to study the phase plane and phase trajectory properties using graph theory in MHD channel flow. In consequence, this paper will adopt mathematical tools such as graph isomorphism and equivalence class to study nonlinear second-order dynamical system describing the flow of MHD constant/variable cross-section channel flow to analyze the properties in a more comprehensive and in-depth manner.

Based on the previous results, this paper will study the equations describing the quasi-1D, steady, inviscid, compressible MHD constant/variable cross-section channel flow in the dimensionless velocity-Mach number phase plane. Then, mathematical methods such as labeled graph and equivalence class will be adopted to discuss the characteristics of channel flow. This paper is organized as follows: Part 2 will derive the dimensionless MHD equations under the assumption of low magnetic Reynolds number by novelly selecting dimensionless parameters, and then obtain the second-order nonlinear dynamical system that can comprehensively consider the coupling effects of compressibility and electromagnetism even under the variable cross-section conditions; In Part 3, by analyzing the second-order dynamical system, the distribution of different regions in the phase plane is obtained which can reflect the effects of acceleration/deceleration of velocity and Mach number; Finally, Part 4 will establish equivalence classes based on the isomorphism of labeled graphs, and study several cases through equivalence relations to cover all possible cases of MHD constant/variable cross-section channel flow.

Section snippets

Dimensionless equations for quasi-1D MHD channel flow

For flow problems in aerospace, the low electric conductivity of the gas will result in a dimensionless magnetic Reynolds number Rem much less than one [10]. The magnetic Reynolds number Rem is physically characterized by the induced magnetic field ratio to the applied magnetic field. Thus, the induced magnetic field can be ignored, and the magnetic field is considered as the applied magnetic field for low Rem problems. In this section, based on the simplified MHD equations under low Rem

Theoretical analysis of MHD channel flow

A detailed analysis of the dynamical system is of particular importance to obtain the evolution behavior of the MHD channel flow parameters. There is almost no possibility to integrate a nonlinear dynamical system analytically. Even if analytical solutions exist for a small bunch of nonlinear systems, the systems’ properties and characteristics are usually implicit in the solutions. For general nonlinear systems, Poincaré started to analyze the phase trajectories in the phase plane to acquire

Phase planes classification

In the case of MHD wind tunnels or MHD generators in practical applications, according to the definition of the interaction number, N will vary continuously even under the same operating conditions. Hence it is important to study the variation of the acceleration/deceleration regions in the phase plane with N. The analysis in the previous section shows that there are three critical values (N1*,N2* and N3*) and two limiting values (0 and ) of the interaction number. When N transition across

Conclusions

In this paper, a second-order dimensionless dynamical system with respect to the velocity κ and the Mach number Ma with a parameter of the magnetic interaction number N is obtained to theoretically analyze quasi-1D, steady, inviscid, compressible, MHD channel flow at a low magnetic Reynolds number with constant or variable cross-section. The flow behaviors of MHD channel are acquired for all possible cases by investigating the characteristics of fixed points, phase trajectories and

Acknowledgements

This work was supported by the Foundation for Innovative Research Groups of the National Natural Science Foundation of China through grant 11721202 and the National Numerical Wind Tunnel Project through Grant NNW2018-ZT3A05.

References (23)

  • I.B. Chekmarev

    On the quasi-one-dimensional steady flow of a compressible conducting gas in a pipe of constant cross-section in the presence of transverse magnetic and electric fields

    J. Appl. Math. Mech.

    (1960)
  • O.D. Makinde et al.

    Hermite-Pade approximation approach to MHD Jeffery-Hamel flows

    Appl. Math. Comput.

    (2006)
  • E. Gurijanov et al.

    AJAX - New directions in hypersonic technology

    Space Plane and Hypersonic Systems and Technology Conference

    (1996)
  • V.A. Bityurin et al.

    Thermodynamic analysis of the AJAX propulsion concept

    Aerospace Sciences Meeting and Exhibit

    (2000)
  • J. Tang et al.

    The influence of energy-bypass on the performance of AJAX

    44th AIAA Aerospace Sciences Meeting and Exhibit

    (2006)
  • S.O. Macheret et al.

    External supersonic flow and scramjet inlet control by MHD with electron beam ionization

    39th Aerospace Sciences Meeting and Exhibit

    (2001)
  • M. Nishihara et al.

    Low-temperature m=3 flow deceleration by lorentz force

    Phys. Fluids

    (2006)
  • N. Harada et al.

    Basic studies on an MHD accelerator

    33rd Plasmadynamics and Lasers Conference

    (2002)
  • N. Harada

    MHD acceleration studies at Nagaoka university of technology

    32nd Plasmadynamics and Lasers Conference

    (2001)
  • S.R. Pate et al.

    Development of an MHD-augmented, high enthalpy, shock tunnel facility

    Aiaa J.

    (1974)
  • D.W. Bogdanoff et al.

    Experimental demonstration of MHD acceleration

    34th Plasmadynamics and Lasers Conference

    (2003)
  • Cited by (0)

    View full text