Applied Thermal Engineering 127 (2017) 58–69

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Numerical investigation on the optimization of local transpiration

cooling effectiveness
Lin Shen, Jianhua Wang ⇑
Department of Thermal Science and Energy Engineering, University of Science and Technology of China, Jinzhai Road 96, Hefei 230027, PR China

h i g h l i g h t s

An optimization method to improve local effectiveness is suggested.

Step coolant allocation can improve local effectiveness of stagnation region.

Step coolant allocation can reduce coolant demand.

The optimization feasibility is analyzed through a non-uniform porosity strategy.

a r t i c l e i n f o a b s t r a c t

Article history: With the development of active Thermal Protection System (TPS) of hypersonic vehicles, optimizing
Received 7 February 2017 active TPS design and reducing weight including the system and coolant loaded are the critical issues
Revised 28 July 2017 to be considered. This paper presents an optimization method of local transpiration cooling effectiveness
Accepted 4 August 2017
under supersonic condition with a freestream total temperature of 2310 K, and a freestream Mach num-
Available online 5 August 2017
ber of 4.2. The numerical investigations of transpiration cooling are conducted by performing two strate-
gies of coolant allocation over a nose cone surface, (1) constant coolant velocity allocation; (2) step
Keywords:
coolant velocity allocation. The aerodynamic and aerothermal performances including Mach number
Transpiration cooling
Cooling effectiveness
and pressure distributions, cooling performances and coolant demand are analyzed by the two strategies.
Porous matrix The analysis indicates that the step strategy can effectively improve the local cooling effectiveness in the
Non-uniform porosity stagnation region and reduce the coolant demand. To accomplish the optimization strategy, a non-
uniform porosity strategy of the nose cone is analyzed, and the coolant flow characteristics in the porous
matrix at different coolant injection conditions are studied. The study indicates that the non-uniform
porosity can improve the efficiency of coolant transportation, and increase coolant mass flow rate in stag-
nation region.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction investigations, McConarty et al. [7] discussed and compared the

wing’s weight of different passive and active TPSs. Their results
With the aerodynamic performance improvements of super- indicated that from the view point of the cooling performance
sonic and hypersonic vehicles, one of the most severe problems and total weights of cooling systems, the active TPSs with transpi-
faced to us is the extremely high thermal loads, especially for some ration and convection cooling are more effective than passive TPSs.
critical structures such as nose cones, leading edges of wing and According to McConarty et al. [7], the coolant weights contribute a
rudder [1,2]. Hence, it is necessary to develop more effectively proportion from 17% to 82% in a total TPS. Hence, it is valuable to
active TPSs [3]. Transpiration cooling as an active TPS can provide reduce coolant loaded.
much high cooling effectiveness due to its high specific area for Transpiration cooling performances using gaseous or liquid
heat convection [4–6]. Besides, for long duration flights, it is coolants have been investigated with different coolant flow rates
another critical issue to reduce the total weight of vehicles through and in different porous media [8–10]. In the subsonic experi-
decreasing coolant loaded. Through experimental and numerical ment conducted by Liu et al. [11], the cooling effects of different
gaseous coolants and flow rates were studied, and their studies
indicated that the cooling effectiveness could be increased
⇑ Corresponding author.
through a higher coolant flow rate, or a higher specific heat of
E-mail addresses: [email protected], [email protected] (J. Wang).

http://dx.doi.org/10.1016/j.applthermaleng.2017.08.017
1359-4311/Ó 2017 Elsevier Ltd. All rights reserved.
 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 59

coolant. It is well known that liquid water as transpiration 2. Numerical simulation

cooling medium can provide a much higher cooling effect due
to phase change latent heat, and the nose-cone transpiration 2.1. Physical model and numerical method
cooling experiments conducted by [12,13] in supersonic wind-
tunnel have presented the feasibility and advantages of the The physical model used in this work is a nose-cone structure,
cooling processes with liquid coolant phase change. According and its schematic diagram is shown in Fig. 1. The external radius
to the numerical [14] and experimental [15] researches on of the nose cone is 3 mm with an included angle of 14°, and the
transpiration cooling with phase change, one can find that when total length and width of the model is 40 mm and 70 mm, respec-
the phase change occurs in the porous matrix, with a relatively tively. The numerical model is set as a symmetric model and the
low driven forces for coolant injection, the cooling effectiveness boundary conditions are shown in the Fig. 1. The freestream inlet
can reach to a relatively high level. However, most investigations is set as velocity inlet with a turbulence intensity of 5%, and the
of transpiration cooling focused on the overall cooling effect of freestream outlet is set as supersonic.
the specimens. However, according to the experimental and The numerical simulation is conducted by Ansys CFX. A high
numerical results, in the stagnation region of nose cone, due to resolution scheme was used in the discretization of the convection
highest aerodynamic pressure and heating [24], the cooling terms. The gaseous Nitrogen is used as coolant, and it is deemed as
effectiveness here is relatively low [13,15,16]. Hence, it is an ideal gas. The SST turbulence model is applied in the numerical
necessary to search for an effective method to increase the local simulation, which has been used in aerodynamic problems [20].
cooling effectiveness in stagnation region through the optimiza- The convergence criteria of normalized root mean square (RMS)
tion of cooling system designs. In this paper, an optimization of residuals were less than 1E-5.
strategy of coolant allocation over different regions of a nose As shown in Fig. 2, the structured mesh with refined boundary
cone surface is suggested and its influences on local effectiveness grids near the specimen surface is generated by the commercial
and coolant consumption are investigated. software Ansys ICEM CFD. The location of the first node of the
From the view point of the feasibility of the optimization prism layers away from the wall was carefully selected to insure
strategy, the optimization methods can be basically divided into the dimensionless wall space y+ less than 1.5. The mesh indepen-
two categories. The first one optimizes the structure of the nose dence of numerical results was checked by three strategies of node
cone, to increase the coolant flow in the stagnation region. For numbers that Mesh 1: 570,000, Mesh 2: 700,000, Mesh 3: 870,000.
example, in the wind tunnel experiment conducted by Zhao An uncooled case was simulated and the stagnation temperature
et al. [17], through an unequal-thickness configuration that the by three mesh strategies was calculated as shown in Table 1. It is
thickness of porous nose cone was relatively small in the stagna- found that the calculated stagnation temperature with different
tion region, the cooling effectiveness in stagnation region had mesh sizes is almost the same. Hence, in order to save computa-
been increased effectively. The other method is to control the tional resource and guarantee computational accuracy, the mesh
performances of porous material. In the numerical research of with a total node number of 700,000 is applied in the following
Huang et al. [18], the surface temperature of a nose cone numerical simulation.
decreases with an increase in the thermal conductivity of porous
matrix. In the numerical research of Gulli et al. [19], the influence 2.2. Numerical validation
of a uniform porosity on the coolant flow characteristics in a
porous blunt nose cone was investigated, and they indicated that In order to evaluate the reliability of the numerical strategy, a
a lower porosity leads to an increase of coolant injection supersonic experiment was carried out in the arc-heated wind tun-
pressure. nel FD04 in China Academy of Aerospace Aerodynamics [13]. The
In a word, the improvement of local cooling effectiveness in schematic diagram of the wind tunnel system and experimental
stagnation region is one of the most efficient optimization of active apparatus is presented in Fig. 3. In the previous work [13], with
TPS design, and it can be accomplished by an optimized allocation the same numerical strategy, the coolant mass flow of numerical
of coolant flow rate in different regions. In this paper, an optimiza- simulation are compared with the experimental result. And it can
tion method of coolant allocation is suggested based on different be found that the relative error of numerical and experimental
requirements of coolant allocation over the specimen’s surface. result is quite low, and it is mainly caused by the ideal settings dur-
And the feasibility of the optimization method is discussed through ing the numerical simulations. In the following investigations, the
a non-uniform porosity strategy. The aim of this work is to provide heat transfer and supersonic flow characteristics are the objects of
the engineers of optimal designs of transpiration cooling TPS with the numerical simulations, so another supersonic experiment
a relatively comprehensive reference, which considers the local without coolants injection is conducted. When the total specific
transpiration cooling effectiveness and coolant demand, enthalpy of freestream was at 1500 kJ/kg, the total temperature
simultaneously. 1385 K, and the Mach number 4.2, and without coolant injection,

Fig. 1. Diagram of Physical model.

60 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69

Fig. 2. Structured mesh with refined boundary grids.

Table 1
Stagnation temperature in three mesh strategies.

Mesh 1 Mesh 2 Mesh 3

Fig. 4a. Temperature distribution in the supersonic experiment.
(570,000) (700,000) (870,000)
Stagnation temperature/K 2285 2307 2310
where the R is the radius of the nose cone, the Ma is the freestream
Mach number.
Based on Eq. (1), with a Mach number of 4.2, the stand-off dis-
the temperature distribution over the specimen’s surface was
tance of the shock wave from the stagnation point is 1.509 mm.
recorded by an infrared thermal imaging system (ITIS), which has
The numerical calculated value using the present numerical
a measurement accuracy of ±1.0%.
method is 1.512 mm, with a relative error of 0.199%.
Fig. 4a exhibits the surface temperature distribution recorded
by the ITIS. It can be found that the local temperature is quite high
2.3. Motivation
in the stagnation region. The temperature data of the surface cen-
terline is selected to be compared with the numerical results as
In order to investigate the extremely thermal environment,
shown in Fig. 4b. It can be found that they are quite consistent.
where the freestream Mach number is 4.2, the total temperature
In the leading edge, the numerical result is slightly underestimated
is 2310 K, the heat flux distribution along the nose cone surface
but in the downstream region, it is slightly overestimated. The
is necessary to be obtained. The specimen surface temperature
main reason is that in the numerical simulation, some ideal set-
was set by a no-slip wall with a constant temperature of
tings are applied. Despite this, for the stagnation temperature,
800 K, then the heat flux distribution is simulated as shown in
the relative error between the numerical and experimental result
Fig. 5.
is just 2.69%.
If the heat from the freestream is absorbed by gaseous N2 to be
In order to investigate the supersonic flow characteristics, the
heated from 300 K to 800 K, the required mass flow rate of N2 along
stand-off distance of the shock wave from the stagnation point is
the specimen surface is calculated, as shown in Fig. 5. It can be
studied. According to Billig [21], the stagnation-point shock-
found that the required mass flow rate in the downstream region
detachment distance for a wedged cylinder can be calculated as:
is quite small, therefore a conservative design is suggested that
D ð 4:67 Þ the region with transpiration cooling function is limited in a range
¼ 0:386e Ma2 ð1Þ
R from the stagnation point to 40 mm downstream, where the wall

Fig. 3. Schematic diagram of wind tunnel and experimental apparatus.

 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 61

Fig. 4b. Numerical and experimental surface temperature distribution.

Fig. 5. Heat flux and required coolant mass flux along the nose cone surface.

heat flux is higher than or equal to 15% of the stagnation heat flux. stream region. It is noted that in the extreme ‘‘step3” case, the
And based on the calculated mass flow rate, an optimization strat- transpiration region is limited in the stagnation region and no cool-
egy of the coolant allocations are discussed as followings. ant penetrates the downstream region. The coolant mass flux along
the specimen surface is shown in Fig. 6(b) which is quite consistent
2.4. Optimization strategy with the distribution of the calculated mass flux shown in Figs. 4a
and 4b. In all the cooling cases, in order to eliminate the influence
According to the aforementioned studies of the thermal envi- of total coolant mass flow on the cooling effects, there is an
ronment, two strategies of coolant allocation i.e. constant method indicator that the total coolant mass flow is the same so that the
and step method are performed in the nose-cone outside surface comparison of cooling effects between different cases of coolant
and their influences on the outflow field distribution and cooling allocation is much more convincing. The total coolant mass flow
performance are investigated. As Fig. 6(a) shows, for the constant is calculated by integrating the mass flux over the x-axis, as shown
method, the coolant velocity is the same both in the stagnation in Table 2. In all cases, the mass flow is almost the same with quite
region and downstream region of the nose-cone surface. But for small relative errors. Besides, an uncooled case is simulated to be
the step method, which includes three cases, the coolant velocity compared, in which the specimen surface is set as an adiabatic
is higher in the stagnation region but it is reduced in the down- and no-slip wall.
62 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69

Fig. 6. (a) Velocity and (b) mass flux distribution along specimen surface in different coolant allocation.

Table 2
Total coolant mass flow in different coolant allocation.

Constant method Step method 1 Step method 2 Step method 3

Mass flow (g/m * s) 12.428 12.430 12.431 12.429

3. Analysis of numerical results pushing pressure increases so that the high pressure area
increases. But the pressure along the centerline of the nose-cone
3.1. Outflow field distribution surface is almost the same in all cases as Fig. 7(b) shows.
According to the Mach number contours shown in Fig. 8(a), one
The pressure and Mach number distribution of flow field are can find that with the increase of the coolant mass flow in stagna-
simulated as shown in Figs. 7 and 8. One can find that the high tion region, the low Mach number area increases. The shock wave
pressure area in the front of the stagnation region enlarges gradu- is gradually pushed away from the leading edge of the specimen. In
ally. It is because that the high pressure of driving force pushes the order to investigate the shock wave location in different cases, the
coolants out of the stagnation region. And the pushing pressure position with Ma = 1 is defined as the indicator to represent the
and the aerodynamic pressure jointly form the high pressure area. shock wave location. It is noted that the stagnation point of the
With the increase of coolant mass flow in stagnation region, the specimen is the ‘‘x = 0.003” point. According the distribution of
 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 63

Fig. 7. Pressure contours and distribution along specimen surface at all cases.

Mach number along the symmetric line in front of the stagnation 3.2. Cooling performance
point, as shown in Fig. 8(b), one can find that with the increase
of mass flow in stagnation region, the location of shock wave In this section, the cooling performance with different coolant
become further away from the stagnation point. It is indicated that allocations is investigated. Fig. 9(a) presents the temperature con-
the increase of inversely jetting coolants in the stagnation region tours in different cases. In the uncooled case, the highest tempera-
has a positive effect on pushing the bow shock wave away from ture area occurs in the front of the specimen’s stagnation region.
the specimen surface. With the increase of coolants mass flow in the stagnation region,
Synthetically considering the pressure and Mach number distri- the high temperature area is gradually away from stagnation
bution, one can find that the penetrating coolant makes the high region. In the step3 case, there is an apparently low temperature
pressure area and low Mach number area in the front of the stag- layer covering on the surface of the specimen. It is formed by the
nation region become larger but it makes little influence on the gaseous coolants penetrating from the specimen and can effec-
surface pressure. Besides, the coolants penetrating from the speci- tively protect the specimen from the high enthalpy freestream.
men can effectively push the shock wave away from the specimen In order to conduct the quantitative comparison of surface tem-
surface. And with the increase of the coolant mass flow in stagna- perature in different cases, the temperature distribution along the
tion region, the standoff distance of shock wave from specimen centerline of the specimen surface is calculated as shown in Fig. 9
stagnation point becomes further. (b). By comparing the temperature distribution in different cases of
64 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69

(a)

4 uncooled
constant method
step method 1
3
step method 2
step method 3
Mach number

-0.008 -0.007 -0.006 -0.005 -0.004 -0.003

X [ m]
(b)
Fig. 8. (a) Mach number contours and (b) distribution in the front of the stagnation point at all cases.

constant and step allocation, it can be found that with the increase protection area is predictably limited based on the temperature
of coolants mass flow in the stagnation region, the temperature distribution tendency along the freestream flow direction.
here decreases gradually. Besides, even though the coolants mass According to the heat flux distribution shown in Fig. 10, one can
flow in the downstream region is reduced, the temperature here find that in the stagnation region, the ‘‘step” allocation enhances
still decreases gradually. In the ‘‘step3” case, the temperature in the heat transfer of the coolants and porous matrix, and with the
stagnation region is even lower than that in some parts of the increase of coolant mass flow here, the heat flux here increases,
downstream region. Synthetically considering all the above phe- therefore. However, in the downstream region, due to the decrease
nomena, one can find that at the same total coolants mass flow, of coolant mass flow, the heat transfer here is reduced. But even
the increase of coolants mass flow in stagnation region can effec- though the heat transfer in the downstream region is reduced,
tively decrease the surface temperature both in the stagnation the temperature here is still lower as Fig. 9(b) shows. It is reason-
region and downstream region even though the coolants mass flow able that the gaseous layer block the heat transfer in the down-
in the downstream region is reduced. It is reasonable that the stream region.
inverse jetting coolants forms an effective protecting film on The cooling effectiveness with different coolant allocations is
the specimen surface, and the gaseous film can protect not only the calculated. An adiabatic case without cooling is simulated and
stagnation region but also the downstream region. However, the the numerical cooling effectiveness is defined as:
 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 65

Fig. 9. Temperature (a) contours and (b) distribution around the specimen surface.

T adia  T w 3.3. Coolant demand

gnum ¼ ð2Þ
T adia  T c
In order to investigate the influence of the ‘‘step” method on the
coolant demand, another coolant allocation i.e. ‘‘step method 4” is
conducted. The ‘‘step method 4” is aimed to keep the cooling effec-
where Tadia and Tw are the surface temperature of specimen in the tiveness in stagnation region as almost high as the ‘‘step method 2”
adiabatic case and cooled case, and Tc is the coolant inlet case. The detailed allocation in two cases is shown in Table 3. The
temperature. total coolant mass flow in the ‘‘step4” case has been saved of
As shown in Fig. 11, with the increase of coolants mass flow in 38.34% compared with the ‘‘step method 2” case. And in the two
stagnation region, the cooling effectiveness of the overall surface cases, the surface temperature and cooling effectiveness are stud-
increases. And the increasing amplitude of cooling effectiveness g ied as shown in Fig. 12.
in stagnation region is larger than that in downstream region. It can be found that in the two cases, the temperature and cool-
Hence, g is gradually becoming larger in stagnation region than ing effectiveness in the stagnation region are almost same.
that in downstream region. It is significant that in hypersonic Although the cooling effectiveness in the downstream region is
flight, the aerodynamic heating and pressure are the highest in reduced with the decrease of coolant mass flow here, the overall
stagnation region and the method is required that can effectively cooling effectiveness is still kept in a relatively high level. It is sig-
increase the local cooling effectiveness in stagnation region but nificant that by reasonably increasing coolant mass flow in stagna-
without extra coolants consuming in downstream region. tion region and decreasing it in downstream region, the cooling
66 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69

Fig. 10. Heat flux distribution along the centerline of the nose cone surface.

Fig. 11. Cooling effectiveness distribution along the centerline of the nose cone surface.

Table 3 ant flow characteristics in the porous media. Most studies are focus
Coolant allocation in different regions. on the structure optimization, such as an un-equal thickness con-
figuration, a multi-channel design in the coolant chamber. How-
Stagnation region Downstream region Total mass flow
velocity (m/s) velocity (m/s) (g/m * s) ever, a structure optimization is usually limited by the leading
edge scales and chamber circuit designs. So, the following feasibil-
Step method 2 15.895 2.5973 12.431
Step method 4 20 0 8.986 ity analysis is focus on the porous material property. In the case,
the most important parameters influencing the coolant flow per-
formance in a porous material include the porous material proper-
ties and coolant inlet conditions. Permeability K is a direct
parameter that represents the transport capacity of a porous mate-
effectiveness in stagnation region can be controlled in a relatively rial. According to the study of Ergun [23], the permeability K is a
high level and the total coolant demand can be effectively reduced, function of pore diameter dp and porosity e.
simultaneously.
dp  e3
2

K¼ ð3Þ
150ð1  eÞ2
4. Feasibility analysis
And in practical machining process, the material porosity is a
In order to accomplish the coolant optimized allocations, the direct and controllable parameter. So, in the following simulations,
TPS optimization design should be conducted to influence the cool- the porosity is the variable of the porous material property, and
 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 67

Fig. 12. Temperature and cooling effectiveness distribution along specimen surface.

Fig. 13. Schematic diagram of porous specimen and coolant zone.

with a constant pore diameter, a high porosity means a high per- the required driving force to inject a constant quality of coolants is
meability or a high transport capacity. Based on the aforemen- low. But the coolant is mostly concentrated in the downstream
tioned optimized coolant allocation, a non-uniform porosity region as shown in Fig. 14(b). So, in order to increase the coolant
strategy of the porous specimen is performed that in the stagnation mass flow in the stagnation region, the total coolant injection flow
region, the porosity is kept the same value of 0.33 but in the down- must be increased, but this is an un-economic method. In a non-
stream region, the porosity is decreased to 0.2 and 0.1, respec- uniform porosity strategy, the DP is relatively high, which means
tively. Besides, a uniform porosity of 0.33 is simulated to be the driving pressure to inject coolants is high. With the increase
compared. of the differential porosity between the stagnation and down-
Basically, for the coolant chamber inlet, there are two coolant stream region, the DP of the overall specimen surface increases.
injection methods that are constant flow injection and constant For the stagnation region, the increase of the driving pressure leads
pressure injection. The numerical simulations were conducted to to more coolants penetrate here. But in the downstream region, the
analyze the feasibility of the optimized coolant allocation by apply- increase of the driving pressure is caused by the decrease of mate-
ing different coolant injection conditions and porous material rial porosity, which directly influences the transport capacity of
porosity. The surface temperature, pressure and heat flux calcu- porous material. Hence, the downstream region, which has a
lated in the ‘‘step2” case are applied as a boundary condition into relatively low porosity, will sustain high pressure. But it is favor-
the external surface of the porous nose cone. The local thermal able that a lower porosity means a higher pressure resistance. By
equilibrium (LTE) model, which assumes that the fluid and solid applying a non-uniform porosity in the porous specimen, the
temperature in the porous media are equal, has been proven to coolant flow in the stagnation region has been effectively
be an effective method for numerical simulations of transpiration increased. The distribution tendency of the mass flux is quite con-
cooling [22], so it is used in the following simulations. sistent with the aforementioned optimized strategy of the coolant
allocation.
4.1. Constant flow injection According to the cooling effectiveness analysis of different cool-
ant allocations, the increase of coolant flow in the stagnation
At a constant coolant injection flow of 5 g/s and 10 g/s, respec- region is favorable for the overall cooling performance. In order
tively, the differential pressure DP between the internal and exter- to evaluate the efficiency to increase the coolant flow of the stag-
nal surface of the porous specimen and the mass flux along the nation region by applying the non-uniform porosity strategy, a
external surface are investigated as shown in Fig. 14. It is noted parameter gm is suggested and defined as the coolant mass flow
that the differential pressure between the internal and external ratio of the stagnation region and total region.
surface is calculated along the porous thickness direction ‘‘h” as
_
m
shown in Fig. 13. One can find that at a same coolant injection flow, gm ¼ _ s  100% ð4Þ
with a uniform porosity, the DP is relatively low, which means that mt
68 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69

Fig. 15. (a) Differential pressure and (b) coolant mass flux along the specimen
surface.

the coolant optimized allocation, thus improve the local cooling

effectiveness of the stagnation region. But it is noted that the
accomplishment of the coolant allocation is at the cost of the
Pin/Ps

increase of the required driving force to inject coolants.

4.2. Constant pressure injection

At a constant coolant injection pressure of 650 kPa, the differen-

tial pressure DP between the internal and external surface of the
porous specimen and the mass flux along the external surface
are investigated as shown in Fig. 15. It can be found that at a same
injection pressure, a non-uniform porosity strategy can decrease
the pressure compression in the stagnation region. It is beneficial
from the view point of structure stability and mechanical property
Fig. 14. (a) Differential pressure, (b) coolant mass flux along the specimen surface
and (c) at different injection flows. of the porous nose cone. Besides, the corresponding total coolant
mass flow is quite different. The total coolant flow with a non-
Fig. 14(c) presents the mass flow ratio of the stagnation region uniform porosity of ‘‘0.33&0.2” and ‘‘0.33&0.1” is 5 g/s, 3.1 g/s,
over the total coolant flow and the pressure ratio of the coolant respectively. But the total coolant flow in the uniform porosity is
inlet and freestream stagnation point. It can be found that a non- 50.5 g/s, which is ten times or more of the coolant flow with a
uniform porosity strategy can effectively increase the mass flow non-uniform porosity. According to the mass flux along the sur-
ratio in the stagnation region. With the increase of the differential face, at a same injection pressure, even though the increase of dif-
porosity, gm increases, which implies that the efficiency of increase ferential porosity makes the total coolant injection flow reduce, the
mass flow ratio in stagnation region enhances. However, the reduced coolant flow is mostly from the downstream region. It is
required injection pressure is increased, accordingly. The results indicated that at a constant pressure injection, a non-uniform
indicate that the non-uniform porosity strategy can accomplish porosity strategy can save the coolant consumption, and an
 L. Shen, J. Wang / Applied Thermal Engineering 127 (2017) 58–69 69

increase of differential porosity between stagnation and down- [2] T.A. Jackson, D.R. Eklund, A.J. Fink, High speed propulsion: Performance
advantage of advanced materials, J. Mater. Sci. 39 (2004) 5905.
stream region can further reduce the coolant consumption and
[3] D.E. Glass, Ceramic matrix composite (CMC) thermal protection systems (TPS)
guarantee the high quantity of mass flow in the stagnation region, and hot structures for hypersonic vehicles, in: 15th AIAA Space Planes and
simultaneously. Hypersonic Systems and Technologies Conference, 2008, pp. 1–36.
[4] E.R.G. Eckert, N.B. Livingood, Comparison of effectiveness of convection-,
transpiration-, and film-cooling methods with air as coolant, 1954.
[5] J. Polezhaev, The transpiration cooling for blades of high temperatures gas
5. Conclusion
turbine, Energy Convers. Manage. 38 (10) (1997) 1123–1133.
[6] D.E. Glass, A.D. Dilley, H.N. Kelly, Numerical analysis of
An optimization method of local transpiration cooling effective- convection/transpiration cooling, J. Spacecr. Rock. 38 (1) (2001) 15–20.
ness under supersonic condition is suggested that two strategies of [7] W.A. McConarty, F.M. Anthony, Design and evaluation of active cooling
systems for Mach 6 cruise vehicle wings, NASA Contractor Report, 1971.
coolant allocation over the nose cone surface are performed. One is [8] T. Langener, J. Von Wolfersdorf, M. Selzer, et al., Experimental investigations
constant velocity allocation and the other is step velocity allocation of transpiration cooling applied to C/C material, Int. J. Therm. Sci. 54 (2012)
that the coolant flow is increased in stagnation region but 70–81.
[9] H. Hald, A. Herbertz, M. Kuhn, et al. Technological aspects of transpiration
decreased in downstream region. The numerical investigation on cooled composite structures for thrust chamber applications, in: 16th AIAA
flow field distribution and cooling performance with different cool- DLR DGLR International Space Planes and Hypersonic Systems and
ant flow allocations are conducted and their reduction perfor- Technologies Conference, 2009, pp. 19–22.
[10] D. Greuel, A. Herbertz, O.J. Haidn, et al., Transpiration cooling applied to C/C
mance of coolant consumption is studied. Then, a numerical liners of cryogenic liquid rocket engines, in: 40th Joint Propulsion Conference,
investigation of the coolant flow characteristic in the porous spec- Reston, VA, AIAA, 2004.
imen with a uniform and non-uniform porosity at different coolant [11] Y.Q. Liu, P.X. Jiang, S.S. Jin, et al., Transpiration cooling of a nose cone by
various foreign gases, Int. J. Heat Mass Transf. 53 (23) (2010) 5364–5372.
inlet conditions was conducted and the feasibility of the coolant
[12] A.V. Foreest, M. Sippel, A. Gülhan, et al., Transpiration cooling using liquid
optimized allocation was analyzed. Through the numerical investi- water, J. Thermophys. Heat Transf. 23 (4) (2009) 693–702.
gations, the following conclusions can be drawn. [13] L. Shen, J. Wang, W. Dong, et al., An experimental investigation on
transpiration cooling with phase change under supersonic condition, Appl.
Therm. Eng. (2016).
1. The penetrating coolant enlarges the high pressure area and low [14] F. He, J. Wang, Numerical investigation on critical heat flux and coolant volume
Mach number area in the front of the stagnation region and the required for transpiration cooling with phase change, Energy Convers.
step coolant allocation can effectively push the shock wave Manage. 80 (2014) 591–597.
[15] J. Wang, L. Zhao, X. Wang, et al., An experimental investigation on
away from the specimen stagnation region. transpiration cooling of wedge shaped nose cone with liquid coolant, Int. J.
2. Compared with the constant allocation, the step allocation Heat Mass Transf. 75 (2014) 442–449.
improves the cooling effectiveness in the leading edge and even [16] J. Wang, J. Messner, H. Stetter, An experimental investigation of transpiration
cooling. Part I: Application of an infrared measurement technique, Int. J.
enhances the cooling effectiveness in the downstream region Rotating Mach. 9 (3) (2003) 153–161.
even though the coolant flow here is reduced moderately. [17] L. Zhao, J. Wang, J. Ma, et al., An experimental investigation on transpiration
3. With a same high level of cooling effectiveness in the stagnation cooling under supersonic condition using a nose cone model, Int. J. Therm. Sci.
84 (2014) 207–213.
region, the step allocation can effectively reduce the coolant [18] Z. Huang, Y. Zhu, Y. Xiong, P. Jiang, Investigation of transpiration cooling for
demand. sintered metal porous struts in supersonic flow, Appl. Therm. Eng. 70 (1)
4. The feasibility of the coolant optimized allocation is preliminar- (2014) 240–249.
[19] S. Gulli, L. Maddalena, S. Hosder, Integrated analysis for the design of reusable
ily validated through a non-uniform porosity strategy of porous
TPS based on variable transpiration cooling for hypersonic cruise vehicles, in:
specimen that it can effectively increase the mass flow ratio of 10th International Energy Conversion Engineering Conference, 2012, pp. 4161.
stagnation region over the total region but its accomplishment [20] F.R. Menter, Two-equation eddy-viscosity turbulence models for engineering
applications, AIAA J. 32 (8) (1994) 1598–1605.
is at the cost of an increase of the required injection pressure.
[21] F.S. Billig, Shock-wave shapes around spherical-and cylindrical-nosed bodies, J.
Spacecraft Rock. 4 (6) (1967) 822–823.
[22] J. Wang, H. Wang, A discussion of transpiration cooling problems through an
analytical solution of local thermal non-equilibrium model, J. Heat Transf.
References ASME 128 (10) (2006) 1093–1098.
[23] S. Ergun, Fluid flow through packed columns, Chem. Eng. Prog. 48 (1952) 89–
[1] S.H. Choi, S.J. Scotti, K.D. Song, H. Ries, Transpiring cooling of a scram-jet 94.
engine combustion chamber, in: 32nd AIAA Thermophysics Conference, [24] C. Dittert, M. Selzer, H. Böhrk, Flowfield and pressure decay analysis of porous
Atlanta, GA, 1997. cones, AIAA J. (2017).