Available online at www.sciencedirect.

com

Applied Thermal Engineering 28 (2008) 881888 www.elsevier.com/locate/apthermeng

Heat transfer in a 155 mm compound gun barrel with full length integral midwall cooling channels
Bin Wu
b

a,*

, Gang Chen a, Wei Xia

a Department of Military Engineering, Artillery Academy, Hefei 230031, China School of Mechanical Engineering, South China University of Technology, Guangzhou 510640, China

Received 4 July 2006; accepted 7 July 2007 Available online 20 July 2007

Abstract When ring, large amounts of heat ow into the gun bore surfaces and result in wear and erosion of the gun bore. Moreover, the chamber surface temperature will reach the cook-o temperature of propellant during long sustained ring, which will impact on user safety and facilities. For large-caliber gun, a serious limitation on the weapons availability for action is imposed with high-energy propellants used and ring rates increasing. An eective method for solving this problem is to adopt barrel liquid-cooling technique. In this paper, heat transfer in a 155 mm midwall cooled compound gun barrel was analyzed theoretically. For the reference, heat transfer in a naturally cooled monobloc gun barrel was also discussed. Finite element analysis (FEA) method was employed to validate the results obtained by theoretical analyses. The present study showed: (1) natural air cooling is ineective for transferring the heat out of the barrel because the combined convection and radiation heat transfer coecient is relatively small; (2) forced midwall cooling has great heat extraction capability and is able to keep the chamber temperature below the cook-o temperature by increasing the heat transfer coefcient; (3) an optimal ow rate should be selected to balance the cooling eciency and the pressure loss. 2007 Elsevier Ltd. All rights reserved.
Keywords: Heat transfer; Gun barrel; Midwall cooling

1. Introduction Gun bore surfaces receive large amounts of heat resulting from the combustion of ammunition propellants and from the friction between the projectile driving band and the gun bore when ring. Before another round is red the barrel has some time to cool down, but only a small amount of heat is transferred to the environment by means of convection and radiation, which leads the temperature of the barrel to increase. During long-burst ring, heat accumulates and causes the barrel to reach a high temperature, limited by the cook-o temperature. Once the cooko temperature is reached, propellant from a new round will self-ignite. In rapid-re small-calibre weapon, such as

Corresponding author. E-mail address: mewubin@tom.com (B. Wu).

machine guns and machine cannons, the barrel can become red-hot under long-burst ring condition, and subsequently the ring must be stopped immediately to allow barrel cooling. In addition to cook-o, gun bore erosion has a close relation to this intense thermal condition. In general, the major contributors to erosion damage are thermal eects, chemical attack of propellant gases, mechanical wear of projectile passage, and mechanical loading from gas pressurization [1]. No matter what the specic mechanism of erosion, it is a function of the rate of heat transferred to the gun bore surface. Wear and erosion of the gun barrel reduces the accuracy of the projectile and ultimately diminishes the life of the gun. The mission requirements of future war ghting require the development of new generations of large-calibre guns capable of providing increased range, rate, accuracy, and energy on target. High-energy propellants with ame temperature up to 3700 K are used for improving gun

1359-4311/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2007.07.010

882

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

Nomenclature A Abo Ajo AC r rbi rbo rji rjo cp q k a E e d l V n T Tf T0 T1 Tbi,j area the heat exchange area of cooling channels the heat exchange area of the outside surface of the jacket the cross-section area of the cooling channels radius inner side radius of barrel outer side radius of barrel inner side radius of jacket outer side radius of jacket specic heat at constant pressure density thermal conductivity thermal diusivity Youngs modulus Poissons ratio barrel wall thickness length of barrel the volume of gun barrel round per minute temperature the reference temperature of the coolant the initial temperature of gun bore surface before rst round the ambient temperature the initial temperature of gun bore surface before round (j + 1)th Tc Tbo Tjo Tmax DTbi,j (DTj)V DT j h h1 hc hr Qin Qout Q0out Q00 out t th s H1 um qV j the cook-o temperature the outside surface temperature of gun barrel the outside surface temperature of the jacket the maximum bore temperature the temperature increment resulted from red round jth the temperature increment of dV the average bulk temperature increment heat transfer coecient the combined convective heat transfer coecient the natural convection heat transfer coecient the radiation heat transfer coecient the thermal energy transferred into the gun bore the thermal energy transferred out of the gun barrel heat exchange between cooling channels and coolant heat exchange between the outside surface of jacket and the environment time the heating period time-constant the total heat transfer per unit area per round the average coolant velocity the volume ow integer

performance. However, unacceptable erosion occurs due to the large thermal energy exchange between the combustion of the ammunition propellant and the gun bore. Additionally, ring at sustained high rates will quickly make the gun too hot to load or re and impact on user safety and facilities as a result of the possibility of cook-o. A Thermal Warning Device (TWD) is used to indicate to the gun crew when it is safe to shoot. It does not work for modern large caliber gun that are capable of ring at rates of 810 rounds per minute. Generally, it takes more than 12 h for the barrel to cool down from the cook-o temperature to ambient temperature under natural air cooling condition. So, other methods of cooling barrel are needed to retain barrel life and enable artillery weapons to increase individual mission durations and reduce their recovery time between missions. Wu [2] briey reviewed cooling technologies for gun barrels, which are either passive or active. Passive cooling technologies, such as chromium/tantalum plating and wear-reducing additives, are adopted to reduce the thermal energy input by setting up a thermal barrier between the hot propellant gas and the steel of the barrel. Lesquois et al. [3] studied thermal eects induced by mechanical friction in condition of high speed tribology and found

that the hard chromium coating was a good thermal barrier and mechanical protector for the substrate steel. Matson and co-workers [4] have demonstrated the eectiveness of tantalum as protection against high temperature wear and erosion. Franco and Peter [5] showed that using silicon dioxide as an additive can lead to a reduction of the steel temperature by approximately 150 C. Boisson et al. [6] showed that in the cross-section near the end of the barrel additives reduce the inner wall maximum temperature value to about 250 C, and about 150 C in the section near the forcing cone. Lawton [7] showed that the wear-reducing additive gradually reduced the surface temperature uctuation from about 950 C to about 600 C, and reduced heat transfer per round from about 950 kJ/m2 to about 600 kJ/ m2 over a period of 50 rounds, which is expected to increase the number of rounds to cook-o from about 53 to 90 at 3 rounds/min. It is well known that chromium plating and wear-reducing additives reduce wear and erosion of the gun bore and consequently extend barrel life. However, it has a minor eect on the heat entering into the gun bore surface; the majority of the heat is still transferred into the gun bore surface and accumulates with the number of rounds. Active cooling technologies, such as nned-barrel cooling and

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

883

forced liquid/air cooling, reduce the temperature increment of each round red by increasing the thermal dissipation rate. Finned barrels with longitudinal or radial ns have been employed in small-calibre guns, such as machine guns and machine cannons. In addition, many mortars also have nned barrels because their barrel walls are relatively thin. Barrel liquid-cooling technology was rst used on the machine gun with an external water-cooled jacket around the barrel by Hiram Maxim in late 1880s and then applied to other small-calibre barrels. Bofors invented the rst internal cooling system in 1948. Barrel-cooling technology has been widely employed in navy guns such as the Italian OTO Melara 76 mm, the French Creusot-Loire 100-mm Compact, and the Russian AK630. The 155-mm XM297 Crusader cannon incorporates the unique active cooling channels within its structure, which maintains chamber temperature to below the safe temperature limit. This enables continuous ring of ammunition in the gun barrel in combat [8]. A barrel-cooling system to provide higher maximum ring rates is being developed for the AS90 by BAE Systems RO Defence [9]. Barrel cooling is one of the optional subsystems available to increase the eciency of the G6 by Denel [10]. A cooling system is one of features of the ST Kinetics 120 mm Super Rapid Advanced Mortar System (SRAMS), which reduces temperature build up thereby enabling continuous ring at a maximum rate of up to 18 rpm [11]. Barrel liquid-cooling has three modes: internal cooling, midwall cooling, and external cooling. Because internal cooling of the gun bore is only executed after the projectile exits, and the rapid cooling of gun bore surface will induce thermal stress, this mode may not meet the requirements of sustained high rates of the re mission. External cooling has found wide application in navy guns. But for large calibre self-propelled howitzers (SPH), the strength, weight and size of barrel with an external cooling jacket may not be desirable. This is the reason why midwall cooling is being developed for the SPH. Midwall cooling is also used in electromagnetic rail guns, in which the thermal conditions are much more extreme [12]. Laboratory heat transfer experiments on a 155 mm compound gun tube with full length integral midwall cooling channels were carried out by Bass and de Swardt [13]. The fact that natural cooling is dramatically slower than forced midwall cooling is proved. They also carried out numerical modeling and close correlation between experimental data and numerical predictions has been achieved. The purpose of this work was to analyze theoretically the heat transfer in a 155 mm compound gun barrel with full length integral midwall cooling channels. Finite element analysis (FEA) method was also employed to validate the analytical results and simulate the heat transfer characteristics under the condition of sustained ring at high rates. For the reference, the heat transfer in conventional monobloc gun barrel under natural air cooling was also discussed.

2. Theoretical analysis of heat transfer process 2.1. Heat transfer in monobloc gun barrel Heat transfer in a 155 mm monobloc gun barrel under natural air cooling condition was rstly analyzed. It should be noted here that the gun barrel is simplied as a cylinder to make the analysis easier. For a sustained ring of N rounds, the initial temperature of gun bore surface of round jth, Tbi,j, can be given by the following equation: T bi;j T 0 T bi;1 T 0 T bi;j T bi;j1 T 0 DT bi;1 DT bi;j T0
j X 1

DT bi;j

for j 1 to N

where T0 is the initial temperature of gun bore surface before rst round, Tbi,j the initial temperature of gun bore surface before round (j + 1)th, DTbi,j the temperature increment resulted from red round jth. In [7], this initial bore temperature was called the quasisteady temperature, which is the long-term bore temperature that transmits heat to the ammunition when it is loaded and it is assumed that self-ignition occurs when the quasi-steady bore temperature reaches the so-called cook-o temperature of the propelling charge, Tc, namely T bi;j T c T 0
j X 1

DT bi;j

This means that round (j + 1)th cannot be red and ring has to stop to allow the barrel to cool down. For large caliber artillery guns, the barrel temperature increases by between 2 C and 3 C with each shot at rates of 24 rounds per minute. If the initial temperature of the barrel is 27 C and the cook-o temperature is assumed to be 200 C in this case, we got the total temperature increment Pj 1 DT bi;j 173 C, which indicates that only about 5886 shots can be red. However, with the aid of automated or assisted replenishment of ammunition, modern 155 mm artillery guns are capable of ring at higher rates and consequently in less than half an hour the barrel can reach the cook-o temperature. Obviously, a serious limitation on the weapons availability for action is imposed. The energy balance of gun barrel for round jth is as follows (see Fig. 1): Z Qin Qout qcp DT j V dV Qout qcp V DT j 3
V

where Qin is the thermal energy transferred into the gun bore, Qout the thermal energy transferred out of the gun barrel, q the density of the steel, cp the specic heat, V the volume of gun barrel, (DTj)V the temperature increment of dV, DT j the average bulk temperature increment. As is well known, each time the gun is red the gun steel at the gun bore experiences a large temperature uctuation.

884
dV

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

r
r + dr

Jacket
rbi = 77.5

Barrel (Liner)

rbo = 127.5

Fig. 1. Schematic of monobloc gun barrel.

Cooling channel

Due to short period of heating phase, this temperature uctuation cannot penetrate far from the gun bore surface along the radial direction. For example, it may be 1000 C at the surface but 1 mm from the surface it is only 50100 C. Since the heat ux induces an acute temperature gradient in the radial direction when a round is red, the average bulk temperature increment is employed to evaluate the total temperature increment of gun barrel. For monobloc gun barrel under natural air cooling condition, Qout can be given by Newtons law Qout h1 AT bo T 1 Dt 4

Fig. 2. Schematic of mid-wall cooled compound gun barrel.

where h1 is the combined convective heat transfer coecient (h1 = hc + hr, where hc is the natural convection heat transfer coecient, and hr the radiation heat transfer coefcient), Tbo the outside surface temperature of gun barrel, A the outside surface heat exchange area, T1 the ambient temperature, Dt = 60/n (s), having a rate of re, n (rpm). For a 155 mm gun, the maximum heat transfer capacity Qout can be calculated approximately as 24.9 kJ (here we consider unit length and let h1 = 30 W/(m2 K), constant ring rate n = 10 rpm), and Qin approximates 486.7 kJ (Heat transfer per round typically is about 1.0 J/mm2). Substituting Qin and Qout into Eq. (3), we obtain DT j 4 C. As we know, when ring rate increases, the period between two shots will decrease and the barrel has less time to cool down. So, the temperature increment induced by a shot at high rate is larger than that at low rate. This means that fewer shots can be red before the chamber of the barrel reaches the cook-o temperature. From the aforementioned fact it can be concluded that the most eective cooling method is to greatly increase heat transfer coecient by using forced liquid/air cooling. Due to the dierence of thermal properties between liquid and air, it is clear that forced air cooling is not the rst choice. 2.2. Heat transfer in midwall cooled compound gun barrel Midwall cooled compound gun barrel is composed of an outer jacket and a barrel (liner) with an array of axial semicircular grooves on its exterior (see Fig. 2). Cooling liquid is passed between the barrel and the jacket. For small to medium caliber gun, the outer jacket may be loose removable because the stresses in the barrel are relatively low. However, much higher stresses occur in the barrel of larger

caliber guns like the 155 mm artillery gun. The outer jacket should be an integral part of the barrel to withstand the internal pressure with the liner. The problem of manufacturing midwall cooled compound gun barrel is not considered in this work. For midwall cooling, the thermal energy is transferred out of the barrel through two ways: heat exchange between cooling channels and coolant through forced convective heat transfer, Q0out , and heat exchange between the outside surface of jacket and the environment through radiative and convective heat transfer, Q00 . So we have out Qout Q0out Q00 out Q0out hf Abo T bo T f Dt Q00 h1 Ajo T jo T 1 Dt out 5 6 7

where hf is the forced convective heat transfer coecient, Abo the heat exchange area of cooling channels, Tbo the outside surface temperature of the barrel, Tf the reference temperature of the coolant, Ajo the heat exchange area of the outside surface of the jacket, and Tjo the outside surface temperature of the jacket. The heat transfer in the gun barrel during sustained ring is transient. It takes some time for heat to transfer from the gun bore surface to the outside surface of the barrel and then to the jacket. So, only after some rounds red can a certain temperature dierence occur between the outside surface of the barrel and the coolant. Let us rst omit Q00 and assume that the heat input into the gun bore durout ing round jth is completely transferred out of the barrel by coolant, we obtain Qin Q0out hf Abo T bo T f Dt 8

In other words, for arbitrary round after jth, the bore temperature will increase fast to its maximum temperature during heating phase and then decrease back to its original value before next round is red. This heat transfer regime is here called steady-like heat transfer. In fact, the heat input into the gun bore surface in less than 20 ms during heating phase is balanced with the heat transferred out of the barrel by coolant in 6 s. Therefore, the heat transfer process is always transient. The heat transfer process during a shot can be approximated as thermal energy that

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

885

owed into the gun bore surface and rstly stored in a thin layer during heating phase then transferred along the radial direction to the outside surface of gun barrel during cooling phase by means of heat conduction. Here, a heat ux transformation method is presented to compute Tbo and then hf. The essence of this method can be expressed by Z th Z Dt qin dt qi dt 9
0 0

where, th is the heating period when ring, qin the heat ux input into the gun bore in the heating period, qi the transformed heat ux input into the gun bore in Dt. We can transform steady-like heat transfer into steady heat transfer by transforming heat ux. Therefore, Tbo can be computed according to steady heat conduction equation. Generally, factors aecting the forced convective heat transfer coecient (HTC) can be classied into three sorts: the rst is the coolant velocity; the second is the characteristic length, which is determined by the geometry of the channels; the third is the thermal physical property of coolant, which varied with the used coolant. Design of cooling channels is important for enhancing the heat transfer and increasing pressure loss as small as possible at the same time. Furthermore, the eect of the cooling channels on the stresses in the barrel is signicant. This problem was discussed in detail by de Swardt [14] and not considered in this paper. So, when the geometry of the cooling channels and coolant are determined, the forced convective heat transfer coecient will be only dependent on the coolant velocity, namely,   q hf gum g V hqV 10 AC where um is the average coolant velocity, qV the volume ow, AC the cross-section area of the cooling channels. For a specic geometry of the cooling channels, the value of AC is a constant. Tbo, hf and qV can be obtained according to the following three equations: 2pklT bi T bo lnrbo =rbi H 1 2prbi l hf 2prbo lT bo T f Dt H 1 2prbi l=Dt hf 2prbo l T bo T f Dt qqV Dt cp T fo T fi 11 12 13

sumed 30 mm. The gun barrel is assumed to be made of steel and to bear the typical geometrical, mechanical and thermal properties as listed in Table 1. The water is used as the coolant and therefore assumed that the coolant temperature will not exceed 100 C. The inlet and outlet temperatures of the water are assumed to be 27 C and 93 C in this analysis. Combining Eqs. (11)(13), we obtain Tbo = 103.5 C, hf = 2523.5 W/(m2 K) and qV = 132.2 lpm (litres per minute). In the above analysis, the heat transferred out of the barrel through the heat exchange between the outside surface of the jacket and the environment is omitted. In fact, if taking this amount of heat into account, the gun should be safer given the same coolant volume ow. Combining and rewriting Eqs. (12) and (13), we have qV T fo T fi H 1 2prbi l C1 qcp Dt H 1 2prbi l hf T bo T f C2 2prbo l Dt 14 15

Comparing both Eqs. (14) and (15), we know that they have the same function form f x; y xy C 16

Bass and de Swardt [13] observed that the cooling capability increased markedly with an increase in ow rate from 8 to 12 l/m but less so when increased from 12 to 16 l/m and no signicant dierence in cooling between 16 and 20 l/m. However, they did not give out the reason why these phenomena occur in their forced cooling tests, whereas this fact could be deduced from Eq. (16) and well explained. Generally, for internal forced liquid cooling, the coolant pressure loss between the barrels coolant inlet and outlet ports increases with increasing the coolant velocity. Therefore, an optimum coolant volume ow should be selected in order to balance the cooling capability and the pressure loss. 3. Finite element analysis of heat transfer 3.1. The mathematical model One-dimensional (radial) heat ow analysis is adopted to approximate the near bore transient temperature. Although it has been shown that there is a dierence between the 1D and 2D calculations, 2D conduction calculations are quite dicult to realize because a square mesh is needed if we want a good accuracy [6]. So, 1D heat transfer analysis is performed in the present study by using MSC.Marc. The nite element mesh in thermal analysis

In [15], the total heat transfer per unit area per round, H1, is given as 0.9136 MJ/m2 for a 155 mm gun, which is used in our work for the reference. In this case, the actively liquid cooled 155 mm gun is also assumed to be red at a constant rate, 10 rpm and the barrel wall thickness is asTable 1 Geometrical, mechanical and thermal properties of the gun barrel Inner radius rbi (mm) 77.5 Barrel thickness d (mm) 30 Youngs modulus E (GPa) 201 Poissons ratio e 0.3 Density q (kg m3) 7833

Specic heat cp (J kg1 K1) 460

Thermal conductivity k (W m1 K1) 40

Thermal diusivity a (m2 s1) 1.11 105

886

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

is shown in Fig. 3. The computational domain is partitioned into 1932 eight-node quadrilateral axisymmetric elements. Since high gradient of the temperature is localized to a region adjacent to the gun bore surface, mesh renement is employed in this region. Moreover, any possible temperature dependence of the thermal and mechanical properties is neglected. The governing equation for transient heat conduction is 1 oT rT a ot
2

qin 1:927 108 expt=4:74 W=m ; qin k

 oT   or rrbi 19

This value is induced from the following equation presented in [15]: H H 1 1 et=s 20 where a heat transfer per round H1 = 0.9136 MJ/m2 and a time-constant s = 4.74 ms. 2. At the outer surface of the barrel: r = rbo  oT  k  hf T bo T f 21 or rrbo 3. At the inner surface of the jacket: r = rji  oT  k  hf T f T ji or rrji 4. At the outer surface of the jacket: r = rjo  oT  k  h1 T jo T 1 or rrjo

17

where a = k/qcp is the thermal diusivity of the gun barrel. The initial condition is T r; 0 T 0 T 1 27  C rji 6 r 6 rjo and t 0 for rbi 6 r 6 rbo ; 18

The boundary conditions for the gun barrel are: 1. At the inner surface of the barrel: r = rbi

22

23

It should be noted that in the thermal analysis h1 is assumed to be 40 W/(m2 K) and remain unchanged for the entire period of time considered. The ring rate assumes to be 10 rpm and constant for 3 min ring mission so that one round is red every 6 s. The heating and cooling periods are set to be 0.02 s and 5.98 s, respectively. 3.2. Validation of the numerical model Finite element calculations of the transient temperature during the heating phase for the inner surface and the outer surface of the rst round are shown in Fig. 4. The calculated result of the inner surface temperature agreed well

800 700 600

Temperature T [ C]

500 400 300 200 100 0 0 5 10 15 20

Tbi Tbo

Time t [ms]

Fig. 3. The nite element mesh used in the present work.

Fig. 4. Transient temperatures of the inner surface and the outer surface during heating phase.

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

887

Temperature T [ C]

with the measured and theoretical values provided by Lawton [15]. The maximum bore temperature, Tmax, can also be computed according to the following equation provided in [15]: r H1 a T max T i 1:082 24 ps k Generally, when t = s, the bore temperature reaches its maximum. The simulated result showed that s is equal to 4.4 ms and the maximum bore temperature is 672.8 C, whereas that computed by Eq. (24) is 701.8 C. The relative error is 4%. Therefore, the accuracy of the present numerical model was validated. It can be observed from Fig. 4 that the outer surface temperature remain unchanged during the heating phase because heat cannot be transferred that far in such a short period. Fig. 5 shows the temperature gradient induced by thermal input during the heating phase. For example, the temperature at the position 1 mm below the surface is 67 C. It can be seen that the thermal energy input during the short heating period could not be transferred far from the bore surface. In fact, it was rst stored in a thin layer and then transferred to the outer surface during the cooling phase under the control of heat conduction. In other words, the heat transfer process includes three successive stages: thermal input, thermal store and thermal output. 3.3. FEA results of forced liquid cooling Figs. 6 and 7 show the simulated results of the inner surface temperature and the outer surface temperature of the barrel under natural air cooling and forced liquid cooling, respectively. It can be observed that the inner surface temperatures uctuated sharply, whereas the outer surface temperatures changed softly. Obviously, heat transfer process in gun barrel is always transient. However, due to large dierence of heat transfer coecient between natural air
800 700 600

900 800 700 600 500 400 300 200 100 0 0 30 60 90 120 150 180

Tbi Tbo
Time t [s]

Fig. 6. Inner and outer surface temperatures under natural air cooling.

900 800 700

Tbi Tbo

Temperature T[ C]

600 500 400 300 200 100 0 0 30 60 90 120 150 180

Time t [s]

Fig. 7. Inner and outer surface temperatures under forced liquid cooling.

500 400 300 200 100 0 0 10 20 30 40

Radius r [mm]

Fig. 5. The temperature distribution through the barrel wall during the heating phase.

cooling and forced liquid cooling, the temperature increment, DTbi, induced by a shot was steadily changing smaller and smaller under forced liquid cooling, but changed little under natural air cooling. So, with a number of rounds red the accumulation of temperature will lead the quasi-steady temperature to the safe temperature limit under natural cooling. In fact, the initial temperature of the 24th round had reached 200 C, which suggested that the last six rounds could not be red. However, the heat transfer process in midwall cooled compound barrel had nearly been controlled. Fig. 8 shows the correlation between the heat transfer coecient and the temperature dierence between the outer surface temperature and the reference temperature of the coolant. It was observed that the numerical results validated the theoretical analysis. When hf equals 2523.5 W/ m2 K (labelled as hfs), the calculated (Tbo Tf) (labelled as DTbf) is about 36.8 C. Therefore, the relative error between the results obtained by theoretical analysis method

Temperature T [ C]

888
80 70

B. Wu et al. / Applied Thermal Engineering 28 (2008) 881888

Temperature difference bf [ C]

FEA Theory

60 50 40 30 20 10

(d) The correlation between the heat transfer coecient and the volume ow links the experiment and the simulation. For a certain ring mission, a required heat transfer coecient can be rstly determined by nite element analysis and then the corresponding volume ow can be given out by experiments. (e) According to the relationship between the heat transfer coecient/volume ow and the temperature difference, an optimal ow rate should be selected to balance the cooling eciency and the pressure loss.

Acknowledgements
0 0 1000 2000 3000 4000 5000 6000

Heat transfer coefficient hf [W m-2 K]

Fig. 8. Correlation between heat transfer coecient hf and temperature dierence DTbf.

The authors thank Prof. Fu-qiu Ma for his great support in the study of barrel cooling and acknowledge the help of Dr. Pu-qing Chen for conducting much of the FEA work. References
[1] P.J. Cote, C. Rickard, Gasmetal reaction products in the erosion of chromium-plated gun bores, Wear 241 (2000) 1725. [2] B. Wu, Barrel cooling: a key technology for improving gun performance, Journal of Battleeld Technology 8 (3) (2005) 14. [3] O. Lesquois, J.J. Serra, P. Kapsa, S. Serror, C. Boher, Degradations in a high-speed sliding contact in transient regime, Wear 201 (1996) 163170. [4] D.W. Matson, E.D. McClanahan, S.L. Lee, D. Windover, Properties of thick sputtered Ta used for protective gun tube coatings, Surface and Coatings Technology 146147 (2001) 344350. [5] P. Franco, H. Peter, Gun barrel erosion: study of thermally insulating layers, in: 19th International Symposium of Ballistics, Interlaken, Switzerland, 711 May 2001, pp. 315321. [6] D. Boisson, G. Legeret, J.F. Barthelemy, Experimental investigation of heat transfer in a 120 mm testing gun barrel based on a space marching nite dierence algorithm for the inverse conduction method, in: 19th International Symposium of Ballistics, Interlaken, Switzerland, 711 May 2001, pp. 163169. [7] B. Lawton, The inuence of additives on the temperature, heat transfer, wear, fatigue life, and self-ignition characteristics of a 155 mm gun, Journal of Pressure Vessel Technology 125 (2003) 315 320. [8] http://www.army-technology.com/projects/crusader/. [9] http://www.army-technology.com/projects/as90/. [10] http://www.denel.co.za/liw/weapons/G6.htm. [11] http://defence-data.com/dsei/pageds1019.htm. [12] T.W. Hurn, J. DAoust, L. Sevier, J. Wesley, Development of an advanced electromagnetic gun barrel, IEEE Transactions on Magnetics 29 (1) (1993) 837842. [13] M. Bass, R.R. de Swardt, Laboratory heat transfer experiments on a 155 mm compound gun tube with full length integral midwall cooling channels, S7/P30, in: Gun Tubes Conference 2005, 1014 April 2005, Keble College Oxford. [14] R.R. de Swardt, Finite element stress analysis of autofrettaged midwall cooled compound gun tubes, S2/P8, in: Gun Tubes Conference 2005, 1014 April 2005, Keble College Oxford. [15] B. Lawton, Thermo-chemical erosion in gun barrels, Wear 251 (2001) 827838.

and numerical method is about 15%. As we know, only when the heat transfer coecient value is larger than hfs, the heat transfer process of the barrel can reach the steady-like regime. It was observed that small temperature dierence existed between the theoretical and numerical results and it decreased with increasing heat transfer coefcient. It was also found that there existed an optimum heat transfer coecient, namely, the corresponding optimum ow rate. 4. Conclusion Heat transfer in a 155 mm compound gun barrel with full length integral midwall cooling channels under forced liquid cooling condition and in a 155 mm monobloc gun barrel under natural air cooling condition is investigated comparatively. On the basis of the results obtained from the theoretical analyses and nite element simulation, the ndings are summarized as follows. (a) For a 155 mm gun, about 43 rounds could be red before the chamber surface temperature reaches the propelling charge cook-o temperature at 10 rpm under natural convection condition, which imposes a serious limitation on its availability for action. (b) Forced liquid cooling greatly improves the thermal dissipation rate by increasing the heat transfer coecient and is able to keep the chamber temperature below the cook-o temperature. (c) The temperature dierence between the outer surface of the barrel and the reference temperature of the coolant can be calculated by presented heat ux transformation method and then used to compute the heat transfer coecient.