8
Scientia Iranica B (2012) 19 (6), 1511–1518 Sharif University of Technology Scientia Iranica Transactions B: Mechanical Engineering www.sciencedirect.com Investigation of the common nose cone shapes in different gas mixtures in high Knudsen numbers Y. Amini a,, H. Emdad a , K. Akramian b , F. Bordbar b a School of Mechanical Engineering, Shiraz university, Shiraz, Iran b Department of Mechanical Engineering, Islamic Azad University, Neyriz Branch, Neyriz, Iran Received 15 January 2012; revised 23 July 2012; accepted 26 September 2012 KEYWORDS DSMC; Knudsen number; Slip; Transitional; Free molecular; Nose cone shapes. Abstract The numerical simulation of different types of nose cone shapes involving conical, Bi-conic, parabolic, spherical Blunt cone and tangent ogive in high Knudsen numbers flow and for three different mixtures O 2 –N 2 , He–Ar and He–Xe are investigated in this article. A computer code based on the method of Direct Simulation Monte Carlo (DSMC) has been developed to study numerically the supersonic flow around these cone shapes in rarefaction condition. In order to verify the computer code, the flow field in the slip regime is solved and compared with the results of Navier Stokes equation with the slip boundary condition. The results obtained in the supersonic flows regime show that the parabolic and tangent ogive nose shapes have the least mean shear stress distribution and lowest tip temperature. © 2012 Sharif University of Technology. Production and hosting by Elsevier B.V. 1. Introduction The nose cone is the typical shape for almost all supersonic vehicles and outer space spaceships. These vehicles normally fly in high altitude and eventually may depart from earth atmo- sphere. The air properties in this flight region are characterized by low density, so molecular mean free path is taking a signif- icant value. Thus, the flight Knudsen number (Kn = λ L where λ and L is the mean free path and characteristic length, respec- tively) is increased. For examining the flow fields around these nose cone shapes, those governing equations need to be considered, which also include the effect of high rarefied fluid. The general governing equation in the rarefied gas flow is the Boltzmann equation. Practically, the level of non-equilibrium thermodynamics or gas rarefaction can be represented by Kn number. In general, for problems with Kn < 0.01, equilibrium thermodynamics will be Corresponding author. Tel.: +98 711 2303051; fax: +98 711 6473511. E-mail addresses: [email protected] (Y. Amini), [email protected] (H. Emdad), [email protected] (K. Akramian), [email protected] (F. Bordbar). established and the flow field can be represented by continuum governing equations such as Navier Stokes equations [1,2]. With increasing flow Kn number, the non-equilibrium thermodynamics effects are becoming an important factor. In the flight regime with Kn in the range of 0.01 and 0.1 (commonly known as the slip flow regime), the compressible Navier Stokes equations with the slip boundary condition is valid. In addition, in this range the rarefaction effects are incorporated via the velocity slip as well as the temperature jump boundary conditions. Finally, the Kn number in the range of 0.1–10 is known as the transition flow regime [1,2]. For the analysis of the transition flow regime, the Boltzmann equation, which is valid over the whole range of the Kn number, should be considered. Due to complexities associated with collision term, analytical and numerical solutions of this equation currently are not feasible. Various alternatives have been proposed to approximate the Boltzmann equation. The first one is the extension of the hydro- dynamic equations which is known as the Burnett equations. In this method, the higher order non-linear constitutive relations are adopted for the continuum Navier Stokes equations, which cause departure from thermodynamic equilibrium. The Bur- nett equations are highly non-linear partial differential equa- tions in which its numerical analysis is normally prompted with some instability. The second approach is the Direct Simulation Monte Carlo (DSMC) method which is developed by Bird [3]. This method is a kinetic model based on particle simulation. Peer review under responsibility of Sharif University of Technology. 1026-3098 © 2012 Sharif University of Technology. Production and hosting by Elsevier B.V. doi:10.1016/j.scient.2012.10.029 Open access under CC BY-NC-ND license. Open access under CC BY-NC-ND license.

Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

Scientia Iranica B (2012) 19 (6), 1511–1518

Sharif University of Technology

Scientia IranicaTransactions B: Mechanical Engineering

www.sciencedirect.com

Investigation of the common nose cone shapes in different gasmixtures in high Knudsen numbersY. Amini a,∗, H. Emdad a, K. Akramian b, F. Bordbar b

a School of Mechanical Engineering, Shiraz university, Shiraz, IranbDepartment of Mechanical Engineering, Islamic Azad University, Neyriz Branch, Neyriz, Iran

Received 15 January 2012; revised 23 July 2012; accepted 26 September 2012

KEYWORDSDSMC;Knudsen number;Slip;Transitional;Free molecular;Nose cone shapes.

Abstract The numerical simulation of different types of nose cone shapes involving conical, Bi-conic,parabolic, spherical Blunt cone and tangent ogive in high Knudsen numbers flow and for three differentmixtures O2–N2, He–Ar and He–Xe are investigated in this article. A computer code based on the methodof Direct Simulation Monte Carlo (DSMC) has been developed to study numerically the supersonic flowaround these cone shapes in rarefaction condition. In order to verify the computer code, the flow field inthe slip regime is solved and compared with the results of Navier Stokes equation with the slip boundarycondition. The results obtained in the supersonic flows regime show that the parabolic and tangent ogivenose shapes have the least mean shear stress distribution and lowest tip temperature.

© 2012 Sharif University of Technology. Production and hosting by Elsevier B.V.Open access under CC BY-NC-ND license.

1. Introduction

The nose cone is the typical shape for almost all supersonicvehicles and outer space spaceships. These vehicles normallyfly in high altitude and eventually may depart from earth atmo-sphere. The air properties in this flight region are characterizedby low density, so molecular mean free path is taking a signif-icant value. Thus, the flight Knudsen number (Kn =

λL where

λ and L is the mean free path and characteristic length, respec-tively) is increased.

For examining the flow fields around these nose cone shapes,those governing equations need to be considered, which alsoinclude the effect of high rarefied fluid. The general governingequation in the rarefied gas flow is the Boltzmann equation.Practically, the level of non-equilibrium thermodynamics or gasrarefaction can be represented by Kn number. In general, forproblemswith Kn < 0.01, equilibrium thermodynamics will be

∗ Corresponding author. Tel.: +98 711 2303051; fax: +98 711 6473511.E-mail addresses: [email protected] (Y. Amini),

[email protected] (H. Emdad), [email protected](K. Akramian), [email protected] (F. Bordbar).Peer review under responsibility of Sharif University of Technology.

1026-3098© 2012 Sharif University of Technology. Production and hosting by Els

doi:10.1016/j.scient.2012.10.029

established and the flow field can be represented by continuumgoverning equations such as Navier Stokes equations [1,2].

With increasing flow Kn number, the non-equilibriumthermodynamics effects are becoming an important factor.In the flight regime with Kn in the range of 0.01 and 0.1(commonly known as the slip flow regime), the compressibleNavier Stokes equations with the slip boundary condition isvalid. In addition, in this range the rarefaction effects areincorporated via the velocity slip as well as the temperaturejump boundary conditions. Finally, the Kn number in the rangeof 0.1–10 is known as the transition flow regime [1,2].

For the analysis of the transition flow regime, the Boltzmannequation, which is valid over the whole range of the Knnumber, should be considered. Due to complexities associatedwith collision term, analytical and numerical solutions of thisequation currently are not feasible.

Various alternatives have been proposed to approximate theBoltzmann equation. The first one is the extension of the hydro-dynamic equations which is known as the Burnett equations. Inthis method, the higher order non-linear constitutive relationsare adopted for the continuum Navier Stokes equations, whichcause departure from thermodynamic equilibrium. The Bur-nett equations are highly non-linear partial differential equa-tions inwhich its numerical analysis is normally promptedwithsome instability. The second approach is the Direct SimulationMonte Carlo (DSMC) method which is developed by Bird [3].This method is a kinetic model based on particle simulation.

evier B.V. Open access under CC BY-NC-ND license.

Page 2: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

1512 Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518

Nomenclature

Kn Knudsen numberL Characteristic length of the flow fieldλ Molecular mean free pathM Mach number, = U

√γ RT =

U√γ κ

m T

P PressureT Temperatureρ Densityn Number densityc Molecule velocityd Molecular diameterm Molecular massκ Boltzmann constant = 1.380658 × 10−23

1V Cell volumeFN Number of real molecules represented by a

simulated moleculesPw Wall pressureτw Wall shear stressqw Wall heat fluxγ Ratio of specific heatsm Molecular massU Velocity of oncoming flow

Boyd et al. investigated the experimental and numerical so-lutions for low density nozzle and plume flow of nitrogen byusing DSMC method [4]. Also, Campbell solved shock interfer-ence predictionwith DSMCmethod [5]. Jain and Ramachandrandeveloped an object-oriented code based on Direct SimulationMonte Carlo method for modeling of rarefied flow around ge-ometries of arbitrary shapes [6]. Wang and Bao investigate theAerothermodynamics of hypersonic small nose cone with localrarefied gas effects [7]. Lilley and Macrossan used an alterna-tive technique instead of acceptance–rejection method for im-plementing the stream boundary condition [8]. Sengil and Edisused highly efficient volume generation reservoirs in molecu-lar simulations of gas flows [9]. Scanlon et al. used OpenFOAMDSMC to solve flow around a cylinder [10]. Liu et al. used anobject-oriented serial implementation of the DSMC simulationpackage to investigate hypersonic flow over a wedge at M =

10 [11].According to the authors’ knowledge the flow field in

high Knudsen number around nose cone shapes has not beeninvestigated to date. The purpose of the current study is toconsider the flow around five commonnose cone shapes in highKnudsen number.

2. Direct Simulation Monte Carlo (DSMC) method

The DSMC which is a method was first introduced by Bird[3,12]. This method is based on the direct simulation of gasflow and the kinetic theory. In DSMC method, for modelinggas flow, the enormous number of simulated particles is used,while each simulated particles represents large number ofmolecules. The DSMCmethod can bementioned as an unsteadyand explicit timemarchingmethod; thus, it can be used directlyfor unsteady gas flow. For steady flow, the simulation willbe continued until the change of flow field variables becomestrifling. The major approximation of this method is that duringeach step the deterministicmolecularmotions andprobabilisticintermolecular collisions are decoupled from each other.

The DSMC method is made of four essential processes.In the first step, the particles travel deterministically basedon their velocities and if they cross the boundary, theboundary condition is enforced. The Maxwellian diffusive andthe specular reflection model are used for the solid surfaceboundary and symmetric boundary, respectively. The secondstep is indexing and tracking of the particles. During the firststep the position of particles are changed, and therefore theircell numbers will be altered. Moreover, the new particles areentered from inlet boundary. Therefore, before proceeding tothe third and fourth step, the cell number of particles mustbe specified. The third step is the modeling of the collisionbetween particles. In DSMC method, the simulation of collisionof particles is done statistically. It is noteworthy that onlythe particles located in the same cell are allowed to beselected as collision partner. In the present work, No TimeCounter (NTC) technique of Bird (1994), in conjunction withthe sub-cell technique, is used for modeling the collision ofparticles [3].Moreover, the Variable Hard Sphere in conjugationwith Larsen–Borgnakke phenomenological model [13] is usedto simulate inelastic binary collision. Sampling is the last stepof DSMC method. The macroscopic properties of a flow fieldare sampled from field cells, and the value represents the flowproperty at the cell center. For a steady state problem, the flowfields sampling is started when the simulation has reached asteady state. There are two types of averaging processes; thefirst one is the time average and the other one is the ensembleaverage. Time averaging is used for steady problems and inthis method, macroscopic properties is obtained over manytime steps. In ensemble averaging, sampling process takes overa large number of instantaneous averages in the same spaceelement at the same time and this method is used in unsteadyproblem to increase the sample size. The flow density andvelocity can be calculated through the following equations:

N =

Nti=1

Ni, ρ =1Nt

NFN1V

m,

U =1N

Nti=1

Nij=1

cu,j, V =1N

Nti=1

Nij=1

cv,j,

W =1N

Nti=1

Nij=1

cw,j,

(1)

where Ni is the number of molecules in the cell at currenttime step and Nt the number of time steps. 1V ,m and FNare the volume of the cell, mass of the simulated moleculeand the number of real molecules represented by a simulatedmolecules, respectively. Here, cu,j, cv,j, cw,j are the componentsof velocities vector of jth simulated molecule in x, y and zdirections, respectively. ρ is the flow density and U, V , and Ware the mean flow velocities in the center of cell in x, y and zdirections, respectively. The flow temperature and pressure arecalculated as follows:

Tx =mκ

1N

Nti=1

Nij=1

c2u,j −

1N

Nti=1

Nij=1

cu,j

2 ,

Ty =mκ

1N

Nti=1

Nij=1

c2v,j −

1N

Nti=1

Nij=1

cv,j

2 , (2)

Tz =mκ

1N

Nti=1

Nij=1

c2w,j −

1N

Nti=1

Nij=1

cw,j

2 ,

Page 3: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518 1513

Figure 1: Geometry of all nose cone shapes.

T =13

Tx + Ty + Tz

, P =

1Nt

NFN1V

κT , (3)

where Tx, Ty and Tz are the thermal temperatures in the x, y andz directions, respectively, T is the flow temperature and P is thepressure of the flow.

Pressure (Pw), shear stress (τw) and heat transfer rate (qw)on the surface are calculated by Eqs. (4)–(6):

Pw =FN

ts1A

Nsj=1

mcrn,j − c in,j

, (4)

τw =FN

ts1A

Nsj=1

mcrt,j − c it,j

, (5)

qw =FN

ts1A

Nsj=1

12mc2j

i

Nsj=1

12mc2j

r

. (6)

In the above equations, the superscript ‘‘i’’ denotes the valuesof the incident to the wall and ‘‘r ’’ recognizes the valuesreflected from the wall and cn and ct are the normal andtangential components of molecular velocity, respectively. Nsis the total number of molecules colliding with solid surfaceduring sampling time ts.

3. Results

In the current study, the flow field around conical, Bi-conic, parabolic, spherical Blunt and tangent ogive nose coneshapes are investigated. The geometry of these bodies is shownin Figure 1. The length of forebody is 0.3 m. The diameterand length of aftbody circular cylinder are 0.2 m and 1.5 m,respectively. The simulation is done for two Knudsen numbers(Kn = 0.05, 0.5) and three different gas mixtures, such asO2–N2, He–Ar and He–Xe.

In order to validate the results, the slip regime is investigatedby the DSMC method and the numerical approximation ofthe Navier–Stokes equations with Maxwell’s velocity slipand Smoluchowski temperature jump boundary conditions.The relations for velocity slip and temperature jump are asfollows [1]:

Us − Uw =2 − σu

σuKn

∂Us

∂n+

32π

(γ − 1)γ

Kn2ReEc

∂T∂s

Ts − Tw =2 − σT

σT

γ + 1

KnPr

∂T∂n

(7)

where Uw and Tw and the non-dimensional wall velocity andtemperature, respectively. γ is the ratio of specific heats, σv andσT are the tangential momentum and energy accommodationcoefficients and Ec, Re and Pr are the Eckert, Reynolds and

Prandtl numbers, respectively. Also, n and s denote the unitoutward normal and the unit tangential vector of wall.

Finally, the flow field in the transition regime is solved byDSMC method.

3.1. Slip regime

For the selected Knudsen number (Kn = 0.05), both thetangential momentum and the energy accommodation coeffi-cient are selected equal to 0.85. Also, the incident Mach num-ber, ambient temperature andwall temperature are consideredM = 5, T = 250 K, T = 500 K, respectively. The pressure atthe far field boundaries will be calculated from the followingequation:

n =1

√2πd2L Kn

p = nkT . (8)

The mixture components are nitrogen and oxygen and theirconcentration fraction are 0.8 and 0.2, respectively.

The velocity in x direction and pressure contours for DSMCmethod and Navier Stokes Equations (NSE) are compared inFigure 2. In this figure, the NSE solution is shown in the upperhalf and the DSMC results are shown in the lower half. As canbe seen, the pressure contours are in the excellent agreementbut the velocity contours illustrate some differences. The sourceof this discrepancy is related to the inaccuracy of Maxwell’svelocity slip boundary condition. As mentioned before, thisboundary condition is valid for the Kn between 0.01 and 0.1and, therefore, with increasing Kn from 0.01, the accuracy ofthis boundary condition is reduced. It is worth to mention thatthe NSE solution contours are smoothwhile the DSMC contoursshow some wiggles. This is mainly due to the statistical natureof DSMC. For both methods, the pressure and slip velocity atthe wall surface are compared in Figure 3. The results showthat the predicted pressure field is in good agreement, but thereis some discrepancy in the velocity field. For all of the nosecone shapes, maximum difference in velocity occurred nearthe end portion of the nose cone. The tangent ogive has thegreatest discrepancy with respect to the other nose shapes. Themaximumdifferences between x component of the velocity andpressure predicted by DSMC and FVM are about 10% and 1%,respectively.

3.2. Transitional regime

3.2.1. Oxygen and nitrogen mixtureIn the second case for Kn = 0.5, the wall surface is consid-

ered to be adiabatic and all other parameters are selected thesame as the first case (Kn = 0.05). Figure 4 illustrates the Machnumber and temperature contours of DSMC results for all nosecone shapes. As seen from this figure, it is clear that there is notany strong shock in this regime as compared with the previoushigh dense gas example (Kn = 0.05). Furthermore, as it is ap-parent from Figure 4, the Mach number and temperature fieldare smoothly changed from the ambient values to the surfacevalues. Figure 5 represents the pressure at the wall surface. Asit is shown, the stagnation point is formed at the front of thespherical blunt cone shape, whereas at this point the pressure isabout 27 times greater than the ambient pressure. The pressureat the tip of the tangent ogive, parabolic, bi-conic and conicalis about 13, 9, 8, and 6 times more than the ambient pressure,respectively. Moreover, the parabolic and Bi-conic have theminimum mean pressure where the spherical blunt has the

Page 4: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

1514 Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518

Figure 2: Comparison between solution of Navier–Stokes equations with slip BC and DSMC results.

Page 5: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518 1515

Figure 3: Comparison between solution of Navier Stokes equations with slip BC and DSMC results on the solid surface.

maximum value on the surface. The pressure contour showssmooth changes of pressure for the tangent ogive and theparabolic nose shapes but, for the other nose shapes, there isa pressure jump at the end of the nose cone. Figure 6 representsthe slip velocity at the wall surface. Regarding this figure, thespherical blunt cone’s stagnation point is formed at the tip of

the cone and also all nose cone shapes have the same slip ve-locity after x > 0.4. In other words, the shape of the nose conedoes not have a noticeable effect on the slip velocity distribu-tion on the cylindrical body region.

The tangent ogive and parabolic nose shapes have a smoothchange in the slip velocity distribution, but the other nose

Page 6: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

1516 Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518

Figure 4: Mach number and temperature contours of DSMC method Kn = 0.5.

Figure 5: Pressure contours of DSMC method on the cone surface for all nosecone shapes at Kn = 0.5.

shapes have a constant slip velocity on the nose cones and alsoa jump at the end portion of nose cones. Figure 7 representsthe wall shear stress at the wall surface. Similar to the pressureand slip velocity figures, wall shear stress for the tangent ogiveand the parabolic nose has a smooth change, but the conicaland spherical blunt cone have a constant wall shear stress on

Figure 6: Slip velocity contours of DSMC method on the cone surface for allnose cone shapes at Kn = 0.5.

the cone surface while the bi-conic wall shear stress graphrepresents a jump. In addition, as can be seen from Figure 7, thetangent ogive and parabolic cone shapes have the lowest meanshear stress on the surface where the bi-conic and conical havethe highest value of mean shear stress on the wall.

Page 7: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518 1517

Figure 7: Wall shear contours of DSMCmethod on the solid surface for all nosecone shapes at Kn = 0.5.

Figure 8: Temperature contours of DSMC method on the solid surface for allnose cone shapes at Kn = 0.5.

Figure 9: Temperature of mixture components (Oxygen, Nitrogen) on the conesurface.

Figure 8 represents the temperature distribution at thewall surface. The variation of temperature on the wall issimilar to other properties, i.e. smooth change for tangent ogiveand parabolic nose, but constant value for other shapes. Asshown in Figure 8, the parabolic nose shape has the minimumtemperature on the tip and the tangent ogive has the minimummean temperature on the cone surface. Finally, in Figure 9the temperature of two components of mixture, i.e. oxygenand nitrogen, is plotted. Since the differences in molecularweights of these gases are small, the temperature of the twocomponents have a low difference.

3.2.2. Helium and argon mixtureIn this section, just the parabolic nose shape is examined

for the fluid mixture of helium and argon. In this case, theconcentration fraction of both components is considered 0.5.

Figure 10: Temperature of mixture components (Helium, Argon) and overalltemperature on the cone surface.

Figure 11: Concentration fraction of mixture components (Helium, Argon) onthe cone surface.

Figure 10 shows the temperature distribution of mixturescomponents on the solid surface. As can be seen on the solidsurface, the temperature of heavy component (Ar) is largerthan the temperature of light component (He). The maximumtemperature for He and Ar is 2200 K and 900 K, respectively.Regarding Figure 11, the fraction of heavy component (Ar)increases on the surface and reaches to the maximum valueabout 63%, where the fraction of light component decreasesand reaches the minimum value about 37%. In other words,the heavy component near the surface has a high density andtemperature, whereas other side of the light component has alow density and temperature.

3.2.3. Helium and xenon mixtureSimilar to the previous sections, in this section the

concentration fraction of helium and xenon is also set equal to0.5. In Figures 12 and 13 the temperature and concentrationfraction of the components are plotted. Since the xenonis much heavier than the helium (almost 32 times), thedifference between the temperature and concentration fractionof components increase. The maximum value of temperaturefor xenon reaches 6500 K while the maximum temperature ofHe is about 1600 K at the tip of nose cone.

4. Conclusion

In the present study, the flow field around five special nosecone shapes (conical, Bi-conic, parabolic, spherical Blunt, tan-gent ogive) at two different Knudsen numberswas investigated.The results show that the spherically blunt cone shape has themaximum tip pressure and the parabolic and Bi-conic have the

Page 8: Investigation of the common nose cone shapes in different ... · ScientiaIranicaB(2012)19(6),1511–1518 SharifUniversityofTechnology Scientia Iranica TransactionsB:MechanicalEngineering

1518 Y. Amini et al. / Scientia Iranica, Transactions B: Mechanical Engineering 19 (2012) 1511–1518

Figure 12: Temperature of mixture components (Helium and Xenon) andoverall temperature on the cone surface.

Figure 13: Fraction of mixture components (Helium and Xenon) on the conesurface.

minimum tip pressure. Also, the spherical blunt shape hasmax-imum mean pressure on the surface. In addition, the parabolicnose shapehas theminimumtip temperature and tangent ogivehas the minimummean surface temperature. Furthermore, thetangent ogive and parabolic cone shapes have the minimummean shear stress on the surface. Besides, the heavy componentof fluid mixture moves close to the surface and consequentlyits density increases near the cone surface. Finally, the temper-ature of heavy component gas is greater than the light compo-nent gas at the region close to the cone surface and stagnationpoint.

References

[1] Karniadakis, G., Beskok, A. and Aluru, N., Microflows and NanoflowsFundamentals and Simulation, Springer press, New York, NY (2005).

[2] Shen, Ch.,Rarefied GasDynamics Fundamentals Simulations andMicro Flows,Springer press, Berlin (2004).

[3] Bird, G.A., Molecular Gas Dynamics and the Direct Simulation of Gas Flows,Clarendon Press, Oxford, United Kingdom (1994).

[4] Boyd, I.D., Penko, P.F., Meissner, D.L. and Witt, K.J.D. ‘‘Experimentaland numerical investigations of low-density nozzle and plume flows ofnitrogen’’, AIAA Journal, 30, pp. 2453–2461 (1992).

[5] Campbell, D.H. ‘‘DSMC analysis of plume-free stream interactionsand comparison of plume flow field predictions with experimentalmeasurement’’, In 26th Thermo-physics Conference, AIAA, Honolulu p. 16(1991).

[6] Jain, S. and Ramachandran, P. ‘‘Development of object-oriented directsimulation Monte Carlo code for modeling of rarefied flow around ge-ometries of arbitrary shapes’’, Applied Mechanics and Materials, 110(116),pp. 2491–2496 (2012).

[7] Wang, Z. and Bao, L. ‘‘Aerothermodynamics of hypersonic small nose conewith local rarefied gas effects’’, Chinese Journal of Computational Physics,27(1), pp. 59–64 (2010).

[8] Lilley, C.R. and Macrossan, M.N. ‘‘Methods for implementing the streamboundary condition in DSMC computations’’, International Journal forNumerical Methods in Fluids, 42(12), pp. 1363–1371 (2003).

[9] Sengil, N. and Edis, F.O. ‘‘Highly efficient volume generation reservoirsin molecular simulations of gas flows’’, Journal of Computational Physics,228(12), pp. 4303–4308 (2009).

[10] Scanlon, T.J., Roohi, E., White, C., Darbandi, M. and Reese, J.M. ‘‘An opensource, parallel DSMC code for rarefied gas flows in arbitrary geometries’’,Computers & Fluids Journal, 39, pp. 2078–2089 (2010).

[11] Liu, H., Cai, C. and Zou, C. ‘‘An object-oriented serial implementation ofa DSMC simulation package’’, Computers & Fluids Journal, 57, pp. 66–75(2012).

[12] Bird, G.A. ‘‘Forty years of DSMC, and now in rarefied gas dynamics’’, In 22ndInternational Symposium, Sydney, Australia, 9–14 July 2000, AmericanInstitute of Physics (2001).

[13] Borgnakke, C. and Larsen, P.S. ‘‘Statistical collision model for Monte Carlosimulation of polyatomic gas mixtures’’, Journal of Computational Physics,18, pp. 405–420 (1975).

Yasser Amini received his B.S. and M.S. degrees in Mechanical Engineeringfrom Shiraz University, Iran, in 2006 and 2009, respectively. He is currentlya Ph.D. student in the school of Mechanical Engineering at Shiraz University.His research interests include: CFD, mesh-less methods, SPH method, fluidstructure interaction, rarefied gas flow dynamics and DSMC method.

Homayoon Emdad received his Ph.D. and M.S. degrees in AerospaceEngineering from Kansas University, U.S.A. in 1988 and 1984, respectively, andhis B.S. degree in Aerospace Engineering from Washington University in 1982.He is currently an Associate Professor in the school of Mechanical Engineeringat Shiraz University. His primary research interests are computationalfluid dynamics, flow control, fluid structure interaction, mesh-less methods,multiphase flow, and aerodynamics.

Kourosh Akramian was born in 1976, in Iran. He graduated from HormozganUniversity with a Bachelor’s degree. He also graduated from Power & WaterIndustry University with a Master’s degree of Mechanical Engineering. He iscurrentlyworking as a university Lecturer in the Islamic AzadUniversity branchof Nayriz. His research focuses on CFD and rarefied gas dynamics.

Farhad Bordbar received his B.S. and M.S. degrees in Aerospace Engineeringfrom Isfahan University of Technology and Shiraz University in 2004 and 2009,respectively. He is currently working as a university Lecturer in the IslamicAzad University branch of Nayriz. His research focuses on CFD and rarefied gasdynamics.