Course Project in Aerospace Engineering

Transonic and Supersonic Aircraft Aerodynamics

CFD Investigation of Supercritical

Airfoils with Modified Trailing Edge
Shapes at Transonic Speeds
A Comprehensive Study on Shock-waves and Corresponding
Aerodynamics in Transonic Flight Conditions

DHANASEKARAN GUNASEKARAN, HARI PRASAD P, JACOPO LEON

CUCCHIARO, NAGAHARISH GOWNDA NAGANATHAN
This page intentionally left blank.
 CFD Investigation of Supercritical
Airfoils with Modified Trailing Edge
Shapes at Transonic Speeds
A Comprehensive Study on Shock-waves and
Corresponding Aerodynamics in Transonic Flight
Conditions

Dhanasekaran Gunasekaran, Hari Prasad P, Jacopo Leon

Cucchiaro, Nagaharish Gownda Naganathan

Master’s Programme, Aeronautics

Examiner: Raffaelo Mariani
School of Engineering Sciences
This page intentionally left blank.
Abstract
This study investigates the aerodynamic effects of trailing edge geometries on
a NASA SC(2)-0710 supercritical airfoil under transonic flow conditions. The
geometry is modified to produce sharp, blunt, concave, convex, and diverging
trailing edge profiles. Numerical simulations are conducted using ANSYS
Fluent with the Reynolds-averaged Navier-Stokes (RANS) equations and the
SST k–ω turbulence model. The flow conditions correspond to freestream
Mach numbers ranging from 0.7 to 1.1 at a fixed angle of attack of 1.5◦ ,
consistent with experimental studies used for validation. Aerodynamic
performance is evaluated in terms of drag and lift coefficients, with a primary
emphasis on pressure coefficient (Cp ) distributions and shock behaviour. The
results reveal how trailing edge modifications alter shock strength, location,
and wake development, thereby influencing aerodynamic efficiency in the
transonic regime.

Keywords
Supercritical Airfoils, Transonic Flow, Trailing Edge Shapes, Computational
Fluid Dynamics
This page intentionally left blank.
Acknowledgments
The authors express deepest gratitude to Professor, Dr. Raffaelo Mariani, for
the unwavering support, guidance, and encouragement throughout the course
of this research and for making the lectures interesting and
thought-provoking.
This page intentionally left blank.
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1 Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Meshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.3 Mesh Independent Study . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.4 Numerical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.5 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.6 Validation and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.0.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.1 Effect of TE Geometry at Fixed Thickness . . . . . . . . . . . . . . . 13
4.2 Effect of Trailing Edge Thickness . . . . . . . . . . . . . . . . . . . . 15
4.3 Shock Study with Cp distribution and Velocity contours . . . . . . 17
4.3.1 Sharp Trailing Edge . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.2 Blunt Trailing Edge . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.3.3 Convex Trailing Edge . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3.4 Concave Trailing Edge . . . . . . . . . . . . . . . . . . . . . . . 22
4.3.5 Diverging Trailing Edge . . . . . . . . . . . . . . . . . . . . . . . 24
4.4 Performance Summary: Cd and Cl /Cd Trends . . . . . . . . . . . . . 26
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
This page intentionally left blank.
1 Introduction
The transonic regime, typically defined within the Mach number range of 0.7
to 1.2, represents a highly complex and aerodynamically sensitive phase of
flight. Within this range, the flow around an aircraft or airfoil simultaneously
exhibits both subsonic and supersonic characteristics, resulting in the
formation of shock waves, boundary layer interactions and localized flow
separation. These non-linear phenomena have a significant impact on
aerodynamic performance, influencing lift, drag, stability and control. As
aviation advanced into higher speed regimes particularly during the mid-20th
century optimizing flight through the transonic envelope emerged as a key
challenge for both commercial and military aircraft design.
The development and application of Computational Fluid Dynamics (CFD)
have proven instrumental in addressing the intricacies of transonic
aerodynamics. CFD allows for detailed resolution of flow features such as
shock-induced separation, shock-boundary layer interactions and trailing
edge wake development, offering insights that are often difficult to capture
experimentally. With the continuous improvement of numerical methods,
turbulence modelling and computational resources, CFD has evolved into a
powerful and reliable tool for analysing and optimizing aerodynamic shapes
specifically tailored for transonic performance.
One of the most prominent areas where CFD has contributed significantly is
in the design evolution of airfoils optimized for transonic flight. Initial
strategies to reduce transonic wave drag led to the creation of rooftop airfoils,
which featured flattened upper surfaces to delay shock formation. These were
succeeded by peaky airfoils, characterized by sharp leading-edge curvature
and high thickness-to-chord ratios, enabling better pressure recovery and
delayed drag rise. However, the most notable advancement came with the
development of the supercritical airfoil, pioneered by NASA in the 1960s. This
airfoil type marked by a blunt leading edge, flattened upper surface,
aft-loaded camber and a relatively thin trailing edge was specifically designed
to weaken shock waves, delay the onset of drag divergence and improve
overall aerodynamic efficiency in the transonic regime.
Among various supercritical profiles, the NASA SC(2)-0710 airfoil has been
widely studied and employed due to its favourable cruise performance near
Mach 0.8. While significant research has focused on optimizing its thickness
distribution, camber and leading-edge geometry, trailing edge design remains
a relatively under explored area despite its crucial aerodynamic role. The
shape of the trailing edge affects pressure recovery, wake development and
shock positioning, all of which have a direct influence on lift, drag and
pitching moment characteristics. Traditional sharp trailing edges offer clean
pressure equilibrium but may raise structural and aeroelastic concerns,
whereas blunt trailing edges enhance structural robustness but induce base
drag due to separated wake flow. Additionally, concave, convex, converging
and diverging trailing edge geometries further influence flow characteristics
by altering local pressure gradients, shock wave formation and boundary layer
behaviour near the trailing edge.
Understanding these aerodynamic effects is particularly important in the
context of high-speed aircraft, where reducing drag and maintaining
structural reliability are critical design objectives. While experimental studies
have demonstrated the sensitivity of shock position and pressure distribution
to trailing edge geometry, CFD provides a more systematic and controlled
platform for investigating a wide range of geometric variations.
This study presents a numerical analysis of the aerodynamic effects of various
trailing edge modifications applied to the NASA SC(2)-0710 supercritical
airfoil. Using CFD, the analysis focuses on transonic flow conditions across
Mach numbers ranging from 0.7 to 1.1. Key parameters such as pressure
coefficient distributions, shock wave location, lift and drag coefficients are
used to evaluate the impact of each trailing edge configuration. The goal is to
deepen understanding of how subtle changes in trailing edge geometry
influence overall airfoil performance in the transonic regime, with
implications for both aerodynamic efficiency and structural design.
2 Background
The transonic flight regime, typically ranging from Mach 0.7 to 1.1 presents
unique aerodynamic challenges due to the simultaneous presence of subsonic
and supersonic flow over the same surface. This results in phenomena such as
shock wave formation, shock-boundary layer interaction and flow separation,
all of which contribute to increased drag, flow instability and control issues.
Supercritical airfoils were introduced to address these issues by delaying the
onset of shock waves and reducing the associated drag rise. While significant
work has been done on optimizing thickness distribution, camber and
leading-edge shape, the influence of trailing edge geometry on aerodynamic
performance especially in transonic conditions remains less thoroughly
investigated.
One of the earliest detailed experimental investigations into trailing edge flow
in the transonic regime was conducted by Brown and Viswanath (1984) [1],
who studied supercritical airfoils with blunt trailing edges. Their research,
carried out at high Reynolds numbers, revealed the importance of
viscous-inviscid interaction near the trailing edge and highlighted the
limitations of common turbulence models in predicting skin friction and
displacement thickness in such flow regions. Complementing this, a more
recent experimental study by Jacquin et al. (2021) [2] investigated the
transonic buffet phenomenon on a supercritical airfoil using schlieren
imaging and particle image velocimetry (PIV). They observed unsteady shock
oscillations across the chord length, confirming the dynamic nature of
shock–trailing-edge interactions and the importance of trailing edge geometry
in flow stability and drag control.
In addition to these physical experiments, computational methods have
become essential for examining trailing edge modifications under transonic
flow. Yoo (2001) [3] carried out a CFD-based investigation comparing a
baseline supercritical airfoil (DLBA-186) with a modified version (DLBA-283)
featuring a divergent trailing edge. Using RANS simulations with the
Spalart–Allmaras turbulence model, Yoo found that the divergent trailing
edge significantly reduced wave drag, delayed shock formation and improved
the lift-to-drag ratio demonstrating the aerodynamic benefits of even modest
trailing edge changes. Similarly, Rezaei, Sobhani and Nejat (2018) [4] studied
the effects of adding Gurney flaps to a supercritical airfoil at transonic speeds
using k-ω SST models. Their findings showed that such small modifications
near the trailing edge could increase lift and suppress shock intensity,
although these benefits were dependent on Mach number and angle of attack.
Recent studies have also focused on optimizing airfoil performance through
geometry modifications using modern computational techniques. For
instance, a 2025 study [5] applied multi-objective optimization (via NSGA-II
and Kriging surrogate modelling) to an RAE2822 airfoil, varying both leading
and trailing edge deflection angles to improve transonic performance. The
research demonstrated that appropriate trailing edge deflections could delay
the drag divergence Mach number and significantly enhance the lift-to-drag
ratio. A separate study by Dai et al [6]. introduced a novel approach by
integrating a jet on the lower surface of the trailing edge. Their simulations
showed a substantial improvement in lift-to-drag ratio over 26% highlighting
the effectiveness of both passive and active trailing edge modifications.
Morphing airfoils have also emerged as a promising approach to optimizing
transonic performance through adaptive geometry. Cheng et al. (2022) [7]
investigated a morphing version of the RAE2822 supercritical airfoil, where
both camber and trailing edge geometry could be actively adjusted. Their
results indicated that such adaptability offered advantages across a wide flight
envelope, from low-speed take-off to high-speed cruise. Similarly, Lv et al.
(2019) [8] demonstrated that continuous deformation of the trailing edge
introduced unsteady aerodynamic effects, including lift hysteresis, which
must be considered in both design and control strategies.
Experimental work by Livya and Pillai (2022) [9] extended the investigation
of trailing edge effects to various turbulence intensities. Although conducted
at subsonic conditions with extended trailing edges on a NACA 0020 airfoil,
their findings showed that certain trailing edge geometries could maintain
performance in adverse pressure gradients, suggesting broader implications
for wake behaviour and separation control in transonic flows. This aligns with
observations made by Koti and Khan (2018) [10], who used CFD to analyse
the RAE2822 airfoil and documented the influence of trailing edge pressure
distributions on shock behaviour and drag rise. Their study emphasized the
importance of grid independence and turbulence modelling when simulating
flow near the trailing edge, a region known for complex wake development.
Furthermore, other experimental studies have explored non-conventional
trailing edge configurations such as serrated edges. One study published in
Experiments in Fluids (2015) [11] compared blunt and serrated trailing edges
at transonic speeds, showing that serrations can dampen pressure wave
amplitude and modify wake dynamics. These findings reveal the potential of
geometric tailoring at the trailing edge to control unsteady flow phenomena.
Despite the wide range of work conducted on supercritical airfoils, it is clear
that trailing edge geometry remains a relatively under explored parameter
compared to leading-edge and camber optimization. Many studies focus on
passive performance gains, such as shock control and drag reduction, while
others have demonstrated the utility of active or morphing geometries for flow
adaptability. However, systematic investigations that isolate and compare
different trailing edge profiles, such as sharp, blunt, concave, convex, and
divergent shapes, on a well-established supercritical airfoil such as the NASA
SC(2)-0710 are limited in the current literature.
Thus, the present study aims to fill this gap by numerically analysing a range
of trailing edge modifications applied to the NASA SC(2)-0710 airfoil in the
transonic regime. By examining how these geometric changes influence
pressure coefficient distributions, shock wave position, and integrated
aerodynamic coefficients across Mach numbers from 0.7 to 1.1, this research
seeks to deepen the understanding of trailing edge sensitivity in transonic
flows and contribute valuable insights for future airfoil design and
optimization.
3 Methodology
The analysis of the supercritical airfoil has been done by designing the
computational domain and then using the Computational Fluid Dynamics
methods to analyse the effects of the changes in the trailing edge of a
supercritical airfoil at different transonic speeds.

3.1 Geometry

The selected airfoil for this study is the NASA SC(2)-0710, a member of the
NASA Supercritical Airfoil family. This airfoil features a maximum thickness
of 10% of the chord, located at 37% of the chord length. It is designed for a lift
coefficient of 0.7 and is optimised for transonic flow conditions. The airfoil’s
flattened upper surface profile serves to delay shock wave formation, thereby
reducing wave drag.
The airfoil geometry was modelled using the CAD software, Autodesk Fusion
based on the official coordinate data. A chord length of 635 mm was selected
for the analysis, consistent with values reported in experimental studies used
for validation purposes.

Figure 1: Computational Domain - Circular

The computational domain was established with a diameter ten times the
chord length as shown in Fig. 1. This approach is commonly employed to
minimise the influence of domain boundaries on the flow field around the
airfoil, thereby ensuring the accuracy of the aerodynamic results.

(a) Sharp (b) Blunt

(c) Convex (d) Concave

(e) Diverging

Figure 2: Comparison of different trailing edge geometries.

3.2 Meshing

Figure 3: Refined Mesh (Zoomed in view)

A circular mesh is generated around the airfoil, extending in all directions to
minimize far-field reflection effects as shown in Fig. 3. The mesh was
generated using ANSYS Meshing software. The mesh is refined near the airfoil
surface with quadrilateral elements, ensuring high aspect ratio and
orthogonality in near wall regions. Inflation layers were applied with a growth
ratio of 1.2 and 50 steps to maintain y + ≤ 1. Additional refinement is applied
in the trailing edge regions to capture pressure gradients and shocks
effectively. The final mesh contains around 220000 to 250000 elements.

3.3 Mesh Independent Study

A mesh independence study was conducted using the sharp model for three
different mesh densities: coarse, intermediate, and fine. The coarse mesh,
consisting of approximately 75,000 elements, produced large errors in both
drag and lift predictions but required minimal computational time as shown
in figure 4 and figure 5. Refining the mesh to about 220,000 elements
(intermediate mesh) improved accuracy at the expense of increased
computational cost. The fine mesh, which contained the largest number of
elements, provided the most accurate results but with significantly longer
computation times.
The difference in lift and drag coefficients between the intermediate and fine
meshes was found to be less than 5%, indicating near convergence. Although
the ideal error margin for research-grade simulations is typically between
1–2%, further refinement was constrained by the limitations of the student
license. Therefore, the intermediate mesh was selected as the optimal
configuration, balancing computational efficiency and accuracy.

Figure 4: Cd vs Number of Elements

 Figure 5: Cl vs Number of Elements

As shown in Figure 4 and Figure 5, both the coefficient of drag (Cd ) and the
coefficient of lift (Cl ) exhibit noticeable variation when transitioning from the
coarse to the intermediate mesh. However, the change between the
intermediate and fine meshes is minimal, confirming that mesh independence
has been nearly achieved.

3.4 Numerical Model

A 2-equation turbulence model, k-ω SST (Shear Stress Transport) model, has
been used for the simulations. This model has been widely used in
aeronautics, external aerodynamics and turbulent flows. The equation
involves 2 transport equations, one for the turbulent kinetic energy (K) and
the other for the specific rate of dissipation (ω).
The main advantage of this model is that it captures the near wall conditions.
It accurately resolves the boundary layer flows and captures the adverse
pressure gradients and flow separation. Additionally, it has the benefits of the
k- model in the free stream region, which is less sensitive to inlet turbulence
values.
The Shear Stress Transport (SST) incorporates a shear stress limiter that
accounts for the principal transport shear stress,improving the prediction of
flow separation and complex boundary layer [Link] feature enables
better accuracy in flows with strong adverse pressure gradients and separation
compared to other Reynolds-Averaged Navier-Stokes (RANS) turbulence
models.
The K-omega SST model is broadly favoured in the aerospace industry due to
its high accuracy-to-cost ratio and its reliability in predicting aerodynamic
forces and flow behaviour around complex geometries such as airfoils.
However, it can generate excessive turbulence levels in regions with strong
acceleration or stagnation, although this effect is less severe than in some
other models.
The equation for continuity is given as:

∂ρ  
+ ∇ · ρV~ = 0 (3.1)
∂τ
The equation for momentum is given as:

∂ V~
ρ = −∇p + ∇ · σ + ρf~ (3.2)
∂τ
The equation for energy is given as:
  
∂h ∂p
ρ + ∇ · hV~ =− + ∇ · (k∇T ) + φ (3.3)
∂τ ∂τ
Where,

∂Vi
φ = (σ · ∇)V~ = σij (3.4)
∂xj
The turbulence model that is used for the simulation is k-ω SST. The
turbulence kinetic energy (k) and the specific dissipation rate are:

  
∂ ∂ ∂ui ∂ ∗ ρa1 k ∂k
(ρk) + (ρuj k) = ρij − β ∗ ρkω + µ+σ (3.5)
∂t ∂xj ∂xj ∂xj ω ∂xj

  
∂ ∂ ω ∂ui ρ ∂k ∂ω ∂ ρa1 k ∂ω
(ρω)+ (ρuj ω) = α ρij −βρω 2 +σd + µ+σ
∂t ∂xj k ∂xj ω ∂xj ∂xj ∂xj ω ∂xj
(3.6)
For the Shear Stress Transport (SST) model there are some defined constants:
β ∗ = 0.09, a1 = 0.31

3.5 Boundary Conditions

The free-stream Mach numbers vary from 0.7 to 1.1 with increments of 0.1 at
a fixed angle of attack of 1.5◦ , following the conditions reported in the
experimental literature for the SC(2)-0710 airfoil. The far-field boundaries
are prescribed as pressure far-field conditions, while the airfoil surface is
modelled as a no-slip adiabatic wall.
3.6 Validation and Analysis

The baseline SC(2)-0710 airfoil case was validated against published

experimental data [12] for surface pressure coefficient distributions, as shown
in Figure 6. The validation was performed at a Mach number of 0.78 and an
angle of attack of 1.5◦ . For the CFD simulations, the k–ω SST turbulence
model was employed due to its superior capability in predicting adverse
pressure gradients and flow separation for transonic flows.
As illustrated in Figure 6, the CFD results show good agreement with the
experimental pressure coefficient (Cp ) distribution over the airfoil surface.
The suction peak on the upper surface is well captured and the overall
pressure recovery trend matches the experimental data. Minor discrepancies
can be observed near the trailing edge, likely due to mesh resolution
limitations and turbulence model assumptions. The sudden dip observed in
the experimental Cp distribution can be attributed to geometric discrepancies
arising during the reconstruction of the airfoil. Although the airfoil data for
the experimental profile were available, the reconstruction process did not
yield a perfectly accurate geometry. As a result, the experimental model used
for comparison was based on a closely matching airfoil with a maximum
thickness of approximately 10%. While the overall shape followed the same
aerodynamic characteristics as the NASA SC(2)-0710 airfoil, these minor
geometric deviations likely contributed to localized variations in the surface
pressure distribution, particularly near the suction peak and pressure
recovery regions.

Figure 6: Experimental vs CFD Validation

Overall, the comparison demonstrates that the CFD setup provides a reliable
prediction of the aerodynamic behaviour for the SC(2)-0710 airfoil under the
given flow conditions. The validated model can therefore be confidently used
for subsequent aerodynamic and stability analyses.
4 Results
4.0.1 Overview

This section presents the aerodynamic results from the CFD study of the
supercritical airfoil with five trailing-edge (TE) shapes (blunt, sharp, concave,
convex, diverging) and three trailing-edge thickness ratios (0.5%, 0.7%, 0.9%
chord). Results are grouped into three sets of plots:

1. Mach sweep of total drag coefficient Cd comparing the five TE geometries

at each thickness.
2. Cd versus Mach for each TE geometry showing the effect of thickness.
3. Pressure coefficient Cp distributions for all five TE geometries at
representative Mach numbers.

4.1 Effect of TE Geometry at Fixed Thickness

Figure 7: Cd variations with Mach for 0.5% T.E thickness

Figures 7–9 show Cd vs Mach for the five trailing edge shapes at thickness
ratios 0.5%, 0.7% and 0.9%. Each figure compares how TE geometry alters
drag generation across the transonic range.
The general trend is that Cd remains relatively low and slowly varying up to a
critical Mach number (Mdd ), followed by a rapid rise as transonic shocks
strengthen and shock-induced separation develops. For the 0.5% thickness
ratio, however, this slow-variation regime is already past; the profiles exhibit
drag growth almost immediately as Mach increases. This is because the
airfoil’s aft section is too sharp to provide smooth pressure recovery after the
shock. Even at lower Mach numbers, a small shock forms that induces early
separation near the trailing edge. As seen in Fig. 7, the diverging TE performs
poorest at this thickness, showing higher Cd across the sweep relative to other
geometries because it increases local flow turning, causing stronger shocks
and more flow deceleration, increasing separation.

Figure 8: Cd variations with Mach for 0.7% T.E thickness

Figure 9: Cd variations with Mach for 0.9% T.E thickness

At the 0.7% thickness ratio (Fig. 8), the trend is more complex. The concave
TE exhibits the lowest drag for much of the Mach sweep, while the convex
geometry achieves the highest drag-divergence Mach as it delays the location
where the shock forms. Although the convex TE delays the drag rise, once the
increase begins, it grows more steeply compared to other geometries.
For the 0.9% thickness ratio (Fig. 9), the blunt TE performs surprisingly well.
It achieves the highest Mdd and exhibits the lowest drag over most of the Mach
range. A blunter trailing edge can weaken the shock strength near the aft
region by promoting a more favourable pressure recovery, and the boundary
layer remains more attached. Interestingly, the diverging TE performs better
here than the concave TE, in contrast to the 0.5% case.

4.2 Effect of Trailing Edge Thickness

Figures 10–13 present CD across Mach 0.7–1.1 for the three thickness ratios
for each geometry. These highlight how trailing edge thickness modifies drag
behaviour.

Figure 10: Cd variations with Mach for blunt T.E

For the blunt TE (Fig. 10), the lowest drag is obtained at 0.9% thickness, not
at 0.7% as generally expected from literature. This anomaly may be the result
of inaccuracies in the modelling: the domain was solved for a steady state,
which would fail to capture all of the unsteady aspects of transonic flow.

Figure 11: Cd variations with Mach for convex T.E

For the convex TE (Fig. 11), both 0.7% and 0.9% thicknesses exhibit similar
drag characteristics, with 0.7% performing better over a wide mach range. It
is to be noted that the apparent dip in the Cd is due to curve fit error, and only
the values at 0.1 increments have to be considered to compare. The convex
geometry promotes a gentler pressure recovery after the shock because the
surface curves outward, allowing the boundary layer to remain attached over a
wider range of conditions. This smoother deceleration of flow prevents abrupt
separation and spreads out the drag rise over a broader Mach range once it’s
past a sharp curve, as seen in 0.5%.

Figure 12: Cd variations with Mach for concave T.E

In contrast, the concave TE (Fig. 12) shows a clear optimum at 0.7%

thickness, consistent with theoretical expectations and prior studies. The drag
rise occurs earliest at 0.9%, highlighting the adverse effect of excess thickness
on separation tendencies.

Figure 13: Cd variations with Mach for diverging T.E

The diverging TE (Fig. 13) depicts similar characteristics to the convex case,
as the 0.5% thickness gives rise to stronger shocks, thereby increasing the
drag. However, here the 0.9% thickness marginally outperforms 0.7% across
parts of the Mach sweep, likely due to delayed separation recovery in the aft
region. A diverging geometry diffuses the flow downstream, creating a longer
but weaker pressure recovery zone.
4.3 Shock Study with Cp distribution and Velocity contours

Figures 14–31c show Cp (x/c) for the five TE geometries at selected Mach
numbers (M = 0.7 and 0.8). These Mach numbers are selected since they are
the primary region of shock growth. The corresponding velocity contours are
depicted alongside to better explain the phenomenon involved. Each figure
overlays the three thickness ratios in order to better understand the shock
behaviour, which is complemented by the velocity contours.

4.3.1 Sharp Trailing Edge

For the sharp TE geometry, it is noted from Fig. 14 that the presence and
strength of shocks vary based on the Mach number.

Figure 14: Chord-wise variations of Cp over airfoils for sharp TE with M = 0.7 and 0.8

At Mach 0.7, Fig. 15a, the Cp distribution exhibits a strong suction peak near
the leading edge, followed by a shock located at approximately 20% chord.
The shock is relatively weak, resulting in smoother pressure recovery
downstream and a smaller supersonic region, as confirmed by the velocity
contours. Flow deceleration across the shock is moderate and the boundary
layer remains largely attached.
At Mach 0.8, Fig. 15b, the aerodynamic behaviour changes significantly. The
suction peak at the leading edge is weaker, but the shock is stronger and
displaced further downstream. The Cp distribution reveals a larger pressure
rise across the shock and a higher plateau in the post-shock region, signifying
stronger shock–boundary layer interaction. Velocity contours corroborate
this, showing an extended supersonic bubble and a more abrupt deceleration
across the shock.

(a) Velocity Contours with sharp TE at M (b) Velocity Contours with sharp TE at M
= 0.7 = 0.8

4.3.2 Blunt Trailing Edge

Figure 16: Chord-wise variations of Cp over airfoils with blunt TE at M = 0.7

At Mach 0.7, the Cp distributions for all blunt trailing edge configurations
show a leading-edge suction peak followed by shock formation on the upper
surface. For the 0.5% blunt trailing edge case, the suction peak is strongest
and the shock occurs around 20–25% chord, producing a well-defined
supersonic bubble with sharp deceleration behind the shock. The 0.7% blunt
trailing edge reduces the suction peak and shifts the shock slightly
downstream, while the 0.9% blunt trailing edge results in the weakest suction,
a broader shock profile, and smoother post-shock recovery. These results
indicate that increasing trailing edge bluntness at Mach 0.7 decreases shock
intensity and wave drag potential but reduces lift due to weaker suction. The
effect of the trailing edge moves upstream towards the leading edge affecting
the position and strength of shocks.
 (a) Velocity Contours with blunt TE (0.5% (b) Velocity Contours with blunt TE (0.7%
thickness) at M = 0.7 thickness) at M = 0.7

(c) Velocity Contours with blunt TE (0.9%

thickness) at M = 0.7

Figure 18: Chord-wise variations of Cp over airfoils with blunt TE at M = 0.8

At Mach 0.8, shock behaviour becomes more pronounced and the influence of
trailing edge thickness is amplified. For the 0.5% blunt trailing edge, the
suction peak remains high, but the shock is stronger and located further
downstream, leading to a larger supersonic region and steeper pressure
recovery. The 0.7% blunt trailing edge produces a weaker suction peak and a
more gradual shock, reducing wave drag relative to the thinner trailing edge.
The 0.9% blunt trailing edge case exhibits the weakest suction and broadest
shock, with significantly smoother recovery but reduced aerodynamic
efficiency due to lower lift generation. As the Mach increased compared to the
M = 0.7 case, the effects of the trailing edge on the upstream and leading edge
gets minimized due to the presence of shocks.

(a) Velocity Contours with blunt TE (0.5% (b) Velocity Contours with blunt TE (0.7%
thickness) at M = 0.8 thickness) at M = 0.8

(c) Velocity Contours with blunt TE (0.9%

thickness) at M = 0.8

4.3.3 Convex Trailing Edge

Figure 20: Chord wise variations of Cp over airfoils with convex TE at M = 0.7

At Mach 0.7, all three trailing edge configurations exhibit a pronounced

leading-edge suction peak, followed by shock formation on the upper surface.
For the 0.5% convex trailing edge, the shock occurs around 25% chord and is
sharp, generating a strong supersonic region with rapid post-shock
deceleration. The 0.7% case shows the shock shifted slightly downstream with
a weaker suction peak and more gradual recovery, indicating reduced wave
drag. The 0.9% trailing edge further attenuates the suction peak and broadens
the shock profile, leading to smoother downstream recovery and diminished
shock intensity but reduced aerodynamic loading. Velocity contours confirm
this progression, showing that thicker convex trailing edges suppress shock
strength and shrink the supersonic bubble.

(a) Velocity Contours with convex TE (b) Velocity Contours with convex TE
(0.5% thickness) at M = 0.7 (0.7% thickness) at M = 0.7

(c) Velocity Contours with convex TE

(0.9% thickness) at M = 0.7

Figure 22: Chord-wise variations of Cp over airfoils with convex TE at M = 0.8

At Mach 0.8, shock effects intensify and the influence of trailing edge
geometry becomes more pronounced. In the 0.5% convex case, the suction
peak is strong, and the shock is displaced downstream, producing an extended
supersonic bubble with steep pressure gradients. The 0.7% convex trailing
edge weakens the suction peak and shifts the shock slightly further
downstream. The 0.9% convex trailing edge configuration exhibits the
weakest suction and broadest shock, with the smoothest recovery and
minimal wave drag contribution but at the expense of lift.

(a) Velocity Contours with convex TE (b) Velocity Contours with convex TE
(0.5% thickness) at M = 0.8 (0.7% thickness) at M = 0.8

(c) Velocity Contours with convex TE

(0.9% thickness) at M = 0.8

4.3.4 Concave Trailing Edge

Figure 24: Chord-wise variations of Cp over airfoils with concave TE at M = 0.7

At Mach 0.7, the 0.5% concave trailing edge exhibits a strong suction peak
near the leading edge, followed by a shock around 25–30% chord. The Cp
curve shows a distinct shock discontinuity with moderate downstream
recovery. The 0.7% concave case displays a more pronounced shock with a
larger drop in suction and a wider supersonic region, as confirmed by the
velocity contours. This configuration introduces stronger shock–boundary
layer interaction, leading to higher unsteadiness in the Cp trace. The 0.9%
concave trailing edge weakens the suction peak but shifts the shock slightly
downstream, smoothing the recovery and reducing the post-shock adverse
pressure gradient.

(a) Velocity Contours with concave TE (b) Velocity Contours with concave TE
(0.5% thickness) at M = 0.7 (0.7% thickness) at M = 0.7

(c) Velocity Contours with concave TE

(0.9% thickness) at M = 0.7

Figure 26: Chord-wise variations of Cp over airfoils with concave TE at M = 0.8

At Mach 0.8, transonic effects intensify, and the differences between trailing
edge thicknesses become more significant. The 0.5% concave trailing edge
shows a sharp suction peak and a strong shock, producing an extended
supersonic bubble and steep pressure recovery. The 0.7% case experiences the
strongest shock among the three, with Cp distributions indicating a large
adverse pressure jump. The 0.9% trailing edge alleviates this effect,
weakening the suction peak and distributing the shock over a larger
chord-wise extent, resulting in smoother recovery but reduced lift capability.

(a) Velocity Contours with concave TE (b) Velocity Contours with concave TE
(0.5% thickness) at M = 0.8 (0.7% thickness) at M = 0.8

(c) Velocity Contours with concave TE

(0.9% thickness) at M = 0.8

4.3.5 Diverging Trailing Edge

Figure 28: Chord-wise variations of Cp over airfoils with diverging TE at M = 0.7

At Mach 0.7, all configurations exhibit a distinct shock on the upper surface
near the mid-to-aft chord. For the 0.5% divergence case, the shock is stronger
and located further upstream, as seen in both the velocity contour and the
sharp Cp drop. Increasing the trailing edge divergence to 0.7% and 0.9%
weakens the shock and shifts it slightly downstream, with a smoother recovery
in pressure. This results in a reduction in peak suction levels on the upper
surface, indicating less severe compressibility effects. The thicker trailing edge
cases also show a more diffused wake structure, reducing adverse pressure
gradients.

(a) Velocity Contours with diverging TE (b) Velocity Contours with diverging TE
(0.5% thickness) at M = 0.7 (0.7% thickness) at M = 0.7

(c) Velocity Contours with diverging TE

(0.9% thickness) at M = 0.7

Figure 30: Chord-wise variations of Cp over airfoils with diverging TE at M = 0.8

At Mach 0.8, the shocks intensify across all trailing edge divergences due to
higher compressibility effects. The 0.5% divergent trailing edge again
produces the strongest shock, characterized by a sharp Cp discontinuity and a
narrow high-velocity region in the contour plots. For the 0.7% and 0.9%
divergence cases, the shocks are weaker and more smeared, with the 0.9%
case showing the smoothest transition in pressure recovery. However, the
thicker trailing edges induce a slightly larger wake, which may increase base
drag, even though shock-induced drag is mitigated.

(a) Velocity Contours with diverging TE (b) Velocity Contours with diverging TE
(0.5% thickness) at M = 0.8 (0.7% thickness) at M = 0.8

(c) Velocity Contours with diverging TE

(0.9% thickness) at M = 0.8

4.4 Performance Summary: Cd and Cl /Cd Trends

Table 1 (attached in the appendix) summarizes the lowest drag coefficient

(Cd ) and highest lift-to-drag ratio (Cl /Cd ) across all Mach numbers for each
trailing edge geometry, providing a quantitative comparison that
complements the earlier qualitative shock and pressure analyses.
At M = 0.7, the concave (C-shaped) trailing edge with 0.7% thickness achieves
the lowest drag (Cd = 6.8 × 10−4 ) and the highest aerodynamic efficiency
(Cl /Cd ≈ 55.6). The C-shaped aft cavity moderates the adverse pressure
gradient downstream of the leading suction peak, maintaining an attached
boundary layer and smooth pressure recovery. This results in high lift with
minimal drag, as no significant shocks form at this subcritical Mach number.
At M = 0.8, weak shocks begin to appear on the upper surface. The concave
0.7% TE still performs best in terms of Cl /Cd (≈ 24.6), while the minimum
drag shifts to the blunt 0.9% TE (Cd = 1.54 × 10−3 ). The thicker blunt TE
diffuses the wake and smooths base pressure recovery, mitigating early
shock-induced drag rise.
For Mach numbers 0.9–1.1, the flow becomes fully transonic, and wave drag
due to strong shocks dominates. In this regime, the concave 0.7% TE
continues to produce the lowest drag (Cd ≈ 0.004–0.0058), while the
divergent 0.5% TE achieves slightly higher Cl /Cd values compared with other
geometries. The divergent shape helps diffuse the post-shock pressure rise
and delays shock-induced separation, reducing drag growth relative to more
abrupt geometries but increasing the lift.
In summary, the concave 0.7% trailing edge is optimal in the pre and
early-transonic regime (up to M ≈ 0.8), offering excellent lift-to-drag
performance through smooth pressure recovery and attached flow. Beyond
M = 0.9, divergent trailing edges can mitigate drag rise in the presence of
strong shocks, though at the cost of reduced lift and lower overall efficiency.
5 Conclusion
This study investigated the aerodynamic performance of supercritical airfoils
with modified trailing edge geometries at transonic speeds. It was found that
trailing edge shape and thickness significantly influence both drag and
lift-to-drag characteristics. In particular, a concave (C-shaped) 0.7% trailing
edge exhibited the lowest drag at lower Mach numbers (up to M ≈ 0.8),
thanks to its ability to moderate adverse pressure gradients and maintain
attached flow, while preserving strong lift. For higher Mach numbers
(M ≥ 0.9), where shock-induced separation dominates, divergent trailing
edges with 0.5% thickness were slightly more effective in maintaining higher
lift-to-drag ratios, as the diverging shape diffused post-shock pressure rises
and partially delayed boundary layer separation. Blunt trailing edges also
improved performance relative to sharp geometries, particularly in smoothing
base pressure and reducing shock-induced drag rise.
These findings highlight that careful tailoring of trailing edge geometry can
optimize airfoil performance across different transonic regimes, balancing low
drag and high aerodynamic efficiency. Future work could extend this analysis
to different airfoil types, three-dimensional wing effects, and variations in
trailing edge curvature, providing deeper insight into supercritical airfoil
aerodynamics and more refined strategies for drag reduction and lift
enhancement in practical applications.
Bibliography
[1] Brown, P. W., & Viswanath, P. R. (1984). An experimental study of a
supercritical trailing-edge flow. NASA Technical Reports Server (NTRS).
Retrieved from [Link]
[2] Jacquin, L., Molton, P., & Garnier, E. (2021). Experimental investigation
of the transonic buffet cycle on a supercritical airfoil. Experiments in
Fluids, 62, Article 145. Retrieved from
[Link]
[3] Yoo, J. H. (2001). Aerodynamic performance improvement by divergent
trailing edge modification to a supercritical airfoil. Journal of Mechanical
Science and Technology, 15(8), 1152–1158. Retrieved from
[Link]
[4] Rezaei, H., Sobhani, M., & Nejat, A. (2018). Aerodynamic Loads
Alteration by Gurney Flap on Supercritical Airfoils at Transonic Speeds.
arXiv preprint arXiv:1809.06975. Retrieved from
[Link]
[5] Meng, Q., Hu, X., & Zhang, J. (2025). Multi-Objective Optimization of
Transonic Variable Camber Airfoil with Leading- and Trailing-Edge
Deflections Using Kriging Surrogate Model. Aerospace, 12(8), 659.
Retrieved from [Link]
[6] Dai, C., Zhao, Q., & Chen, Z. (2022). Optimization of supercritical airfoil
with jet on the lower surface of trailing edge. Engineering Mechanics,
39(9), 228–234. Retrieved from [Link]
cn/en/article/doi/10.6052/[Link].1000-4750.2022.09.0785
[7] Cheng, Z., Jiang, Y., Liu, J., & Liu, X. (2022). Aerodynamic Characteristics
of Morphing Supercritical Airfoils for Aircraft with All-Stage High
Performance. Applied Sciences, 12(3), 1128. Retrieved from
[Link]
[8] Lv, C., Liu, H., & Wang, H. (2019). Research on the Aerodynamic
Characteristics of Morphing Supercritical Airfoil. In Proceedings of the
International Conference on Intelligent Manufacturing and Industrial
Automation (pp. 577–584). Springer. Retrieved from
[Link]
[9] Livya, A. J., & Pillai, P. S. (2022). Effect of turbulence intensity on
aerodynamic characteristics of extended trailing edge airfoil. Aircraft
Engineering and Aerospace Technology, 95(2), 245–253. Retrieved from
[Link]
AEAT-12-2021-0382/full/html
[10] Koti, V. S., & Khan, M. I. (2018). Numerical Analysis of Transonic Airfoil.
International Journal of Engineering Research and Technology (IJERT),
7(7). Retrieved from
[Link]
[11] Li, Y., Liu, Y., & Wu, J. (2015). Investigation of wave phenomena on a
blunt airfoil with straight and serrated trailing edges. Experiments in
Fluids, 56, Article 200. Retrieved from
[Link]
[12] Harris, & Charles D. (1971). Wind-tunnel investigation of effects of
trailing-edge geometry on a NASA supercritical airfoil section. Retrieved
from [Link]
6 Appendix
Table 1: Complete aerodynamic data for all trailing edge configurations and Mach numbers.
Columns: Lift [N], Drag [N], Cl , Cd , Cl /Cd .

Geometry Mach Lift (N) Drag (N) Cl Cd Cl /Cd

Sharp 0.7 1160.01 30.2548 0.051828079 0.001370913 37.80552772
Sharp 0.8 1435.19 82.1913 0.049094052 0.002811547 17.46158048
Sharp 0.9 594.557 172.602 0.016069706 0.004665093 3.444670398
Sharp 1.0 793.531 303.244 0.017372541 0.006638832 2.616806928
Sharp 1.1 875.737 345.245 0.015844837 0.006246568 2.536566786
Blunt 0.5 0.7 1145 32.4117 0.051882513 0.001468647 35.32674929
Blunt 0.5 0.8 1444.07 86.7998 0.05009794 0.003011275 16.63678949
Blunt 0.5 0.9 562.675 170.679 0.015423545 0.0046785 3.296685591
Blunt 0.5 1.0 741.833 298.097 0.016470916 0.006618647 2.488562448
Blunt 0.5 1.1 816.5 337.405 0.014982436 0.006191241 2.419940428
Blunt 0.7 0.7 751.832 19.6652 0.035697469 0.000891074 40.0611618
Blunt 0.7 0.8 1151.9 53.7566 0.041874295 0.001864934 22.45350203
Blunt 0.7 0.9 250.142 167.04 0.007184794 0.004578751 1.569160032
Blunt 0.7 1.0 506.314 274.216 0.011779657 0.006088417 1.934765104
Blunt 0.7 1.1 572.966 311.194 0.011016818 0.005710281 1.92929544
Blunt 0.9 0.7 596.395 16.8533 0.028986884 0.000763661 37.95780241
Blunt 0.9 0.8 975.678 40.998 0.036306982 0.00142231 25.52676968
Blunt 0.9 0.9 188.949 167.35 0.005555501 0.004587249 1.211074611
Blunt 0.9 1.0 426.799 266.499 0.010164523 0.005917076 1.717828596
Blunt 0.9 1.1 481.279 301.571 0.00947273 0.005533702 1.711824961
Convex 0.5 0.7 1185.33 30.9956 0.053625506 0.00140448 38.18175075
Convex 0.5 0.8 1457.46 85.0061 0.050482968 0.002949047 17.11839969
Convex 0.5 0.9 510.387 169.982 0.013968276 0.004659395 2.997873332
Convex 0.5 1.0 741.668 296.117 0.01644136 0.006574685 2.500707002
Convex 0.5 1.1 819.325 336.992 0.015010634 0.006183663 2.427466388
Convex 0.7 0.7 650.624 19.6161 0.030790429 0.000931385 33.05875222
Convex 0.7 0.8 929.586 42.5643 0.033681494 0.001547313 21.76772909
Convex 0.7 0.9 297.652 158.593 0.008521294 0.004555245 1.870655587
Convex 0.7 1.0 497.155 268.737 0.01152852 0.006252305 1.843883134
Convex 0.7 1.1 586.069 310.021 0.01123169 0.005960991 1.884198623
Convex 0.9 0.7 608.192 15.8903 0.029411216 0.000772324 38.08143872
Convex 0.9 0.8 1007.91 41.4972 0.037317293 0.001544196 24.16616304
Convex 0.9 0.9 526.71 170.852 0.015408306 0.00502341 3.067299881
Convex 0.9 1.0 429.854 266.928 0.010185664 0.006357081 1.602254799
Convex 0.9 1.1 473.07 302.62 0.00926421 0.005956291 1.555365716
Concave 0.5 0.7 1187.9 31.8223 0.053826408 0.00144194 37.32916854
Concave 0.5 0.8 1451.06 86.2003 0.050340439 0.002990477 16.83358411
Concave 0.5 0.9 524.428 170.753 0.014375152 0.004680529 3.071266683
Concave 0.5 1.0 737.04 297.849 0.016364497 0.00661314 2.47454247
Concave 0.5 1.1 814.74 338.011 0.01495014 0.006202361 2.410394928
Concave 0.7 0.7 797.368 15.0173 0.037859548 0.000680467 55.63755632
Concave 0.7 0.8 1185.53 50.4353 0.043096825 0.001749711 24.63082967
Concave 0.7 0.9 218.811 156.373 0.006284878 0.004286357 1.466251549
Concave 0.7 1.0 496.004 260.866 0.011539789 0.005792007 1.992364528
Concave 0.7 1.1 551.736 295.829 0.010608614 0.005428339 1.954302005
Concave 0.9 0.7 1195.81 31.5638 0.05812055 0.001430226 40.63730669
Concave 0.9 0.8 1469.47 86.2337 0.054681997 0.002991635 18.27829668
Concave 0.9 0.9 526.71 170.852 0.015486389 0.004683243 3.306766394
Concave 0.9 1.0 740.061 296.399 0.017625082 0.006580946 2.678198748
Concave 0.9 1.1 816.818 337.106 0.016076946 0.006185755 2.599027267
Divergent 0.5 0.7 1451.8 49.7251 0.064864962 0.002221667 29.19652248
Divergent 0.5 0.8 1841.93 134.014 0.063007552 0.004584264 13.7443103
Divergent 0.5 0.9 1504.09 251.16 0.040652594 0.006788361 5.988573021
Divergent 0.5 1.0 906.223 315.619 0.019839674 0.006909754 2.871256166
Divergent 0.5 1.1 993.823 358.806 0.017981385 0.006491929 2.769805968
Divergent 0.7 0.7 896.078 27.9769 0.041869269 0.00124998 33.4959583
Divergent 0.7 0.8 1411.21 82.4438 0.050484413 0.002820184 17.90110447
Divergent 0.7 0.9 457.701 170.959 0.012937265 0.004620685 2.799858447
Divergent 0.7 1.0 601.967 284.607 0.013782193 0.006230818 2.211939729
Divergent 0.7 1.1 662.693 323.752 0.012539281 0.005857692 2.140652173
Divergent 0.9 0.7 596.395 16.8533 0.028505374 0.000752988 37.85632211
Divergent 0.9 0.8 1306.01 73.714 0.047792017 0.002521561 18.9533464
Continued on next page
 Table 1 – continued from previous page
Geometry Mach Lift (N) Drag (N) Cl Cd Cl /Cd
Divergent 0.9 0.9 335.019 168.59 0.009686642 0.004556656 2.12582252
Divergent 0.9 1.0 501.024 279.988 0.011734035 0.006129695 1.914293337
Divergent 0.9 1.1 561.741 316.999 0.010872755 0.005735509 1.895691256