Aerodynamic Design and Cross-Country Flight Performance Analysis of Diana-2 Sailplane
Aerodynamic Design and Cross-Country Flight Performance Analysis of Diana-2 Sailplane
Aerodynamic Design and Cross-Country Flight Performance Analysis of Diana-2 Sailplane
of Diana-2 sailplane
Krzysztof Kubrynski
Warsaw University of Technology, Institute of Aeronautics and Applied Mechanics
Warsaw, Poland, [email protected]
Presented at the XXVIII OSTIV Congress, Eskilstuna, Sweden, 8-15 June 2006
Abstract
The aerodynamic design philosophy and computational analysis of cross-country flight performance of
Diana-2 sailplane in different thermal conditions are presented. To address the second issue, a continuous
spectrum of thermal models is created. For specified thermal model (thermal strength vs distance from core
center, interthermal descent, length and strength of cloud streets), the sailplane mass and aerodynamic
characteristics optimum flight parameters (circling speed and bank angle, interthermal glide speed and speed
in cloud streets) is found applying optimization technique. Finally expected cross-country speed for different
thermal conditions and water ballast weight, as well the optimum water ballast amount, can be predicted. The
main goal of the analysis is the optimization of sailplane parameters and helping pilot to make decisions
regarding optimum balasting technique in flight. An estimation of the influence of bugs and rain on the
sailplane aerodynamic characteristics is also presented.
wC = wT − w SINK (1)
Fig. 1 Assumed cross-country flight pattern. The final equation for altitude balance takes form of:
2
a) a)
b) b)
3
gust (one that decreases the angle of attack) leads to a)
some loss of lift and some loss of altitude. An
upward gust (one that increases the angle of attack)
leads to additional lift in the case of a monotonic lift-
angle of attack relationship, and an increase in speed
and climb rate. However, in the case when the lift
curve has a local minimum as noted, an increase in
angle of attack leads to a loss of lift and altitude. A
variation of angle of attack due to a gust can reach up
to 5 deg. Thus, there can be a significant problem in
circling at such a value of the lift coefficient due to
both the loss of climb efficiency and the danger of
stall. This means that circling at lower CL is
necessary for efficiency in case of an airfoil with lift
characteristics containing local minimum.
b)
Wing sections
Basic airfoils designed for Diana-2 wing have
characteristics slightly different to more typical
sailplane airfoils. Comparison of one of the Diana-2 Fig. 6 Computational characteristics of Diana-2
airfoils to another typical modern sailplane wing airfoil and typical modern airfoil – clean and wet
section is presented on the Fig. 6. The main conditions simulated, a) no flaps, b) flaps down.
properties of new airfoil are monotony of the lift
characteristics and a much higher maximum lift
coefficient. Moreover, there is much less sensitivity The airfoil is equipped with 17% chord
to bugs and rain. That problem is related not only to performance flap with a deflection range of -20 to
an increase of drag in such conditions but also to +280. The laminar flow extends to 70-75% of the
significant loss of lift due to separation, especially at upper surface (flaps down and higher CL conditions)
larger flaps deflection. Wet conditions were and 92% of the lower surface (flaps up and low to
computational simulated by forcing transition at 7% moderate CL). The transition on the lower surface is
of the airfoil chord and specifying a low value of the enforced by pneumatic turbulator in order to prevent
critical amplification factor. It is observed in the laminar separation.
figure, both problems have been significantly reduced The characteristics of the airfoil suggest that safe
in the case of present design. The main drawback of and efficient circling at a CL of about 1.5 (flaps
the new airfoil is the reduction of low drag-bucket deflection +210) is possible. Such characteristics of
width and a higher sensitivity to the incorrect flap the airfoil have been obtained thanks to a unique
settings in flight. However, minimum profile drag is pressure distribution along the chord (with lower
expected to be slightly lower than for typical modern pressure gradients in recovery region and reduced
glider airfoils. stability of laminar boundary layer) and lower
4
thickness. The Diana-2's airfoil sections are even by ~10%). The aerodynamic design of the wing
thinner then original SZD-56’s, ranging from 12.8% includes three main stages6: planform design, airfoils
at the root to 12.2% at the tip (and much less at the design, and finally reduction of adverse interference
winglets). effects (mainly in wing-fuselage intersection and
wing-winglet juncture). The main objective of
aerodynamic design itself is to achieve the best
possible performance at various soaring situations,
including a low stall speed and low sink rates at all
speeds. Good stalling characteristics, effective
ailerons, good flying qualities in thermals, and low
sensitivity to bugs and rain are additional
requirements.
At lower flying speeds, the wing drag is about
90% of total glider drag (about 65% is induced drag).
Fig. 7 Old and new of Diana wing airfoils. At high speeds, wing drag is about 60% of total
glider drag (the greater part of which is profile drag).
Considering this, it is obvious that the main objective
during the aerodynamic design process of a glider
should be the reduction of both profile and induced
drag. The first is possible by designing a wing with
the maximum extent of laminar flow. Due to the fact
that flow conditions (such as Reynolds number and
lift coefficient) vary along the span, the airfoils
should be designed specifically for each spanwise
station in order to satisfy the actual requirements.
Minimization of induced drag can be achieved by the
proper distribution of load along the span by using
the optimum planform and adding winglets. The
process of the Diana-2 wing design includes a
number of considerations and must take their mutual
interactions into consideration. To name just a few,
improvement of high-speed characteristics (lowering
contribution of parasitic drag), internal volume, and
strength characteristics. Given these considerations,
Fig. 8 Drag polars of new and old Diana airfoils. the Diana-2’s wing area is slightly larger than the
original Diana (8.65m2 vs. 8.16m2). A few computer
programs have been used. They allowed for flow
Even though the profiles are thinner than the analysis, design and optimization of two-dimensional
original Diana’s airfoils, the stiffness and strength wing profiles,7 as well as the entire three-dimensional
characteristics are significantly better, and they glider configuration8. The wing planform is totally
maintain a higher cross-section area. This has a great curved. This reduces both induced drag and wing
impact on wing weight, structure, and the volume profile drag, as well as allowing for the proper wing
available for water ballast. Flaperons were used stall progression. The wing profile changes along the
instead of separate flaps and ailerons. Figure 7 shows span. The optimum load distribution along the
a comparison between new and old airfoil shapes, wing/winglet span was determined by applying
while Fig. 8 summarizes the expected aerodynamic Munk’s induced-drag analyses. The optimum wing
characteristics. Expected drag reduction of about 20 planform (more precise: local chord value)
to 25% over the entire range of lift coefficients is minimizing induced and profile drag is possible to
seen. evaluate by division of the above optimum load by
optimum sectional lift coefficient. Modification to
Aerodynamic design of the wing such a planform is undertaken in order to achieve
As noted above, a new wing with entirely new proper stall progression along the span. All the
aerodynamics has been designed. The fuselage and design parameters, such as planform and profiles at
tail are the same as that of the old SZD-56 Diana. various span stations, are subsequently updated to
The proven inner wing structure is also retained (but achieve the properties dictated by the iterative design
design stresses in the monocoque skin were reduced process.
5
The final planform and wing sections were The design of the glider and modified wings,
obtained as a result of using three-dimensional including geometry definition, has been performed
optimization methods during detailed aerodynamic using Unigraphics NX system. The wing mock-up
design. In its final stage, adverse interference effects was made using CNC technology that enabled a
between wing and fuselage,9 as well as in concave precision of better than 0.1mm, which is crucial to
corner of wing/winglet intersection, are removed. benefit from subtle aerodynamic design.
The wing planform and the continuous family of Wing span 14.942 m
wing sections used along wingspan are shown in Wing area 8.657 m2
Figs. 9 and 10 (expect for near wing-fuselage Empty mass 182 kg
intersection sections). The entire internal volume of Wing panel mass 46 kg
the Diana-2 wing is used as an integral water tank. Water ballast (wings) 248 kg
The wings can carry 50% more ballast than Max. mass 500 kg
previously, so the Diana-2 can fly at the highest as Min. wing loading 28 kg/m2
well as the lowest wing loading of any current 15- Max. wing loading 58 kg/m2
meter glider. When flying fully ballasted, nearly half
the glider's weight is water ballast. The optimum
balancing is achievable for any pilot weight through Sailplane performance analysis
the 5.6-liter tail tank. The location of the center of The performance of Diana-2 was calculated using
gravity that minimizes trim drag is at 39% of mean a rather unusual methodology. As performance of
aerodynamic chord. old Diana had been measured and described in Mr.
Richard Johnson’s flight-test results, the evaluation
of drag characteristics of the glider was possible. The
wing drag characteristics obtained for the old Diana
wing (calculated using a panel method with boundary
layer interaction) allowed for the estimation of the
drag characteristics of the remaining parts: fuselage,
fin and tail. By adding the drag of the new wing,
calculated using the same method, to these values,
allows the drag characteristics of the new Diana to be
estimated. It should be noted that the entire procedure
was based on introducing numerical corrections of
the wing characteristics to Johnson's flight-test
results, which were used to predict the flight
characteristics of the Diana-2.
Experimental characteristics are not always
smooth, so smoothing the parasitic drag
characteristics was necessary. Flight tests had been
completed for only one wing loading, so the same
parasitic drag characteristics were used for other
wing loadings. Weight changes were considered only
in the numerical calculations of the wing
characteristics. The prediction for maximum L/D
(with water ballast) exceeds the almost mythical 50:1
boundary for racing class sailplanes. Final speed
polars for the clean wing at different sailplane
Fig. 10 Geometry of the wing and winglet airfoils. weights are presented on Fig. 11.
6
Fig. 11 Diana-2 calculated speed polars.
Fig. 13 Simulation of clean wing, bugs and rain in
As the airfoils applied on new wing show performance calculations (mass: 350 kg).
considerable high sensitivity to the proper flap
settings, a detailed analysis of flap position influence However, the optimum flap settings change
on the speed polars was of primary importance. significantly. Rainy conditions are expected to lead
Computational optimum of flap position for various to much worse sailplane performance. On the other
sailplane weights and speeds are presented in Fig. 12. hand, the estimated loss is still much lower then for
any other high performance glider. The other feature
worth mentioning is the very low sensitivity of wet
wing to flap setting.
7
rate is observed. Nevertheless, the distribution of the logs results indicates that the wide thermal model is
final average cross-country speed is very flat near the most relevant; however, very often lengths and
optimum. The most significant benefit from heavy strengths of cloud streets were greater then assumed
ballast or lighter sailplane, which corresponds to high in calculations. Hence the real average cross-country
or low allowable wing loadings, is expected mainly speed is sometimes higher than predicted. Quite
for very strong and very week thermal conditions. unusual and slightly surprising is the significant
underestimation of calculated cross-country speed in
strong thermal conditions (climb rate about 2.5 m/s).
Commonly applied for average cross-country
performance analysis is the mixed weather model of
Quast2, which consists of some portions of wide and
narrow thermals of strong and weak types.
Obviously, the main objective of the analysis
presented is the determination of the amount of water
ballast optimal for various thermal strengths. A
family of mixed thermals – narrow and wide - with
the same vertical velocity at 60m-radius position
from core center, was specified. The same as in
Quast’s model, the assumption was made that 84% of
the total flight distance is dominated by wide
thermals, whereas 16% by narrow thermals. Mean
value of the interthermal descent is 10% of the
maximum vertical velocity inside thermal core,
which means approximately 20% of average climb
ratio. Figure 16 shows the relevant results.
It is worth noting that the mixed weather model
similar to Quast’s leads to a different optimum of
sailplane’s mass compared to the model including
wide thermals only, but the expected average cross-
country speeds in both cases are rather similar (the
model including some amount of narrower thermals
predicts a few km/h lower speed). Comparison of the
presented computational results to those extracted
from flight logs data suggests some overestimation of
the narrow thermal presence in Quast’s model.
The most important conclusions from the
Fig. 14 Sample computational cross-country flight analyses concerns the optimal amount of water
characteristics of Diana-2 at two thermal strengths. ballast for various average climb rates. Generally,
the optimum is higher then the values usually
presented in literature. A maximum sailplane mass
Figure 15 presents the optimum sailplane mass vs of 500 kg is expected to be optimum for average
average climb rate relationships for both kinds of climb rate above 1.75 m/s in the case of wide
weather conditions and wide thermals. Additional thermals and no cloud streets. In the case of cloud
boundaries of a 1% penalty on the average cross- streets of 20% relative length and 30% strength, the
country speed (thin line) and constant thermal maximum sailplane mass should be used at climb
strength conditions (broken oblique lines) are rates above 1.5 m/s. The minimum practical
marked. Solid points represent the measured values achievable mass of about 280 kg (in the case of 80 kg
extracted from flight logs (year 2005: Polish Open pilot + equipment) achieves an optimum for 1 m/s
Class Nationals and European Championships). The and 0.9 m/s respectively. Expectation is that
average climb rate was defined as a sum of all the removing of 50kg water ballast should improve the
altitude gained in circling and difference of altitudes average climb rate by about 0.15 m/s. The values of
at start and finish divided by the total time spent masses should be slightly lower in the presence of
circling. The average cross-country speed includes some fraction of narrow thermals. It is interesting,
the real kilometers flown (not the task distance only). that mixed model affects mainly high mass
Analogical analysis has been performed for characteristics with little influence on the light
narrow and middle thermals. Examination of flight sailplane.
8
a) a)
b) b)
Fig. 15 Optimum Diana-2 mass vs. climb rate for Fig. 16 Optimum sailplane mass and expected cross-
wide thermals and two weather models: a) no cloud country speed for mixed thermals and two weather
streets, b) with cloud streets. models a) no cloud streets, b) with cloud streets
On the other hand, the dependency between mass Any further significant improvements in sailplane
and final cross-country speed is very flat near the flight performance in a way of drag reduction
optimum. Over- or under-ballasting the sailplane by (especially profile drag) is almost impossible without
50 to 70 kg (about +/- 15% of sailplane optimum the use of new flow technologies (e.g. active flow
mass) leads to the lost of about 1% of average speed. control of boundary layer).
Some moderate improvement of overall sailplane
Conclusions performance can still be achieved by careful
The design process of a modern high-performance aerodynamic design of every detail of sailplane and
sailplane is rather demanding issue. A very high new solutions for sailplane structure and technology.
level of sailplane aerodynamics has been achieved A relatively broad scale of possible improvement in
through the use of advanced computational design final flight performance is still possible in the area of
tools and extensive wind-tunnel investigations. low-speed characteristics (climb performance).
9
References
6
Kubrynski K., “Wing-Winglet Design Methodology
1
Horstmann, K.H., Neue „Modellaufwindverteilungen for Low Speed Applications”, AIAA Pap. 03-0215, 41
und ihr Einfluss auf die Auslegung von Segelflugzeugen“ Aerospace Science Meeting, Reno, 2003
7
OSTIV Publication XIV, 1976 Drela M., “XFOIL: An Analysis and Design System
2
Thomas, F., Fundamentals of Sailplane Design, for Low Reynolds Number Airfoils”, in Low Reynolds
College Park Press, 1999 Number Aerodynamics, Ed. T. J. Mueller, Lecture Notes in
3
Cochrane J.H., “MacCready Theory with Uncertain Eng. 54 (1989)
8
Lift and Limited Altitude”, Technical Soaring 23(3) (July Kubrynski K., “Subsonic Aerodynamic Design Via
1999), also: Optimization”, Ed. K.Fuji, G.S.Dulikravich, Notes on
http://www.gsb.uchicago.edu/fac/john.cochrane/research/Soaring Numerical Fluid Mechanics, Vol. 68, 1999, Vieweg
4 9
Almgren, R., Tourin, A., “The Mathematics of Gliding Boermans L.M.M, Kubrynski K., Nicolosi F., “Wing-
Racing,” RCA, Nov. 7, 2004 Fuselage Design of High-Performance Sailplanes” –
(http://www.math.toronto.edu/almgren/optsoar/) Boundary Layer Separation in Aircraft Aerodynamics, ed.
5
Boermans L.M.M, van Garrel A., “Design and Henkes R.A.W.M., Bakker P.G. – Delft 1997
Windtunnel Test Results of a Flapped Laminar Flow 10
Johnson, R.H., “A Flight Test Evaluation Of
Airfoil for High-Performance Sailplane Applications” – The SZD-56-1 Diana Sailplane” - Soaring-April 1999
ICAS 94-5.4.3
10