1. AME 436 Energy and Propulsion Lecture 14 Propulsion 4: Nonideal performance of airbreathing propulsion devices

2. 2 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Outline  Component performance - definitions  AirCycles4Propulsion spreadsheet  Non-ideal performance  Diffuser  Compressor  Burner  Turbine  Nozzle  Fan  Heat loss

3. 3 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Component performance  Diffusers (1  2) & nozzles (7  9) subject to stagnation pressure losses due to viscosity, shocks, flow separation  More severe for diffusers since dP/dx > 0 - unfavorable pressure gradient, aggravates tendency for flow separation  If adiabatic, T t = constant, thus  d  T 2t /T 1t = 1 &  n  T 9t /T 7t = 1  If also reversible,  d  P 2t /P 1t = 1 and  n  P 9t /P 7t = 1  If reversible,  d =  d (  -1)/  = 1,  n =  n (  -1)/  = 1, but if irreversible  d (  -1)/  < 1 or  n (  -1)/  < 1 even though  d = 1,  n = 1  Define diffuser efficiency  d =  d (  -1)/  = 1 if ideal  Define nozzle efficiency  n =  n (  -1)/  = 1 if ideal  For nozzle - also need to consider what happens if P exit ≠ P ambient, i.e. P 9 ≠ P 1

4. 4 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Compressors, fans & turbines  Compressors (2  3) & fans (2'  3') subject to losses due to  Flow separation around compressor blades (again, unfavorable dP/dx)  Gas leakage between blade tips and compressor walls  How to define compression efficiency  comp ?  If irreversible dS >  Q/T; if adiabatic (  Q = 0) dS > 0  For same P ratio, more work input (more  T) during compression, less work output (less  T) during expansion through turbine  Define  comp similar to piston/cylinder (Lecture 6):

5. 5 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles  Fan efficiency  fan defined similarly but not as problematic since smaller pressure ratios  Turbines (4  5 and 5  6) less problematic since dP/dx < 0, but need to bleed compressor air through turbine blades for cooling  Define turbine efficiency  turb :  Fan turbine: replace state 4 with 5 and replace state 5 with 6 Compressors, fans & turbines

6. 6 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Component performance - combustors  Combustor (4  5) and afterburner (6  7) - stagnation pressure losses due to turbulence (required for mixing of fuel and air) and heat addition at M > 0  Define combustor (burner) or afterburner efficiency in terms of stagnation pressure ratio  b or  ab (= 1 if ideal)  For heat addition at finite M, stagnation pressure will always decrease; recall (lecture 12)

7. 7 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles AirCycles4Propulsion.xls  See also Lecture 9 (AirCycles4Recips.xls)  Thermodynamic model is exact, but heat loss, stagnation pressure losses etc. models are qualitative  Constant  is not realistic (changes from about 1.4 to 1.25 during cycle) but only affects results quantitatively (not qualitatively)  Wall heat transfer model   T ~ h(T w -T gas,t ), where heat transfer coefficient h and component temperature T w are specified - physically reasonable  Unlike AirCycles4Recips.xls, wall temperature T w is specified separately for each component since each experiences a constant T, not T averaged over the cycle  Increments each cell not each time step  Doesn't include effects of varying area, varying turbulence, varying time scale through Mach number, etc. on h

8. 8 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles AirCycles4Propulsion.xls  Work or heat in or out going from step i to step i+1 computed according to dQ = C P (T i+1,t - T i,t ) and dW = - C P (T i+1,t - T i,t ) (heat positive if into system, work positive if out of system)  Diffuser  Mach number decrements from flight Mach number (M 1 ) to M 2 = 0 after diffuser in 25 equal steps  Stagnation pressure decrements from its value at M 1 (P 1t ) to  d P 1t =  d  /(  -1) P 1t in 25 equal steps  Static P and T calculated from M and P t, T t  Sound speed (c) calculated from T, then u calculated from c and M  No heat input or work output in diffuser but may have wall heat transfer (page 7)  Fan-flow diffuser exactly same as main flow diffuser except mass flow rate  times higher (  d,main flow =  d,fan flow assumed)

9. 9 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles AirCycles4Propulsion.xls  Compressor  Mach number = 0, thus P = P t, T = T t  Compression in two steps (1) Wall heat transfer at constant P according to formula on page 7 (step i,a to step i,b) (2) Compression according to the Pv  relations with efficiency  comp (step i,b to step i+1,a)  Compression ends at pressure =  c P 2t where  c is the specified pressure ratio  Pressure increments in 25 equal steps  Work per step = - C P (T i+1,a,t - T i,a,t )  Fan compressor follows exactly same as rules except that mass flow rate is  times higher and  fan may be different from  comp

10. 10 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles  Combustor and afterburner  Enthalpy input due to combustion added in 25 equal steps (1/25 of total stagnation temperature change per step)  Heat addition ends at T 4t = T 1  (i.e the turbine inlet temperature limit) or T 7t = T 1 ,ab (afterburner temperature limit)  Wall heat transfer at constant P according to formula on page 7  Stagnation pressure drops in 25 equal decrements; P after combustion P 4t =  b P 3t or for afterburner P 7t =  ab P 6t  Mach number = 0 throughout AirCycles4Propulsion.xls

11. 11 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles AirCycles4Propulsion.xls  Turbine  Expansion to pay for compressor work done in two steps: (1) Wall heat transfer at constant P according to formula on page 7 (step i,a to step i,b) (2) Expansion according to usual Pv  relations with efficiency  turb (step i,b to step i+1,a)  Each of 25 expansion steps provides (after including losses due to irreversibility) 1/25 of the work required to drive turbine (calculated in compressor analysis)  Work per step = - C P (T i+1,a,t - T i,a,t ),  Mach number = 0 throughout  Expansion through fan turbine to pay for fan work - same as main turbine, except equate fan work to turbine work

12. 12 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles AirCycles4Propulsion.xls  Nozzle  No heat input or work output  Static (not stagnation) pressure decrements from value after afterburner to specified exhaust pressure P 9 in 25 equal steps  Stagnation pressure decrements from its value after afterburner (P 7t ) to  n P 7t =  n  /(  -1) P 7t in 25 equal steps  T t changes due to wall heat transfer (page 7)  M calculated from P (static pressure) and P t (stagnation pressure)  T (static temperature) calculated from M and T t (stagnation temperature)  Sound speed (c) calculated from T  u calculated from c and M  Same rules apply to fan nozzle except »Go directly from state 3' (after fan compressor) through nozzle to state 9’ (no combustor or turbine) »Mass flow is  times higher »Assume P exit = P ambient always, i.e. P 9 ' = P 1

13. 13 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - diffuser - effect of  d  Obviously thrust &  overall decrease, TSFC increases as  d decreases, but can have pretty bad diffuser (  d ≈ 0.55) before thrust  0 Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

14. 14 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - diffuser - effect of  d  T 2 same, regardless of  d, since air decelerated to M = 0  s 2 increases & P 2 decreases with decreasing  d  T 3, T 4, T 5 don’t change since  c,  & compressor work unchanged  If P 2 is too low (can be < P 1 ), after compressor/combustor/turbine, P 5 is too low, can only expand back to u 9 = u 1, so no net thrust! Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

15. 15 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of  comp  Obviously thrust &  overall decrease, TSFC increases as  comp decreases, but (again) can have pretty bad compressor (  c ≈ 0.58) before thrust disappears Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

16. 16 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of  comp  State 2 always same since air decelerated isentropically to M = 0  T 3 & s 3 increase as  comp decreases (more work required to drive compressor for same  comp ), T 4 same (same  limit)  T 5 lower as  comp decreases (more turbine work to drive compressor)  If  comp too low after paying for compressor work in turbine, P 5 & T 5 are too low, can only expand back to u 9 = u 1, so no net thrust! Basic turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

17. 17 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - burner - effect of  burner  Obviously thrust &  overall decrease, TSFC increases as  burner decreases, but can have really lousy  burner ≈ 0.13 before thrust disappears (though (  burner ) (  -1)/  ≈ 0.56!) Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

18. 18 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - burner - effect of  burner  States 2 & 3 are the same, independent of what burner does  T 4 unchanged (same  limit) but as  burner decreases, P 4 decreases, s 4 increases  T 5 same for all (same turbine work required to drive compressor)  If  burner too low after paying for compressor work in turbine, P 5 & T 5 are too low, can only expand back to u 9 = u 1, so no net thrust! Basic turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

19. 19 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - turbine - effect of  turb  Obviously thrust &  overall decrease, TSFC increases as  turb decreases, but again can have low  turb ≈ 0.57 before thrust disappears Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

20. 20 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - turbine - effect of  turb  States 1 - 4 are unchanged, independent of  turb  T 5 also independent of  turb (same work required to drive compressor) but s 5 increases as  turbine decreases  If  turb too low after paying for compressor work in turbine, P 5 is too low, can only expand back to u 9 = u 1, so no net thrust! Basic turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

21. 21 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - nozzle - effect of  n  Obviously thrust &  overall decrease, TSFC increases as  turb decreases, but again can have low  turb ≈ 0.56 before thrust  0  Interesting that  d,  comp,  turb,  nozzle at Thrust = 0 all ≈ same! Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

22. 22 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - nozzle - effect of  n  States 1 - 5 unchanged, independent of  turb  If  nozzle too low, can only expand back to u 9 = u 1, so no net thrust!  Note that even for zero-thrust cycle, there is still some decrease in T after turbine to create u 9 ; if u 9 = 0 then thrust is negative!) Basic turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

23. 23 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - nozzle - effect of P exit ≠ P amb  Thrust &  overall decrease, TSFC increases if P 9 ≠ P 1, but performance very insensitive to P 9 /P 1 (note horizontal scale is logarithmic) Basic turbojet No fan No afterburner M 1 = 0.8  c = 30  = 5

24. 24 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - nozzle - effect of P exit ≠ P amb  States 1 - 5 are unchanged, independent of P 9 /P 1  If P 9 too low, negative (P 9 - P 1 )A 9 term in the thrust equation balances the positive a [(1+FAR)u 9 - u 1 ], so no net thrust, even though u 9 increases as P 9 decreases  If P 9 too high, u 9 = 0; analysis fails at higher P 9 although even at P 9 = P 5, there is still net thrust (although the [(1+FAR)u 9 - u 1 ] term is zero, the (P 9 - P 1 )A 9 term generates thrust) Basic turbojet No fan No afterburner M 1 = 0.8  c = 10.87  = 5

25. 25 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbofan cycle - effect of  fan  Obviously thrust &  overall decrease, TSFC increases as  fan decreases, but can have low  fan ≈ 0.33 before analysis fails  Actually Thrust ≠ 0 at this point but fan work is so large that turbine exit pressure = ambient pressure; beyond this point analysis fails since turbine work insufficient to supply greedy fan Turbofan  = 2 No afterburner M 1 = 0.8  c = 30  c ' = 2  = 5

26. 26 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbofan cycle - effect of  fan  States 1 - 5 are same, independent of  turb  If  fan too low, fan work is so large that turbine exit pressure P 6 = ambient pressure P 9 ; beyond this point analysis fails since turbine work insufficient to supply greedy fan Turbofan  = 2 No afterburner M 1 = 0.8  c = 10.97  c ' = 2  = 5

27. 27 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of heat loss Turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

28. 28 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of heat loss  For small heat loss, T 2 & T 3 decrease (since I chose T w = 300K for diffuser & compressor), and since P 3 /P 2 is set by choice of  c, s 3 decreases  Heat addition is greater - need to add more heat to get to  limit limit (starting from lower T, plus wall heat loss during heat addition (combustion))  Since heat addition at constant pressure and same maximum temperature, T 4 is the same with heat loss  Heat loss during turbine & nozzle expansion also result in decreased T; since P 9 is fixed, lower T 9 causes lower s 9  Note that s decreases in nozzle, even though T 9 < T w – why?  Heat transfer formula uses stagnation temperature, not static T  Since heat loss occurs at walls and gas must decelerate to M = 0 at walls, T t is a better temperature to use in heat transfer estimate  Actually, whether T t or T provides a better estimate of heat transfer depends on wall “accommodation coefficient" – beyond scope of this course

29. 29 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of heat loss  For ridiculously high heat transfer coefficient h  Gas temperature follows T w during compression (constant-T compression, like Stirling cycle)  Heat added at constant P  During turbine & nozzle expansion, gas follows ≈ constant-P path until T t = 900K (= T w for turbine & nozzle), then remainder of turbine work extracted isothermally (again like Stirling cycle)  Gas expands adiabatically through nozzle (no change in T t during nozzle expansion - adiabatic since T t = wall T)  Effect of heat loss on performance (plot next page)  Barely affects specific thrust since the pressures (thus velocities, thus propulsive efficiencies) don't change much  Much more thermal energy must be added, thus TSFC increases, thermal &overall efficiency decrease

30. 30 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbojet cycle - effect of heat loss Turbojet No fan No afterburner M 1 = 0.8  c = 10.97  = 5

31. 31 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Turbofan, no afterburner M 1 = 0.8,  c = 10.97,  = 5  c ' = 2,  = 2  diffuser = 1 or 0.9  comp = 1 or 0.9  burner = 1 or 0.9  turb = 1 or 0.9  nozzle = 1 or 0.9  fan = 1 or 0.9 h = 0 or 0.02 (all wall temperatures = average gas temp for that component) Drag coefficient = 0 or 0.2 Combination of all non-ideal effects Spec. Thrust TSFCThermal Eff. Prop. Eff. Overall Eff. Specific impulse (sec) Ideal0.93142.51930.55040.57700.31765247.4 Non-ideal0.38006.28030.19760.64450.12742104.9

32. 32 How accurate is aircycles4recips.xls?  Real engine: GE90-85B turbofan, no afterburner  Static sea level (M 1 = 0, T 1 = 300K, P 1 = 1 atm) or cruise (M 1 = 0.83, T 1 = 219K, P 1 = 0.235 atm)  Reported properties:  c = 39.3;  c ' = 1.65;  = 8.4; Inlet area = 7.66 m 2 ; C D = 0 (drag not included in reported performance)  Guessed properties  Turbine inlet temp. = 1700K   = 5.67 (static) or 7.76 (cruise)   diffuser = 1 (static) or 0.97 (cruise);  comp = 0.9;  burner = 0.98;  turb = 0.9;  nozzle = 0.98;  fan = 0.9  h = 0.01 (all wall temperatures = average gas T for that component)  Results (very sensitive to  diffuser because of huge fan air flow!) AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Spec. Thrust (static) TSFC (static) Spec. Thrust (cruise) TSFC (cruise) Spreadshee t 0.9881.095 0.7742.355 Actual0.7881.183 ??? (air flow not reported) 2.213

33. 33 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Example - numerical For a nonideal turbofan with bypass ratio (  ) = 8,  = 1.35, compressor pressure ratio (  c ) = 30, fan pressure ratio (  c ') = 1.8, flight Mach number (M 1 ) = 0.8, turbine inlet temperature = 1800K, ambient pressure (P 1 ) = 0.25 atm and ambient temperature (T 1 ) = 225 K, with P 9 = P 1 and FAR << 1, and component efficiencies as follows: diffuser (  d ) = 0.97, compressor (  c ) = 0.90, combustor (  b ) = 0.98, turbine (  t ) = 0.90, nozzle (  d ) = 0.98, fan (  f ) = 0.90, determine the following (this is the same as the example in lecture 13 except non-ideal components are presumed here): (a)T, P and M after the diffuser (station 2) (b)T, P and M after the compressor (station 3) (c) T, P and M after the combustor (station 4) M 4 = 0; P 4t = P 4 = (  b )P 3t = 0.98(10.04 atm) = 9.84 atm ; T 4t = T 4 = 1800K

34. 34 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Example - numerical (continued) (d) T, P and M after the compressor turbine (station 5). Balance turbine & compressor work: (e) T, P and M after the fan turbine (station 6) For the fan stream, T 2 ' = T 2 = 250.2K and P 2 ' = P 2 = 0.335 atm (same as main stream) Then we need to compute the fan stream work and equate it to the turbine work:

35. 35 Example - numerical (continued) (f) T, P and M after the nozzle (station 9, and 9' for the fan stream) (note that for the fan stream, nothing happens between stations 3 and 6) (g) Specific thrust (ST) (recall FAR << 1 and P 9 = P 1 by assumption) AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles

36. 36 Example - numerical (continued) (h) TSFC and overall efficiency Recalling that only 1/(1+  ) of the total air flow is burned, note that compared to the ideal cycle, TSFC suffers slightly more (2.96 vs. 2.32 = 28% increase) than ST (0.552 vs. 0.728 = 24% decrease) (i) Propulsive efficiency AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles

37. 37 Example - numerical (continued) (j) Thermal efficiency Note that  th  p = (0.430)(0.630) = 0.271 =  o as advertised. Note that compared to the ideal cycle where  th  0.629,  p = 0.549 and  o = 0.345, for the non-ideal case the thermal and overall efficiencies decreased (as expected) but propulsive efficiency actually increased slightly due to lower exit velocities. (k) Specific impulse AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles

38. 38 In an ideal  -limited afterburning turbojet, how would the T-s diagrams be affected if (a) The turbine is irreversible, but all other components are still ideal The cycles are the same up to the end of the first heat addition process. The turbine work (thus  T) does not change, but the irreversibility during expansion increases the entropy, so that afterburning heat addition occurs along a lower constant-pressure curve up to the same temperature limit. Although not shown, since the afterburner  T is the same for the modified cycle, the heat addition, thus area under the T-s curve, is the same for the modified afterburner. (b) A new compressor is used with a higher pressure ratio, but this compressor is irreversible (all other components are still ideal) The compression process ends at a higher pressure and (since it is irreversible) a higher entropy than the base cycle. Heat addition then occurs along this constant-pressure curve up to the turbine temperature limit. More work, thus more  T, is required from the turbine, so the afterburning heat addition occurs along a lower constant pressure curve. Example - graphical AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Same  T (a) (b)

39. 39 In an ideal turbofan, how would the T-s diagrams be affected if (a ) The fan is removed and the redesigned turbine that drives the compressor is irreversible The cycle is the same up to the end of heat addition. The  T and thus the T after the compressor turbine is the same but the entropy is higher. There is no fan turbine. Expansion through the nozzle continues to ambient pressure. (b) There are substantial pressure losses in the combustor The cycle is the same up to the end of the compression process. Heat addition occurs at decreasing pressure but up to the same temperature limit as the base cycle. The compressor and fan work requirements are unchanged, so the temperatures at states 5 and 6 are unchanged. Example - graphical AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles (a) (b) (a) 2 3 5 6 5'5' 6'6' 4'4' 4

40. 40 AME 436 - Spring 2016 - Lecture 14 - Nonideal effects in propulsion cycles Summary - non-ideal effects  Irreversible diffusion, compression, expansion (due to viscosity, shocks, flow separation, etc.) all reduce thrust &  o  Stagnation pressure loss in diffuser - less expansion possible at end of cycle  More  T (thus more work) to drive compressor  More  P needed to extract turbine work required to drive compressor  Surprisingly, component efficiency at thrust = 0 nearly same for diffuser, compressor, turbine, nozzle! (0.55 - 0.6 for cases shown)  Effect of combustor pressure loss relatively small  P exit ≠ P ambient (P 9 ≠ P 1 ) reduces thrust &  o but effect is small  Heat transfer to/from gas during cycle  Modeling of heat losses different for steady flow device »Each component has its own T (unlike unsteady flow engines - use 1 wall temperature) »Stagnation vs. static temperature  As with unsteady flow engines, more work required to compress hot gas than cold gas, lowers  o - best performance at h = 0