Defence Science Journal, Vol. 65, No. 4, July 2015, pp. 261-264, DOI : 10.14429/dsj.65.

7960
 2015, DESIDOC

Numerical Characterisation of Jet-Vane based Thrust Vector Control Systems

M.S.R. Chandra Murty and Debasis Chakraborty*
Directorate of Computational Dynamics, Defence Research and Development Laboratory, Hyderabad-500 058, India
*
E-mail: [email protected]

Abstract
Computational fluid dynamics methodology was used characterise jet vane based thrust vector control
systems of tactical missiles. Three-dimensional Reynolds Averaged Navier-Stokes equations were solved along
with two-equation turbulence model for different operating conditions. Nonlinear regression analysis was applied
to the detailed CFD database to evolve a mathematical model for the thrust vector control system. The developed
model was validated with series of ground based 6-Component static tests. The proven methodology is applied to
a new configuration.
Keywords: Jet vanes, CFD, nonlinear regression analysis, thrust vector control

1. INTRODUCTION advent of powerful parallel computers and advanced numerical

At the initial stage of missile launching, when the speed methods, CFD tools are increasingly being used in the design
is not very high, the control and stability requirements of and development of the jet vane TVC. The literature on
the missiles are met by thrust vector control (TVC) system. the numerical simulation of the jet vane flow field is very
The TVC system re-orients the direction of thrust vector and limited. Rogers7, et al. conducted numerical studies of jet
provides necessary lateral force for quick change of flight path. vane flow field using upwind flux difference splitting Navier
Among many TVC systems (secondary injection of gas/liquid Stokes code. Sung and Hwang8 used commercial software to
into the nozzle gas flow, the flex nozzle, jet vane, jet tabs, etc.), study the aerodynamic characteristics of jet vane arranged in
jet vane system is preferred for tactical missiles1 due to its X-formation within TVC shroud. These studies revealed that
small torque, small space requirements, and the capability to when erosion of jet vane is less than 1 per cent of the cross-
control pitch, roll, and yaw simultaneously2,3. section of the vane, the simulated aerodynamic characteristics
The basic principle of providing control forces using matches with the experimental results. Chandra Murty9, et. al.,
vanes is the same as lift generation of supersonic wing at an have presented CFD based jet vane TVC characterisation of
angle of attack. As the vane is deflected at a typical angle to the a tactical missile. A comprehensive regression analysis was
exhaust gas, an oblique shock increases the surface pressure done with CFD database to arrive at a mathematical model
on the windward side whereas expansion reduces surface to estimate the forces generated by vanes depending on the
pressure on leeward side of the vane. The pressure difference chamber pressure and vane deflection angle. The mathematical
between the windward and leeward sides of the vane provides model may be incorporated into OBC of the missile that
a force normal to the chord of the vane. Its two components (i)
generates TVC control and actuation signals based on trajectory
Lift (side force) - useful to control the missile, and (ii) Drag -
requirements. The developed model was validated extensively
resulting in thrust loss. The smaller is the ratio of the thrust loss
for a number of ground tests and flight tests for different duty
to the side force, the better is the performance.
cycles of vane deflections. In the present work, CFD based
Jet vanes are exposed to hot exhaust gas of the rocket
methodology has been used to characterise a new TVC system
nozzle, and with temperature as high as 3000 K and velocities
up to Mach 3.5 that exert extreme mechanical and thermal of a tactical missile.
loads on the vane. The degree of erosion on the vane surfaces
directly affects its ability to directionally control the thrust 2. SYSTEM DESCRIPTION
vector of the missile. As erosion reduces the vane area, the lift Jet vane TVC system mounted in X configuration on the
and side forces provided by the vane reduce with time. Hence, rear of the missile is shown in Fig. 1. Primary objective of
due to complexity of modelling of erosion, the characterisation the TVC is to direct the missile into the line-of-attack from
of jet vane is mostly done through experimental4 and semi- its vertical launch within short warning time. Validated CFD
empirical methods5-6. methodology was applied on this new configuration and we
Traditionally, experiments took major role in TVC have evolved a nonlinear TVC mathematical model that is to
development procedure of conventional missile. With the be integrated into OBC control logic.
Received 01 October 2014, revised 03 July 2015, online published 30 July 2015

261
 Def. SCI. J., Vol. 65, No. 4, july 2015

TVC System 4. RESULTS AND ANALYSIS

Computed side force obtained from two different
grids (namely 2.9 million and 5.5 million) are compared in
Table 1. Very close agreement of results between the two
grids demonstrate the grid independence of the results and 2.9
Million grids is taken as baseline grid to carry out the remaining
parametric simulations. Different CFD studies were performed
for various combinations of chamber pressures and vanes
deflections. Mach number contours for different vane deflection
Figure 1. Schematic of the missile rear section. angles (β/βmax = 0.42, 0.625, 0.83, and 1.04) are shown in
Fig. 3. With the increase in vane deflection angle, plume was
3. SOLUTION METHODOLOGY getting more and more distorted. Maximum Mach number was
The commercial CFD software Ansys10 Fluent 14.5 was estimated around 4 occurring in the plume expansion region
used in the present analysis. Density based coupled solver and Nozzle exit Mach number was 3. Complex flow structure
with first-order upwind Roe scheme11 was considered for was observed near the jet vane flow field, particularly when
discretisation of inviscid fluxes. First-order schemes are found the trailing edges of adjacent jet vanes were closer. Three-
to be adequate in the estimation of overall parameters like dimensional interactions of oblique shocks originated from
pitch/roll forces, and moments etc. for jet vane applications9. vanes leading edge is presented through Mach number contour
Temporal discretisation was achieved, by a first-order, implicit plot in Fig. 4. Shock interaction and reflection into the plume
Euler scheme. Turbulence was modelled using k-ε turbulence region is clearly observed. For a fixed chamber pressure of
model along with wall functions. Computational domain 110 ksc, the surface pressure distribution on jet vanes for vane
consists of rocket motor blast tube, nozzle, TVC and external deflection angles of (β/βmax = 1.04) is presented in Fig. 5. Non-
domain covering free stream up to about 2 m in all directions. uniformity in the surface pressure is attributed to the shape
Commercial grid generator12 ICEMCFD 14.5 was used to of the jet vane as well as flow expansion/compression near
generate unstructured grid in the computational domain. Typical the jet vane root cavity. The pressure difference between the
computational grid is presented in Fig. 2. Necessary care was windward and leeward surfaces provides the net side force.
taken in placing the first gird point on the wall boundaries to Under large vane deflections, flow decelerates on windward
maintain proper wall y+ values. Grid spacing near the walls through an oblique shock. Rapid expansion of flow on leeward
was about 3 micron (wall Y+ = 1 to 60) while for the far region side was also seen in Mach contour plot across jet vanes (Fig. 4).
it was about 10 mm. More elements were clustered around
the jet vanes to account for large gradients in flow variables. 5. DEVELOPMENT OF ENGINEERING
Nearly 2.9 million grid points were used in the domain which CORRELATIONS
are adequate to capture all the relevant features of the flow, Characterisation of TVC system involves many parameters
as shown in the grid-independence results presented later. At viz., chamber pressure, jet vane angles, rotational offset
inflow of blast tube, rocket chamber conditions were imposed. angle, etc. Enormous combinations of these parameters can
Total pressure was varied between 4 MPa to 14 MPa for exist; therefore it is impractical to perform CFD simulations/
different cases, total temperature was fixed as 2944 K. Plume ground test for all the combinations. A mathematical model
gas specific heat, dynamic viscosity, thermal conductivity, and is generated for TVC forces and moments which can be used
molecular weight were given as 1870.9 J/Kg K, 9.5133 x 10-5
Ns/m2, 0.27542 W/mK and 25.12, respectively. External free Table 1. Comparison of computational grids
stream boundaries were given atmospheric. No slip, adiabatic Grid No. Size, cells Side force, Fz/Fmax
wall condition was imposed on all walls. Maximum residual
(= ϕnj +1 − f (ϕnj +1 , ϕnj )) < 10-4 is taken as convergence criterion.
Grid-1 2.9 Million 0.420
Grid-2 5.5 Million 0.419

Figure 3. Contours of Mach number for different vane angle

Figure 2. Computational grid. deflection configurations.

262
 Murty & Chakraborty: Numerical Characterization of Jet Vane based Thrust Vector Control Systems

side force, f/f, max

mach Number static Pressure,ksc jet vane deflection angle, deg

Figure 4. 3-D interaction of oblique shocks originated from Figure 6. Variation of side force with vane angle for a fixed
vanes leading edge. chamber pressure.

Windward

side force, f/f, max

Leeward

chamber pressure, pc/pc, max

Figure 5. Surface pressure on vane for (β/β max = 1.04) vane Figure 7. Variation of side force with chamber pressure for
deflection angle. fixed vane angles.

in On Board Computer (OBC) of the missile, for generating Table 2. Model constants
actuator commands of vane deflections necessary for missile Model constant Predicted value
maneuver. a 1.0952
Figure 6 shows the variation of side force with vane angle
b 0.8744
for a fixed chamber pressure. Vane angle (β/βmax) is varied from
0.42 to 1.04 while keeping the rocket motor chamber pressure c 12.1235 (Yaw, Pitch), 6.3064 (Roll)
(Pc/Pc, max) as 0.73. A near-linear relationship between side force
Effective Jet vane deflection angle δe is given by
and vane angle (β/βmax) up to about 0.83 is obtained. Figure 7
(ξ1 × δ1 + ξ 2 × δ 2 + ξ3 × δ3 + ξ 4 × δ 4 )
shows the variation of side force with chamber pressure for a δe = (2)
4
fixed vane angle. Vane characteristics at different vane angles
a, b and c are model constants predicted using CFD based
show different slopes.
nonlinear regression analysis.
Large database is generated by CFD simulations for
Net yaw force is given by
different chamber pressures, viz., (Pc/Pc, max = 0.27, 0.53,
= F Z × cos(θ) + F Y × sin(θ)
1 1
0.73 and 1.0) and vane angles i.e., (β/βmax = 0.42, 0.625, 0.83 F Z (3)
and 1.04). A theoretical model is developed by performing Net pitch force is given by
nonlinear regression analysis9 using CFD database. The model = F y × cos(θ) − F z × sin(θ)
1 1

gives required side force/roll moment for any combination of F Y (4)

chamber pressure, jet vane angles, and rotational offset angle. Net roll moment is given by
Table 2 presents model constants estimated from nonlinear R = FR1 × r (5)
regression analysis. Equations (1)-(4) were used to calculate
yaw, pitch, and roll forces. where r is centre of pressure location in radial direction (Ycp)
= 41.4 mm.
Pca × δeb
F Y ,Z ,R =
1
(1)
c 5.1 Error Estimation for Correlation Against CFD
Subscripts Y, Z and R refer to pitch, yaw and roll Data
configurations, respectively and Pc is rocket motor chamber The correlations developed using nonlinear regression
pressure. analysis have some standard data fitting errors. The error band

263
 Def. SCI. J., Vol. 65, No. 4, july 2015

of the correlation need to be ascertained against experimental 4. Harrisson, V.; deChamplain, A.; Kretschmer, D.;
data and with CFD data as well. For all the simulations, the Farinaccio, R. & Stowe, R. Force measurements
predictions using correlation and CFD data is presented in evaluating erosion effects on jet vanes thrust vector
Table 3. A very important observation of the comparisons is control system. In 39th AIAA/ASME/SAE/ASEE Joint
that the correlations are deviating largely at lower chamber Propulsion Conference and Exhibit, American Institute
pressure (Pc/Pc, max ~ 0.27) and especially at higher vane angles of Aeronautics and Astronautics. AIAA Paper No. 2003-
(β/βmax = 1.04). The model constants in the correlations need 5237, 2003.
to be refined with a series of wisely designed experiments. doi: 10.2514/6.2003-5237
Nevertheless, it proves a basic structure of the theoretical 5. Giladett, L.V. & Wineman, A.R. Investigation of vanes
model which will be helpful in the process of charactering the immersed in the jet of a solid-fuel rocket motor. Report
TVC system no. NACA R M L 52F12, 1952.
6. Rahaim, C.; Cavalleri, R.; McCarthy, J. & Kassab, A. Jet
Table 3. Comparison of predicted side force between CFD vane thrust vector control - A design effort. In 32nd Joint
and correlation Propulsion Conference and Exhibit, American Institute
of Aeronautics and Astronautics. AIAA Paper No. 1996-
Simulation Pc/Pc, max β/βmax Side force Side force, % 2904, 1996.
no. (F/Fmax), correlation Error
doi: 10.2514/6.1996-2904
CFD
7. Roger, R.; Chan, S. & Hunley, J. CFD analysis for the
1 0.533 1.042 0.416 0.418 0.36 lift and drag on a fin/mount used as a jet vane TVC for
2 0.733 1.042 0.597 0.592 0.77 boost control. In 33rd Aerospace Science Meeting Exhibit,
3 0.733 0.625 0.383 0.379 0.99 American Institute of Aeronautics and Astronautics.
4 1.000 0.625 0.526 0.532 1.22 AIAA Paper 2904, 1996.
doi: 10.2514/6.1995-83
5 1.000 1.042 0.822 0.832 1.22
8. Sung, H.G. & Hwang, Y.S. Thrust-vector characteristics
6 0.533 0.625 0.275 0.267 2.76 of jet vanes arranged in X-formation within a shroud. J.
7 0.733 0.833 0.505 0.487 3.46 Propuls. Power, 2004, 20(3), 501–508.
8 0.733 0.417 0.257 0.266 3.48 doi: 10.2514/1.10381
9 0.267 0.625 0.118 0.125 6.43 9. Murty, M.S.R.C.; Rao, M.S. & Chakraborty, D. Numerical
10 0.267 1.042 0.180 0.196 8.84
simulation of nozzle flow field with jet vane thrust vector
control. Proc. Inst. Mech. Eng. Part G J. Aerosp. Eng., 2010,
6. CONCLUSIONS 224(5), 541–548. doi: 10.1243/09544100JAERO677
Jet vane thrust vector system of a tactical missile was 10. Ansys Fluent 14.5 theory and users guide. India, Ansys
numerically analysed by solving 3-D Reynolds Averaged Inc, 2013.
Navier-Stokes equations along with k-ε turbulence model. 11. Roe, P.L. Characteristic based schemes for the Euler
Simulations are performed for different combinations of equations. Annu. Rev. Fluid Mech., 1986, 18, 337–65.
chamber pressure and jet vane angles. Various forces and doi: 10.1146/annurev.fl.18.010186.002005
moments like side/roll forces, moments, bending loads on vanes, 12. ICEM CFD 14.5 Modelling and meshing guide. India,
hinge-moments, drag forces etc are predicted. Engineering Ansys Inc., 2013.
correlations are developed by performing nonlinear regression
analysis on CFD database. The model can predict yaw force, CoNTRIBUTORs
Mr M.S.R. Chandra Murty obtained his BTech (Mech
pitch force, and roll moment for a given chamber pressure and
Engg) from Regional Engineering College (REC), Warangal. He
vane deflection angles. Theoretical model is agreeing well with is currently working as Scientist, in Computational Combustion
CFD data except for lower chamber pressures and higher vane Dynamics Division, Defence Research and Development
angles. Laboratory (DRDL), Hyderabad. His areas of work include
heat transfer and computational fluid dynamics related to
REFERENCE aerospace propulsion.
1. Lloyd, R. & Thorp, G. A review of thrust vector control In the current study M.S.R. Chandra Murthy has done
systems for tactical missiles. In 14th Joint Propulsion simulation planning, grid generation, simulations with different
Conference, American Institute of Aeronautics and grids, chamber pressures and vane deflections, post processing of the
results and preparation of the figures and draft manuscript.
Astronautics. AIAA Paper No. 1978-1071, 1978.
doi: 10.2514/6.1978-1071
Dr Debasis Chakraborty obtained his PhD in Aerospace
2. Woodberry, R.F.H. & Zeamber, R.J. Solid rocket thrust Engg from Indian Institute of Science, Bengaluru. Presently, he
vector control. NASA- SP-8114, 1974. is working as Technology Director, Computational Dynamics
3. Facciano, A.B.; Seybold, K.G.; Westberry-Kutz, T.L. Directorate, DRDL, Hyderabad. His current interests are CFD,
& Widmer, D.O. Evolved sea sparrow missile jet vane aerodynamics, high-speed combustion, and propulsion.
control system prototype hardware development. J. In the current study Debasis Chakraborty has done overall
Spacecraft Rockets, 2002, 39(4), 522–531. planning and guidance of the work, review of results and
doi: 10.2514/2.3865 preparation of the final manuscript.

264