Professional Documents
Culture Documents
硏 究 論文 DOI: http://dx.doi.org/10.6108/KSPE.2012.16.2.042
ABSTRACT
The goal of heat transfer studies is the accurate prediction of temperature and heat flux distribution on material boundaries. To this
purpose, general-purpose computational fluid dynamics (CFD) code is used: FLUENT. Mass fluxes and pressure ratio are calculated
for two types of nozzle. The comparative studies reveal that the computational results are in agreement with the experimental data.
Also, heat transfer coefficients from FLUENT for one type of nozzle are very similar and agree well with the experimental data in the
diverging part of the nozzle, but the calculated results are large in the converging part. The heat transfer coefficients from Bartz
equation are over-predicted. We can consider various reasons for these differences, ie, laminarization by the highly accelerated flow
in the nozzle, turbulent flow model and grid generation.
second rock
The purpose of the heat transfer study is to more accurately predict the temperature and heat flux distribution. Commercial for this CFD
Code-in FLUENT use with 2 The mass flow rate and pressure ratio were calculated for the nozzles of different types, and they matched well with the experimental results. In
addition One Type of nozzle FLUENT The heat transfer coefficient calculation results at the nozzle wall surface were predicted to be slightly larger than the experimental results at
the nozzle shrinkage section, but agree well with the expansion section. Bartz The calculation of the heat transfer coefficient using the equation was predicted to be larger than the
overall experimental results. The reason why the calculation result differs from the experimental result is laminarization by rapid acceleration flow in the nozzle, Turbulence models
Key Words: Turbulent Flow ( turbulence), Bartz Equation expression), Acceleration Parameter ( Acceleration parameter),
One. Introduction
Reception date Oct. 18, 2011, Modification completion date March 5, 2012, Confirmation date March 9, 2012
The thermal analysis of a solid rocket nozzle defines and transfers the heat
* Regular member, Defense Science Research Institute One Technology Research Division 6 part
† Corresponding Author, E-mail: godbless1001@empal.com transfer between the high-temperature combustion gas and the nozzle heat-resistant material.
Volume 16, Issue 2 2012. 4. Comparative study of heat transfer coefficient of rocket nozzle 43
It is divided into calculating the thermal reaction of materials. Heat transfer 45 °- 15 ° Nozzle[ 2] of 2 Branch, mass flow rate, pressure ratio, The heat
between combustion gas and nozzle wall is convection transfer coefficient at the nozzle wall was calculated and compared with the
(convection), copy( radiation), Particle deposition ( particle impingement) Etc. The experimental results.
nozzle flow field is subsonic, Transition, Divided into supersonic regions, In the
supersonic flow region, convective heat transfer is dominant. The heat transfer
coefficient is usually obtained by analyzing the boundary layer flow in the case 2. Governing equation
where there is no surface retreat due to abrasion of the heat-resistant material. Precise
prediction of the heat transfer coefficient transferred from the hot combustion gas to In this study, axis symmetry 2 Dimension Reynolds Average
the solid surface is necessary not only for the nozzle, but also for the O-ring of the Navier-Stokes To solve the compressive fluid flow solution
assembly and the insulation design inside and outside of the solid rocket motor. The based on the equation CFD Code-in
heat transfer coefficient required for the design of a solid rocket motor has been FLUENT Was used, Standard to simulate turbulence ••• ( 2- Equation) Model
obtained mainly by experiments or by semi-experience. However, this method is and standard wall function were applied. Axis symmetry 2 Dimensional
applicable only when the flow conditions are sufficiently similar to each other. Another continuous equation, momentum equation in axial and radial directions, And
applying the radius empirical equation to directly obtain the heat transfer coefficient. CFD •• • • •••
•• • • •• • • • • • • • •• • • • • • • • • • • • •
(One)
The heat flux is first calculated from the solution, and then the heat transfer coefficient
•• • • • ••• • •
• • • • • ••• • • • • • ∇ • •• • • •
turbulence models and wall functions. Many problems, including heat transfer, are
• • •• ••• ••• •
•• •
deposition, flow separation, Reattachment, It is very difficult to accurately predict the • • •
• •• •
•• •
•• • • •
•• • •
•
pressure gradient.
• • • • •
•• • • • • • • • •• •• • • • • • • • • • • •• •• • • • • • • • • • •
•• • • • ••• ••• •
• • • • • ••• • • • • •
•
In the previous study, the half-width 30 °, Half of the enlarged part 15 ° Phosphorus •• • •• • •• •• •
(3)
• • • ••• • •
nozzle [ One] And the reduced half 45 °, Enlarged half-width 15 ° Phosphorus nozzle [ 2] • • • ••• • • • • • • ∇ • •• • • • •
• •• • •• • •
The convective heat transfer coefficient was experimentally determined for various •• • •
• • • • • • • • ∇ • •• • • • • •
flow conditions. In addition, Various commercials are used to predict the convective •
•
• •
heat transfer coefficient between flow and solid surfaces under various flow
conditions. CFD Code was used and compared with the experimental results [ 3, 4]. •
•• • • • • • • ∇ • • •• • •• • • • •••
••
(4)
∇∙ ••• ∇•• • •
••
• • • • • • •• ••• ∙
•• •
••
• • •
In this study, it was used to predict the heat transfer coefficient of the wall
From here • Is the axial coordinate, • Is the radial coordinate, • • Is the axial
surface of the solid rocket nozzle. CFD Code-in
FLUENT Was used, Existing experimental results and velocity, • • Is the radial velocity, • Is the density, • Pressure, • Is the viscosity coefficient, •
Bartz Compared to the expression FLUENT The applicability of was studied. The • • • • Is the external volume forces in the axial and radial directions, respectively. And
••• ••• •• of 2 Eggplant models were applied. Supersonic nozzle flow Back Etc( 1964 ∼ 1972) The
∇ • •• • • • • • • • (5)
•• •• • experimental study of the study contributed to understanding the compressive
turbulent flow field with severe pressure gradient and heat transfer along with the
ego, • ••• Is the effective thermal conductivity ( ••• • • •• • • Each gas-dynamic flow characteristics in the nozzle. In other words,
3.1 30 °- 15 ° The nozzle analysis shape Fig. One And the central wall of the nozzle and
And Current Enthalpy h Is defined as follows for each of the ideal gas and
the reduced wall 30 °, The wall of the enlarged part 15 ° And the inner diameter of the
the incompressible fluid.
nozzle neck 45.8 mm to be.
•
• • • •
• •• • • • • •
• •• • • •
•
(7) 2 A dimensional axisymmetric grid was used, and the total number of grids in the
flow region was 151 × 40 It is a dog. Initial stagnation pressure / stagnation temperature
Fig. 2 It is shown in. When calculating, the viscosity model of the inner wall of
to be. From here • ••• The 298.15 K to be.
the nozzle inviscid Was applied. Fig. 2 in z
The boundary conditions define the flow and thermal variables at
Is the axial coordinate, L Silver nozzle full length ( 150.5 mm),
the model boundary. Properly defining the boundary conditions FLUENT
• Is the density, • Is speed. Oswatitsch Etc[ 5] This derived prediction equation is
It is a very important factor in the analysis using. The boundary
derived by ignoring the wall boundary layer and assuming that the flow in the
conditions applied in this study are for the inlet and outlet flow
wall is parallel to the wall.
boundaries. pressure inlet, pressure outlet, pressure far-field And the
boundary between the wall and the axis wall, symmetry Condition, For
Fig. 2 Ratio of Local to 1-d Mass Flux along the Fig. 3 Ratio of Static to Stagnation Pressure along
Nozzle Wall the Nozzle Wall
All. Fig. 2 As can be seen from, it can be seen that the calculation results are Compared to the result Fig. 5 It is shown in. Initial stagnation pressure /
different from the experimental results, but show a similar trend. The mass stagnation temperature at the inlet and outlet respectively 75.3 psia / 558.3
flow rate ratio upstream of the nozzle neck is the maximum, You can see it K, 1.0 psia / 164.3 K to be. Fig. 5 As can be seen from the pressure rise in
descending / rising rapidly from downstream. Stagnant pressure in the the inlet curve due to the change of flow direction in the nozzle inlet region
combustion chamber ( • •) For nozzle wall can be seen. The back pressure gradient along the wall surface caused by
the change of direction of the flow affects the boundary layer structure, and
thus the heat transfer to the wall surface. Pressure ratio at the inner wall of
Static pressure at • •) The ratio of was calculated and compared with the
the nozzle under another condition ( • • • • •) Was calculated and compared with
experimental results Fig. 3 It is shown in. The inner wall of the nozzle no-slip
the experimental results
condition Was applied. Viscous models are standard ••• ( 2- Equation) Model and
standard wall function were applied. Fig. 3 in L Silver nozzle full length ( 150.5 mm) to
be. Fig. 3 As can be seen from the results, the calculation results of the pressure
ratio are in good agreement with the experimental results, except for the nozzle Fig. 6 It is shown in. Initial stagnation pressure / stagnation temperature at the
exit region where the static pressure rises rapidly at a lower stagnation pressure. inlet and outlet respectively 75.2 psia / 833.3 K,
1.0 psia / 245.2 K to be. Fig. 6 in L Silver nozzle full length ( 185.0
mm) to be.
3.2 45 °- 15 ° The nozzle analysis shape Fig. 4 Is equal to, The nozzle central axis and shock wave caused by flow separation
the reduced wall 45 °, The wall of the enlarged part 15 ° And the inner diameter of the
2 A dimensional axisymmetric grid was used, and the total number of grids in the
flow region was 214 × 40 It is a dog. Nozzle inner wall silver no-slip condition Was
applied. Viscous models are standard ••• ( 2- equation) The model and standard wall
Pressure ratio at the inner wall of the nozzle ( • • • • •) Was calculated, Experiment
with calculation results in the nozzle inlet area Fig. 4 Schematic of 45 ° -15 ° Nozzle
46 Heechul Ham ․ Kang Yun-gu Journal of the Korean Society for Promotion and Engine
Fig. 5 Ratio of Static to Stagnation Pressure in the Fig. 6 Ratio of Static to Stagnation Pressure
Inlet Region along the Nozzle
Except for the nozzle exit area, the calculation results of the pressure ratio agree well
with the experimental results. It also shows the rise / fall of the static pressure on the
wall related to the change in the direction of flow in the downstream region adjacent to
words, 3.2
The number of grids used in the section 214 × 40 And 2 Fold big 428 × 80, And
shown in, it can be seen that the pressure ratio at the wall of the nozzle is Fig. 7 Ratio of Static to Stagnation Pressure along
the Nozzle Wall on 107 × 20, 214 × 40, and 428 × 80
independent of the lattice size. Fig. 7 in L Silver nozzle full length
mesh
• ••
• • •
Heat transfer coefficient at the nozzle wall, • To 2 Calculated by type Back Etc[ One]
• • • (10)
• •
• • •
Was compared with the experimental results. First, Commercial Code FLUENT The
•••
Degree, It is the temperature outside the boundary layer. Recovery factor • • • • ,
Was used. The analysis shape Fig. One Is equal to, 2 Was calculated.
A dimensional axisymmetric grid was used, The total number of grids in the flow region Back Etc[ One] In order to create a high temperature and high pressure
is 151 × 40 It is a dog. The inner wall of the nozzle flow, a method of mixing compressed air and methanol was used. At this
no-slip condition Was applied. Viscous models are standard ••• ( 2- equation) The time, the mixing ratio of methanol was sufficiently lean to allow the fluid to
model and standard wall function were applied. Where the first grid dimension be regarded as air. Accordingly, the property values used in this calculation
of the boundary layer grid near the wall is 0.1 mm to be. The resulting were also used for air. Compare the calculated result with the experimental
dimensionless vertical distance from the nozzle wall to the first grid point, result Fig. 9 Wow Fig. 10 It is shown in.
•
• The 30 ∼ 300 It was made to have a value between. The density of air is the
ideal gas, Viscosity coefficient sutherland law Was applied. Wall thermal Figure 9, 10 As can be seen from FLUENT
conditions Fig. 8 Same as Back Etc[ One] The experimentally determined wall The calculation result was slightly predicted more than the experimental result in the
temperature was applied. nozzle reduction section, but it was in good agreement with the expansion section
outlet respectively 75.2 psia / 843.3 K, 4.4 psia / 373.8 K, And 254.0 psia /
• • •• • • •• • •• • ••
• •• • • • • • • • •
• • • • • •
• • •
• • • • •• • • •• • • •• •• • (11)
• • • • • •• • •
••
From here • • • • • • •• • • • • • • • • • • • • • • • • • • Is the nozzle neck diameter, specific heat, Viscosity
Fig. 9 Comparison of the Predicted Heat Transfer
coefficient, Prandtl number, combustion chamber pressure, Characteristic speed, Nozzle Coefficients with the Experimental Data
neck curvature radius, Nozzle neck area, nozzle area at any location, It is a
• • • • • • •• • • • • • • •
• • • • •• • •
o Indicates the condition in the combustion chamber. FLUENT The heat transfer
coefficient is derived by deriving the variable value from the calculation result.
Fig. 8 Temperature Distribution (° F) Inside the Nozzle Fig. 10 Comparison of the Predicted Heat Transfer
from Thermocouple Measurements by Back et al. [1] Coefficients with the Experimental Data
reported by DeLise and Naraghi [7] • • • • • • • •• • • • • • • •
• • • • •• • •
48 Heechul Ham ․ Kang Yun-gu Journal of the Korean Society for Promotion and Engine
• • •• • • •• • •• • ••
Is doing. Compared to the experimental results, the calculation results are each at
• •• • • • • • •• • • •
• • • • ••• •
•
• • • • •• • • •• •• •• (12)
the nozzle neck position. 13%, 17% It was greatly predicted. • • • • • •• • •
••
FLUENT In the case of the calculation result, the maximum heat transfer coefficient
occurs upstream of the nozzle neck where the mass flow rate becomes maximum, You
Here, the first four terms on the right side have a constant value
can see it descending downstream of the nozzle neck. Bartz The calculation result
regardless of the nozzle position under the same conditions, and only the
using the equation was largely predicted as a whole compared to the experimental
area ratio remains to be calculated locally.
result, respectively at the nozzle neck position 88%, 87% It was overstated. • •• • ••
•• • • •• •
• • The FLUENT Combustion chamber calculated from
Enter the value from inside (replaced by the nozzle inlet because the length of the
combustion tube is short in this study). This value has a substantially constant
value inside the combustion chamber, and may be slightly different from the
5. Bartz Understanding Expression
nozzle inlet value.
using the equation in the neck of a specific nozzle and the result of
Assuming that the value is constant at the stagnation temperature because the
Wow • You only have to consider. • Is divided into the value at the stagnation
parameters, K Is presented as the following equation. In other words,
temperature and the value at the reference temperature, and the value at the
• ••
stagnation temperature is • • • • ••• • • •
• • Included in, standards
The value at temperature • Included in. Here, the reference temperature is defined as •• •• •
• • • (13)
• •• •
the average temperature of the wall temperature and the outside temperature of the ••• •
[6] In the way presented in • If you calculate 0.94 ∼ 1.0 Since it has a value between, From here • • • • • • • • Is the viscosity coefficient, density, and flow velocity outside the
it does not significantly affect the value of the heat transfer coefficient. boundary layer, respectively. Acceleration parameter
And • • • • • • ••••• • • • • • • • • • •• •
• Because of
entire boundary layer, and accordingly, "laminarization" starts in the entire
Mass flow rate per unit area, •• Rocket performance parameters light layer. Although the turbulent boundary layer is not clearly laminar near
the wall, it exhibits behavior similar to laminar flow. This phenomenon mainly
Phosphorus Characteristic Speed • • And combustion chamber pressure • • Related to, • • The
occurs in the rapid acceleration flow in the nozzle, resulting in a decrease in
• • The nozzle has a constant value over the entire area inside the nozzle. Therefore, as
heat transfer at the nozzle wall.
in the experimental results, the heat transfer coefficient is maximized in the upstream
of the nozzle neck and the effect of rapidly decreasing in the downstream of the
nozzle does not appear. 4 paragraphic 2 By calculating the acceleration parameters for the kinds of
conditions Fig. 11 It is shown in. Fig. 11 As shown in the figure, the acceleration
Therefore, by combining the above Eq. 11 If you rewrite as: parameters are inside the nozzle except for the nozzle inlet and outlet part. • × •• ••
Greater than
Volume 16, Issue 2 2012. 4. Comparative study of heat transfer coefficient of rocket nozzle 49
Economical numerical calculations are possible, and the wall function has
lost due to the development of a larger and faster computer, and the
argued that the use of wall functions should be avoided on the grounds
high-capacity computers, but the more physical and general the model,
the more rigorous and complex the calculations are. As the new
Fig. 11 Acceleration Parameter, K along the Nozzle turbulence models continue to develop, the usefulness of the wall
It shows the value. In other words, Laminarization occurred inside the understanding of the turbulence structure. As the model of the wall
nozzle except for a part of the nozzle inlet and outlet. It is judged to have function is improved and the scope of its application is expanded, its
influenced the heat transfer to the wall. application in the high-speed flow field as in the nozzle flow is expected.
Figure 9, 10 As can be seen, the laminarization occurs more severely in the nozzle
reduction section where the acceleration parameter is relatively large. FLUENT It can
be seen that the calculated result and the experimental result show a larger difference
than the enlarged part. For the standard wall function applied in this study, the dimensionless
vertical distance from the nozzle wall to the first grid point, • • The 30 ∼ 300 It
6.2 Turbulence model and grid configuration should be in between, and it is desirable to have a value close to the lower limit.
The turbulence model used in this study is a general standard. In this calculation, the first grid dimension of the boundary layer grid near the
• • • (2- equation) The model was applied and a standard wall function was wall is the same for all areas. 0.1 mm It consisted of. Accordingly, it is close to
used for boundary conditions near the wall. the lower limit at the nozzle inlet and outlet, but close to the upper limit at the
2- Although the equation model has been pointed out as a failure to properly nozzle neck, and the dimensionless vertical distance ••≺• • ≺••• As the limit is
model the effect of the wall surface on turbulence near the wall, Compressive slightly out of range. Therefore, it is judged that the difference between the
turbulent heat flow field analysis inside the nozzle with severe pressure calculated result and the experimental result can be further reduced by
gradients and wall heat transfer was applied to perform in a more rational and adjusting the grid so that the dimensionless vertical distance at the nozzle neck
more general way. In the vicinity of the wall, the changes in the flow variables area approaches the lower limit.
are sudden, and in order to integrate them in the numerical calculation, the
grid network near the wall must be specially densely constructed. Especially 2- Changes
making the lattice arrangement more stringent. The wall function is used to
mitigate this dense lattice arrangement by using flow variables inside the 7. conclusion
from the wall surface. By excluding the need for a dense grid arrangement, FLUENT To verify the calculation result using 30 °- 15 ° Nozzle, 45 °- 15 ° Of
nozzle 2 The mass flow rate and pressure ratio were calculated using the branch
nozzle
50 Heechul Ham ․ Kang Yun-gu Journal of the Korean Society for Promotion and Engine
About the nozzle wall FLUENT The heat transfer coefficient calculation results using Transfer in a Conical Supersonic Nozzle, "
are predicted to be slightly larger than the experimental results in the nozzle J. Spacecraft, Vol. 4, No. 8, 1967,
reduction section, but agree well with the expansion section. Bartz The heat transfer pp.1040-1047
coefficient calculation result using the equation was predicted to be larger than the 3. DeLise, JC and Naraghi, MHN,
test result. The reason why the calculation results differ from the experimental "Comparative Studies of Convective Heat Transfer Models
results can be considered by considering laminarization, turbulence model, and for Rocket Engines, "
lattice composition by rapid acceleration flow in the nozzle. In order to obtain a AIAA 95-2499, July 1995
more accurate calculation result in the future, efforts to solve the cause are 4. Wang, Q., "On the Prediction of
required. Convective Heat Transfer Coefficients
Using General-Purpose CFD Codes, "AIAA 2001-0361,
Jan. 2001
5. Oswatitisch, K. and Rothstein, W., "Flow Pattern in a
references Converging-Diverging Nozzle," NACA TM-1215, 1949
1. Back, LH, Massier, PF and Gier, HL, "Convective 6. Bartz, DR, "A Simple Equation for Rapid Estimation of
Heat Transfer in a Rocket Nozzle Convective Heat Transfer Coefficients," Jet
Convergent-Divergent Nozzle, "International Journal of Propulsion, Jan. 1957, pp.49-51
Heat and Mass Transfer, Vol. 7, No. 5, 1964, pp.549-568