Copyright ©1996, American Institute of Aeronautics and Astronautics, Inc.
AIAA Meeting Papers on Disc, July 1996A9637237, AIAA Paper 96-3118
Investigation of a nozzle tap-off liquid rocket engine scheme
Alexander A. Kozlov
Moscow Aviation Inst., Russia
Jose N. Hinckel
INPE, Sao Jose dos Campos, Brazil
Adalberto Comiran
INPE, Sao Jose dos Campos, Brazil
AIAA, ASME, SAE, and ASEE, Joint Propulsion Conference and Exhibit, 32nd, Lake
Buena Vista, FL, July 1-3, 1996
This paper describes a variation of the combustion tap-off cycle Liquid Rocket Engine (LRE) where the turbinedrive gas is bled off from the main combustion chamber. In the 'classical' tap-off scheme, a fuel rich gas mixture isbled from the main combustion chamber near the injector to drive the turbine. One of the main difficultiesassociated with this scheme is the design of a turbine robust enough to handle the large variation of flow propertiesin this region. In the scheme described herein, the turbine gas is bled from the boundary layer in the supersonicsection of the nozzle. In this region the flow conditions are more stable and the gas temperature low enough tosatisfy the turbine requirements. The calculations performed demonstrated that the scheme is feasible. Comparedto the open cycle scheme, a performance gain is obtained, especially in the lower thrust range. (Author)
Page 1
Investigation of a Nozzle Tap-off Liquid Rocket Engine Scheme
Alexander A. Kozlov* Jose N. HinckelJ Adalberto ComiranInstitute National de Pesquisas Espaciais
Caixa Postal 515Sao Jose dos Campos, SP, Brazil
AbstractThis paper describes a variation of the combustion tap-off cycle Liquid Rocket Engine (LRE) where
the turbine drive gas is bled off from the main combustion chamber. In the "classical" tap-off scheme, afuel rich gas mixture is bled from the main combustion chamber near the injector, to drive the turbine.One of the main difficulties associated with this scheme is the design of a turbine robust enough to handlethe large variation of flow properties in this region. In the scheme described herein, the turbine gas isbled from the boundary layer in the supersonic section of the nozzle. In this region the flow conditions aremore stable and the gas temperature low enough to satisfy the turbine requirements. The calculationsperformed, demonstrated that the scheme is feasible. Compared to the open cycle scheme, a performancegain is obtained , especially in the lower thrust range.
c:D:d:F:I:9-
gas velocitydiameterdiameterarea ratioimpulsemass fraction
Nomenclature
fc: oxidizer/fuel ratio R: gas constantL: gas pressure work T: temperaturerh: mass flow rate u:N: power, specific speed V:p: pressure a:P: thrust c.
Subscripts
blade velocityvolume flow rateequivalence ratiopartiallity
pressure ratio: efficiencyspecific heat ratioloss coefficientdensity, reactivityrotational speed
ad:bl:cr:ch:e:
adiabaticbladecriticalchambereffective
eng: engineent: entry
ex: exitf: fuelh: hydraulicm: massmech: mechanicalmid: midleo: oxidizer
A well known engine scheme for liquid rocket mo-tors is the one with a turbopump feed system with-out afterburning of the generator gas (Fig. la). Thepower required by the turbopump and the mass flowrequired to drive the turbine, increase with increas-ing chamber pressure. Since this gas is ejected withlow specific impulse, the specific impulse of the LREdevice decreases.
p: pumpsel: selectionsh: shaftsp: specifict: turbinen: nozzle
th:v.0:
theoreticalvolumetricentry sectionstoichimetric ratiomid sectionexit section
= Isp,ch9ch
+ mt)>,tflt (1)
where: fft, = r9ch =
The chamber pressure for this scheme of LREhas an upper limit of approximately 13 MPa.
A more recent generation of LRE use a schemewith a turbopump feed system with afterburning ofthe generator gas (Fig.lb).
"Visiting Researcher,Professor Moscow Aviation Institute, Moscow, RussiatAIAA Member
Copyright ©1996 American Institute of Aeronautics and Astronautics. All rights reserved.1
Scheme ofr I.KF! wilhoulafterburning ol generator
gdS
I chamber 2 - turbopump. 3 gasregulator of chamber mass ratio.
7 - regulator of gas generator mass ratio
Scheme of IRE with cafterburning ofgenerator gas
generator, 4 solid p rope lan t starter,thrust regulator.
Scheme of I.RE withturbine 2as drive fromsupersonic nozzle
Figure 1: Liquid rocket engine schemes
With this engine scheme, the chamber operat-ing pressure limit is approximately 25 MPa. Thisscheme is more complicated, with two or more fireunits.
A third engine scheme (Fig. Ic) is also possiblewith the drive gas for the turbine being bled off fromthe boundary layer in the supersonic extension of thenozzle. The drive gas is reinjected into the nozzle ata downstream location. This scheme combines theadvantages of the two schemes described above:
• The losses of specific impulse due to the tur-bine gas flow are eliminated.
• This LRE scheme has only one fire unit - andis therefore more reliable.
• In this scheme the chamber pressure may beincreased in comparison with the first scheme,therefore improved the energy and mass char-acteristics of the LRE.
However, the practical treatment of this schemerequires additional computational and experimentalwork to determine the optimal range of main engineparameters: the thrust, the chamber pressure, theexpansion ratio of the gas in the nozzle, and oth-ers. Additionally, constructive details, such as thelocation of bleeding the turbine gas from the nozzleand reinjection into the nozzle must be found alongwith the main parameters of the turbine. We mustcalculate the mass flow of gas to drive the turbineas function of the thrust and chamber pressure. The
main difficulty is the calculation of the efficiency ofthe pump and the turbine. The power requirementsof the pump is obtained from the following expres-sion
= ND N,p,f (2)
where: Npi0 = m0(pex,o - Pent,o)/A>»?p,oWp,f = rh[(pex.f - Pent,f)//3P7p,f
For an engine without afterburning of generatorgas, the exit pressures are obtained from the follow-ing expressions1':
Pex.o = 1.428pch, Pex.f = 1.714pch
The mass flow of the oxidizer and fuel are ob-tained from the following expressions:
mkm . mm° = ~,——~T> mf — ',——7
(3)The pump efficiency is calculated from the formulas
rip = J7h7?v'7mech (4)
where %, T;V and 77mech aie the hydraulic, volumet-ric and mechanical efficiencies respectively. The hy-draulic efficiency of the pump is obtained from thefollowing empirical correlation2':
0.042-0.172)2 (5)
PchP[kN]m0 [kg/s]rhf [kg/s]Nso
Nst
I?ho?7hfTJvo
IJvf
?7mo
?7mf^po
VplNP,o [W
JVp,/ (WNpE\W
5MPa51.0640.409435.3517.140.640.540.80.80.80.80.410.346151701190027100
153.1941.22961.229.20.7480.70.850.850.850.850.540.506345602555060086
255.3232.04779.0537.70.7810.740.850.850.850.850.5650.535550483856793615
15MPa5
15.097.440.6250.6180.70.70.80.80.350.346551003699584031
15
26.1812.90.7570.7140.850.850.850.850.5470.51610580474348180152
25
33.816.650.790.7530.850.850.850.850.570.5438169346117533286879
Table 1: Engine operating parameters
where the specific entry diameter Z?o,sp is obtainedfrom the following correlation:
(6)
dh = dsh + 0.003m, i = o,f. The diameter of the shafto!sh is calculated from the well know correlation:
UVp,0/WT (7)
T is the admissible potential of shaft diameter ontwirl. The following values were used for calcula-tions:
T = 108N/m2 = 4000rad/s = 5m/s
Volumetric and mechanical pump efficiencies r?v andT/mech were calculated as functions the specific speed
Nsp = (8)
r\v = tynech = °-9 - °-231 exP (-0.01936JVsp) (9)To start the calculations the following initial datawere used:
1. Entalpy of propellants: LOX (-398kJ/kg) andkerosene C7.2iH13.29 (-1728kJ/kg)
2. Oxidizer to fuel ratio: km = 2.6
3. Engine thrust: P = 5kN, 15kN, 25kN
4. Chamber pressure: pch = 5MPa, 15MPa
5. Propellant densities: p0 = 1140kg/m3
pf = 800kg/m3
6. Theoretical specific impulse: 7* = 3500m/s
7. The loss coefficients of specific impulse in thechamber and in the nozzle are given by: <fc^ —0.98 and <pn - 0.97 respectively
The results of the calculations of the pumps pa-rameters are shown in Table 1. For calculation of themass flow of turbine drive gas we use the balance ofpower condition:
Np = Nt = (10)
where L^ is the adiabatic work of the turbine gas.Lad is given by the formula.
Lad =rC — 1
[l - (pex/Pent)(K-1)/Kl (11)L J
where the numerical value is obtained by substitut-ing thevalues:
K
(*r)«,tPent/PexL**
1.190.382 10HJ/kg
150.907 10°J/kg
The theoretical value of the discharge velocityis Cth = y^Lad = 1347m/s; the critical velocity isCcr = v/2«:/ (K + 1) (RT)mt = 618m/s; the coeffi-cient of velocity A^o = Cth/Ccr = 2.18.
Pch
P[kN]^pc,S
Fcr [m2J
%e%,«™t [kg/s]
5MPa52.71 • 104
0.07426io-3
0.620.0820.370.0807
156.008 • IO4
0.1507-10-3
0.620.1170.4190.158
259.361 • IO4
0.219 • ID"3
0.620.14150.4410.234
15MPa58.403 • IO4
0.2038 • 10-J
0.620.1360.4370.212
1518.015 • IO4
0.388 • 10~a
0.620.1880.4850.41
2528.69 • IO4
0.584 • ID"3
0.620.2310.5170.614
Table 2: Engine data
Choose an active axial partial turbine (pt = 0,e < 1). Set the maximum of circular velocity at mid-diameter r/^jj* = 450m/s. The coefficient of lossesof velocity in the turbine nozzle for supersonic activeturbine is1':
¥?n = 0.975 + 0.005At,0 - 0.012A20 = 0.925 (12)
The absolute velocity at the entry of the turbinewheel is
?n = 1246m/s (13)
For choosing the efficiency factor of the turbine rje weuse the expression for a single stage active turbine:
= 0.0175 + 2.575u/ci - 2.5 (u/ (14)
For H/CI = 0.361 we have the value ije = 0.62 (firstguess). Further, using the method bellow we takeinto account the partiality of the turbine in the fol-lowing way: The mass flow of the gas at the turbineis given by:
iht = Wp.S/iad'Je (15)
The critical area of the turbine nozzle is obtainedfrom:
(16)
The coefficient of entry of turbine wheel: AI =Ath.o^n = 2.02. Define the gas dynamic functions:
«ty * (4i
9 ( A , ) :
and
_ /« + l\1/(K 1} / K-l 2\1/(K~-(———) Xl(l-^+lX>)
(17)
(18)
The gas pressure loss coefficient of the turbine nozzleis obtained from:
o-c = Pi,o/Pent,o = T(Ai)/rr (A) = 0.552 (19)
Nozzle exit area:
Partiality of the turbine:
= di/4Dmidsinai (20)
Set the entry angle of the turbine wheel blade atQI = 20°. The mid diameter of the de turbinewheel: Z?mid = 2umjd/w = 0.225m. Select the angu-lar velocity ai = 4000rad/s (near the bearing limitpoint) for improvement of turbine efficiency factorand mass characteristic of turbopump.
< 2.14 • 105rad • mm/s (21)The partial turbine efficiency factor is calculated us-ing experimental correlation.
= r?e(l-exP(-7.675e)) (22)
The calculation using the formulas 15 and 22 is re-peated until the mass flow rfit and efficiency factorare constant within a given accuracy. The results ofthe calculation for engine with thrust in the rangeof 5kN to 25kN and chamber pressure in the rangeof 5MPaan
PchQI<Pnrrmax^mid
d 15MPa are shown5MPa
20°0.925
450m/s
15MPa20°
0.925450m/s
in table 2
Further, using the correlation 1 we may cal-culate the increase in the specific impulse for en-gine device of scheme (l.c) as compared to scheme(l.a). The relative mass flow of turbine gas, <?t =f(P,pch) and the relative change of specific impulseA/sp = f(P,pCh) is shown in Figure 2 for engineusing scheme (l.a) at conditions:75p,t = 1500m/s.
= 3327m/s,
O ISMPi
"~ 15MP»O 5MP«
A SMPi 10
£00
0 10 20 30P(kN)
Figure 2. Gain of specific impulseand turbine mass flow
The analysis of Figure 2 shows that the increaseof the specific impulse in scheme (l.c) as comparedto scheme (l.a) as the thrust decreases and thechamber pressure increases may be as high as 7.9%for a thrust of 5kN and chamber pressure of 15MPa.
This behavior is connected with the increase inthe relative mass flow of the drive gas for the turbop-ump. It was assumed that the specific impulse is thesame for both schemes, although preliminary calcu-lations of the gas flow in the nozzle, bleeding theboundary layer in the supersonic section for turbinegas drive and reinjection at a lower section, showeda small increase in the specific impulse. In the cal-culations above it was assumed that the bleeding ofthe turbine driving gas at the nozzle section withpressure of psei = IMPa. Using the gasdynamic cor-relation
_F = (23)
it is easy to calculate the diameter of the section ofthe nozzle at the bleeding point and the thicknessof the boundary layer at drive of the turbine. Thethickness selected ring layer S as a function of theengine thrust and the chamber pressure is shown inFigure 3. The mean velocity and gas density is as-sumed as w = 800m/s, p — 2.62kg/m3 taking intoaccount the temperature distribution in the bound-ary layer. The mass flow of the turbine driver gasmay be decreased, by moving the nozzle bleedingsection downstream. In this case the partiality ofthe turbine increases, the turbine efficiency factorincreases also and the mass flow at turbine drive de-creases. However, in order to keep constant the adi-abatic work Lad it is necessary to move downstreamthe reinjection section, in order to keep constant theratio PseiPbi = 15 =const.
The influence of the location of the blowing sec-tion on the losses of specific impulse in the nozzle isthe subject of future investigation.
Conclusions
1. The scheme of a LRE with bleeding of the tur-bine driver gas from the boundary layer in thesupersonic part of the nozzle and reinjectionof the gas at a downstream location (a noz-zle tap-off engine scheme) is introduced. Thisscheme combines the positive characteristics ofthe LRE without afterburning of generator gaswith those of the engine with afterburning ofgenerator gas.
2. The increase in the specific impulse for thescheme (l.c) as compared to the scheme (l.a),without afterburning of generator gas, variesfrom 1.8 to 8% and increases with an increasehi chamber pressure and decrease of the enginethrust.
3. The relative mass flow at drive turbine is 3.2to 5.5% (for chamber pressure 5 to 15MPa andthrust 25 to 5kN accordingly) and 8.3 to 14%(for chamber pressure 15MPa and the samethrust) and may be decreased further by meansof optimizing the locations of the bleeding andreinjection sections.
Acknowledgment
This work was supported by Conselho Na-cional de Desenvolvimento Cientffico e Tecnologico(CNPq).
References[1] Kozlov, A. A., The choosing scheme, oxidizer
and fuel components and fundamental param-eters of LRE devices on early stages of designwork by means of interactive system "FORDU";Third China-Russian Scientific Conference onAeroengines, Beijing, 1993.
[2] Kurpatenkov, V. D. et alii. Osnovy Teorii irastcheta JRD (Fundamentals of Theoryand Calculation of LRE), Vyschaia Schkola,Moscow, 1953, Vol. I and Vol II, (in Russian).
[3] Kozlov, A. A., Novikov, V. N. and Soloviov,E. V. Sistemy Pitania i Upravlenia JRDUstanovok (Feed and Control Systems ofLRE devices), Mashinostroenie, Moscow, 1988,p. 352, (in Russian).