See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/341149874

Compressible Flow through Convergent-Divergent Nozzle

Conference Paper · February 2020

CITATIONS READS

0 1,283

4 authors, including:

Buddhadev Das Rajdeep Sardar

Jadavpur University Jadavpur University
1 PUBLICATION   0 CITATIONS    4 PUBLICATIONS   1 CITATION   

SEE PROFILE SEE PROFILE

Nirmal K. Manna
Jadavpur University
83 PUBLICATIONS   318 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Heat Transfer Analysis on Microtextured Superhydrophilic Surface View project

Numerical investigation of Compressible flow characteristics in a Core and By-pass Nozzle of Titan Engine View project

All content following this page was uploaded by Rajdeep Sardar on 05 May 2020.

The user has requested enhancement of the downloaded file.

 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

Compressible Flow through Convergent-Divergent Nozzle

Buddha Dev Das, Rajdeep Sardar, Sandip Sarkar and Nirmal K. Manna
Department of Mechanical Engineering, Jadavpur University, Kolkata–700032, India
(email: [email protected], [email protected],
[email protected], [email protected])

ABSTRACT

The convergent-divergent nozzle finds application in many areas of industry and technology. The
present work focuses on the CFD simulation of a convergent-divergent nozzle undergoing
compressible flow, as there exist a few works on this topic. The work performed using a standard
geometry to capture the normal shock under different exit pressure conditions. Finite volume method
is to be utilized for the numerical simulations using ANSYS FLUENT. The results are to be analysed
using pressure plot, Mach number plot, different contour plots for pressure, temperature, density,
velocity and Mach number. The study is performed to contribute towards the fundamental knowledge
and practical applications.

Keywords: Compressible flow, CFD, Convergent-divergent nozzle, Normal shock, Expansion.

1. INTRODUCTION

A nozzle is a device designed especially in the form of a tube of varying cross-sectional area
through which hot gas or liquid flows to generate thrust following Newton’s third law of motion. The
nozzle is often used to control the characteristics of fluid flow (specifically the rate of flow that
emerges from the outlet), pressure and the direction of flow, and to enhance the velocity of a gaseous
substance. In the area of compressible flow, the nozzles are typically categorized as a convergent
nozzle and a convergent-divergent (CD) nozzle. Both types of nozzles have serious applications in
industry and technology. In particular, the CD nozzle plays a vital role in the case of a supersonic
version of the missile, jet engines, wind tunnel, ramjets, scramjets, and rocket science as well [1]. A
CD nozzle is frequently to proselytize chemical energy into kinetic energy in a thermal chamber and
vice versa [2].

In the present work, a typical geometry of converging-diverging (CD) nozzle (based on the
standard literature [3]) is considered. For CD nozzle, the Mach number is less than 1.0 in the
converging part (subsonic section) and Mach number is 1.0 at the throat section. In the diverging part
(supersonic section), the Mach number increases over-unity; in other words, the supersonic flow is
achieved particularly for which CD nozzle is designed. However, depending upon the outlet
condition, Mach number can have different values at the exit, and a phenomenon of normal shock
can develop the diverging part before the outlet. The main objective of the present work is to capture
normal shocks under different exit conditions numerically using computational fluid dynamics
(CFD). On the CD nozzle, the scarcity of works [4-6] is observed in the literature. There are a few
works on nozzle flow [7-12] using CFD simulation; however, almost no work has reported earlier on
the analysis of normal shock using the standard geometry [3]. To understand the shock phenomenon
properly and to enrich the knowledge base in the area motivates us to this study. The dynamics of
expansive flow in CD nozzle are analysed in terms of different involved parameters mainly pressure,
temperature, velocity and Mach number. The abrupt changes of pressure and other parameters are
1
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

captured and presented in terms of different contour plots and line curves. The present CFD
simulation is performed using ANSYS FLUENT software.

2. PROBLEM AND NUMERICAL PROCEDURE

2.1 Theory

A CD nozzle is first invented by a Swedish scientist named De Laval, and now it is known as
de Laval nozzle or converging-diverging nozzle. This type of nozzle is required to gain supersonic
speed. Due to the certain values of backpressure [7] at the exit of the nozzle (or the overall pressure
differential across the nozzle) under a supersonic flow condition, the supersonic flow suddenly
transforms into a subsonic flow resulting normal shock only in the divergent part of the nozzle. The
expansion in the converging section continues as a subsonic flow and becomes sonic flow at the
throat. Afterward, the flow becomes supersonic before the shock occurs. After the sudden shock, the
flow becomes subsonic which continues to the exit of the nozzle. Consequently, a small continuous
decrease in Mach number takes place in the diverging area of CD nozzle, but the density, temperature
and pressure also increase after the shock at the exhaust section. According to the principle of energy
conservation, the stagnation pressure and enthalpy does not vary across the normal shock.

(a)

(b)
Figure 1: Problem description: (a) 2D geometry of the converging-diverging nozzle, (b)
axisymmetric grid generation
2
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

2.2. Geometry and mesh

This paper aims to simulate Anderson’s geometry model [3] to capture the normal shocks at
different exit conditions. The shape of the converging-diverging nozzle is specified by A=A(x),
considering a parabolic profile as given by

𝐴(𝑥) = 1.0 + 2.2(𝑥 − 1.5)2 (1)

where 0 ≤ x ≤ 3 and x ≤ 1.5 indicates the convergent section, x =1.5 represents the throat of the nozzle,
and the divergent section implies for x 1.5 [1]. For the purpose of the simulation, the axisymmetric
geometry is considered. The corresponding geometric model is shown in Fig. 1a in the platform of
ANSYS FLUENT. The generation of computation mesh is depicted in Fig. 1b, showing 2D structured
grids. The structured mesh is employed for more accuracy of this simulation. The mesh consists of
14271 nodes and 14000 elements with refine mesh of 0.0001 sizing for all cases.

2.3. Mathematical Formulation

As discussed earlier, in a C-D nozzle, the normal shock occurs in the diverging portion,
however, the location of the shock and the minimum pressure prevailed during the shock are the most
significant questions regarding the same. To address the same the present work is undertaken using
the facility of ANSYS FLUENT. In the software, there are multifarious governing equations
amalgamated with convergent-divergent nozzle, which should be taken into account along with the
elementary theory of CFD. The set of equations is as follows:

The equation for an ideal gas is denoted by

𝑃
𝜌 = 𝑅𝑇 (2)
where 𝑝 = pressure (Pa), 𝑇 = temperature (K), 𝑅 = universal gas constant (J/mol. k).

As well as, the sound speed is given by

𝑐 = √𝛾𝑅𝑇 (3)
𝑉
Mach number can be obtained by the following equation, = 𝐶 , where v is the flow velocity and c are
the sonic velocity.

The conservation of mass and momentum equations for a Newtonian fluid is known as Navier-
Stokes equations. For the present 2D compressible flow, considering steady-state condition and
neglecting temporal term yields:

The expression of continuity equation is

𝜕𝜌𝑢𝑗
=0 (4)
𝜕𝑥𝑗
and the momentum equations are
𝜕𝜌𝑢𝑖 𝑢𝑗 𝜕𝑝 𝜕𝜎𝑖𝑗
= − 𝜕𝑥 + + 𝛱𝜕𝑖1, (5)
𝜕𝑥𝑗 𝑗 𝜕𝑥𝑗

The steady flow energy equation becomes

𝜕𝑇
𝜕𝜌𝑢𝑗 𝑇 𝜕(𝜌𝛼 )
𝜕𝑥𝑗
= , (6)
𝜕𝑥𝑗 𝜕𝑥𝑗
3
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

Where, 𝑢𝑖 is the velocity component in the 𝑖𝑡ℎ direction, p is the thermodynamic pressure in Pa,
𝜌 is the fluid density in 𝑘𝑔/𝑚3 , and 𝜎𝑖𝑗 is the viscous stress.

In this work, we have approached the simulation using compressible flow through a CD nozzle.
Standard ideal gas (Sutherland Model) was used as the flowing material whose CP =1006.43 J/kgK
and thermal conductivity = 0.0242 W/mK. The solver was set as a pressure-based option; however,
in a steady-state situation, the flow was taken laminar, keeping the energy equation on. All standard
initialization of the software was computed from the inlet section. The CD nozzle is simulated with
the help of axisymmetric conditions in 2D space. The boundary conditions are: at inlet pressure
=101325 Pa, temperature = 300K, and for outlet section pressure = 68740 Pa. The outlet pressure was
varied to capture the shock at different exit pressure conditions. In the case of the wall side of the CD
nozzle, ‘no-slip condition’ was provided, as the wall is stationary. There are some other parameters
like relaxation factor taken as 0.25 (for flow variable only), the reference frame (relative to cell zone),
the number of iterations = 1000, to be set properly for getting the final solution under the normal
shock condition.

3. RESULTS AND DISCUSSION

In our work, the standard air is the working fluid. It flows at 1.0 atmospheric pressure (100 kPa)
from the left-hand side of the C-D nozzle and via throat section, finally exits from the right end. The
simulations have been conducted under the pressure variation in the range of 10 to 68.740 kPa. It has
been observed that from 10 to 30 kPa, the normal shock is not visible. However, when the pressure
difference between the entry and exit becomes around 50 to 68.740 kPa, the shocks occur in the
divergent region due to the over expansion. This may be attributed to the fact that the pressure is one
of the significant parameters on which the formation of shockwave depends. All these matters are
observed in simulation with the help of different contour and plot of various parameters including
density, velocity, pressure and Mach number also as illustrated below in Figs. 2-9 for the different
situations.

(a) (b)
Figure 2: Density contours at exit pressures 10 kPa (a) and 30 kPa (b)

4
Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

(a) (b)
Figure 3: Density contours at exit pressures 50000 Pa (a) and 68740 Pa (b)

(a) (b)
Figure 4: Pressure contours at exit pressures 10 kPa (a) and 30 kPa (b)

(a) (b)
Figure 5: Pressure contours at exit pressures 50000 Pa (a) and 68740 Pa (b)

5
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

The abrupt change in colour in the plot of Mach number (Figs. 6 and 7) indicates the region of
shock, which starts to fall with a definite mutation. It is observed that Mach number increases up-to
throat section and afterward at the shock zone its starts to decrease, where the temperature is
minimum. A schematic view of contours of Mach numbers at different pressure conditions is shown
in the figures below.

(a) (b)
Figure 6: Mach numbers at exit pressures 10 kPa (a) and 30 kPa (b)

(a) (b)
Figure 7: Mach numbers at exit pressures 50000 Pa (a) and 68740 Pa (b)

The Mach number variation in the case of above CD nozzle, at different pressures at the nozzle
exit, is depicted in Fig. 8, and the overall pressure variation along the axis viz. the present of normal
shock is illustrated in Fig. 9. As the exit pressure changes, the location of the normal shock changes.
The normal shock shifts towards right when exit pressure decreases.

6
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

Figure 8: Mach number variation for different outlet pressure

Figure 9: Pressure variation along axis

4. CONCLUSIONS

In the aforementioned geometry, the maximum value of Mach number is in a place where
pressure is minimum and at throat region pressure value should be near about backpressure. Up to
30000 Pa Mach number curves are smooth just before the exit but increasing exit pressure, we have
noticed shockwave occurs after throat, but before outlet boundary, as expected from theoretical
understandings.

REFERENCES

[1] Md Iliyaas, A. and Karuppasamy, K. Computational analysis of CD nozzle for solid propellant
rocket. International Journal of Innovative Works in Engineering and Technology (IJIWET)
2(2016), 314-324

7
 Proceedings of the International Conference on Innovations in Thermo-Fluid Engineering and Sciences
[ICITFES - 2020] NIT Rourkela, India, 10-12 February 2020

[2] Sudhakar, B.V.V.N., Sekhar, B.P.C., Mohan, P.N. and Md Ahmad, T. Modeling and simulation
of convergent-divergent nozzle using computational fluid dynamics. International Research
Journal of Engineering and Technology (IRJET) 03(2016), 2395-0072
[3] Anderson, J.R. Computational fluid dynamics; the basics with application. (McGraw Hill, Inc.
United States of America, 1995)
[4] Malay S.P., Mane, S.D. and Raman, M. Concepts and CFD analysis of De-Laval nozzle.
International Journal of Mechanical Engineering and Technology (IJMET) 7(2016), 221–240
[5] Modesti, D., Pirozzolic, S., Grassoa, F. Direct numerical simulation of developed compressible
flow in square ducts. International Journal of Heat and Fluid Flow 76(2019), 130-140
[6] Mohan, K.G., Dominic, X.F. and Kumar, R.M. Design and optimization of De Laval nozzle to
prevent shock induced flow separation. Advances in Aerospace Science and Applications
3(2013), 119-124
[7] Gopi, G. and Rao, B.R. Effect of back pressure on nozzle flow. International Journal of
Research in Advent Technology Special Issue (March 2019) on Conference Proceedings of 3rd
National Conference on Recent Trends & Innovations in Mechanical Engineering 15-16th
March 2019.
[8] Kumar, R.R. and Devarajan, Y. CFD simulation analysis of two-dimensional convergent-
divergent nozzle. International Journal of Ambient Energy (2018) DOI: 10.1080/ 01430750.
2018.1517683
[9] Yu, Y., Shademan, M., Barron, R.M. and Balachandar, R. CFD Study of Effects of Geometry
Variations on Flow in a Nozzle. Engineering Applications of Computational Fluid Mechanics
6(2012), 412-425
[10] Madhu, B.P., Syed Sameer, Kalyana, K.M. and Mahendra, M.G. CFD Analysis of Convergent-
Divergent and Contour Nozzle. International Journal of Mechanical Engineering and
Technology 8(2017), 670–677
[11] Narayana, K.P.S.S. and Reddy K.S. Simulation of Convergent Divergent Rocket Nozzle using
CFD Analysis. IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) 13(2016),
58-65
[12] Aabid, A., Chaudhary, Z.I. and Khan, S.A. Modelling and Analysis of Convergent Divergent
Nozzle with Sudden Expansion Duct using Finite Element Method. Journal of Advanced
Research in Fluid Mechanics and Thermal Sciences 63(2019), 34-51

View publication stats