HIGH ALTITUDE OPERATIONS WITH PISTON ENGINES POWER PLANT DESIGN OPTIMIZATION PART V: NOZZLE DESIGN AND RAMJET GENERAL CONSIDERATIONS

Transcription

1 HIGH ALTITUDE OPERATIONS WITH PISTON ENGINES POWER PLANT DESIGN OPTIMIZATION PART V: NOZZLE DESIGN AND RAMJET GENERAL CONSIDERATIONS Luca Piancastelli 1 and Stefano Cassani 2 1 Deartment of Industrial Engineering, Alma Mater Studiorum University of Bologna, Viale Risorgimento, Bologna (BO), Italy 2 MultiProjecta, Via Casola Canina, Imola (BO), Italy luca.iancastelli@unibo.it ABSTRACT In stratosheric flights with iston owered aircrafts, the cooling system taes art to the vehicle design otimization rocess. An integrated design of the cooling duct(s) is strictly necessary. At high altitudes, the cooling air is taen from high-ressure areas into a subsonic ramjet: the Meredith cooling duct. A diffuser reduces the airseed and increases the ressure of the cooling air. Then a grou of high erformance finned radiators rejects the heat from coolant, air charge and lubricant. A variable geometry nozzle transforms the added enthaly into seed and thrust. The nozzle is ositioned in a low ressure, high turbulence area. The nozzle design and the duct thrust are discussed in this aer. At first the results from Parts I to IV are summarized and discussed. The resulting data are also exosed and summarized. The ressure recovery and heat rejection are evaluated in function of aircraft seed for a 1-m 2 vertical-radiator circular duct. The nozzle is then otimized and the total thrust is evaluated. Keywords: otimization, HALE, UAV, nozzle, ramjet, meredith effect. 1. INTRODUCTION Proulsion system ofuavs(unmanned Aerial Vehicle) designed to fly subsonically >20,000m (65,000ft) for several hours requires very accurate design of the cooling system. In this flight regime, TurboChargers (TC), intercoolers and after coolers are needed to suly most of the intae ressurization required to comress the small air density into the engine. Volume flow requirements increase with altitude, which translates to larger TCsize. Automotive derived sar ignition and diesel iston engines are available u to 700HP. In these engines the intae is ressurized with three cascaded stages of automotive derived TCs u to about 200HP. Over this ower level the engine requires an axial suercharging system that can be derived from commercial turbines as it has been done on the STRATO 2C. It is also ossible to have an emergency additional ower system as described in the VD007 hybrid arrangement [1] [2] [3] [] [5] [6]. Due to lower air mass requirement and flat rating iston engine are a more efficient choice for stratosheric aircrafts than turboshafts and turbojets. However, modern iston engine roulsion systems lac of the huge technical bacground of turbines. In fact, after WWII the research in this field have been discontinued, while turbines have exerienced enormous develoment that led to extremely reliable and efficient designs. The roof is given by the fact that the record flight altitude of Mario Pezzi with his Caroni of 1939 has been overcome only in 1995 by Grob-Strato2C. Curiously, as it haens in many cases, the Formula 1, Reno Racing and WWII exerience in cooling systems have been mostly neglected by the designers of new remarable iston-owered research stratosheric aircrafts. Also the develoment costs were underestimated. In fact, the nowledge ga left in these 50 years of turbines only cannot be filled without roer trials and errors. The solar-owered alternative cannot fill this ga due to the low wing loading needed by this alication. The stratosheric aircraft needs to ass the trooshere atmoshere with its environmental roblems to reach stratosheric altitudes. Low wing loading reduces aircraft control and overloads structures. As it was introduced in revious aer the most critical art of the cooling duct is the identification of a high ressure area for the intae and a low ressure one for the nozzle. Straight, all external ducts are theoretically ossible, but the external drag revents high thrusts for this solution. In fact, radiator owered, subsonic ramjets suffer from the lac of high temeratures. As it was shown in art IV of this aer, the use of the highest oil temeratures of 150 DEG C oututs air at about 135 DEG C that is the maximum achievable ractical temerature for heated air in the Meredith duct. Theoretically, it is ossible to recover energy also from turbo charging turbine housings, but the high temeratures would require highly ressurized systems or secialized coolants with all the annexed weight, reliability and safety roblems. Two alternative solutions have been devised for Meredith s ram ducts. The first one is to embed long slim radiators into the wings as in the De Havilland Mosquito and the Messerschmitt Bf 109 F-K. This solution maes it ossible to reduce divergent length, which goes with radiator height. However, it faces high ressure loss due to the unfavourable internal surface to erimeter radio (hydraulic diameter). Another solution is to embed the duct inside the fuselage or the nacelle, leaving only the intae exosed to the incoming air. This most favourable solution taes the air from the last 1/3 or the wing and ositions the exhaust on the uer art of the fuselage or the nacelle. This solution is a hybrid from the P51 Mustang and the Formula 1 cooling duct and obtains the best comromise for otimum erformance. An inclined, thin radiator 222

2 acage comosed by intercooler(s), after cooler, liquid coolant and oil coolant offers the maximum temerature rise for the cooling air slowed down by the diffuser. The diffuser will have an extremely long streamline shae. The area ratio from the inlet and the exit of the diffuser is A i/a B=0.35 for best efficiency. Formula 1 radiator with 25 FPI (Fin er Inch) can be used to obtain otimum heat rejection with minimum ressure loss (drag). As it has been demonstrated in Part IV, Reynolds and Prandtl number after the diffuser are roughly the same for Formula 1 in stratosheric aircrafts cruising in the range of Mach. A variable geometry nozzle comletes the subsonic ramjet cooling duct. Unsurrisingly, many high altitude iston engine alications failed to oor design of the cooling installation. Powerlants and their cooling system have always been a roblem u to this day. NACA eole used extensively their wind tunnels and their nowledge to solve cooling roblems even during the aogee of iston engines (WWII). Many aers come from that eriod to revive the nowledge of cooling that is eriodically lost by the designers. This aer describes the solutions and the udates of this last 50 years of extensive wor and otimization of automotive racing cooling systems. These udates can be directly alied to high altitude flying with a few corrections. The mission Since the fuel consumtion follows a cubic low with seed, long endurance requires flying at reasonably low seeds. The dynamic ressure available limits the minimum seed to about 0.M. A more liely seed will be between 0.6 and 0.7 M to avoid excessive wingsan and too low wing loading. In fact, the aircraft should climb through the trooshere with its climatic roblems to reach the calmer stratoshere. Therefore, the aircraft will be more lie a saillane than a owered general-aviation aircraft and will face handling roblems at tae-off and lower altitudes. These roblems will be amlified by the installation of radiators and cooling ducts. Subsonic ramjet heritage Two short documents are the milestones of ramjetor athodyd (Aero THermODYnamic Duct) roulsion. The first was ublished as early as 1913 in the magazine "L'Aerohile" and dealt with a suggestion made by Rene Lorin for a hyothetical flying vehicle described as a "roulseur ar reaction directe" (direct reaction roulsion system). Lorin stated that a heated aero duct consisting of an inlet diffuser able to ram incoming air and an attached combustion chamber with suitably shaed exansion nozzle, would roduce thrust. In 1915, Albert Fonó devised a cannon-launched rojectile with a ramjet roulsion unit for a long range heavy shell. Fonó atented his invention in May 1928 with German Patent No. 55,906. In Soviet Union, Britain, Germany and United States exeriments and theoretical research on suersonic and subsonic ramjets have been carried out during WWII. At the end of the war the subsonic ramjet exerience ended due to the much more efficient suersonic solution. However, several exeriments of subsonic ramjets from 0.2 u to 0.8 Mach were carried out by the above Nations. These exeriments and ost-war refinements are the basis of this aer. This research was guided by the following considerations. The ram-jet engine has simle construction and eculiar thrust characteristic, which, similar to drag, rises nearly roortionally to the dynamic ressure and to the main cross-sectional area. Above a given minimum flight seed it is ossible to obtain thrust still in the subsonic range. As oosed to rocets and turbo-jets, the ram-jet engine construction is inexensive and it is suitable for mass roduction due to the uncomlicated shae of the engine and the absence of moving arts. The engine is largely insensitive to the ind of fuel or heat source used. German exerience is well described in aer [7]. The ramjet was installed on a Dornier Do 17 (Figure-1). Figure-1. Do 17Z with D3=0.5m ramjet-0.3 M-2,000m (192) [7]. Figure shows a well dimensioned ramjet with a very high ratio diffuser. The ramjet had a 10-DEG conical diffuser with Ai/AB=0.158 and a nozzle with AB/A= The maximum combustion chamber temerature was 600 DEG C. Several different ramjets were develoed with D3 increased u to 2 m (Figure-2). Figure-2. This Do 17Z with a 1.5m ramjet encountered serious stability and handling roblems (193) [7]. For measuring uroses, a duct with D3=1m was selected for final tests. The length of the cylindrical combustion chamber was again chosen of 000 mm. It had a very high ratio 10-DEG conical diffuser with A i/a B=0.158 and a slightly curved intae. The nozzle had A B/A =0.565 and a cone-angle of DEG. Tests were carried out at M and altitudes from 1000 to 7000m. 223

3 Figure-5. Velocity attern at the radiator bac face in function of arameter a (see Figure-6). Figure-3. Pressure on ad3=1m ramjet at 000m, V 0=0.3M (continuous line) vs. theoretical ones (dotted line) (193) [7]. Curiously, the calculations were made with unitary diffuser efficiency and 0.95 nozzle efficiency. This aroach is common to RAE theoretical estimations. Figure 3 shows that the estimation for the nozzle is fully accetable, while the theoretical diffuser erformance was overestimated. In radiator heated ducts the diffuser estimation is extremely imortant due to the limited air enthaly increase. The nozzle erformance is almost isentroic for the Meredith s ramjet due to reduced velocity and friction. Figure-6. Volume on the radiator bac (arameter a). Figure-5 shows the brown rofile assures an almost constant distribution of the air velocity in the radiator bac. The most imortant arameter in dimensioning the nozzle is the distance a between the nozzle throat and the radiator (Figure-6). An increase the of ratio between a and the radiator height H B reduces the radiator drag due to the more uniform velocity on the radiator bac (Figure-7). Figure-. Ramjet owered P51 Mustang (195). The oor results on a few aircrafts with ramjets (see Figure-) ended the exeriments on subsonic ramjets. Nozzle In the dimensioning of the nozzle, it is ossible to neglect both the distributed and the concentrated losses. So, the only roblem is to evaluate the influence that the resence of the nozzle has on the ressure dro of the radiator. Figure-7. Radiator drag coefficient as a function of a/h B and diffuser area ratio A i/a B (<1). Finally, the duct can be equied with a variable nozzle. Figures 8 and 9 shows the fla system mounted on the Bf

4 osition 1 to 2. This effect taes lace with extremely high efficiency only if the intae area is smaller than the nozzle throat area. Figure-8. Messerschmitt Bf 109 Friedrich? with the nozzle fully oened. Nozzle design The first value to be evaluated is the arameter a of figure 2. In the examle of arts I-IV of this aer the cooling duct is circular with a diameter (vertical radiator section) of D B=1.128 m (A B=1m 2 ). The diffuser ratio is A i/a B=0.35.From Figure-3 a/hb is then 0.5. a m (1) An additional volume of 0.56 m 3 (the radiator section of the duct is of unitary area) should then be added after the radiator before the subsonic convergent nozzle. The air is heated by the radiator (thicness=27mm). For the airseed of Mach 0. and the coolant temerature of 150 DEG C has the following arameters: T 3=98 K, 3=5725Pa, ro 3=0.0g/m3, V 3=88m/s. The seed of sound at station 3 is (2): M R T 7 (2) 3 3 Figure-9. Bf109 with the nozzle fully oened (ossibly a late G or K due to the absence of the boundary layer byass duct). Cooling duct regulation through variable intae fla roved to be not efficient (Figure-6). and V3 Ma (3) M 3 3 The total arameters at station 3 are then () (5) and (6): 1 2 T T3 1 Ma K () Ma Pa g 3 1 Ma (6) 3 2 m (5) Figure-10. Variable air intae on early P51 Mustang rototyes roved to be ineffective. Figure-11. Nozzle influence on duct air flow. Figure-10 shows that the ability of the subsonic duct to suc air increases by oening the nozzle from The Meredith duct thrust is maximized when the nozzle throat area is the minimum ossible. For nozzle efficiency the condition A >A i (nozzle throat area larger than intae area) should hold. It is then ossible to assume that A c=1.05a i, being a mere 5% increase the minimum technically significant. If a tentative value of Ma is assumed, it is ossible to iterate to the correct value that satisfy the chosen A c. Equation (7) exresses the energy conservation rincile Ma (7) 2 It is then ossible to evaluate V from the equation of isentroic transformation (8). 225

5 V (8) The ideal gas rincile oututs T (9) and ρ (10). V R T (9) Ma (10) R T Equation (11) exresses the mass conservation rincile. V g 3 3 A3 Q 3. 5 (12) s It is then ossible to obtain A (13). Q A (13) V Ma is modified until the desired value for A is obtained. In this case the otimum value for the exit velocity Ma, that matches A c=1.05a i, is0.58 Mach. It is then ossible to evaluate the thrust (1). However, at this seed is lower than 0 (outside air ressure). Therefore, if 3= we have Ma =0.32, V =12 and A =0.63. T V A ( V V0 ) 70 N (1) The thrust coefficient C T is exressed by equation (15). 2T C T 0.22 (15) 2 V A The thrust is then minimal and only an extremely accurate integration of the duct inside the fuselage will avoid having drag instead of thrust from the cooling duct. This is due to the extremely reduced air mass flow (3.5 g/s), and the low vehicle seed (0. Mach). In common automotive alication the maximum oil temerature is et at 110 DEG C. In this case the thrust is even lower. The seed and the maximum ressure are increased to reach higher values of thrust. At Mach 0.7 it is ossible to reach 00 N. If the radiator thicness and the heat rejection is doubled the thrust at an airseed of 0. Mach is increased u to 121 N (72% increase). The advantage is et also at Mach 0.7 with a trust of 500N (25% increase). The radiator thicness it then a critical arameter in the duct. An otimum radiator thicness can be found for every diffuser-ressure-recovery and airseed. CONCLUSIONS In stratosheric flights with iston owered aircrafts, it is easy to face cooling roblems. At low seeds, below Mach 0.5, it is difficult to obtain a significant amount of thrust even using the Meredith ramjet cooling duct. On the contrary it is extremely easy to face overheating and additional drag from the cooling duct. An otimized design of the cooling duct(s) is then strictly necessary to avoid overheating and to obtain thrust. The cooling air is taen from high-ressure areas into subsonic ramjet: the Meredith cooling duct. A diffuser reduces the airseed and increases ressure of the cooling air. Then a grou of high erformance finned radiators rejects the heat from coolant, air charge and oil. A variable geometry nozzle transforms the added enthaly into seed and thrust. The nozzle is ositioned in a low ressure, high turbulence area. The nozzle design and the duct erformance have been discussed in this aer. The ressure recovery and heat rejection are shown in function of aircraft seed and coolant temerature for a vertical 1- m 2 -radiator circular duct. The nozzle has been then otimized and the total thrust has been evaluated. Afterwards the radiator duct erformance as a high altitude ramjet was evaluated. It is extremely imortant to otimize the radiator thicness and to obtain the maximum coolant temerature ossible. 226

6 Symbols Symbol Descrition Unit Value i, 1 0 Diffuser Inlet ressure (station 1 = station 0) Pa - ρ i, ρ 1 ρ 0 Diffuser inlet air density (station 1) g/m 3 - T i, T 1 T 0 Air temerature inlet (station 1) K - V i, V 1 V 0 Velocity (station 1) m/s - D 1, D i D 0 Diffuser inlet diameter (station 1) m - A 1, A i A i Diffuser inlet area (station 1) m 2 - B Diffuser outlet ressure (station B) Pa - ρ B Diffuser outlet air density (station B) g/m 3 - T B Temerature (station B) K - V B Velocity (station B) m/s - A 3, A B Diffuser outlet area=radiator area m 2 1 H B, D B D 3 Radiator height=diffuser outlet area m - 3 Radiator outlet ressure (station 3) Pa - V 3 Air velocity after radiator (station 3) m/s - ρ 3 Air density after radiator (station 3) g/m 3 - T 3 Air temerature after radiator (station 3) K - Total ressure (station 3) Pa - T Total temerature (station 3) K - ρ Total air density (station 3) g/m 3 - M 3 Sound velocity (station 3) m/s Ma 3 Velocity (station 3) M D Nozzle outlet diameter (station ) m - A Nozzle outlet area (station ) m 2 - ρ Nozzle outlet air density (station ) g/m 3 - REFERENCES [1] L. Piancastelli, L. Frizziero, E. Pezzuti Aircraft diesel engines controlled by fuzzy logic. Asian Research Publishing Networ (ARPN). Journal of Engineering and Alied Sciences. ISSN , 9(1): 30-3, EBSCO Publishing, 10 Estes Street, P.O. Box 682, Iswich, MA 01938, USA. [2] L. Piancastelli, L. Frizziero, E. Morganti, A. Canaaro Fuzzy control system for aircraft diesel engines. International Journal of Heat and Technology. ISSN (1): [3] L. Piancastelli, L. Frizziero and I. Rocchi Feasible otimum design of a turbo comound Diesel Brayton cycle for diesel-turbo-fan aircraft roulsion. International Journal of Heat and Technology. 30(2): [] L. Piancastelli, L. Frizziero, N.E. Daidzic, I. Rocchi Analysis of automotive diesel conversions with KERS for future aerosace alications. International Journal of Heat and Technology. ISSN (1). [5] L. Piancastelli, L. Frizziero, E. Pezzuti Kers alications to aerosace diesel roulsion. Asian Research Publishing Networ (ARPN). Journal of Engineering and Alied Sciences. ISSN: (5): EBSCO Publishing, 10 Estes Street, P.O. Box 682, Iswich, MA 01938, USA. [6] L. Piancastelli, L. Frizziero, E. Morganti, E. Pezzuti Method for evaluating the durability of aircraft iston engines. Walaila Journal of Science and Technology. ISSN (): [7] L.Saengcr-Bredt Air-Borne Towing Exeriments with Ram-Jet Ducts, History of German Guided Missiles Develoment, AGARD, First Guided Missiles Seminar, Munich, Germany. 227