1. Multidisciplinary Design Optimization of Low-Airframe-Noise Transport Aircraft 44 th AIAA Aerospace Science Meeting and Exhibit, Reno January 9, 2006 Leifur Leifsson, William Mason, Joseph Schetz, and Bernard Grossman Virginia Tech and Raphael Haftka, University of Florida Work sponsored in part by NASA Langley Research Center Phoenix Integrations Inc. provided ModelCenter software

2. 2 Outline u Introduction u Research objectives u Methodology u MDO formulation u Design studies u Conclusions u Future work (Source: www.airliners.net)

3. 3 Aircraft noise is a growing problem Approach Noise (EPNdB) (Data from “Advisory Circular”, DOT, FAA, November 2001) u 100% increase in noise related restrictions in the last decade u NASA’s goal is to reduce noise by 20 decibels in next 20 years

4. 4 Aircraft Noise Certification u Aircraft must be certified by the FAA and ICAO in terms of noise levels u Certification noise is measured at flyover, sideline, and approach u Based on aircraft max TOGW and number of engines, the noise level is limited u Additionally, regulations limit the hours and the number of operations 2,000 m (1.24 miles) 6,500 m (4.04 miles) Approach Flyover Sideline 450 m (0.28 miles) Threshold Lift-Off Thrust Cutback Brake Release 120 m (394 ft)

5. 5 Research Objectives u Include aircraft noise in the conceptual design phase u Design low-airframe-noise transport aircraft using MDO u Quantify change in performance w.r.t. traditionally designed aircraft Airframe Noise Sources

6. 6 Optimize aircraft without considering aircraft noise Aircraft noise analysis of reference configuration Reference configuration Reference noise level, Re-optimize the reference configuration for a target noise reduction Add a noise constraint New configuration with less noise Design Methodology: Noise as a Design Constraint

7. 7 Optimizer Aircraft Analysis Noise Analysis ModelCenter MDO Framework u Aircraft analysis codes previously developed at Virginia Tech –High-lift system analysis module was added u ANOPP used for aircraft noise analysis u ModelCenter used to integrate the codes u DOT is the optimizer; Method of Feasible Directions optimization algorithm

8. 8 ANOPP Overview u Semi-empirical code u Uses publicly available noise prediction schemes u Continuously updated by NASA u The airframe noise module is component based u Based on airframe noise models by Fink u The general approach: Far-Field Mean Square Acoustic Pressure Acoustic Power

9. 9 ANOPP – Acoustic Power of Each Component u Wing Trailing-Edge (Clean wing) u Leading-Edge Slat –Increment on wing TE noise –TE noise of LE slat u Trailing-Edge Flap u Landing-Gear Turbulent BL thickness

10. 10 MDO Formulation u Objective function –Min Takeoff Gross Weight u Design variables (17-22) –Geometry –Average Cruise Altitude –Sea level static thrust –Fuel weight u Constraints (16-17) –Geometry –Performance Takeoff, Climb, Cruise, Landing u Parameters –Fuselage geometry

11. 11 High-Lift System Configuration c V  ff Section View c’ ss (E a ) (E f ) Planform View crcr b f / 2 b / 2 cbcb ctct bb d / 2 x b s / 2 b a / 2 Slat Flap Aileron (E s ) u High-lift analysis model based on semi-empirical methods by Torenbeek u Model validated by analyzing a DC-9-30 and comparing with published data

12. 12   f  0 CLCL C L max  app C L app  limit =  ts -  gs C L max limit  stall FAA Design Requirement: High-Lift Design Limits and Requirements  s  0

13. 13 MDO Formulation for the High-Lift System DV’s: Constraints: MDO Parameters: Side Constraint: Limited by ANOPP Flap Deflection:

14. 14 Design Studies 1. Approach speed study 2. TE flap noise reduction 3. Airframe noise analysis of cantilever wing vs. SBW Climb Warmup Taxi Takeoff Landing Descent Reserve = 500 nm Mach = 0.85 Range = 7,730 nm Payload = 305 pax Cruise - Climb

15. 15 Study 1: Approach Speed Study

16. 16 Reducing airframe noise by reducing approach speed alone, will not provide significant noise reduction without a large weight penalty TOGW (lb) Approach Speed (knots) S ref (sqft) S ref TOGW Noise (EPNdB) Approach Speed (knots) Total Airframe Noise LE Slat Main Landing Gear TE Flap Clean Wing Nose Landing Gear -1.75 EPNdB +14,240 lb (+2.4%) +714 sqft (+14.3%)

17. 17 Study 2: TE flap noise reduction

18. 18 Eliminate TE flaps by increasing S ref and  without incurring significant weight penalty  (deg)  S flap (sqft) S flap 85.6  f = 30 deg S ref (sqft) S ref TOGW (lb) TOGW 85.6 +5.2 deg +15.2% +1,900 lb TE Flap Noise Reduction (EPNdB)

19. 19 Thus, eliminating any noise associated with TE flaps TE Flap Noise Reduction (EPNdB) Total Airframe Noise Main Landing Gear LE Slat TE Flap Nose Landing Gear Noise (EPNdB) 0 5.07 9.58 No TE Flap TE Flap Noise Reduction (EPNdB) Clean Wing

20. 20 Study 3: Airframe noise analysis of cantilever wing and SBW

21. 21 SBW shows a significant improvement in weight & performance compared to a cantilever wing

22. 22 SBW has a similar or potentially lower total airframe noise than a cantilever wing aircraft u Main landing gear –Cantilever with 6 wheels; SBW with 4 wheels and ½ the strut length u Wing strut modeled as wing TE noise

23. 23 Conclusions u A methodology for designing low-airframe-noise aircraft has been developed and implemented in an MDO framework u Reducing airframe noise by reducing approach speed alone, will not provide significant noise reduction without a large weight penalty u Therefore, more dramatic changes to the aircraft design are needed to achieve a significant airframe noise reduction u Cantilever wing aircraft can be designed with minimal TE flaps without significant penalty in weight and performance u If slat noise and landing gear noise sources were reduced (this is being pursued), the elimination of the flap will be very significant u Clean wing noise is the next ‘noise barrier’ u SBW aircraft could have a similar or potentially lower total airframe noise compared to cantilever wing aircraft

24. 24 Future Work u Important topics –Effects of reduced runway length –Effects on other noise sources Increased drag at approach => Increased engine noise for same speed u SBW’s and BWB’s should be considered in future studies –Clean wing noise model by Hosder et al.