Conceptual Design of Environmentally Friendly Rotorcraft - A Comparison of NASA and ONERA Approaches

Conceptual Design of Environmentally Friendly Rotorcraft – A Comparison of NASA and ONERA Approaches Carl Russell Pierre - Marie Basset Aerospace Engineer Aerospace Engineer NASA Ames Research Center ONERA – The French Aerospace Lab Moffett Field, CA Salon de Provence, France ABSTRACT In 2011, a task was initiated under the US - French Project Agreement on rotorcraft studies to collaborate on design methodologies for environmentally friendly rotorcraft. This paper summarizes the efforts of that collaboratio n. The French and US aerospace agencies, ONERA and NASA, have their own software toolsets and approaches to rotorcraft design. The first step of this research effort was to understand how rotorcraft impact the environment, with the initial focus on air pol lution. Second, similar baseline helicopters were developed for a passenger transport mission, using NASA and ONERA rotorcraft design software tools. Comparisons were made between the designs generated by the two tools. Finally, rotorcraft designs were gen erated targeting reduced environmental impact. T he results show that a rotorcraft design that targets reduced environmental impact can be significantly different than one that targets traditional cost drivers, such as fuel burn and empty weight.

INTRODUCTION NOTATION Because air pollution is becoming increasingly regulated in ATR Average Temperature Response CAMRAD Comprehensive Analytical Model of Rotorcraft industrialized nations, new rotary ­‐ wing aircraft will need to Aerodynamics and Dynamics be designed for minimal environmental impact. In Europe, CREATION Concepts of Rotorcraft Enhanced Assessment total CO emissions by airlines were capped in the year Through Integrated Optimization Network 2012, with other emissions likely to follow. No such FOCA Swiss Federal Office of Civil Aviation regulation has been enacted in the US, but may be in the GHG G reenhouse Gas future. If aircraft operators are limited in the amount of HOGE Hover Out of Ground Effect emissions they can legally produce, they will require designs HOST Helicopter Overall Simulation Tool that are not only efficient in terms of traditional metrics , ISA International Standard Atmosphere such as fuel burn and maintenance costs, but that are also LOSU Level of Scientific Understanding environmentally friendly.

NDARC NASA Design and Analysis of Rotorcraft OEI One Engine Inoperative Direct emphasis on e nvironmental performance, particularly RF Radiative Forcing from an air pollution standp oint, has been largely absent up RSM Response Surface Model to this point in rotorcraft design, but it has been implicit in b Number of blades the design of fuel - efficient engines. Worldwide, aviation c Mean blade drag coefficient d mean accounts for approximately 5% of all anthropogenic sources C / σ Rotor thrust coefficient divided by solidity of radiative forcing, a measure of the atm ospheric effects of T L/D Effective lift - to - drag ratio e various pollutants (Ref. 1 ). If rotorcraft are to become a = weight * speed / power large part of the civil aviation fleet, they have the potential R Rotor radius to make a substantial contribution to aviation’s overall UP Utopian Point climate impact. There are multiple existing metrics that can V Speed for best range br be used to evaluate the effects of combustion emissions on W Weight of component xxx x xxxx the environment. Metrics specifically targeted at evaluating Z Cruise altitude cr aircraft emissions are also becoming available .

α Rotor shaft angle of attack κ Induce d power factor i Collaboration on a task named “Environmentally Friendly μ Advance ratio Rotorcraft Concepts” has been ongoing since September μ Axial advance ratio z 2011 between ONERA, NASA, and the US Army Aeroflightdynamics Directorate under the US - French Project Agreement on rotorcraft studies. The purpose of this task was to introduce environmental metrics in the design and evaluation tools for rotorcraft concepts. This paper provides Presented at the AHS 71st Annual Forum, Virginia Beach, a summary of the work that has been ac complished under Virginia , May 5 – 7 , 20 15. This material is declared a work of this France - US collaboration.

the US Government and is not subject to copyright protection.

One of the first steps was to survey available metrics for In addition to choosing metrics that are relevant to current or measuring the environmental impact of rotorcraft. The future public policy and that have acceptable levels of prima ry focus was on air pollution, but the issue of noise uncertainty, it may be desirable to use metrics that account was also addressed. Second, comparisons were made for all rel evant aircraft emissions, rather than a single between the rotorcraft design and analysis tools in use by species. Figu re 2 shows the radiative forcing (RF) in the year both NASA and ONERA. Baseline helicopter designs were 2000 for the primary emission species produced by aircraft developed as a common starting point fo r developing (Ref. 3 ). RF is a measure of the amount of heat trapped in environmentally friendly rotorcraft. Finally, alternative the atmosphere by a particul ar pollutant, and is expressed in rotorcraft designs were developed to optimize both terms of trapped energy per unit area. The level of scientific traditional cost metrics and environmental performance understanding (LOSU) for each emission species is shown in metrics. This paper includes comparisons of the rotorcraft the right - most column.

design tools and met hods used by NASA and ONERA as well as examples of the environmentally friendly rotorcraft 1.2*.(,1#*%%*$.%, concepts that can be generated.

!" $% &" $%( "$%)" $%+,,-$%.-/0% # ' # * BACKGROUND !"#$%&'()*+,-$.+(.")/0$.%, Rotorcraft Environmental Impact Designing rotorcraft while taking into account 3/4*/05( ,6$)+*.2, environmental impact from an emissions sta ndpoint is a 8.+)(/%*.2, 8.+)(/%*.2, fairly new area of research, though the impacts of fixed - :.+()"/*.";, 3(7(5/.+(, wing aircraft have been studied for decades. There is -7*#/"(,-'/.2(, 1.23.45-64.%/7589.$%+.5%:.;.:% significant uncertainty in many of the metrics that can be /7589.$%.'-4.2.%<.5-7.4%.;.8-+% used to evaluate the effects of emissions. Figure 1 shows the cause and e ffect chain linking aircraft emissions to 8#&/+"%, atmospheric changes and ultimately societal impacts =94>/6:-64.$%./,+?+-.2+$%.8.49?% (Ref. 2 ). Effects near the top of the figure are relatively easy 34,@6/A,8B/,8+623A,8$%76258%7.5:-7% to quantify, but are difficult to link to costs in terms of social welfare and are thus not very useful for evaluating rotorcraft 9/#/2(%, concepts. Effects near the bottom of the figure are much C,8.-54?%/,+-+$%+,/>5:%<.:D54.% more difficult to accurately quantify, but are much more Figure 1. Cause - effect chain for climate change induced relevant from a political and social standpoint. Any metric by aircraft emis sions, adapted from Ref. 2 that is used to evaluate new rotorcraft concepts should balance uncertainty with relevance as much as possible.

RF,$$ $ LOSU 2$ RF$Component$ mW/m 25.3$ Good$ 5+6I)1$J/)K/;=$ 21.9$ Fair$ FG)1=$H6);7<0)1$ NO $ x Emissions$ 110.4$ Fair$ D=*A+1=$E=;7<0)1$ 2.0$ Fair$ B+*=6$C+@)6$ 13.5$ Fair$ ?7,@A+*=$-=6)8),$ 2.5$ Fair$ ?))*$-=6)8),$ 10.0$ Fair$ >/1=+6$5)1*6+/,8$ [10$to$80]$ Poor$$ :1;7<=;$5/6678$ 47.8$ ()*+,$-./+0)1$234)$5/66789$ !"#$ #$ "#$ %#$ &#$ '#$ !"#$"%&'()*+,$-./(0120 ( Figure 2. Components of RF due to various aircraft emission species, adapted from Ref. 3 NO emissions cause changes in RF indirectly through In Europe , after the “Friendcopter” project (2004 - 2008) , x chemical processes in the atmosphere. Increases in NO lead which was aim ed at reducing rotorcraft noise, a wider and x to both increases in atmospheric ozone (a warming effect), more ambitious project was launched called “Clean Sky” and reductions in methane (a cooling effect) (Ref. 4 ). The (2008 - 2015). Targets were fixed for the entire environmental methane reduction has a secondary effect of reducing ozone, impact of aviation: reduc tion of GHG emissions ( - 26% so there are actually three components of RF due to NO to - 40% of CO , - 53% of NO ), halving the perceived noise, x 2 x emissions. Note that the impact of NO emissions is similar and implementing a green life cycle (through the entire life x in magnitude to that of CO , but there is greater uncertainty of a helicopter , which is about 40 years). The ACARE in the NO RF values shown in Fig. 2. The black dot and (Advisory Council for Aeronautics Research in Europe) x error bars show, respectively, the best estimate and upper fixed even further reductions to be reached in 2020 with an d lower bounds of RF due to induced cirrus reported in respect to 2000 levels : 50% reduction of CO emissions, Ref. 3 . The total RF due to aviat ion shown at the bottom of 80% reduction of NO and the same goal for the noise as in x Fig. 2 does not include the effects of induced cirrus Clean Sky ( - 50% i.e. - 3 dB of perceived noise).

cloudiness, due to the low level of scientific understand ing.

Within Clean Sky, there are six “Integrated Technology Demonstrators” (ITDs) and a Technology Evaluato r. One of The contribution of rotorcraft operations to greenhouse gas (GHG) emissions is currently very low when compared with the six ITDs is dedicated to rotorcraft and is called Green RotorCraft (GRC). A good recent overview of the GRC other sectors. Indeed, for the period 1970 - 2004, the transpor t sector contributed to 13.1% of the total anthropogenic GHG project was presented in Ref. 6 . Further information relevant to the present topic is contained in Ref. 5 , which also seems emissions ( 23% for just the year 2004). For the EU - 27 (27 European Union Member States) in 2005, aviation to be the first attempt to define a green metric specifically represent ed 3% of the total GHG emissions. Helicopters for helicopters.

represent only 1% of that , meaning they represent 0.03% of the total EU - 27 GHG emissions and 0.36% of emis sions in Environmental Impact Metrics the transport category . Thus , in a global comparative Multiple metrics were considered to evaluate the assessment, current helicopter operations have a very small environmental performance of rotorcraft. Two of the more contribution, as illustrated in Fig. 3 (reproduced from promising metrics are described here. Ref. 5 focuses on CO Ref. 5 ) .

emissions , and it proposes a metric adapted for rotorcraft operations for quantifying their relative level of CO All#Sectors# Transport# Interna'onal*Avia'on*.*10%* Interna'onal*Avia'on*.*2.5%* Domes'c*Civil*Avia'on*.*2%* Domes'c*Civil*Avia'on*.*0.5%* Marine*Transport*.*15%* Other*Sectors*.*97%* Other*Transporta'on*.*1%* Road*Transporta'on*.*72%* Avia1on# Interna'onal*Avia'on*.*85%* Domes'c*Civil*Avia'on*.*14%* Helicopters*.*1%* Figure 3. A Eu ropean view of Helicopters’ GHG emission contribution s in 2005 (Reproduced from Ref. 5 ) pollution. One reason for this focus , despite the fact there are Rotorcraft Software Toolsets other kinds of GHG emissions by rotorcraft ( e.g., NO , etc.)

x One of the goals of this collaborative research effort was to is that CO emissions are directly proportional to the understand some of the differences between the rotorcraft quantity of fuel burned . Indeed , if the combustion is conceptual design and analysis sof tware tools under considered as complete, these two quantities are connected development at NASA and ONERA. This section gives a by a simple sto i chiometric relationship: brief overview of the tools used by the two organizations.

Quantity of CO emitted ≅ 3.16 x quantity of fuel burned Rotorcraft design software The purpose of rotorcraft design software is to quickly Therefore , all efforts for reducing the CO emissions will be evaluate rotorcraft concepts using reduced - order models that directly beneficial to the reduction of fuel consumption and , allow quick program execution. Both NASA and ONERA to a certain extent , the d irect o perating c ost. The final have been developing their own software codes for rotorcraft proposed metric in Ref. 5 is the hourly fuel consumption design over the past few years. These software tools are divided by a certain quantity of “transported kilogram . ” That briefly describe here, and are extensively documented in allow s t he evaluat ion of the climate impa ct (in this case, the external lit erature.

fuel burned ) w ith respect to the service provided. It is proposed to use the useful load, which seems to be better 10 11 12 NDARC adapted for rotorcraft than the passengers’ weight or the payload weight.

NDARC (NASA Design and Analysis of Rotorcraft) is a conceptual/preliminary design and analysis code for rapidly So the final proposed metric in Ref. 5 is: sizing and conducting performance analysis of new rotorcraft concepts (Refs. 10 - 12 ). NDARC has a modular Kilogram of fuel burned / hour / kilogram of useful load code base, facilitating its extension to new concepts and the implementation of new computational procedures. A typical determined at maxi mum takeoff weight Sea - Level ISA, by NDARC run consists of a sizing task, which can be followed averaging the fuel consumption on a typical mission. This by off - design performance analysis. D uring the sizing metric falls at the top of Fig. 1, since it is directly based on process, mission performance is calculated and the aircraft is measurable quantities.

resized both geometrically and mechanically until convergence criteria are met. The software uses reduced - Another potential metric is the Average Temperature order performance models for various rotorcraft subsystems, Response (ATR) , a re cently developed metric that such as ro tors and engines, in order to facilitate short specifically targets aircraft emissions. ATR uses measurable runtimes. These models require curve - fits to higher - fidelity quantities as well as climate models to assess the relative models or experimental data in order to capture rotorcraft performance of aircraft concepts with respect to climate performance. Engines are represented using a non - change. The metric is measured in terms of global mean dimensionalized model based on curve fit s to real engine temperature change caused by operation of a particular data . Mass flow is used to scale engine performance and aircraft. ATR can be used with a number of different climate size . For this study , CAMRAD II (Ref. 13 ) provided the models, but simple linear climate models are appropriate for rotor performance data for the NDARC model curve - fits.

the conceptual design of rotorcraft (Ref. 7 ).

NDARC has previously been applied to environmentally friendly rotorcraft d esign in Ref. 9 . Environmental impact The ATR metric is based on the radiative forcing generated effects were captured by post - processing NDARC output by each emission species. Many climate change metrics, file s .

such as Global Warming Potentials, rely on RF, but do not specifically target emissions due to aviation (Ref . 8 ). The CREATION total RF for all emitted pollutants is used to cal culate the CREATION (Concepts of Rotorcraft Enhanced Assessment global temperature response. The use of an altitude - sensitive Through Integrated Optimization Network) is a climate model captures the effects of operating a particular computational work shop dedicated to the evaluation of aircraft at a multitude of operating conditions. In addition, rotorcraft concepts with respect to flight performance and ATR includes parameters such as usage rates and operating environmental impact. It is composed of both models and li fetime of the aircraft to determine the total climate impact methods tailored for rotorcraft presizing and evaluation. The that results from adding a particular aircraft to an operator’s models are distributed in seven disciplinar y modules: flight fleet. For the current study, ATR was the primary metric performance, environmental impact, aerodynamics, weights used to evaluate rotorcraft environmental performance. It and structures, power generation, missions and was also previou sly used in Ref. 9 to evaluate the specifications, and architecture and geometry. Within each environmental impact of tiltrotors .

disciplinary module, several modeling levels are available in order to adap t the models used to the available data. Four main modeling levels are currently used: CAMRAD II • Level 0: Response Surface Models (RSM) ba sed on databases or simulations For this study, induced and profile power calculated by • Level 1: S imple ana lytical models based on physics CAMRAD II were used to calibrate the equations used by • Level 2: M ore comprehensive analytical models the NDARC rotor performance model. CAMRAD II is an • Level 3: Numerical models aeromechan ics analysis of rotorcraft that incorporates a combination of advanced technologies, including multibody More information can be found on the CREATION models dynamics, nonlinear finite elements, and rotorcraft in Refs. 14 and 15 . The methods include both surrogate aerodynamics. CAMRAD II finds the equilibrium solution model generation techniques and Multidisciplinary Design for a steady state operating condition, and then produces the Optimization (MDO) methodologies. An example of solution for performance, loads, and vibration. CAMRAD II application of the se different kinds of methods has been has undergone extensive correlation of performance and described in Ref. 16 . Two different MDO approaches have loads measurements on rotorcraft (Refs. 19 - 26 ).

been applied to the presizing o f a 90 - passenger helicopter in 19 20 21 22 23 24 25 26 Ref. 17 . One makes use of a genetic algorithm for For thi s study, rotor performance analysis in CAMRAD II computing first the Pare to front of the multi - objec tive sizing considered a single rotor for a tiltrotor or a conventional problem from a RSM of the modeling chain , and then a helicopter, but included both main rotors for a tandem deterministic algorithm is applied for selecting a global design in order to capture interference effects . The optimal design. The other approach uses Mixture of Expert calculations for calibration of the siz ing code rotor model models for generating the RSM relative to each objective considered an isolated rotor, without interference from other and then a determi nistic algorithm for computing the best components, such as the wing or fuselage. Rotor solution for each objective , and finally the global best performance was calculated using free wake geometry for compromise design. The general architecture of CREATION both hover and high - speed cruise. Airfoil characteristics is given in Fig. 4 , and a more thorough description of its use were obtain ed from tables representing advanced technology is contained in a later section.

airfoils. In hover, rotor thrust was varied from zero to above the point of stall . In cruise, the pitch, thrust, and forward Rotorc raft Comprehensive Analysis velocity were varied through the expected envelope of operations.

NDARC and CREATION are both supplemented by rotorcraft comprehensive analysis codes to provide detailed HOST rotor performance data used by the design codes’ surrogate models. CAMRAD II is used to generate the surrogate HOST (Helic opter Overall Simulation Tool) is the Airbus models for ND ARC, while HOST (Ref. 18 ) provides Helicopters aeromechanics simulation code. ONERA has CREATION’s rotor performance models.

contributed to its development for years in many areas, such as rotor inflow and wake models, interference, and soft blade aeroelastic and dynamic stall model s (e.g., Refs. 27 - 31 ).

NFM: Numerical Flight Mechanics AFM: Analytical Flight Mechanics BP: Balance of Power Level 3: NFM More refined " optimization " More refined " Level 2: AFM optimization " st 1 presizing " loop " Level 1: BP st 1 guess " Level 0: Statistics and reduced models Figure 4. Organization of the CREATION tool This simulation code is used both for detailed studies of The conventional helicopter design is shown in Fig. 6 . This rotor aeroelasticity (vibration and loads) and rotor helicopter is designed to carry the 90 - passenger payload aeroacoustics, as well as for the overall simulation of the over the design mission given in Fig. 5 . The first iteration of rotorcraft flight dynamics. For years this code has been this design was introduced in Ref. 33 as a baseline for upgraded and validated with respect to wind tunnel test and comparison of tiltrotor and compound helicopter designs for flight test experimental databases for different kinds of a mission similar to the one shown above. The helicopter helicopters and tilt - rotor configur ations. configuration was also used as a basis of comparison for the 27 28 29 30 31 current study. The conventional helicopter was chosen as a At the modeling L evel 1 of CREATION, analytical baseline because it is a well - understood configuration, and 18.62 aerodynamic rotor models are used for calculating the was the first configuration that could be simultaneously induced rotor inflow correction factor , κ , and the blade modeled by both CREATION and NDARC.

i mean drag coefficient which are required for estimating respectively the induced power and the blade profile power.

These analytical models are based on physics by including different terms corresponding to the different physical involved phenomena (e.g. stall, compressibility, etc.). They are calibrated with respect to numerical simul ations performed with higher order aerodynamic rotor models. In practice at ONERA, these reference simulations can be done with HOST or FlightLab. Here the computations have been performed with the HOST blade element rotor model including soft blade aeroel asticity and a realistic 119.50 aerodynamic field calculated by FiSuW (Ref. 32 ), the finite th state rotor dynamic inflow model, truncated at the 24 order giving 325 states for a fine representation of the rotor inflow distribution.

Rotorcraft Designs Figure 6 . 90 - passenger conventional helicopter configuration Three rotor craft configurations were studied as part of this R 48.97 research effort: a conventional helicopter, a tandem The second rotorcraft configuration was a tandem compound compound helicopter, and a tiltrotor. All three aircraft were helicopter, shown in Fig. 7 . This compound helic opter designed to fly a mission with a payload of 90 passengers configuration was chosen based on results from Ref. 34 , (approximately 9,000 kg) over a d istance of 1,000 km with additional reserve segments. The mis sion profile is shown in which compared the performance of various compound 9.00 R 9.79 helicopter configurations for a 500 nautical mile passenger Fig. 5 . Note that the cruise altitude was not specified, but was optimized as part of the design process. Cruise speed is transport mission. Of the four configurations studied (the other three being single - main - rotor aircraft), the tandem also not specified, but rather optimized for min imum fuel compound had the lowest fuel burn, empty weight, and burn . The design mission includes two reserve segments: installed power. There were two key reasons why the one at cruise altitude for 185 km, and one at an altitude of 10.00 tandem compound outperformed the other designs: reduced 1,500 m for 30 min. Both reserve segments are flown at V , br download, due to the wing not being in the rotor downwash, speed for best range. Two other sizing conditions are 30.06 and reduced disk loading, which was enabled by placing the 5.00 imposed. First, an OEI hover out of ground effect is two large rotors at the ends of the long fuselage.

required. Second the transmission is sized for maximum 10.00 24.4 gross weight hover out of ground effect at maximum rated 108.90 engine power under sea - level static conditions.

AlFtude" (m)" Cruise"at" V Cruise"at" V (reserve) br $ br $ $ Cruise"Alt."

30"min"Cruise"at" V Climb"at"minimum"" br $ (reserve) power""margin" $ 1,500" (ISA"+20°C)" 3"min"taxi,"2"min"HOGE" 1"min"HOGE" 161.1 Range" 0" 10 00" 1185 " (km)" Figure 7 . 90 - passenger tandem compound helicopter Figure 5 . Design mission 72.0 January'10,'2013' The third rotorcraft configuration studied was a tiltrotor. The DESIGN PROCESS tiltrotor modeled for this stu dy was based on NASA’s Large The underlying principles of the rotorcraft design processes Civil Tiltrotor 2 (LCTR2), which has been extensively at ONERA and NASA are very similar. Both organizations studied and refined over the past decade (Refs. 35 and 36 ).

employ an iterative design process using reduced - order The 90 - passenger tiltrotor c onfiguration is shown in Fig. 8 .

models that are supplemented by more advanced The tiltrotor has previously been shown to perform bet ter comprehensive analysis to calibrate various performance than either a compound helicopter or a lift - offset coaxial models. There a re, however, a couple of key differences design in Ref. 37 . One of the goals of the current study was between the two tools.

to see how updated versions of the tiltrotor and tandem compound helicopters compare in terms of both With respect to vehicle optimization, CREATION employs a conventional cost metrics and e nvironmental performance more formal optimization process, while the NDARC metrics.

optimization process relies on manual variation of parameters. With respect to roto rcraft configurations, NDARC currently allows a more comprehensive suite of aircraft, including arbitrary combinations of components (i.e., wings, rotors, tails).

These differences reflect the focus of the software development for the two tools. Addition al development is currently underway to add more formal optimization to the NDARC design process, and additional capabilities are being added to CREATION to model more arbitrary rotorcraft configurations. In the future, these additions will bring the capab ilities of the two software tools more in line Figure 8 . 90 - passenger tiltrotor with one another. The following sections describe the design processes used by NASA and ONERA.

There is a large amount of commonality between the three configurations. All three aircraft have the same fuselage

R 32.50

NASA Design Process dimensions. They also all have four engines, which are Much of the initial design for the three rotorcraft based on the sa me scalable engine model. Various fixed configurations described in the previous sections had already system weights, such as furnishings and hydraulic systems, been completed in previous studies. Parameters suc h as are either equal or are based on equal scaling factors. All number of blades, design C / σ , fuselage sizing, and tail three designs also use a hover tip speed of 200 m/s. This T volumes were therefore already pre - determined. For the choice of tip speed is based on predict ed noise requirements current study, the main design variables investigated with for a passenger transport rotorcraft. A common set of NDARC wer e wing loading, disk loading, cruise altitude, technology assumptions are used. A technology factor of and cruise speed. CAMRAD II was used to investigate main 0.79 was applied to al l of the weight groups in NDARC.

rotor tip speed in cruise, main rotor twist, and wing lift share This weight - scaling factor was chosen to make NDARC and

10.00

in cruise. Designs were evaluated using empty weight, fuel CREATION resu lts for the baseline helicopter have burn, and installed engine p ower, as these three metrics tend approximately equal empty weights, allowing for a common

43.04

to correlate well to cost. Environmental performance was starting point for the two software tools.

evaluated using both total CO emissions for the design

108.90

mission (which are directly proportional to fuel burn) and One difference between the helicopter and the compound the ATR metric. Noise was only addressed implicitl y and tiltrotor models was a difference in the transmission through the choice of hover tip speed.

weig ht. Both the tiltrotor and the compound helicopter use a slowed rotor in cruise. Based on results from Ref. 38 , a ten The iterative design process used for this study is illustrated percent weight penalty was applied to the transmission to in Fig. 9 . Tasks of the design process utilizing NDARC are represent the shifting mechanism that is needed to slow the contained in the heavier square boxes, while tasks using rotor. Another differ ence was the design C / σ . For the T CAMRAD II are contained in the li ghter rounded boxes. The helicopter, C / σ was set to 0.09, based on typical values for T rounded boxes with dashed lines indicate tasks that used a helicopters . For the compound and tiltrotor configurations, a combination of CAMRAD II and spreadsheet analysis. Data C / σ of 0.15 was used, based on results for maneuver T passed between steps is identified next to the flowchart requirements found by Ref. 39 . Although there are some arrows. The process for each of the different configuration s configuration - dependent differences between the designs, was generally the same, but certain steps only apply to one t he goal was to use a common set of assumptions so that the or two configurations. T he steps in the process are outlined benefits or limitations of the configuration would be clear in below.

the final results .

Final' 1. Sweep aircraft parameters design' Sweep'aircra+' Off#design'analysis,' parameters' green'metrics' Aircraft characteristics such as wing loading, disk loading, and cruise altitude were varied in N DARC Rotor'radius,'solidity,'and' using a baseline rotor model, resulting in an initial blade'number;'cruise'and' hover'flight'condi5ons' configuration.

Determine'op5mal' 2. Determine optimal lift share in cruise li+'share'and'rotor' 5p'speed'in'cruise' The main rotor shaft angle was varied at fixed Rotor'opera5ng' collective, and spreadsheet calculations were used to Updated'rotor' condi5ons' performance' determine the lift share that provides ma ximum lift - to - model'and'li+' drag ratio in cruise. Th e lift share step only applies to Analyze'rotors'with' share' varied'twist' the compound helicopter.

distribu5ons' Point'condi5on' 3. Determine optimal cruise tip speed κ 'and' c d!mean!

To determine optimal tip speed in cruise, the rotor RPM Re#size'aircra+'for' was varied in CAMRAD II. The propulsive efficiency different'twist' (for the tiltrot or) or the L/D (for the tandem compound ) distribu5ons e !

were used to evaluate cruise efficiency. The tip speed Best'rotor'blade' step does not apply to the conventional helicopter.

twist' 4. Analyze rotors with varied twist distributions Generate'updated' rotor'performance' Using the rotor diameter and solidity determined in model' Step 1, rotors with varying blade twists were simulated Figure 9 . Iterative design process. NDARC tasks are in in CAMRAD II at the design mission cruise and hover square boxes, and CAMRAD II tasks are in rounded conditions to develop a set of candidate rotors.

boxes. Tasks using both CAMRAD II and spreadsheet analysis are contained in dashed rounded boxes 5. Re - size aircraft for different twist distributions ONERA Design Process The κ and c determined in Step 4 for each of the i d mean candidate rotors were used in NDARC to re - size the The main components of the design process used by aircraft. The rotor blade twist was chosen based on the ONERA for this study include the following fo ur steps: candidate rotor that minimized fuel burn, empty weight, and engine power.

1. S etting the L evel 1 models of CREATION The analytical L evel 1 (see Fig. 4 ) rotor aerodynamic 6. Generate updated rotor performance model models for the considered rotorcraft concept are calibrated Using the rotor twist dist ribution determi ned in Step 5 , with respect to the available , “most up to date” rotor best various flight conditions were simulated in suited for the studied confi guration. In this study, f or CAMRAD II to generate a math model of the rotor helicopter and compound designs , a rotor with the best up to power consumption. While the κ and c determined i d mean date technology was simulated with HOST. For the tilt - rotor, in S tep 4 were for only two specific flight conditions, a rotor based on the A DYN tilt - rotor (Ref. 40 ) was used as the performance model determined here spanned the reference. From these simulations, the induced power factor expected range of operating conditions for the aircraft.

and the blade mean drag coefficient analytical models are calibrated for the expected flight envelope (in terms of C / σ , T 7. Re - sweep aircraft parameters μ and μ ). The weight models are set up using statistics on z databases for some parts and specific analytical models Using the rotor performance model generated in Step 6 , developed by ONERA for others (e.g., the blades, the wings, aircraft characteristics were swept again to arrive at a the fuselage, the main gear box ) .

revised configuration. Steps 1 - 6 could be repeated multiple times if necessary. For this study, the loop was 2. O ptimization process at L evel 1 only completed once for each aircraft.

Two d if ferent methods for dealing with the optimization 8. Off - design analysis problem of multi ple objectives under constraints for a rotorcraft predesign have been previously studied in Ref. 17 Once t he final aircraft design was determined, NDARC and are briefly described here. Results from both methods was used to analyze different operating conditions and wer e applied to the current study.

missions. Post - processing tools were used to calculate environmental performance metrics.

Method 1 : 2 - step h ybrid approach combining both a genetic Method 2 : A lternate approach using only a determinist ic algorithm and a determinist ic algorithm algorithm A) Multi - objective optimization with a Genetic Algorithm : With this method, t he best solution for each objective is A llows global exploration of the design space giving a calculated with a determinist ic algorithm ( for example Pareto Fro nt Nelder - Mead or Sequential Quadrati c Programming ). Once the best (minimum in this case ) values are known, the B) Selection of a best compromise solution by a Utopian P oint is then defined. The global optimum can then Determinist ic Algorithm : be calculated as before by minimizing the distance with If the different objectives or cost functions are not of the respect to the UP using again a determinist ic algorithm to same nature (e.g. weight, power, noise level, etc.), a find th e optimal design values .

normalization is required. From the Pareto front , the minimum and maximum values of each objective are known This alternate method is quicker than the first one. However and thus their normalization is possible. For the case of M ethod 1 is richer as it provides the Pareto front objectives that will be minimized ( which is the case of the corresponding to the assessment of a wide range of optimal objectives here): solutions. It should be emphasized that optimization proces ses often require the use of Response Surface Models m

F X F − ) (

instead of the complete chain of models. M ethod 1 is much i i

X F = ) (

i m M more time consuming because a genetic algorithm for

F F −

i i widely exploring the design space requires 10 , 000 to 20 , 000 computation s of solutions. Theref ore , M ethod 1 requires where X is a solution on the Pareto front P . T he minim um even more use of the RSM than M ethod 2.

and maxim um values F of the objective i are given as : 3. F urther predesign at upper modeling levels m M F = m i n F X and F = m a x F X

( ) ( )

i i i i Once the L evel 1 optimization has provided a first x ∈ P x ∈ P assessment of the main design variables (e.g. for a helicopter, the main characteristics are th e main rotor, the By this way the solution s are non - dimensionalized to sizes of the fuselage and empennage , and the sizing of the produce comparable values between 0 and 1.

engines), more refined optimization can be performed at upper levels. For example , a more detailed presizing of the Next, a global criterion is defined as being the distance wit h tail surfaces requires the calculation of the equilibrium of th e respect to the “Utopian Point” ( UP ) , which represents the forces and moments at each flight point, which can be done best values for each objective: from L evel 2 using a flight mechanics model of the rotorcraft including interference from the main rotor wake .

UP = ( W , W , F , Obj _i , … ) fuel , min empty , min acou , min min At modeling L evel 3, a more detailed predesign of the blade can be studied by using a rotor blade element model W are various weights, and F is a metric quantifying xxxx acou allowing a more refined desc ription of the blade geometric , the noise level on the ground footprint durin g the landing aerodynamic , and structural properties: twist, chord, airfoil approach. For this study, the objectives were minimum fuel profiles, weights and stiffness, etc. The blade dynamics are burn, empty weight, and acoustic impact, in addition to computed in order to check that the eigenmode s are correctly maximum hover figure of merit and maximum cruise L/D .

e placed. Iteration between blade aerodynamic optimization and blade dynamics assessment are performed until a well - A norm for quantifying the distance from the UP has to be suited blade definition is achieved.

chosen . H ere , the Euclidean norm is used (with O the number of objectives): From these more detailed optimizations, new simulations are generated for a better calib ration of the analytical surrogate O models used at L evel 1. For example, for the rotor

= X F X F ) ( ) (

∑ i

2 aerodynamics model s , blade mean drag coefficient and = i 1 induced power factor are tuned again to better reflect this more realistic (or more defined) rotor.

W hen the engineer has no preferences between the different objectives, an impartial global optimum can be assessed by 4. Repeat previo us steps until solution converges calculating the solution on the Pareto Front that minimiz es the distanc e with respect to the Utopian Point. Another loop of optimization at L evel 1 is repeated, followed by the other steps and so on until the whole definition of the rotorcraft is consistent through all the levels of modeling. For this study, Steps 1 - 4 were repea ted at least two times.

12000" COMP ARISON OF SOFTWARE TOOL RESULTS CAMRAD"II"Ptot" HOST"Ptot" To ensure that NDARC and CREATION were providing 10000" CAMRAD"II"Pind" similar results for a given rotorcraft design, comparisons HOST"Pind" were made between the software tools for the baseline 90 - 8000" CAMRAD"II"Ppro" passenger helicopter and tilt rotor .

HOST"Ppro" 6000" Comparing Comprehensive Analysis Codes Power&(kW)& CAMRAD II and HOST are used to generate the surrogate 4000" rotor performance models for NDARC and CREATION, respectively. Comparisons were made between CAMRAD II 2000" and HOST results for the main rotor power of the 90 - passenger helicopter as a function of speed. The results of 0" these comparisons for free wake rotor simulations are shown 0" 100" 200" 300" 400" 500" in Fig. 1 0 . Results are shown for profile (Ppro) and induced Speed&(km/hr)& power (Pind) as well as the total power (Ptot) . The two Figure 1 0 . Comparison of CAMRAD II and HOST software tools gen erally produce good agreement on the results for 90 - passenger helicopter power consumption required power.

8000" Comprehensive analysis r esults were also compared for an CAMRAD"II"Ptot" example 90 - passenger tiltrotor. For both the tiltrotor and the 7000" HOST"Ptot" helicopter, t he airframe forces and moments were calculated 6000" CAMRAD"II"Pind" from polars that give six forces and moments as a function HOST"Pind" of angle of attack and sideslip angle for the major 5000" CAMRAD"II"Ppro" components (wing, fuselage, tails). The rotors were modeled HOST"Ppro" 4000" with two different wake models in both CAMRAD II and HOST. The results for the components of power are show n Power&(kW)& 3000" in Figs. 11 and 12 .

2000" In Fig. 11 , the CAMRAD II results are for uniform inflow.

1000" The HOST results are for Meijer - Drees inflow modeling. In Fig. 12 , the CAMRAD II results are for free wake, and the 0" 0" 100" 200" 300" 400" 500" 600" 700" HOST results use the FiSuW dynamic wake model. The Flight&Speed&(km/hr)& agreeme nt between HOST and CAMRAD II is generally Figu re 11 . Comparison of CAMRAD II and HOST good for both wake models. There is some disagreement in results for 90 - passenger tiltrotor power consumption – the total power at high speed, apparently due to differences simple inflow model in the calculated parasite power. This discrepancy appears to be due to a difference in trimmed pitch angle .

8000" CAMRAD"II"Ptot" Comparing Rotorcraft Design Software Results 7000" HOST"Ptot" To ensure the engine models for NDARC and CREATION 6000" CAMRAD"II"Pind" were producing similar results, c omparisons were made HOST"Pind" between the two software tools for engine fuel flow. The 5000" CAMRAD"II"Ppro" engine here has a maximum rated po wer of 3,563 kW at sea - HOST"Ppro" 4000" level - static conditions. The data generated in NDARC were for an engine scaled from the generic 4 , 000 HP (2,983 kW) Power&(kW)& 3000" engine model included with the distribution of the NDARC software. The results are shown in Fig. 13 for fuel flow at 2000" maximum rated power as a function of a ltitude for both hover and high - speed flight. As shown, t he agreement is 1000" good between the two software tools.

0" 0" 100" 200" 300" 400" 500" 600" 700" Flight&Speed&(km/hr)& Figure 12 . Comparison of CAMRAD II and HOST results for 90 - passenger tiltrotor power consumption – detailed inflow model Finally, similar 90 - passenger helicopters were generated with both NDARC and CREATION to show whether the two design tools could produce comparable aircraft. Both NDARC 0 km/h software design tools were exercised to generate a 90 - CREATION 0 km/h passenger helicopter for the mission specif ications show n in NDARC 556 km/h Fig. 5 . In NDARC, the assumptions on advanced technology CREATION 556 km/h (through the use of tech factors) were adjusted to match the empty weight of the NASA and ONERA designs. A universal tech factor of 0.79 produced good agreement on empty weight.

The bar chart i n Fig. 14 shows weights of various 300 Fuel Flow (kg/hr) component groups for the two resulting designs. The NDARC design is called H90, and the CREATION design is called HO - 90. The empty weights of the two designs are nearly identical ; h owever, some of the weight models used b y NDARC and CREATION are quite different, so it is not 0 2000 4000 6000 8000 10000 12000 surprising to see differences between the component Altitude (m) weights. These differences in are mostly small (less than Figure 13 . Comparison of NDARC and CREATION 5%), but for the propulsion group, the difference is 22%, due engine model results to a difference in the transm ission weights.

50000" The radar plot in Fig. 15 shows a comparison of several 45000" design and flight performance characteristics of the H90 and 40000" NDARC"H90" HO - 90 designs. The agreement between the two is generally 35000" CREATION"HON90" good, except for hover ceiling, which is significantly higher 30000" fo r the CREATION design. The results shown provide 25000" confidence that NDARC and CREATION produce similar helicopter designs, allowing further exploration and 20000" Weight"(kg)" optimization of green rotorcraft concepts.

15000" 10000" ROTORCRAFT DESIGN RESULTS 5000" With good agreement between NDARC and CREATION, 0" Main" Fuselage" Propulsion" Fuel"for" Payload" Empty" Mission" Max" the next step was to use the tools to optimize different Rotor" Group"" Group" Mission" Weight" Weight" Takeoff" rotorcraft configurations . For the initial designs, Group" Weight" environmental performance was handled implicitly by Figure 14 . Comparison of 90 - passenger helicopter design targeting minimum fuel burn (and therefore minimum CO 2 weights generated in NDARC and CREATION emissions). The following sections describe the aircraft that were produced using this method . A later section describes NDARC&H90& Max#Speed# the impact of applying the Average Temperature Response CREATION&HO.90& (km/h)# metric to the design process.

393# Sea#Level# Power#per# Helicopter Hover#Figure# 373# Engine#(kW)# 0.79# of#Merit# Results are presented here first for the NASA results and 3734# 0.78# 3731# seco nd for the ONERA results.

NASA Results 7519# 3500# 7418# Hover#ceiling# Mission#Fuel# OGE#(m)# Burn#(kg)# 4700# Some modifications were made to the NDARC model of the helicopter presented in Ref. 33 . The scalable generic 40642# 6434# 4,000 HP engine model was used instead of the more 6700# 41128# advanced model used in Ref . 33 , and the technology factor Service# Mission# Ceiling#(m)# Weight#(kg)# of 0.79 was applied across all weight groups. With the modified input parameters, disk loading and cruise altitude Figure 15 . Comparison of 90 - passenger helicopter design were varied to find optimal baseline values (step 1 in the characteristics design process previ ously outlined). A disk loading of 39 kg/m and a cruise altitude of 3, 700 m were chosen as the baseline values for the helicopter configuration.

4500" 30000" With the baseline values for disk loading and cruise altitude chosen, the twist distribution of the main rotor was varied to 4000" achieve the best balance of hover and cruise performance.

25000" 3500" CAMRAD II was used to determine the performance of the main rotor for the two most important sizing conditions 3000" 20000" (Step 2 in the design process). These conditions are OEI 2500" Empty"Weight" hover, which sizes the eng ines, and the primary cruise 15000" Fuel"Burn" segment, which sizes the fuel tank. The engine power is 2000" Engine"Power" minimized by maximizing hover figure of merit, and the 1500" 10000" required fuel is minimized by maximizing lift - to - drag in Power&per&Engine&(kW)& cruise .

1000" Fuel&Burn&or&Empty&Weight&(kg)& 5000" 500" The CAMRAD II results form a Pareto front of crui se L/D e vs. hover figure of merit along which the optimum twist 0" 0" 20" 30" 40" 50" 60" 70" 80" mus t fall. Test points along the Pareto front were selected, Disk&Loading&(kg/m )& and the corresponding induced and profile power factors, κ i Figure 17 . Effects of disk loading on the helicopter and c , were input into NDARC to find the twist that d mean configuration; note that engine power is plotted on the minimized fue l burn, engine power, and empty weight. The secondary axis twist that minimized these values was - 12 deg/R inboard and - 14 deg/R ou tboard .

ONERA Results Once the rotor twist was chosen, rotor performance maps On the ONERA side, before switching to the predesign of were generated by varying rotor thrust and speed in other configurations, it was decided to work on further CAMRAD II. The coeffici ents of the NDARC rotor optimization of the HO - 90 helicopter predesign. The goal performance model were then tuned to match the was to refine the predesign by optimizi ng at the same time CAMRAD II outputs. Finally, cruise altitude and disk both the helicopter presizing parameters and the operational loading of the helicopter were varied to determine their parameters ( i.e. cruising speed and altitude ) . This work was optimum values. Figure 16 shows the results for a sweep on performed in a study dealing with five incremental altitude, and Fi g. 17 shows the results for disk loading. A optimization steps considering operational parameters, cruise altitude of 3, 700 m provided the lowest fuel burn, helicopt er design parameters, and engine presizing (with a empty weight and installed engine power. For the disk fixed fuselage geometry for 90 passengers): loading results, there is a minimum in empty weight at 44 kg/m , but fuel burn and engine power cont inue to Step 1: O ptimization of cruising speed and altitude, decrease below the range of values tested. At low disk fixed helicopte r and engine design (HO - 90ini2) loadings, the blade aspect ratio gets very high. Based on values for helicopters currently flying, an upper limit of 20 Step 2: S tep 1 plus engine predesign optimizatio n was set on the aspect ratio of the main rotor blades. This leads to selec ting a disk loading of 39 kg/m .

Step 3: St ep 2 plus main r otor sizing optimization (ra dius, chord, number of blades) 5000" 30000" 4500" Step 4: Step 3 plus main r otor rotational speeds (hover 25000" and cruise) optimization 4000" 3500" Step 5: S tep 4 plus alternate t reatment of the OEI 20000" 3000" requirement Empty"Weight" Fuel"Burn" 2500" 15000" th For the 5 step, the idea was to de al with the OEI condition Engine"Power" 2000" differently as this requirement is very demanding (Hover 10000" 1500" OGE at 1,500 m ISA+20°C) and leads to engines that are Power&per&Engine&(kW)& over - size d for the design mission and therefore have Fuel&Burn&or&Empty&Weight&(kg)& 1000" 5000" increased fuel consumption. The engine presizing was done 500" in S tep 5 by considering four engines in hover, take - off and landing, but only three in cruise where the fourth one is in 0" 0" 0" 1000" 2000" 3000" 4000" 5000" 6000" idle mode (a penalty of 5% extra fuel consumption was Cruise&Al;tude&(m)& applied with respect to the fuel burned by the three other Figure 16 . Effects of cruise altitude on the helicopter engines). The engine was pre sized such that it must be able configuration; note that engine power is plotted on the to provide the takeoff power and minimize the fuel secondary axis consumption in cruise.

Sign ificant reductions of fuel burn on the whole mission and disk loading of 49 kg/m . Using the new tandem compound therefore of the emitted air pollutants have been obtained baseline design, main rotor tip speed and wing lift share with respect to the initial pr edesign (HO - 90ini2): ~19% with were investigated simultaneously. To accomplish this task, step 1, 19.5% with step 2, about 24% with step 3, about 26% the main rotors were simulated in CAMRAD II at a speed of with step 4, about 51% with step 5. These results are 426 km/h, which is the optimum cruise speed determined by illustrated in Fig. 18. NDARC. Rotor collective was fixe d at 0 degrees at 75% radius , which was shown by Ref. 34 to produce the best L/D e 0%& 10,000& 0.0%& for the tandem compound configuration . Rotor cyclic 10%& 9,000& controls were used to trim the rotors to zero hub moment.

Shaft angle of attack was varied f or different values of tip 7,638& 20%& 8,000& 19.2%& 19.5%& 21.6%& 22.1%& 24.2%& 24.9%& 26.1%& speed, resulting in varied amounts of lift on the rotor. 26.3%& 30%& 7,000& 6,169& 6,151& 5,989& Spreadsheet analysis (described in further detail in Ref. 34 ) 5,948& 5,788& 5,736& 5,644& 5,628& 40%& 6,000& was then used to determine the L/D of the whole aircraft.

e 50%& 5,000& 50.1%& 51.3%& 52.1%& 3,718& 3,809& 60%& 4,000& 3,659& The results for main rotor tip speed and lift share are 70%& 3,000& presented in Fig. 19 . From the results shown, the optimum 80%& 2,000& tip speed appears to be very low, below the range of values tested. In fact, lower tip speeds were run in CAMRAD II, 90%& 1,000& and the results suggested that the L/D would continue to e 100%& 0& improve as tip speed was further decreased; however, those results were not well converged, so they are omitted here.

6&blades& 7&blades& 8&blades& 6&blades& 7&blades& 8&blades& 6&blades& 7&blades& 8&blades& HO90& Step&1& ini2& Because there are additional structural and control HO90ini2& Step&1& Step&2& Step&3& Step&4& Step&5& considerations (not studied here) at high advance ratios, Wfuel&[kg]& Percentage&of&Wfuel&reducGon&wrt&HO90ini2& especially abov e 1.0, the tip speed was limited to 120 m/s.

This tip speed gives an advance ratio just under 1.0.

Figure 18. Reduction of fuel consumption with the 5 Figure 19 shows that the maximum L/D for this tip speed e optimization steps with respect t o the initial design occurs at a shaft angle of attack of 3 degrees, giving a wing (HO90ini2) lift share of 93 percent. The rotors a t this light loading and shaft angle have a very l ow power input of less than The cruise speed remains very close to the minimal imposed 100 kW.

value (~280 km/h) , regardless of the optimization case . The optimal cruise altitude ( Z ) varies significantly for each cr The twist of the main rotors on the tandem compound was optimization step and with the number of blad es, b . For selected using a similar process to that for the helicopter. Of example : the twist values tested, an inboard twist of 9 de g/R and outboard twist of - 18 deg/R gave the best results.

Step 3: Z ( b =6) = 1 , 000 m ; Z ( b =8) = 1 , 663 m cr cr 8.6$ Step 4: Z ( b =6) = 822 m ; Z ( b =8) = 1 , 358 m cr cr V.p$=$80$m/s$ 8.5$ V.p$=$90$m/s$ Step 5: Z ( b =6) = 2 , 125 m ; Z ( b =8) = 2 , 943 m V.p$=$110$m/s$ cr cr 8.4$ V.p$=$120$m/s$ V.p$=$140$m/s$ In all of these cases, it is interesting to note that the optimum 8.3$ V.p$=$150$m/s$ cruise altitude for the heli copter is lower for the ONERA 8.2$ results than for the NASA results.

e% 8.1$ L/D The tiltrotor and compound helicopter configurations are 8$ still being studied on the ONERA side . At the time of this writing , however, their presizing is on going with in the 7.9$ 3°$$ CREATION numerical workshop. NASA results are now 4°$$ 2°$$ 7.8$ 5°$$ 1°$$ presented for the compound helicopter and tiltrotor α$=$6°$$ 0°$$ configurations. 7.7$ 61°$$ 7.6$ Tandem Compound Helicopter 0.84$ 0.86$ 0.88$ 0.9$ 0.92$ 0.94$ 0.96$ 0.98$ 1$ 1.02$ Wing%Li*%Share% Starting with the tandem compound helicopter designed in Figure 19 . Effect of main rotor tip speed and wing lift Ref. 34 , a uniform technology factor of 0. 79 was applied share on tan dem compound cruise efficiency across all weight groups, and the generic 4,000 HP NDARC engine model was implemented. Initial parameter sweeps led to a baseline design with a wing loading of 4 90 kg/m and a 6000" With the lift share and main rotor tip speed and twist determined, wing loading, disk loading , and cruise altitude were varied a second time to arrive at the final sized aircraft.

5000" The altit ude results are shown in Fig. 20 , followed by the wing loading and disk loading results in Figs. 21 - 23 . The 4000" parameter sweeps show that a cruise altitude of 7,3 0 0 m minimizes engine power, fuel burn and empty weight. There 3000" is a sharp increase in all three metrics above 8,2 00 m. The reason for knee in the curve is that above this altitude , the WL"="340"kg/m²" cruise speed, which is speed for best range, V , is also the 2000" br WL"="390"kg/m²" Power&per&Engine&(kW)& maximum sp eed for the installed power. The cruise segment WL"="440"kg/m²" therefore sizes the engines above 8,200 m , causing a more 1000" WL"="490"kg/m²" rapid rise in power and weights. Below this altitude, the OEI WL"="540"kg/m²" hover requirement sizes the engines.

0" 20" 30" 40" 50" 60" 70" 80" A wing loading of 4 40 kg/m was chosen as the bes t balance Disk&Loading&(kg/m )& between minimizing fuel burn, empty weight, and installed Fig ure 21 . Tandem compound engine power as a power; fuel burn benefits from a slightly lower wing function of wing loading and disk loading loading, and empty weight is minimized for a higher wing loading. Disk loading was set to 54 kg/m . A lower disk 35000" loading would be prefe rable to minimize power, fuel burn, and empty weight, but would result in the main rotors hitting 30000" the propellers.

25000" 5000" 35000" WL"="340"kg/m²" 20000" WL"="390"kg/m²" 4500" 30000" WL"="440"kg/m²" 15000" 4000" WL"="490"kg/m²" Empty&Weight&(kg)& 25000" WL"="540"kg/m²" 3500" 10000" 3000" Empty"Weight" 20000" 5000" Fuel"Burn" 2500" Engine"Power" 0" 15000" 2000" 20" 30" 40" 50" 60" 70" 80" Disk&Loading&(kg/m )& 1500" 10000" Power&per&Engine&(kW)& Figure 22 . Tandem compound empty weight as a Fuel&Burn&or&Empty&Weigt&(kg)& 1000" function of wing loading and disk loading 5000" 500" 7000" 0" 0" 4000" 5000" 6000" 7000" 8000" 9000" 10000" 6000" Cruise&Al:tude&(m)& Figure 20 . Effects of cruise altitude on the tandem 5000" compound configuration; note that engine power is plotted on the secondary axis WL"="340"kg/m²" 4000" WL"="390"kg/m²" WL"="440"kg/m²" 3000" Fuel%Burn%(kg)% WL"="490"kg/m²" WL"="540"kg/m²" 2000" 1000" 0" 20" 30" 40" 50" 60" 70" 80" Disk%Loading%(kg/m )% Figure 23 . Tandem compound fuel burn as a function of wing loading and disk load ing Tiltrotor As with the helicopter and tandem compound, the engine model for the tiltrotor was replaced with the scalable generic NDARC engine model, and a uniform technology factor of 0.79 was applied across all weight groups. Initial parameter sweeps l ed to a baseline disk loading of 68 kg/m and a wing 0.84% loading of 4 90 kg/m . The first design parameter investigated was tip speed. CAMRAD II was used to 0.83% simulate the baseline rotor in hover and cruise at 556 km/hr.

The baseline tiltrotor desig n had a cruise tip speed of 0.82% 12 0 m/s. To provide high cruise efficiency, the blade twist for the tiltrotor was the helical sweep angle of the blade in 0.81% axial flight at a speed of 556 km/hr. A different twist Hover%Figure%of%Merit% distribution was therefore implemented for each tip speed.

0.8% Figur e 24 shows how rotor propulsive efficiency in cruise Propulsive%Efficiency% and figure of merit in hover differ for varying cruise tip 0.79% speed (and therefore varying twist distribution). The hover cases all use a tip speed of 200 m/s. As Fig. 24 shows, hover 0.78% performance favors a t wist distribution for a slightly lower tip speed than cruise. A cruise tip speed of 1 10 m/s was Figure'of'Merit'or'Propulsive'Efficiency' 0.77% chosen for the final design.

0.76% Performance maps of the rotor were generated in 60% 80% 100% 120% 140% 160% 180% 200% CAMRAD II for various flight conditions, and a second set Tip'speed'(m/s)' of parameter sweeps on c ruise altitude, wing loading, and Figure 24 . Tiltrotor figure of merit and propulsive disk loading were performed. Figure 25 shows the effects of cruise altitude on engine power, empty weight, and fuel efficiencies for varying cruise tip speed burn. The cruise altitude that minimizes all three met rics is 10, 400 m, but as Fig. 25 shows, there is a s harp rise in the engine power and weights above this altitude, similar to the knee in the curves for the compo und helicopter, shown in 7000" 40000" Fig. 20 . The cause of the sharp rise is the same in both cases: 35000" the cruise segment begins to size the engines above a cer tain 6000" altitude. To avoid generating designs that would land on this 30000" 5000" rise, a lower cruise altitude of 10, 100 m was chosen.

25000" 4000" Wing loading and disk loading were varied using a crui se Empty"Weight" altitude of 10,100 m. Figures 26 - 28 show the effects of wing 20000" Fuel"Burn" 3000" loading and dis k loading on engine power, empty weight, 15000" Engine"Power" and fuel burn. A wing loading of 4 40 kg/m gives the best 2000" results. With this wing loading, engine power and fuel burn Power&per&Engine&(kW)& 10000" are minimized for a disk loading of 59 kg/m , and empty Fuel&Burn&or&Empty&Weigt&(kg)& 1000" weight is lowest for a disk loading of 78 kg/m . For the final 5000" design, a disk loading of 68 kg/m was chosen as a 0" 0" compromise between minimum empty weight and minimum 4000" 6000" 8000" 10000" 12000" installed power and fuel burn. One point to notice about the Cruise&Al:tude&(m)& tiltrotor results is that the weights and power do not continue to dec rease as disk loading decreases. The reason for this is Figure 25 . Effects of cruise altitude on the tiltrotor that the wingspan is determined by the clearance of the configuration; note that engine power is plotted on the rotors with the fuselage. For low disk loading at constant secondary axis wing loading, the wing aspect ratio therefore becomes high, resulting in a very heavy wing. The minimum in the plots represents the best compromise between the effects of wing weight and the rotor’s induced power in hover .

9000" Environmental Performance Metrics As previously noted, c arbon dioxide emissions are generally 8000" independent of operating conditions, so they can be easily 7000" calculated by multiplying the fuel burn by a factor of 3.16.

NO emissions are more challenging, as they depend on 6000" x operating conditions as well as engine technology. While 5000" there is a large amount of published turbofan NO emissions x data and estab lished methods for estimating variation with 4000" altitude, much of the data for the turboshaft engines used by WL"="340"kg/m²" 3000" WL"="390"kg/m²" existing rotorcraft is proprietary. To the authors’ knowledge, Power&per&Engine&(kW)& WL"="440"kg/m²" there is no publicly available data that quantifies NO x 2000" WL"="490"kg/m²" emissions for specific turbosh aft engines. There is a limited WL"="540"kg/m²" 1000" amount of data that has been collected by the Swiss Federal WL"="590"kg/m²" Office of Civil Aviation (FOCA) as part of their efforts to 0" develop an emissions inventory for civil aviation, but this 20" 40" 60" 80" 100" 120" 140" 160" dataset has considerable uncertainty (Ref. 41 ). The FOCA Disk&Loading&(kg/m )& methods and data have been used for some recent studies, Figure 26 . Tiltrotor engine power as a function of wing such as Ref. 42 , that seek to model helicopter NO x loading and disk loading emissions. They were also used by ONERA for the current study to calculate NO emissions. Another method based on x 60000" published t urbofan emis sions data and the DLR fuel flow method (Ref. 43 ) was used in Ref. 9 and by NASA in the 50000" current study to estimate upper and lower bounds on tiltrotor NO emissions.

x 40000" In order to calculate NO emissions with this method, a x 30000" base line engine must be chosen. The GE CF34 - 3B and the Honeywell HTF7000 have fuel flows similar to what is WL"="340"kg/m²" calculated in NDARC for a turboshaft engine. Both engines 20000" Empty&Weight&(kg)& WL"="390"kg/m²" are relatively modern, high bypass ratio, small turbofans.

WL"="440"kg/m²" WL"="490"kg/m²" The HTF7000 and CF34 also represent the upper and lower 10000" WL"="540"kg/m²" bounds on NO emissions for this category of engine, so x WL"="590"kg/m²" they should bracket the expected quantity of NO emissions x 0" for a turboshaft engine.

20" 40" 60" 80" 100" 120" 140" 160" Disk&Loading&(kg/m )& Figure 29 shows how total NO emissions vary for different x Figure 27 . Tiltrotor empty weight as a function o f wing cruise altitudes of the design mis sion. Values are shown for loading and disk loading NO emissions based on both of the baseline engines. The x trends are si milar to those shown in Figs. 16, 20, and 25 for 9000" fuel burn, but the altitudes corresponding to the minima are lower for the tiltrotor and compound helicopter. The 8000" compound helicopter has minimum fuel burn and CO at 7000" 7,3 00 m, but minimum NO at 5, 500 m or 6,40 0 m, x depending on the baseline engine model. Likewise, the 6000" tiltrotor has minimum fuel burn and CO at 10, 400 m but 5000" minimum NO at 6,40 0 m or 8, 200 m.

x 4000" The CO and NO emissions (as well as several other minor 2 x WL"="340"kg/m²" Fuel%Burn%(kg)% pollutants) were used to calculate the ATR metric for the 3000" WL"="390"kg/m²" WL"="440"kg/m²" three rotorcraft configurations. The variation in the ATR 2000" WL"="490"kg/m²" metric wi th altitude is shown in Figs. 30 and 31 . The first WL"="540"kg/m²" plot shows how the metric varies if the low - NO baseline x 1000" WL"="590"kg/m²" engine (CF34) is chosen. The second plot shows the 0" variation for the high - NO baseline engine (HTF7000). ATR x 20" 40" 60" 80" 100" 120" 140" 160" is most easily expressed in relative terms, where multiple Disk%Loading%(kg/m )% designs are compared against a baseline. In this case, th e Figure 28 . Tiltrotor fuel burn as a function of wing baseline is the helicopter with a design cruise altitude of loading and disk loading 3, 700 m; hence, it has a relative ATR of 1. ATR in this case is calculated assuming one operation per day with a 30 - year 140" Helicopter"High"NOx" operating lifetime of the aircraft and a time horizon of 50 0 Helicopter"Low"NOx" Tandem"Compound"High"NOx" years. Temperature impacts after the 30 - year operating 120" Tandem"Compound"Low"NOx" lifetime are discounted at a rate of three percent per year, so Tiltrotor"High"NOx" after 50 years, the temperature impacts are discounted by 100" Tiltrotor"Low"NOx" more than 75 percent. This means that shorter - lived emissions, such as NO , c an have a higher impact than x 80" longer - lived emissions, such as CO . For a full description of how ATR is calculated, see Refs. 7 and 9 .

60" NOx$Emi(ed$(kg)$ The results show that of the three rotorcraft configurations, 40" the tiltrotor has the lowest environmental impact, as measured by the ATR metric. When the low - NO baseline x 20" engine is assumed, ATR for the tiltrotor is minimized at a cruise altitude of 8, 200 m. With the high - NO baseline x 0" engine, ATR is minimized at an al titude of 5, 500 m. Recall 0" 2000" 4000" 6000" 8000" 10000" 12000" that fuel burn and CO emissions for the tiltrotor were Cruise$Al5tude$(m)$ minimized for a cruise altitude of 10, 400 m. A similar effect Figure 29 . NOx emissions as a function of altitude can be seen in the results for the tandem compound. The reason for the difference is that NO emissions lead to the x 1.400# production of ozone, which is a powerful, but short - lived greenhouse gas, whose potency is increased when it is 1.200# released at higher altitudes. The production of NO at high x altitudes can therefore lead to more warming, despite 1.000# decreased CO emissions. T he NO temperature impact 2 x model used for this study only extends down to 4, 900 m, and 0.800# was assumed constant below that altitude. That is why the helicopter results do not show the same discrepancy as the 0.600# tandem compound and tiltrotor results. The conclusion Rela%ve'ATR' reached from Figs. 30 and 31 is that if a rotorcraft is Helicopter# 0.400# designed for minimal environmental impact, it may come at Tandem#Compound# a penalty of increased fuel burn.

Tiltrotor# 0.200# Finally, because the minimum - ATR tiltrotor design with the 0.000# high - NO baseline engine flies at such a low al titude, x 0# 2000# 4000# 6000# 8000# 10000# 12000# another pass was made at sweeping wing loading and disk Cruise'Al%tude'(m)' loading for this configuration. T he results are shown in Figs.

Figure 30 . Variation in the ATR metric with cruise 32 - 34 . The trends in the results are very si milar to those altitude with NO emissions based on the CF34 - 3B x shown in Figs. 26 - 28 for a cruise altitude of 10, 100 m, except the res ults at the lower altitude favor a slightly lower 1.400# wing loading. The best compromise between engine power, fuel burn, and empty weight here appears to be at a wing 1.200# 2 2 loading of 390 kg/m and a disk loading of 59 kg/m ; however, to minimize fuel burn (and ther efore CO 1.000# emissions and the ATR metric), disk loading should be lowered to 49 kg/m . One notable thing about this lower - 0.800# flying tiltrotor design is that its optimum speed, 437 km/hr, 0.600# is much lower than that of the higher - flying version, which Rela%ve'ATR' cruises at 567 km/hr. This result is consistent with those Helicopter# 0.400# found by Ref. 7 , which showed that fixed - wing turbofan Tandem#Compound# aircraft have a reduced environmental impact if they are Tiltrotor# 0.200# designed to fly lower and slower.

0.000# 0# 2000# 4000# 6000# 8000# 10000# 12000# Cruise'Al%tude'(m)' Figure 31 . Variation in t he ATR metric with cruise altitude with NO emissions based on the HTF7000 x 9000" Results Summary 8000" Table 1 contains a comparison of the designs generated by NDARC and CREATION. The first three columns show 7000" three different versions of the opt imized helicopter. The first 6000" is the design generated by ND ARC (H90) , and the second two show the results from CREATION (HO90) for two 5000" different levels of optimization. The optimized helicopters generated by NDARC and CREATION are similar to each 4000" WL"="340"kg/m²" other, but the sizing parameters have departed significantly 3000" WL"="390"kg/m²" from th eir baseline values given in Figs. 14 and 15. As Power&per&Engine&(kW)& WL"="440"kg/m²" shown by column 3 of Table 1, relaxing the OEI requirement 2000" WL"="490"kg/m²" on the helicopter has a large impact on the final design , WL"="540"kg/m²" 1000" particularly on fuel burn . WL"="590"kg/m²" 0" 20" 40" 60" 80" 100" 120" 140" 160" The tandem compound (TC90) and tiltrotor (TR90) designs Disk&Loading&(kg/m )& are sh own in the fourth and fifth columns, respectively. The Figure 32 . Tiltrotor power per e ngine as a function of rightmost column shows the results for a tiltrotor optimized wing loading and disk loading – 5,486 m cruise altitude for minimal environmental impact as measured by the ATR metric (assuming the high - NO baseline engine).

x 60000" Of the three configurations, t he t iltrotor has the lowest environmental impact as measured by either the ATR metric 50000" or CO emissions . The tiltrotor has a better cruise efficiency ( L/D of 10.2 vs. 7.6 for the tandem compound and 6.6 for e 40000" the helicopter), giving it a much lower fuel burn. Us ing the ATR metric, the environmental impacts of the tandem 30000" compound and helicopter are somewhat mitigated by a naturally lower cruise altitude; however, the increased CO WL"="340"kg/m²" 20000" Empty&Weight&(kg)& and NO emissions that accompany the increased fuel burn WL"="390"kg/m²" x WL"="440"kg/m²" of these designs offsets an y benefit. Because the design WL"="490"kg/m²" mission is heavily cruise dominated, maximizing cruise 10000" WL"="540"kg/m²" efficiency is key to minimizing environmental impact. For a WL"="590"kg/m²" design mission that is dominated by hover, the high hover 0" efficiency of the helicopter and tandem compound woul d 20" 40" 60" 80" 100" 120" 140" 160" certainly cause these configurations to be more competitive. Disk&Loading&(kg/m )& Figure 33 . Tiltrotor empty weight as a function of wing The final result presented here is that d epending on the loading and disk loading – 5,486 m cruise altitude metric used to measure environmental impact , there is variation in the size, cruise speed, and cruise altitude of the 9000" optimal rotorcra ft design . Designing for minimum CO 8000" emissions alone can lead to a different solution than one that puts a heavy penalty on NO emissions.

x 7000" 6000" 5000" WL"="340"kg/m²" 4000" WL"="390"kg/m²" Fuel%Burn%(kg)% WL"="440"kg/m²" 3000" WL"="490"kg/m²" WL"="540"kg/m²" 2000" WL"="590"kg/m²" 1000" 0" 20" 40" 60" 80" 100" 120" 140" 160" Disk%Loading%(kg/m )% Figure 34 . Tiltrotor fuel burn as a function of wing loading and disk loading – 5,486 m cruise altitude Table 1. Design comparison Helicopters Compound Tiltrotors HO90 HO90 TR 90 - H90 TC90 TR90 Step 4 Step 5 Min ATR Mission Fuel (kg) 5 , 770 5,736 3 , 718 5 , 358 4 , 066 4 , 427 Empty Weight (kg) 23 , 330 26,685 26,405 27 , 772 28 , 199 30,472 Power per Engine (kW) 3 , 185 3,133 2,810 4 , 019 4 , 362 3,930 Wing Loading (kg/m ) – – – 4 40 4 40 39 0 Disk Loading (kg/m ) 39 40 34 54 68 4 9 Main Rotor Blades 7 7 7 4 4 4 Cruise Altitude (m) 3 , 700 1,065 2,472 7 , 300 10 , 100 5 , 500 Cruise Speed (km/hr) 306 280 280 419 567 437 CO Emitted (kg) 18 , 233 17,816 14 , 534 16 , 931 12 , 849 13,945 * NO Emitted (kg) 39 - 77 25 14 38 - 81 25 - 53 25 - 50 x * NO emissions for the NASA designs were calculated with the modified fuel flow method. ONERA NO calculations used the FOCA method .

x x CONCLUSIONS REFERENCES There were two goals of this collaborative research effort.

1 . Lee, D.S. et al., “Aviation and Global Climate Change The first was to gain an understanding of how targeting in the 21st Century,” Atmospheric Environment , Vol. 43, reduced environmental impact affects the rotorcraft design No. 22 - 23, July 2009.

process. The second goal was for the collaborators to gain a greater understanding of each other’s toolsets and 2 . Fuglestvedt, J. S., Berntsen, T., Godal, O., Sausen, R., methodologies for rotorcraft design and analysis. With Shine, K. P., and Skodvin, T., “Metrics of Climate Change: respect to th e first goal, the results showed that designing for Assessing Radiative Forcing and Emission Indices,” reduced environmental impact can have a significant impact Climatic Change , Vol. 68, No. 3, 2003, pp. 267 - 331.

on the final aircraft design. To be “greener,” rotorcraft that typically would fly at high altitudes, such as tiltrotors, 3 . Sausen, R. et al., “Aviation Rad iative Forcing in 2000: should fly lower and sl ower to reduce the impact of NO x An Update on IPCC (1999),” Meteorologische Zeitschrift , emissions. This can come at a cost of higher fuel burn and Vol. 14, No. 4, August 2005.

CO emissions.

4 . Köhler, M. O., Rädel, G., Dessens, O., Shine, K. P., With respect to the second goal, the software toolsets were Rogers, H. L., Wild, O., and Pyle, J. A., “Impact of able to produce similar results for both a helicopter and a Perturbations to Nitrogen Oxide Emiss ions from Global tiltrotor configuration. R es ults for rotor power and engine Aviation,” Journal of Geophysical Research , Vol. 113, fuel consumption were very similar betw een t he NASA and 2008, Paper D11305.

ONERA tools . Also, the rotorcraft design software tools NDARC and CREATION were able to produce very similar 5 . Verlut, F. and Dyrla, N., “Definition by Eurocopter of baseline helicopter designs. The optimized helicopter a Green Metric to Assess gas Emitted by Helicopters in desig ns from both tools diverged from one another, but were Operation,” 36th European Rotorcraft Forum, Paris, France, still similar. In the future, the capabilities of NDARC and September 6 - 9, 2010.

CREATION will likely be more closely matched, allowing for easier collaboration between the two sides.

6 . Antifora, A. and Toulmay, F., “Clean Sky – The Green RotorCraft Integrated Technology Demonstrator – ACKNOWLEDGEMENTS State of Play Three Years after Kick - off,” 37th European Rotorcraft Forum, Gallarate, Italy, Sepember 13 - 15, 2011.

The authors would li ke to acknowledge the CREATION project team at ONERA, and especially the contributions of 7 . Da llara, E. S., Kroo, I., and Waitz, I., “Metric for Arnault Tremolet and Julian Roche. The authors would also Comparing Lifetime Average Climate Impact of Aircraft,” like to thank Wayne Johnson for his contributions and AIAA Journal , Vol. 49, No. 8, August 2011.

guidance throughout this project. This research effort was carried out under Task V of the US/French Project 8 . Intergovernmental Panel on Climate Change, Climate Agreement Concerning Rotary Wing Aeromechanics and Change 2007 – The Physical Science Basis, Contribut ion of Human Factors Integration Technologies.

Working Group I to the Fourth Assessment Report of the 1 9 Intergovernmental Panel on Climate Change , 2007, 19 . Johnson, W. "Rotorcraft Aeromechanics Cambridge Univ. Press, Cambridge, U.K., 2007. Applications of a Comprehensive Analysis." HeliJapan 1998: AHS International Meeting on Rotorcraft Technology 9 . Russell, C. and Johnson, W., “Application of Climate and Disaster Relief, Gifu, Japan, April 1998.

Impact Metrics to Civil Tiltrotor Design, ” 51st AIAA Aerospace Sciences Meeting, Grapevine, TX, January 2013. 20 . Johnson, W. "Rotorcraft Aerodynamic Models for a Comprehensive Analysis." American Helicopter Society 10 . Johnson, W. "NDARC. NASA Design and Analysis 54th Annual Forum, Washington, D.C., May 1998.

of Rotorcraft." NASA TP 2009 - 215402, December 2009.

21 . Johnson, W. "Calculation of Tilt Rotor Aeroacoustic 11 . Johnson, W. "NDARC — NASA Design and Model (TRAM DNW) Performance, Airloads, and Analysis of Rotorcraft. Theoretical Basis and Architecture." Structural Loads." American Helicopter Society American Helicopter Society Specialists' Conference on Aeromecha nics Specialists' Meeting, Atlanta, GA, Aeromechanics, San Francisco, CA, January 2010. November 2000.

12 . Johnson, W. "NDARC — NASA Design and 22 . Yeo, H. "Calculation of Rotor Performance and Analysis of Rotorcraft. Validation and Demonstration." Loads Under Stalled Conditions." American Helicopter American Helicopter Society Specialists' Conference on Society 59th Annual Forum, Phoenix, AZ, May 2003.

Aeromechanics, San Francisco, CA, January 2010.

23 . Yeo, H., Bousman, W. G., and Johnson, W., 13 . Johnson, W., "Technology Drivers in the “P erformance Analysis of a Utility Helicopter with Standard Development of CAMRAD II," American Helicopter and Advanced Rotor,” Journal of the American Helicopter Society Aeromechanics Specialist Meeting, San Francisco, Society , Vol. 49, No. 3, July 2004.

California, January 1994.

24 . Yeo, H., and Johnson, W., “Assessment of 14 . Basset, P. - M., T remolet, A., Cuzieux, F., Schulte, C., Comprehensive Analysis Calculation of Airloads on Tristrant, D., Lefebvre, T., Reboul, G., Costes, M., Richez, Helicopter Rotors,” Journal of Aircraft , Vol. 42, No. 5, F., Burguburu, S., Petot, D., and P aluch, B., “The Sept. – Oct. 2005.

C.R.E.A.T.I.O.N. project for rotorcraft concepts evaluation: The first steps”, 37th European Rotorcraft F orum, Gallarate, 25 . Yeo, H., and Johnson, W., “Prediction of Rotor Italy, 13 - 15 September 2011. Structural Loads with Comprehensive Analysis,” Journal of the American Helicopter Society , Vol. 53, No. 2, April 2008.

15 . Basset, P. - M., Tremolet, A., Cuzieux, F., Reboul, G., Costes, M., Tristrant, D., and Petot, D., “C.R.E.A.T.I.O.N. 26 . Harris, F.D. "Rotor Performance at High Advance the Onera multi - level rotorcraft concepts evaluation tool: the Ratio; Theory versus Test." NASA CR 2008 - 215370, foundations”, Future Vertical Lift Aircraft Design October 2008.

Conference, AHS Specialists meeting, San Francisco, CA, USA, 18 - 20 January 2012. 27 . Basset, P. - M., “Modeling of the Dynamic Inflow on the Main Rotor and the Tail Components in Helicopter 16 . Tremolet, A. and Basset, P. - M., “Some Meta - Flight Mechanics”, 22nd European Rotorcraft Forum, Modeling and Optimization Techniques for Helicopter Pre - Bright on, UK, September 1996.

Sizing”, 38th European Rotorcraft Forum, Amsterdam , The Netherlands, 4 - 6 September, 2012. 28 . Basset, P. - M., Tchen - Fo, F., "Study of the Rotor Wake Distortion Effects on the Helicopter Pitch - Roll Cross - 17 . Basset, P. - M., Tremolet, A., B artoli, N., and Couplings,” 24th European Rotorcraft Forum, Marseilles, Lefebvre, T., “Helicopter pre - sizing by multidisciplinary – France, September 1998.

multi objectives optimization”, OPT - i, International Conference on Engineering and Applied Sciences 29 . Basset, P. - M., Heuzé, O ., Prasad, J.V.R., Hamers, Optimi zation, Kos Island, Greece, 4 - 6 June 2014. M., “Finite State Rotor Induced Flow Model for Interference and Ground Effect,” AHS 57th Annual Forum, May 2001.

18 . Benoit, B., Dequin, A. - M., Basset, P. - M., Gimonet, B., von Grünhagen, W., and Kampa, K., “HOST, A General 30 . Basset, P. - M. and Brocard, M., “A Fenestron Model Helicopter Simulation Tool for Germany and France,” AHS for Improving the Helicopter Yaw Dynamics Flight 56th Annual Forum, Virginia Beach, VA, May 2000 . Simulation”, 30th European Rotorcraft Forum, Marseilles, France, 14 - 16 September 2004.

31 . Basset, P. - M., Chen, C., Prasad, J.V.R., and Kolb, S., 42 . Goulos, I, Ali, F., Tzanidakis, K., Pachidis, V., and “Prediction of Vortex Ring State Boundary of a Helicopter d’Ippolito, R., “A Multidisciplinary Approach for the in Descending Flight by Simulation”, Journal of the Comprehensive Assessment of Integrated Rotorcraft - American Helicopter Society , Vol. 53, No. 2, pp. 139 - 151, Powerplant Systems at Mission Level,” J. Eng. Gas April 2008. Turbines Power 137(1), ASME, August 2014.

32 . Basset, P. - M. and Ormiston, R. “Comparison and 43 . Deidewig, F., Döpelheuer, A., and Lecht, M., Validation of the France/US Finite State Rotor Dynamic “Methods to Assess Aircraft Engine Emissions in Flight,” Inflow Models,” 36th European Rotorcraft Forum, Paris, 20th International Council on Aeronautical Sciences France, September 6 - 9, 2010. Congress, Sorrento, Italy, 1996.

33 . R ussell, C. and Johnson, W., “Conceptual Design and Performance Analysis for a Large Civil Compound Helicopter,” AHS Future Vertical Lift Aircraft Design Conference, San Francisco, CA, January 18 - 20, 2012.

34 . Russell, C. and Johnson, W., “Exploration of Co nfiguration Options for a Large Civil Compound Helicopter,” AHS 69th Annual Forum, Phoenix, AZ, May 21 - 23, 2013.

35 . Acree, C.W. Jr., Yeo, H., and Sinsay, J., “Performance Optimization of the NASA Large Civil Tiltrotor,” International Powered Lift Conferen ce, London, UK, July 22 - 24, 2008.

36 . Acree, C. W., Jr., “Integration of Rotor Aerodynamic Optimization with the Conceptual Design of a Large Civil Tiltrotor,” American Helicopter Society Aeromechanics Conference, San Francisco, CA, January 20 - 22, 2010.

37 . Johnson, W., Yamauchi, G., and Watts, M., “NASA Heavy Lift Rotorcraft Systems Investigation,” NASA/TP - 2005 - 213467, December 2005.

38 . Acree, C. W. Jr. and Snyder, C., “Influence of Alternative Engine Concepts on LCTR2 Sizing and Mission Profile,” AHS Futu re Vertical Lift Aircraft Design Conference, San Francisco, CA, January 18 - 20, 2012.

39 Yeo, H., Sinsay, J., and Acree, C.W. Jr., “Selection of Rotor Solidity for Heavy Lift Tiltrotor Design,” Journal of the American Helicopter Society , Vol. 55, No. 1, Janu ary 2010.

40 Lefebvre, T., Beaumier, P., Canard, S., Pisoni, A., Pagano, A., Sorrentino, A., van der Wall, B., Yin, J., D’Alascio, A., Arzoumanian, C., Voutsinas, S., and Hermans, C., “Aerodynamic and Aero - acoustic Optimization of Modern Tilt - rotor Blades w ithin the ADYN Project,” ECCOMAS 2004, Jyväskylä, Finland, July 2004.

41 . Rindlisbacher, T. “Guidance on the Determination of Helicopter Emissions,” Swiss Confederation Federal Office of Civil Aviation (FOCA), Ref. 0 / 3/33/33 - 05 - 20, Bern, Switzerland, Marc h, 2009.