NASA Tec hn ica I Paper 2400

1985 Parametric Study of a -Configured Transport Using Conceptual Design 0p t imiza t ion

P. Douglas Arbuckle and Steven M. Sliwa Langley Research Center Hampton, Virginia

runsn N a t io na I Aeronaut ics and Soace Admlnlstration

Scientific and Technical Information Branch Summary impact of choosing unaugmented longitudinal flying- qualities design criteria; and (3) the sensitivity of a Constrained-parameter optimization is used to transport design to a variety of economic and techno- perform optimal conceptual design of both canard logical assumptions (refs. 3 to 6). The basic OPDOT and conventional configurations of a medium-range code has been enhanced and modified so the program transport. A number of design constants and de- can be used to analyze both canard and conventional sign constraints are systematically varied to com- transport configurations. pare the sensitivities of canard and conventional con- This “new” OPDOT program has been used to figurations to a variety of technology assumptions. conduct a study identifying some of the critical de- Main-landing-gear location and canard surface high- sign parameters of a statically stable medium-range lift performance are identified as critical design pa- canard-configured transport and a similarly config- rameters for a statically stable, subsonic, canard- ured tandem-wing transport. A conventional, stati- configured transport. cally stable medium-range transport was also studied in parallel to provide a performance benchmark with Introduction which canard configurations and tandem-wing con- As new technologies or major configuration figurations could be compared. Each configuration changes are proposed for incorporation into was evaluated and optimized on the basis of an eco- design, a debate is usually waged on the relative nomic performance index of interest to the airlines. merits of changing current design practice. Often This performance index-the income required per this debate is not held when obvious technological flight for a fixed return on investment-is a straight- advances are being proposed. However, proposals forward measure of the cost of operating a transport are periodically made which cause exhaustive debate aircraft. The results of the study, in which a canard within the technical community. Such is the case configuration was compared with a conventional con- with the current debate concerning the relative mer- figuration, are presented in this paper. its of canard configurations versus conventional aft- tail configurations. Historically, early canard designs Symbols suffered pitch divergence problems (;.e., they were unstable); the sometimes fatal consequences of fly- AR aspect ratio ing these aircraft caused most designers to avoid ca- nard configurations (ref. 1). The recent success of B* maximum Brequet range factor canard-configured ultralight and homebuilt aircraft has revived the canard argument. Both sides in this CAS aircraft purchase price, U.S.dollars debate have sought to prove the general superior- CL lift coefficient ity of either a canard configuration or a conveiitiooal aft-tail configuration. The intuitive appeal of using CL, lift-curve siope, per radiao a lifting surface for trim is undeniable, but analyses cm pitching-moment coefficient and comparisons of canard and conventional config- urations are far more complex. The current consen- cm,o wing-body zero-lift pitching- sus is that canard and conventional configurations moment coefficient can only be fairly compared with one another for a c.g. center of gravity given mission or missions. For example, a conven- tional medium-range transport can hardly be com- DOC direct operating cost, U.S. dollars pared with a canard-equipped two-seat sport aircraft per hour such as the Rutan VariEze. To compare canard and conventional configura- FARE income required per flight for a tions, a systematic methodology must be developed. fixed ROI, US. dollars At the Langley Research Center, a preliminary trans- hc, cruise altitude, ft port aircraft design tool has been developed to reli- ably assess the potential payoffs of various new tech- IOC indirect operating cost, US. dollars nologies (ref. 2). This tool is the computer program per hour OPDOT (Optimum Preliminary Design of Trans- k factor for converting hourly income ports). The OPDOT program has been success- to income per flight fully used to perform studies evaluating the follow- ing: (1) the sensitivity of a transport design to re- L lift, normalized by dynamic pres- laxed static-stability augmentation systems; (2) the sure, ft2 lift-drag ratio st stabilizer, horizontal trimming length, ft surface cruise Mach number T1 thrust line

wing-body zero-lift pitching mo- W wing ment, normalized by dynamic pres- sure, ft3 Survey of Canard Research mean aerodynamic chord Over the years, a variety of research has been nosewheel steering traction margin, conducted on canard configurations. Reference 1 percent MAC gives a fairly comprehensive historical overview of the published research to date. Many classes of aircraft annual return on investment, have been the subjects of canard research, including percent fighters, supersonic bombers, and general aviation surface area, ft2 aircraft. After initial use by the Wright brothers, the canard arrangement has not been widely considered minimum static margin, 1OO(x,, - or used until recently. x&), percent, References 7 and 8 are representative of the wind installed thrust, lb tunnel research conducted on canard configurations. Reference 7 describes a study of various planforms at annual tax rate supersonic speeds. The planforms examined are typ- annual utilization, hour ical of fighters or supersonic bombers. This paper, one of many written about supersonic canard con- passenger volume, ft3 figurations during that time period, concludes that aircraft weight, lb the trim drag characteristics of a canard configura- tion are superior to a conventional configuration and longitudinal distance from aircraft that any interference effects created by the canard are c.g., ft minor or can be alleviated with proper design. Ref- erence 8 is a compilation of wind tunnel data supple- longitudinal position of c.g., percent menting previous NASA reports which conclude that MAC close-coupled canard configurations provide substan- longitudinal position of aftmost c.g., tial improvements in fighter maneuverability. percent MAC References 9 to 13 describe typical theoretical re- search examining canard configurations. Reference 9 main-landing-gear longitudinal compares canard and conventional configurations by position, percent MAC using minimum drag with trim and static-stability longitudinal position of stick-fixed constraints and concludes that conventional configu- neutral point, percent MAC, where rations have generally lower drag than canard config-

dCm/dCL 10 urations. Further, it notes that most canard configu- rations (except for those with near-zero span ratios) horizontal-stabilizer longitudinal are more sensitive (in induced drag) to static-margin position, percent 1, variations. vertical distance from aircraft c.g., References 10 and 11 are similar in scope and ft content. The research described in these reports compares canard and conventional configurations by , deg using drag with static-stability constraints and in- coefficient of rolling friction cluding weight effects. Reference 10 concludes that the superiority of conventional configurations has a Subscripts: sound theoretical basis, although conventional con- av available figuration performance could be improved. Also, this paper notes that analysis techniques which account max maxi mum for “roll up’’ of the canard-surface flow onto the wing k are not required for typical “long-coupled” canard configurations (those with a streamwise gap of about req required 400 percent MAC), since the vortex position on the

2 wing does not affect interference drag by more than and concludes that although conventional configura- a few percent. tions have a lower minimum drag, canard configura- Reference 12 builds on the work of references 9 tions are less sensitive to off-optimum conditions. and 14 and concludes that a close-coupled three- Reference 22 is a preliminary design investigation surface configuration (configuration with both a ca- of a 30-passenger commuter configuration with either nard and an aft tail) is superior to either close- a conventional aft tail, a canard surface, or both (a coupled canard or conventional configurations. This three-surface configuration). This paper concludes paper also notes the theoretical basis for modifying that a three-surface configuration is superior to either the classic Prandtl-Munk theory to properly account a canard or a conventional configuration. The paper for the effect of the canard-surface downwash on the also notes that it is impractical to design a high-wing- wing. This modification results in an induced-drag loading canard configuration with inherent static reduction of approximately 5 to 10 percent from that stability. predicted with the classic theory. This induced-drag The research described in the present paper, reduction is also indicated from wind tunnel testing which compares conventional aft-tail configurations (ref. 15). with canard configurations, is unique because of the Reference 13 describes theoretical research of two following: interacting lifting surfaces. The effects on minimum 1. This research is the first comprehensive design induced drag are examined for various lift distribu- study of subsonic, medium-range transport aircraft tions and span ratios. The effect of gap ratio is also with canards. examined. No consideration is given to static sta- 2. The choice of a commercial-transport-class air- bility or to other constraints. This paper concludes craft enables, for the first time, the use of economic that canard configurations require a nonzero gap ra- performance measures as the design figure of merit tio sufficiently large so the canard surface can “carry during trade studies. its share” at, the minimum-drag condition. 3. The research includes all effects normally re- References 16 to 22 describe wind tunnel and quired at the preliminary design stage, such as in- theoretical or design studies of canards applied to duced drag, parasite drag, trim drag, weight, stabil- general aviation or commuter-class aircraft. Ref- ity, and near-optimal mission profiles. Only refer- erence 16 examines a modified version of a popu- ences 20 and 22 include nearly this level of detail. lar homebuilt kit aircraft described in reference 17. 4. A systematic optimization scheme is used to The impact on performance of loss of laminar flow find the “best” canard configuration for the given is noted, as is the effectiveness of the canard sur- mission and technology constraints. face in limiting angle of attack and increasing stall- 5. This is the first reported research in which departure resistance. Reference 18 describes the stall conventional aft-tail and canard configurations are characteristics and vehicle stability fcr :rarious angles designed for the same mission with the same con- of attack of a canard-configured general aviation pro- straints. Fair comparisons are then possibie for the totype. The sensitivity of the configuration to c.g. class of aircraft considered. movement and power effects is significant. References 19 to 21 describe aerodynamic eval- This research indicates that there are two critical uations of canard and conventional general aviation design problems for canard configurations. One is configurations. Reference 19 is a study of the generic well understood and recognized in the literature but effects on drag of stagger and decalage (difference severely limits the performance of this class of air- between the angles of incidence of the wing and the craft. The other design problem is unique to this canard) of a close-coupled canard configuration. Ref- class of aircraft. Both problems are detailed in this erence 20 examines a 6-passenger and a 12-passenger paper. general aviation configuration with either a conven- tional aft tail or a canard, and it concludes that Method of Calculation general aviation configurations with “stall-proof” General canards cannot outperform conventional configura- tions. In this study, minimum drag is measured as The general optimization scheme for OPDOT is various geometric parameters are changed. The val- shown in figure 1. Nominal values for a set of inde- ues for aspect ratio and stagger are comparable to pendent design variables are used as input along with jet transport values, but wing loading is limited to the required design constants for specifying fixed ge- 60 lb/ft2 and the gap ratios used may not be achiev- ometries, mission economic factors, mission profile able by transport designs. Reference 21 is a paramet- data, and the nonlinear aerodynamic terms. The ric study of a general aviation canard configuration nine independent design variables chosen for this

3 study are shown in table I along with the allowable and an initially feasible solution to start the opti- ranges, which act as directly applied side constraints. mization scheme may be difficult to locate. A direct, Wing area and wing aspect ratio are selected as sequential-search simplex algorithm is used to over- design variables since they have the most impact come these difficulties (refs. 3, 23, and 24). During upon the aircraft sizing. Horizontal-stabilizer siz- the iteration, the optimizer routine which contains ing is accomplished by including horizontal-stabilizer the sequential simplex algorithm sends the values of area, horizontal-stabilizer aspect ratio, horizontal- the independent design variables and the design con- stabilizer longitudinal position, center-of-gravity po- stants to the performance-index evaluation routines. sition, and main-landing-gear longitudinal position A schematic representation of the calling sequence for as independent design variables. Finally, fuselage the performance-index evaluation routines is shown length and installed thrust are used as independent in figure 2. design variables to match the aircraft size to the mis- sion and to the wing planform. Evaluation of Unaugmented Performance Index The set of independent design variables is incre- Aircraft weight is estimated by iteration. Most rnented by the optimizer logic in an attempt to im- component weight relations, including fuel weight, prove the design. The value of the performance index are functions of gross weight as well as geometry. The (parameter to be optimized) is determined from the design mission is simulated and repeated until the hy- values of the independent design variables and from pothesized gross take-off weight at the beginning of a information in the communications module. This in- weight iteration approaches the sum of the individual dex is selected from a list of possible performance in- aircraft component weights, the payload, and the fuel dices. The performance index used for these studies weight. The convergence criterion for the weight iter- is the income required per flight for a fixed return on ation is chosen to ensure numerical accuracy, which is investment (FARE). The merits of using this index important for optimizer performance. However, the are described in references 3 and 4. relations used to compute weights are from indus- The constraint functions (involving inequality re- try statistics and are only expected to be accurate to lationships) represent operational, flying-qualities, within 10 percent. Industry statistics for the aircraft and performance constraints and are based upon cer- component weights come from references 25 to 28 and tification regulations, mission definition, and practi- are functions of all the independent design variables, cal considerations. It should be noted that no flying- the gross take-off weight, and approximately 20 of qualities constraints are considered in this study. the design constants input through the communica- Constraints are integrated into the optimization pro- tions module. For canard configurations, the statis- cess by adding a penalty to the performance index tical weight relations are modified such that the wing for each constraint violation. Each penalty term is weight and the horizontal-stabilizer weight reflect the proportional to the square of the violation times a division of lift load between these two surfaces. The weighting factor. The performance index plus these fuel weight is calculated by summing estimates of the penalty terms form an augmented performance func- required fuel for the following mission segments: taxi, tion. If the weighting factor is sufficiently large, take-off and climb, cruise, descent, and reserve. minimizing the augmented performance function is The mission profile as modeled is shown in fig- equivalent to finding the minimum performance in- ure 3. It consists primarily of a multiple-step cruise- dex while satisfying all the constraints. climb approximation of an optimal fuel profile. The cruise portion is broken into 10 equally spaced seg- Optimization Technique ments, and Breguet-type relationships are used for The numerical optimization logic, which iterates calculating the amount of fuel burned during each the independent design variables to minimize the segment (ref. 25). Comparisons of this discrete augmented performance function, is a subject of in- flight path with continuously optimal flight profiles tense research in nearly all fields of engineering. Pre- (ref. 29) show differences of less than 5 percent. vious studies and the authors’ experience indicate Thus, the Breguet-type relationships are good ap- that various gradient methods suffer from numerical proximations for comparative design studies. Engine difficulties when analytical equations are not avail- performancc is computed based on the TF39-GE-1 able to provide the gradients and also from initial- engine used on the C-5A aircraft (ref. 29). The base- ization problems when the number of active con- line engine is “rubberized” as required with standard straints is large with respect to the number of in- scaling laws to achieve the required installed thrust deprndcnt design variables. When aircraft design for the prescribed mission. is posrd as a numerical optimization problem, the Parasite drag for all flight phases is calculated analytical gradients generally will not be available from a component buildup including compressibil-

4 ity and Reynolds number effects using the methods ing hourly income to income per flight, converts this of references 25 and 28 to 31. Induced drag for all result to income required per flight (FARE). phases of flight is estimated by using nonlinear cor- The absolute accuracy of the performance index rections to parabolic drag polars for airfoil-section (in this case FARE) is only expected to be approx- camber (ref. 32) and by adding terms for the tail in- imately 10 percent. In reality, the accuracy may be duced drag and wing-tail interference drag (ref. 9). even less, since any systematic computation of an air- Wind tunnel results and theoretical analysis (refs. 10, craft purchase price based on the costs incurred by 12, and 15) indicate the total induced drag for canard the manufacturer is impossible. However, with the configurations similar to the type considered here will level of detail included in OPDOT, the relative accu- be up to 10 percent less than that indicated by con- racy is quite good for comparing “similar” configu- ventional theoretical methods. Therefore, the total rations. This observation has been borne out by pre- induced drag for canard configurations is empirically vious studies (refs. 3 to 6) and by comparing results reduced 10 percent from that estimated with the from OPDOT with an independent study (ref. 44). above methods (refs. 9 and 32). Calculations of sta- bility and control derivatives for all flight phases are OPDOT Baselines typical of those used in preliminary design (refs. 33 To provide a basis for performing the trade stud- and 34) and include empirical adjustments from aero- ies, a baseline mission is chosen. Table I1 lists the dynamic wind tunnel and flight data (refs. 35 to 39) design constants chosen for the baseline mission that for compressibility, elasticity, and the use of super- are used along with the independent design variables critical airfoil sections. Ground-effect calculations and constraint functions listed in table I. Note that used in the take-off control-power calculations are each of the baseline configurations (see table 111)- from reference 33. A trim routine is used for deter- conventional, canard, and tandem wing-have at mining the wing and tail loads in cruise, take-off, and least a 10-percent-stable static margin. approach phases of flight. This trim routine is simpli- Several design constants have different values for fied from the routine used in reference 2 by assuming the canard and conventional baseline configurations. that the resultant drag vector acts through the air- (See table 11.) Specifically, these dcsign const,ants craft c.g., yielding no contribution to aircraft pitching are the wing and horizontal-stabilizer sweep angles, moment. This assumption permits the “drag terms” the horizontal-stabilizer height above centerline, and in the longitudinal trim equation to be ignored; the the wing longitudinal position along the fuselage. resulting trim equation is referred to as the “simple Originally, both sweep angles and the wing posi- trim equation” in subsequent discussions. Through- tion were treated as independent design variables. out the flight envelope, the simple trim routine yields However, initial studies showed that these three de- results within 10 percent of the results from a more sign parameters were insensitive to constraint and exact, computationally intensive routine described in mission changes (typically the values for these pa- reference 2. rameters varied by less than 5 percent). There- The cost data are approximated from industry fore, the wing and horizontal-stabilizer sweep an- statistics for manufacturing, maintenance, and other gles and the wing longitudinal position are made de- components of direct operating costs as well as the sign constants at their nominal values, which differ indirect operating costs (refs. 25, 27, and 40 to 43). for the canard and conventional configurations. The The direct operating cost is an augmented form of horizontal-stabilizer height is chosen from geometric the industry standard (ref. 3). FARE is calculated in considerations. For the conventional configuration, the following manner: the horizontal-stabilizer root is placed about 2 ft below the vertical-stabilizer-fuselage juncture. For the canard configuration, the horizontal stabilizer is placed such that it is above the wing yet below the passenger-cabin floor. Although this position is not Desired return on investment (ROI) in percent, mul- aerodynamically optimal, it is expected to have lit- tiplied by the aircraft purchase price CAS minus the tle adverse impact for the long-coupled (longitudinal 10 percent investment tax credit is the annual return separation distances in excess of 300 percent MAC) on investment in dollars per year. Dividing this term canard configuration considered (ref. 15). by the quantity 1 minus the tax rate tr times the an- The sensitivity of a transport design to variations nual utilization U yields the pretax hourly ROI, or in mission definition and design constraints is deter- profit. Adding direct operating cost (DOC) and indi- mined for these studies by using OPDOT in the man- rect operating cost (IOC) gives the income required ner shown in figure 4. Design variables, design con- per hour, and multiplying by k, a factor for convert- stants, and design constraints are input to OPDOT,

5 which generates an optimal conceptual design. Then, results from appendix A, the allowable CL,~~~of the a design constant or design constraint of interest is canard surface is increased to the value allowed for changed and OPDOT generates a new resized de- wing CL,max,which is close to the maximum achiev- sign, optimized to minimize FARE. This process is able CLfor a given lifting surface with a triple-slotted repeated until a set of designs have been generated . for a given design parameter. The sensitivity of the design to the varied design parameter is determined Results and Discussion by examining the variation of optimum, or converged, A study of the effect of varying landing and take- FARE values as the design parameter is changed. off field lengths is conducted to compare conven- For the purpose of comparison, baseline concep- tional and canard configurations, since landing and tual designs are developed for the conventional, ca- take-off field length is one of the most critical mis- nard, and tandem-wing configurations. Table I11 sion constraints with respect to configuration sizing shows the values of the design variables for these (refs. 3 and 32). The sensitivity of all configurations baseline conceptual designs. Figures 5 to 7 show, re- to changes in take-off and landing field-length con- spectively, graphics output from OPDOT of the base- straints is shown in figure 8. As expected, designing line conventional configuration, the baseline canard for shorter field lengths increases FARE, decreasing configuration, and the baseline tandem-wing config- the economic efficiency of the design. The sensitiv- uration. The tandem-wing configuration is similar ity of the tandem-wing configuration to variations in to the canard configuration. However, for a tandem- field length approximates the sensitivity of the ca- wing configuration the two lifting surfaces are geo- nard configuration for all field lengths. This is ex- metrically identical. This means that for each lifting pected because a tandem-wing configuration is essen- surface, the aspect ratio, the taper ratio, the thick- tially a special type of a canard configuration. Note ness ratio, and the area are equivalent. It has been that the tandem-wing configuration requires about argued that economic savings will result from this 2 percent more FARE (about $800 per flight) than design because of reduced tooling requirements and the canard configuration. This would seem to indi- production-learning-curve considerations, but these cate that a tandem-wing configuration is a less effi- savings have not been considered in this study. For cient design than a canard configuration, unless the the conventional configuration, two baselines were postulated production savings are large enough to initially developed. One baseline conventional con- offset this lower efficiency. The canard configuration figuration was sized with zlg (the main-landing-gear suffers incrementally less in FARE penalties than position) as a design variable free to move longitu- does the conventional configuration for the shorter dinally along the fuselage to a location determined field lengths. A canard configuration requires at to bc optimum by OPDOT as long as it remained least 2 percent less FARE than a conventional trans- 20 percent MAC behind "5,. The other conven- port design for all field lengths. This result can be tional baseline was sized with qgheld constant at misleading because of several unrealistic (but neces- 65 percent MAC. This is the position selected during sary) assumptions that are made inherently within the initial Integrated Application of Active Controls OPDOT to obtain satisfactory convergence for ca- (IAAC) studies (ref. 44). The differences between nard designs. these two baseline conventional configurations were The first of these inherent assumptions is that very small over a wide range of design conditions; the canard-surface CL,~~~is permitted to exceed a therefore, only the conventional configuration with value of 0.8 without any cost, weight, or drag penal- constant qgis considered in this paper. ties. Recall that the weight and drag equations are Initial OPDOT runs indicated that the program only modified to account for the lift load carried by would not converge to a constrained solution for a the canard surface. Addition of a flap system to the canard configuration. It was observed that the ap- canard surface, necessary for CL,max exceeding ap- proach CL for canard configurations was abnormally proximately 1.0, will invoke cost, weight, and drag low. Analysis of a generic canard configuration was penalties. Typical industry practice is to constrain performed in order to gain insight into the problem. the CL,~~~of the horizontal stabilizer to less than 0.8 Appendix A indicates that by considering a simple for all trimmed-flight conditions. The impact of vary- trim equation and a simple static-stability equation, ing canard-surface CL,maxis shown in figure 9. Note it is possible to show that the allowable CL,max on the that the selected value for canard-surface CL,,~~~of horizontal stabilizer of a canard configuration must 3.15 for the baseline canard configuration provides exceed that of the wing. The canard-surface CL,~~~approximately the minimum FARE. Achieving this constraint, in OPDOT originally had the same value ~~,,,n;rxrequires a triple-slotted flap system on the as for a conventional configuration. In view of the canard. The data of figure 9 indicate a large penalty

6 I in optimum FARE as canard-surface CL,~~~is re- has been suggested that relaxing minimum-static- I duced, because the horizontal-stabilizer CL,,, effec- margin requirements will benefit a canard configura- tively sizes the overall canard-configuration design. tion more than it will a conventional one (ref. 11). I Clearly, a complex flap system is required for eco- Specifically, if negative static margins are allowed nomically viable canard transport designs. by providing relaxed static-stability augmentation I A second assumption inherent to OPDOT is that systems, the performance of a canard configuration the main landing gear can be moved longitudinally might be significantly improved. Changes in the min- away from the wing structure with no weight, cost, imum static margin are accomplished by changing l or drag penalties. The weight and cost penalties are the static-margin constraint and allowing OPDOT I associated with designing and producing a transport to resize each configuration to take advantage (or, aircraft for which the main gear is dislocated from the for changes to more stable static margins, minimize wing structure and, therefore, cannot take structural the disadvantage) of the new static-stability require- : advantage from attaching to it. The drag penalty is ment. The impact of relaxing the static-stability re- associated with adding pods or to house a quirement on canard, conventional, and tandem-wing landing-gear system not located in the wing (such as configurations is shown in figure 13. The percent I for a C-130 transport). The sensitivity of the canard change in FARE is used as the measure of sensitivity configuration to main-landing-gear position is shown to static-margin variations. For minimum static mar- in figure 10. Notice the extreme sensitivity of FARE gins less than 5 percent stable, the canard configu- as the main gear is moved toward the wing. This ration shows less sensitivity to static-margin changes is due to the sensitivity of the take-off control-power than does the conventional configuration. calculation to the distance between the main gear Examination of the canard configuration shows and the aircraft c.g., especially for canard configura- that as the minimum static margin is decreased to tions. For a heuristic discussion of take-off control- values less than 5 percent stable, the CL of the canard I power requirements, see appendix B. OPDOT can- surface decreases, which causes the wing CL to in- not find a satisfactory canard configuration with a crease to maintain overall CL. The resulting increase main-landing-gear position aft of -70 percent MAC. in wing induced drag is greater than the decrease in ~ This position is still slightly forward of the wing-root canard-surface induced drag, causing a slight increase , so that the main landing gear will be in overall induced drag. This drag increase causes I remotely located from the wing structure, as stated compromises in the design, decreasing the sensitivity above. of the canard configuration to relaxation of the static- I With the assumption that future studies might margin constraint; therefore, unstable static margins I establish the cumulative weight penalties inherent will not be as beneficial for canard configurations as in providing an adequate canard-surface flap system for conventional configurations. and rer?lotely locating the main landing gear from The tandem-wing configuration shows little sen- the wing, an attempt is made to document the irn- sitivity t~ changes in minimum static margin down pact of additional weight and drag on the canard to 2 percent static stability. Forcing the minimum configuration. Figure 11 shows the increase in FARE static margin below 2 percent stable has an extremely as the weight is increased from the baseline config- adverse economic effect on the design. This occurs uration. Note that each data point represents a re- because it is necessary to move the aircraft c.g. well sized design after the weight penalty is added. If aft of the baseline position to achieve negative static the weight penalty exceeds approximately 2500 lb margins. This c.g. movement, coupled with the need (approximately 1 percent of maximum gross take- for aircraft trim, requires adjustments of CLbetween off weight), FARE for the canard configuration ex- the two lifting surfaces similar to those that caused ceeds FARE for the baseline conventional configura- compromises in the canard configuration. However, I tion. Cost penalties due to design and certification for the tandem-wing configuration with its limited of a high-lift canard surface and the main landing degrees of freedom in the design variables, this CL gear dislocated from the wing structure will further adjustment is far more critical. penalize the canard configuration. The sensitivity of Variations in canard-surface sweep angle are canard and conventional transport designs to drag made to verify the observed insensitivity of the de- variations is documented in figure 12. The drag vari- sign to this parameter. (See “OPDOT Baselines” ations are obtained by varying only the wing parasite section.) The effect of varying canard-surface sweep drag. Variations in drag affect both conventional and angle is shown in figure 14. Varying the sweep ap canard configurations equally. pears to have little impact on FARE. Reduction in Up to this point, all sensitivity studies have been horizontal-stabilizer sweep angle increases stabilizer conducted with statically stable configurations. It CL, and allows a size reduction of the horizontal sta-

7 bilizer and a potential reduction in FARE. However, fuselage in this area may need to be strengthened. the increases in compressibility drag that result as This structural reinforcement will create a weight the sweep angle is reduced balance the beneficial ef- penalty. Other penalties may also exist and would fects of increasing stabilizer CL-. require identification and assessment. Given these design problems, it was difficult dur- Concluding Remarks ing the course of this study to find an economically The research described in this paper compares a feasible design for a canard configuration. If both the canard configuration and a conventional aft-tail con- CL,~~~limit on the control surface and the landing- figuration for a subsonic, medium-range transport gear placement problems are ignored, the canard con- aircraft. The choice of a commercial transport al- figuration has approximately 2-percent better perfor- lows the use of an economic performance measure as mance. This benefit is traceable to the criticality of a design figure of merit during trade studies. The achieving high lift during take-off and landing and is design program contains an iterative optimization observable even though it is generally accepted that scheme which includes induced drag, parasite drag, a canard configuration will have higher induced drag. trim drag, weight, variable flight paths, and mission A special case of tandem-winged transports was parameters. This level of detail in the optimization studied, and this configuration was shown to be eco- scheme, when applied uniformly to a canard con- nomically inferior to the canard configuration and to figuration and to a conventional aft-tail configura- the conventional aft-tail configuration. It seems un- tion, allows for a fair comparison for this class of likely that the hypothetical benefits of manufacturing aircraft. Two design parameters which are particu- two “identical” lifting surfaces would be substantial larly sensitive for this class of aircraft are identified enough to permit economically successful medium- and studied. These parameters are the maximum lift range transports with tandem wings. coefficient CL,~~~of the horizontal stabilizer and the The ability to design medium-range transports main-landing-gear longitudinal position. with unstable static margins will apparently not ben- The requirement of achieving a large CL,max is efit a canard configuration as much as previously imposed on the horizontal stabilizer of a statically thought because of compromises forced by configu- stable canard configuration for aircraft trim. This ration induced drag penalties. In fact, conventional is particularly critical for transport aircraft since an configurations show more incremental economic ben- efficient design must achieve a very high CL during efits from unstable static margins than do canard approach and landing. Even unstable canard con- configurations. figurations will require a much higher canard-surface The results of this study tend to indicate that CL,,~~~than that of current transport horizontal sta- conventional aft- tail configurations are economically bilizers. The required large CL,max will dictate use of superior to canard configurations for this class of a high-lift system that must provide control-surface aircraft. Even when constraints are impractically modulation as well as contribute to the overall air- reduced, the resulting canard configuration is only craft lift. Such a high-lift control surface suitable for slightly better than a comparable conventional de- a commercial transport has neither been designed nor sign. A possible alternative which could be used to demonstrated. solve these design problems while maintaining the The main-landing-gear position for a canard con- benefits of both configurations would be a configura- figuration will be forward of the wing-root leading tion with three lifting surfaces. Further research is edge. This requirement is dictated by the control- needed to evaluate this transport configuration with power requirement for nose-gear unstick during take- a wing, a canard, and an aft-tail. off. The main gear in this position will require a pod or a to house the stowed gear, creating a drag penalty. Further, since the landing loads will not NASA Langley Research Center be easily transferred to the wing structure, the Hampton, VA 23665 November 29, 1984

8 Appendix A where L is lift normalized by dynamic pressure; thus, equation (A3) becomes

Approximation of Required Horizontal- - sst Xst -CL,st - ZwCL,w = 0 Stabilizer CL,max for a Statically Stable SW Canard Configuration or zstsst - CL,w The required horizontal-stabilizer CL of a trimmed, zwsw CL,st ! statically stable canard configuration may be approxi- mated as follows. Figure A1 is a schematic of a generic From equation (A2), static stability for a trimmed canard configuration (fig. Al) can be expressed as

zw ALw > Z,t ALst (-45) Since AL = CL,S Aa substitution into equation (A5) gives

CLQ,w sw5, > CL, ,st sst zst

Figure Al. Schematic of trimmed canard configuration. or

canard-configured transport. A simple trim equation Substituting the result of equation (A4) yields which ignores drag terms is

Zs+,Lst - Zw Lw + M, = 0 (Ai)

where LSt, L,, and M, are normalized by dynamic or pressure. Assume that Cm,o= 0; therefore, M, = 0. Then, from equation (Al),

ZstLst - ZwLw = 0 (A'4 Typically, Cm,, is negative (pitch-down condition). By examining figure All we see that this condition requires Dividing equation (A2) by Sw gives even more CL,st for trim. If we assume CL,,., CL,,, , then for a statically stable canard transport coiifiguratiGn, CL,st > CL,w ('48) for all trimmed-flight regimes. Therefore,

Note that CL,st,rnax > CL,w,max L CL = - S

9 Appendix B same position relative to the aircraft c.g. to maintain nosewheel traction, this contribution to the moment about the c.g. remains about the same as for the con- Qualitative Examination of Take-Off Control- ventional configuration. Also, the thrust contribution Power Requirement and the wing-body zero-lift pitching-moment contribu- Figure B1 is a schematic of the physics of the tion will be about the same as for a conventional con- take-off control-power requirement for a conventional figuration. However, since the wing is aft of the c.g., configuration. From an examination of the moments the pitching-moment contribution due to wing lift will about the c.g., it is seen that to rotate the aircraft, the be downward. Therefore, the canard surface will have horizontal stabilizer must provide a downward load to to provide enough lift to balance the pitch-down contri- overcome the pitch-down moments generated by wing- butions of wing-body zero-lift moment, reaction at the body zero-lift moment (M, < 0) plus the reaction at landing gear, & wing lift, less the pitch-up contribu- the landing gear. Note that some pitch-up moment tion of engine thrust. The lift demands placed on the is already being provided by wing lift and the thrust canard surface far exceed the downward load demands vector. placed on a conventional configuration’s horizontal sta- A schematic for a canard configuration is shown in bilizer. This is why the take-off control-power require- figure B2. Since the main landing gear must be in the ment is critical for canard configurations.

Lw Lst t !

w-Lw -Lst

Figure B1. Schematic of conventional configuration Figure B2. Schematic of canard configuration during during take-off roll. take-off roll.

10 References B. Laschka and R. Staufenbiel, eds., 1982, pp. 1470- 1488. (Available as ICAS-82-6.8.2.) Burns, B. R. A,: Were the Wrights Right? Air Int., '. 17, ~~t~~,Burt: Development of a Small High-Aspect- vol. 25, no. 6, Dec. 1983, pp. 285-292, 308-309. Ratio Canard Aircraft. 1976 Report to the Aerospace 2. Sliwa, Steven M.; and Arbuckle, P. Douglas: OPDOT: A Profession, Tech. Rev., vol. 13, no. 2, SOC.Exp. Test Computer Program for the Optimum Preliminary Design Pilots, 1976, pp. 93-101. of a fiansport Airplane. NASA TM-81857, 1980. 18. Chambers, Joseph R.; Yip, Long P.; and Mod, 3. Sliwa, Steven M.: Use of Constrained Optimization in the Thomas M.: Wind-Tunnel Investigation of an Advanced Conceptual Design of a Medium-Range Subsonic Bans- General Aviation Canard Configuration. NASA TM- port. NASA TP-1762, 1980. 85760, 1984. 4. Sliwa, Steven M.: Sensitivity of the Optimal Preliminary 19. Selberg, B. P.: Aerodynamic Investigation of Closely Design of a Transport to Operational Constraints and Coupled Lifting Surfaces With Positive and Negative Performance Index. AIAA-80-1895, Aug. 1980. Stagger for General Aviation Applications. AIAA-83- 5. Sliwa, Steven M.: Economic Evaluation of Flying- 0057, Jan. 1983. Qualities Design Criteria for a Bansport Configured With 20. Keith, Michael W.; and Selberg, Bruce P.: Aerodynamic Relazed Static Stability. NASA TP-1760, 1980. Optimization, Comparison, and Trim Design of Canard 6. Sliwa, Steven M.: Impact of Longitudinal Flying Quali- and Conventional High Performance General Aviation ties Upon the Design of a Transport With Active Con- Configurations. AIAA-83-0058, Jan. 1983. trols. A Collection of Technical Papers-AIAA Atmo- 21. Keith, Michael W.; and Selberg, Bruce P.: Aerodynamic spheric Flight Mechanics Conference, Aug. 1980, pp. 133- Canard/Wing Parametric Analysis for General Aviation 141. (Available as AIAA-80-1570.) Applications. AIAA-84-0560, Jan. 1984. 7. Hall, Charles F.; and Boyd, John W.: Effects of Canards 22. Srivatsan, R.; and Roskam, Jan: Unconventional Com- on Airplane Performance and Stability, NACA RM muter Configurations: A Design Investigation. SAE A58D24, 1958. Tech. Paper Ser. 830710, Apr. 1983. 8. Gloss, Blair B.; Ray, Edward J.; and Washburn, for Karen E.: Effect of Canard Vertical Location, Size, 23. Olsson, D. M.: A Sequential Simplex Program Solv- ing Minimization Problems. Qual. Technol., vol. 6, a.nd Deflection on Canard- Wing Interference at Subsonic J. no. 1, Jan. 1974, pp. 53-57. Speeds. NASA TM-78790, 1978. 9, ~~~~~~hli~,~il~~~D.: ~~l~~l~~i~~~,and ~~~~~~i~~~ 24. Olsson, Donald M.; and Nelson, Lloyd s.: The Neldei- With an Ideal Minimum, of fiimmed Drag for Conven- Mead Simplex Procedure for Function Minimization. tional and Canard Configurations Having Various Levels Technometrics, vol. 17, no. 1, Feb. 1975, pp. 45-51. of Static Stability. NASA TN D-8391, 1977. 25. Nicolai, Leland M.: Fundamentals of Aircraft Design. 10. McGeer, Tad; and Kroo, Ilan: A Fundamental Com- School of Eng., Univ. of Dayton, Dayton, Ohio, c.1975. parison of Canard and Conventional Configurations. J. 26. Oman, B. H.: Vehicle Design Evaluation Program. Aircr., vol. 20, no. 11, Nov. 1983, pp. 983-992. NASA CR-145070, 1977. 11. Kroo, I. M.; and McGeer, T.: Optimization of Canard 27. Anderson, R. D.; Flora, C. C.; Nelson, R. M.; Raymond, Configurations-An Integrated Approach and Practi- E. T.; and Vincent, J. H.: Development of Weight and cal Drag Estimation Method. iCAS Proceedings--2082, Cost Estimates for Lifting Surfaces With Active Controls. B. Laschka and R. Staufenbiel, eds., 1982, pp. 1459- NASA CR-144937, 1976. 1469. (Available as ICAS-82-6.8.1.) 28. Kyser, Albert C.: An Elementary Analysis of the Effect 12. Butler, G. F.: An Analytical Study of the Induced of Sweep, Mach Number, and Lift Coeficient on Wing- Drag of Canard-Wing-Tail Aircraft Configurations With Structure Weight. NASA TM-74072, 1977. Various Levels of Static Stability. Aeronaut. J., vol. 87, 29. Aggarwal, R.; et al.: An Analysis of Fuel conserv- no. 868, Oct. 1983, pp. 293-300. ing Operational Procedures and Design Modifications for 13. Laitone, E. V.: Prandtl's Biplane Theory Applied to Bomber/Transport Aircraft- Volume II. AFFDL-TR-78- Canard and Tandem Aircraft. J. Aircr., vol. 17, no. 4, 96, Vol. 11, U.S. Air Force, July 1978. (Available from Apr. 1980, pp. 233-237. DTIC as AD A062 609.) 14. Laitone, E. v.: Positive Tail Loads for Minimum In- 30. Perkins, Courtland D.; and Hage, Robert E.: Airplane duced Drag of Subsonic Aircraft. J. Aircr., vol. 15, Performance Stability and Control. John Wiley & Sons, no. 12, Dec. 1978, pp. 837-842. Inc., c.1949. 15. Feistel, T. W.; Corsiglia, v. R.; and Levin D. B.: Wind- 31. Hoerner, Sighard F.: Fluid-Dynamic Drag. Hoerner Tunnel Measurements of Wing-Canard Interference and Fluid Dynamics (Brick Town, N.J.), c.1965. a Comparison With Various Theories. SAE 1981 fians- 32. Loftin, Laurence K., Jr.: Subsonic Aircraft: Evolution actions, Section 2, Volume 90, 1982, pp. 202G-2039. and the Matching of Size to Performance. NASA RP- (Available as SAE Paper 810575.) 1060, 1980. 16. Yip, Long P.; and Coy, Paul F.: Wind-Tunnel Investiga- 33. USAF Stability and Control Datcom. Contracts AF33 tion of a Full-scale Canard-Configured General Aviation (616)-6460 and F33615-76-C-3061, McDonnell Douglas Aircraft. Proceedings of the 13th Congress of the Inter- Corp., Oct. 1960. (Rev. Apr. 1978.) national Council of the Aeronautical Sciences and AIAA 34. Roskam, Jan: Methods for Estimating Stability and and Technology Conference, Volume 2, Control Derivatives of Conventional Subsonic Airplanes.

11 Pub. by the author (Dep. Aerosp. Eng., Univ. of Kansas, 40. American Airlines: A New Method for Estimating Cur- Lawrence, Kansas), c.1973. rent and Future llfansport Aircraft Operating Economics. 35. Harris, Charles D.: Aerodynamic Characteristics of a 14- NASA CR-145190 (Rev.), 1978. Percent- Thick NASA Supercritical Airfoil Designed for a 41. Maddalon, Dal V.: Estimating Airline Operating Costs. Normal-Force Coeficient of 0.7. NASA TM X-72712, NASA TM-78694, 1978. 1975. 42. Stoessel, Robert F.: A Proposed Standard Method for 36. Bartlett, Dennis W.: Wind-Tunnel Investigation of Sev- Estimating Airline Indirect Operating Ezpense. Rep. No. eral High A spect-Ratio Supercritical Wing Configurations LW70-500R, Lockheed-Georgia Co., May 1970. on a Wide-Body-Type Fuselage. NASA TM X-71996, 43. Assessment of the Application of Advanced Technolo- 1977. gies to Subsonic CTOL llfansport Aircraft. NASA CR- 37. Heffley, R. K.; et al.: Aircraft Handling Qualities Data. 112242, 1973. NASA CR-2144, 1972. 44. Boeing Commercial Airplane Co.: Integrated Application 38. Seckel, Edward: Stability and Control of Airplanes and of Active Controls (IAAC) Technology to an Advanced Helzcopters. Academic Press, Inc., c.1964. Subsonic Pansport Project-Initial ACT Configuration 39. McRuer, Duane; Ashkenas, Irving; and Graham, Design Study, Final Report. NASA CR-159249, 1980. Dunstan: Aircraft Dynamzcs and Automatic Control. Princeton Univ. Press, 1973.

12 TABLE I. DEFINING PARAMETERS FOR OPDOT

(a) Independent design variables

Independent design variable Lower limit Upper limit Wing area, S,, ft2 ...... 1000 4000 3 15 200 3000 2 15 - 500 100 10 000 130 000 100 220

12.5 47.5 70.5 97.5

-250 0 0 200

Available constraint function Baseline function limit Thrust for cruise-climb ...... Tav/Treq 2 1 Second-segment c!imb gradient ...... Tav/Treq 2 1 Missed-approach climb gradient ...... Tav/Tieq 2 1 Landing field length, ft ...... 5 8000 Take-off field length, ft ...... 5 8000 Nosewheel steering traction ...... zkg 5 zlg - NG Passenger volume ...... Vp,req/Vp,av 5 1 Cruise altitude, ft ...... 30000 5 h,, 5 46000 Cruise wing lift coefficient ...... 5 0.7 Static margin, SM, (cruise and approach), percent MAC . . 5 10 Horizontal-stabilizer longitudinal position, ft aft of nose . . 2 20 Horizontal-stabilizer lift coefficient in approach ...... -0.80 5 CL,~~5 3.15 Nose-gear unstick ...... Lst,av/Lst,req 2 1

13 TABLE I1 . KEY DESIGN CONSTANTS USED FOR DESIGN OPTIMIZATION [Supercritical airfoils and curved windshield assumed]

Mission: Cruise Mach number ...... 0.80 Divergence Mach number ...... 0.84 Design range, n.mi...... 3000 Number of seats ...... 200 Cargo, lb ...... 7500 Maximum lift coefficient ...... 3.15 Geometry: Wing longitudinal position. percent lf : Conventional configuration ...... 50 Canard configuration ...... 72 Wing sweep angle. deg: Conventional configuration ...... 25 Canard configuration ...... 20 Horizontal-stabilizer sweep angle. deg: Conventional configuration ...... 35 Canard configuration ...... 22 Height of horizontal stabilizer above fuselage centerline. ft: Conventional configuration ...... 7.0 Canard configuration ...... -2.5 Wing thickness ratio ...... 0.14 Wing taper ratio ...... 0.38 Wing incidence angle. deg ...... 2 Wing geometric twist. deg ...... 5 Horizontal-stabilizer thickness ratio ...... 0.10 Horizontal-stabilizer taper ratio ...... 0.4 Vertical-stabilizer sweep. deg ...... 35 Ratio of area to vertical-stabilizer area ...... 0.30 Ratio of chord to horizontal-stabilizer chord ...... 0.25 Ratio of flap span to wing span ...... 0.6 Maximum flap deflection. deg ...... 45 Fuselage diameter. ft ...... 16.67 Distance of thrust vector below c.g., ft ...... 6 Number of engines ...... 2 Wing dihedral. deg ...... 5 Economics: Fuel cost. dollars per gallon ...... 1.00 Load factor ...... 0.55 Utilization rate. hours per year ...... 3200 Depreciation period. years ...... 14 Residual value. percent ...... 12 Taxrate ...... 0.48 Year of study ...... 1982 Assumed annual inflation rate ...... 0.07 Number of prototype aircraft ...... 2 Aircraft fleet size ...... 250 Initial production rate. per month ...... 0.5 Full production rate. per month ...... 5 Engineering rate (1974). dollars per hour ...... 19.55 Tooling rate (1974). dollars per hour ...... 14.00

14 TABLE I1 . Concluded Labor rate (1974). dollars per hour ...... 10.90 Engines for test aircraft ...... 3 Ratio of manufacturer’s weight to take-off weight ...... 0.75 Miscellaneous: Maximum dynamic pressure. psf ...... 245.6 Pressurized volume. ft3 ...... 6290 Number of pilots ...... 3 Number of attendants ...... 8 Air conditioning flow rate. lb/min ...... 441 channels ...... 5 Generator capacity. kV-A ...... 750 Maintenance complexity factor ...... 1.6 Hydraulics volume flow rate. gal/min ...... 79 Number of inertial platform systems ...... 1 Ratio of auxiliary-power-unit-on time to engine-on time ...... 0.1 Ratio of first class to economy seating ...... 0.15 Maximum speed. knots ...... 483 Airfoil design lift coefficient ...... 0.5 Baseline engine ...... TF39-GE-1 Elevator servo time constant. sec ...... 0.1

16 TABLE 111. BASELINE CONFIGURATION VARIABLES

Canard Conventional ITandem-wing Variable configuration confieurat ion configuration Wing area, ft2 ...... 1748.2 2015.6 {t-air Wing aspect ratio ...... 8.76 11.22 13.00 Horizontal-stabilizer area, ft2 ...... 769.6 682.0 950.1 Horizontal-stabilizer aspect ratio ...... 7.92 3.79 13.00 Aftmost center of gravity, percent MAC ...... - 182 33 Installed thrust, Ib ...... 66445. 68892. Fuselage length, ft ...... 160.0 164.0 Horizontal-stabilizer longitudinal position, percent If . . 19.03 97.36 Main-landing-gear longitudinal position, percent MAC . . . -114 65*

*A design constant for this configuration.

16 DESIGN Sw, ARw, $,T, DESIGN Technology level VARIABLES SSt, ARSt, c.g., etc. CONSTANTS Nonlinear aerodynamics

Etc. DOC t ROI PERFORMANCE \ FARE INDEX e I b COMMUNI CAT1 ONS L/ D 0 - W I - MODULE Fuel CONSTRAINT ). Etc. FUNCTIONS II - I Engine-out performance t Field length

Geometry constraints Control power Etc.

NO t”’

Figure 1. General optimization scheme for OPDOT.

*

1 ,dl A&MAINTENANCE MANUFACTUR I NG

L CRUISE STEP ] FEATURES 0 INDUSTRY STATISTICS FOR COSTS AND WEIGHTS ENG I NE 0 DATCOM-TYPE STABILITY AND CONTROL DERIVATIVES .MULTIPLE-STEP, SUBOPTIMAL CRUl SE-CLIMB STABILITY AND CONTROL 0 GENERALIZED INTERFERENCE DRAG

Figure 2. Performance-index evaluation routines for OPDOT.

17 Climb to CL for 98 percent at 98 percent Cruise-climb at (LI DImax for Mcr (7-10) Bxfor Mcr(2-5) Climb to CL for - - - B* at Mcr(I.11 I

4 5 6 7 8 9

- ORIGIN

IFigure 3. Mission profile from OPDOT.

F G-, DES IG-, OPDOT I OPT1 MAL (Optimize design based on ' * DESIGN CONSTANTS DESIGNS an economic index) DESIGN CONSTRAINTS

CHANGE A DESIGN CONSTANT OR DESIGN CONSTRAINT

INCOME REQU I RED 40 CONFl GURA ION 1 FIXEDFOR AROI, 351%c\I

$/FLIGHT 30 CONFIGURATION 2 1 I I I I 5000 6000 7000 8000 9000 10000 I FIELD LENGTH, ft I

Figure 4. Methodology for conducting sensitivity studies with OPDOT. Figure 5. Three-view depiction of baseline conventional configuration.

19 Figure 6. Three-view depiction of baseline canard configuration.

20 Figure 7. Three-view depiction of baseline tandem-wing configuration.

21 45 - x lo3 0 CANARD (GEAR FREE, CANARD-SURFACE CL,max = 3.15)

0 TANDEM WING (GEAR FREE, CANARD-SURFACE CL,max =3.15)

0 CONVENT1ONAL ( GEAR FIXED 1 40 X BASELl NE ( ALL CONFl GURATl ONS 1 < REQUIRED FARE, $/ FL I GHT 35

30 I I I I 1 - 5 6 7 8 9 10 x lo3 FIELD LENGTH, ft

Figure 8. Effect of take-off-landing field length on required FARE. Static margin of 10 percent.

3 45 - x 10

40

REQUIRED FARE, $/ FL I GHT 35

30 I 1 I I I I I 1.0 1.5 2.0 2.5 3.0 3.5 4.0 CANAR D-S U R FAC E C ,max

Figure 9. Effect of canard-surface CL,~~~on required FARE. Static margin of 10 percent.

22 STATIC MARGIN 10%

40

REQU I RED FARE, $/ FL I GHT 35

-160 -140 -120 -100 -80 -60 GEAR POSIT1ON, percent MAC

Figure 10. Effect on required FARE of main-landing-gear position for canard configuration. Static margin of 10 percent.

3 45 x 10

x CONVENT1 ONAL ( GEAR FIXED )

0 CANARD (CANARD-SURFACE C =?.I51 L,max 40

REQUl RED FARE, $/ FL I G HT 35

30 I I I I 1 .. 0 2 4 6 8 10 x lo> CHANGE IN WEIGHT WITH RESPECT TO BASELINE, Ib

Figure 11. Effect of aircraft weight increase on required FARE for canard configuration. Static margin of 10 percent.

23 Figure 12. Effect of aircraft drag on required FARE. Static margin of 10 percent.

O CANARD (GEAR FREE, CANARD-SURFACE CL,max = 3.15)

0 TANDEM WING (GEAR FREE, CANARD-SURFACE CLtmax = 3.15)

0 CONVENTIONAL (GEAR FIXED)

1.5 1.0 CHANGE .5 IN REQUIRED FARE 0 WITH R ES PECT -* TO -1.0 BASELINE, percent -1.5 -2.0 -2.5 I -20 -15 -10 -5 0 5 10 15 20 MINIMUM STATIC MARGIN, percent

Figure 13. Effect of minimum static margin on percent change in required FARE.

24 45 -x 103

x BASELINE 40 - REQU I RED FARE, $/FLIGHT 35

3( I 1 I !A 0 5 10 15 20 25 30 CANARD QUARTER-CHORD SWEEP, deg

Figure 14. Effect of canard-surface sweep angle on required FARE. Static margin of 10 percent.

25 . Report No. 2. Government Accession No. 3. Recipient’s Catalog No. NASA TP-2400 5, Report Date March 1985 6, Performing Organization Code

~~ 505-34-03-05 ‘. Author(s) P. Douglas Arbuckle and Steven M. Sliwa 8. Performing Organization Report No. L-15856 1. Performing Organization Name and Address 10. Work Unit No. NASA Langley Research Center

~ Hampton, VA 23665 11. Contract or Grant No.

13. Type of Report and Period Covered 2. Sponsoring Agency Name and Address National Aeronautics and Space Administration Technical Paper Washington, DC 20546 14. Sponsoring Agency Code

5. Supplementary Notes

16. Abstract Constrained-parameter optimization is used to perform optimal conceptual design of both canard and conven- tional configurations of a medium-range transport. A number of design constants and design constraints are systematically varied to compare the sensitivities of canard and conventional configurations to a variety of tech- nology assumptions. Main-landing-gear location and canard surface high-lift performance are identified as critical design parameters for a statically stable, subsonic, canard-configured transport.

__~.~~__ L7. Key Words (Suggested by Authors(s)) 18. Distribution Statement Conceptual design Unclassified Unlimited Constrained optimization Sensitivity to design parameters Subject Categories 05, 08 Canard configuration Tarideni-wing configuration Econoniic evaluation ___-. ~ LS. Security ClusiF.(of this report) 20. Security Classif.(of this page) 21. No. of l’agrs 22. I’rire

Unclassified Unclassified 26 __.~~~ A03 ~ -~~- I For sale by the National Technical Information Service, Springfield, Virginia 22161 NASA-Langley, 1985