ICAS2022 0630 Paper
ICAS2022 0630 Paper
ICAS2022 0630 Paper
BOLTZMANN METHOD
Milind Dhake1 , Chakradhar Thantanapally1 , Salvatore Arcidiacono1 & Santosh Ansumali1,2
1 SankhyaSutra Labs Limited©, Bangalore, India
2 Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore, India
Abstract
This work investigates the transonic flow over a 65◦ swept-back delta wing with a sharp leading edge. The
behaviour of vortex breakdown in cross-flow shocks makes the problem setup highly complex in contrast to
subsonic vortical flows. It has been reported that increasing the Angle of Attack (AoA) leads to a complex
interaction of the strong vortex and shock waves over the wing. To study the onset of the vortex breakdown
on the wing in the presence of shocks, we carried out simulations at AoA of 18.6◦ and 23.6◦ . The Mach and
Reynolds numbers of the simulations are 0.85 and 6 × 106 respectively. The simulations are carried out using
the SankhyaSutra Labs©(SSL’s) in-house higher order Entropic Lattice Boltzmann Method (ELBM) transonic
solver [1]. These results are compared with experimental results reported in the literature[2]. The computed
lift coefficient and pressure distribution show good agreement with experiments. The topological aspects of
vortex breakdown and its onset in the presence of a system of shocks is discussed.
Keywords: transonic flow, shock vortex interaction, delta wing, vortex breakdown
1. Introduction
Vortex breakdown in a delta wing is accompanied by a sudden change in the pressure distribution
which has a detrimental effect on the aerodynamic characteristics such as lift distribution and stall.
This forms the motivation for a thorough study of the vortical flows over slender, sharp edge delta
wings [3, 4, 5, 6]. Vortex breakdown takes place in the rear portion of the delta wing at higher
incidences in subsonic flows as well and is fairly well understood[4, 7, 8]. Transonic flow conditions
in delta wings at a moderately high AoA gives rise to complex interactions between shock waves
and the leading-edge vortex system. The experimental database from Langley National Transonic
Facility (NTF) for transonic flow over a 65◦ swept-back delta wing can be found in [9]. NTF carried out
experiments for a range of Mach and Reynolds numbers and created this comprehensive database
which can be used for validations of numerical simulations. The existing literature on transonic flows
shows that an increase in Mach number affects the size and shape of the vortex systems [10, 11, 12,
13]. Moreover, the appearance of shock waves further complicates the flow structures. These shock
waves are caused by the localised supersonic flow regions over the wing. An analysis of the onset of
the vortex breakdown with increasing AoA is explained by Lucy et al [9]. The flow features of interest
here are the terminating shock surface at the rear end and the cross-flow shocks present along the
leading edge of the delta wing [3]. The pressure distribution on the symmetry plane with increasing
AoA, reported from wind tunnel experiments [17], show that the vortex breakdown is observed beyond
22 degrees in addition to a system of shocks already present towards the trailing end of the wing.
The current investigation focuses on capturing the vortex breakdown and shock locations mentioned
above. Considering the highly complex nature of this flow, an accurate understanding of these flow
features requires a simulation like Large Eddy Simulation (LES) on a well-resolved mesh. Hence, to
carry out these simulations, an in-house higher order ELBM solver [1] was used. The lattice model
consists of 167 discrete velocities with the smallest and largest lattice velocities being 0.5 and 3,
Transonic Flow over Delta Wing using ELBM
respectively. The crystallographic LBM used by the solver improves the capability of LBM by optimally
sampling the domain on a Body-Centred Cubic (BCC) lattice instead of the standard Simple Cubic
(SC) lattice [14]. The BCC lattice also improves the discretization of geometries in the computational
domain. Furthermore, higher-order LB methods have been introduced to extend the applicability
over a wider range of Mach and Knudsen numbers. These higher-order models are characterized
by improved stability, accuracy and thermal properties [15, 16]. The recently introduced Essentially
Entropic LBM [18] ensures the stability of under-resolved simulations, based on the second law of
thermodynamics, without any explicit turbulence modelling. This assures an LES like accuracy. The
use of ELBM ensures a seamless transition from LES to Direct Numerical Simulation (DNS) based
only on increasing the grid resolution without any empirical modelling of small scales.
Figure 1 shows the nine level multi-resolution mesh used for the simulation, where smallest grid
resolution is 0.159 mm with a total of 1.2 billion grid points. The grid resolution changes across
consecutive levels by a factor of two. Simulations are run for 28 flow convection times based on the
root chord length. The total memory used for the simulation is 3.393 TB and 2560 processor cores
were used on 40 computing nodes. Each node is configured with two AMD EPYC 7742 processors,
containing 512 Gigabytes of Samsung DDR4 3200 MT/s memory and interconnected with high speed
Nvidia Mellanox HDR InfiniBand.
3. Results
Qualitative and quantitative comparisons are made with experimental data [2, 17] to understand the
phenomenon of shock-vortex interactions with the current solver. The sharp leading edge and the
slender body cause the flow to separate early which leads to a primary vortex that dominates the
upper surface of the wing. This primary vortex can be observed from the streamlines plotted in
2
Transonic Flow over Delta Wing using ELBM
Figure 2 – Streamlines coloured by Mach number capturing the vortical flow and localised
supersonic regions over the delta wing
Figure 2. The streamlines in blue show the low-velocity vortical structures at the sharp leading edge
and supersonic structures at the core of the primary vortex.
Simulation results are validated by comparing the pressure distribution with experimental data given
by [2]. Figure 3 shows the pressure distribution at selected chord wise locations over the wing.
These locations are taken in accordance with the experimental data. Simulations show good match
3
Transonic Flow over Delta Wing using ELBM
in the suction peak close to apex of the wing i.e x/Cr = 0.2. Although values further downstream
exhibit slight under-prediction, overall pressure distribution matches well when compared with the
experiments. The pressure distribution along the leading edge denoted by the corresponding root
Figure 4 – Coefficient of pressure on upper surface near leading edge along root chord
chord location is plotted in figure 4. These leading edge suction pressure values, which is subset of
the earlier figure 3, agree well with the reference data. The computed value of normal force coefficient
from the simulation is 0.975 as compared to the experimental value of 1.0295.
plotted on the symmetry plane of the wing in figure 5. Clustering of contours in figure 5b near
4
Transonic Flow over Delta Wing using ELBM
the upstream of the sting tip indicates the presence of a normal shock approximately located near
x/Cr = 0.6. Another shock towards the rear end of the wing can also be seen. The location of the
normal shock in figure 5a is observed further downstream towards the rear end of the sting.
To understand the shock structures in the flow, mach contours along the spanwise direction on planes
at different x/Cr locations are plotted. Fig 6 shows the Mach number contours for both angles of
attack. A specific color palette was chosen to identify the supersonic region. It can be observed in
the fig 6 that the velocity near the core of the primary vortex drops between the slice x/Cr = 0.6 and
x/Cr = 0.8 with AoA 23.6◦ , however with AoA of 18.6◦ it continues to be supersonic region and does
not show a vast reduction of velocity. The size of the primary vortex is also larger in the downstream
with AoA 23.6◦ than compared with AoA 18.6◦ . This shows the breakdown of vortex in the region
very close to x/Cr = 0.6 in the 23.6◦ case, along with the presence of normal shock evident from the
pressure distribution in figure 5 very close to the sting tip. Hence, a shock vortex interaction can be
inferred from this data, which may also has a role to play in the breakdown of the vortex.
Figure 7 – Iso-contours of vorticity plotted along the span at different streamwise locations
In order to observe the primary vortex closely, we have plotted the iso-contours of the streamwise
component of vorticity on planes along the wing span at different chord sections as shown in Figure 7.
The contours in figure 7 are coloured with the Mach number in order to trace the supersonic patches
in the flow. This highlights the vortical flow structures along with their corresponding Mach number
5
Transonic Flow over Delta Wing using ELBM
showing possible interactions in the shock-vortex system. It also shows the evidence of cross-flow
shocks present in the flow. The secondary and tertiary separation underneath the primary vortex
is well captured and are more dominant in the rear portion of the wing. As the local flow becomes
supersonic, shock waves appear between the primary and secondary vortices. These shock waves
gives rise to complex interactions between the shock-vortex system.
4. Summary
Flow past a sharp leading edge delta wing in the transonic regime is simulated using the SankhyaSu-
tra Labs© ELBM solver. The surface pressure distribution along span at different chord locations is
compared with experimental results for validation. The coefficient of pressure near the leading edge
at various chord locations is also reported against experimental calculations. Primary vortex struc-
tures over the wing are visualised using velocity streamlines. Mach contours on the symmetry plane
show evidence of normal shock upstream of the sting tip. Mach contours at various chord sections of
the wing are plotted to show the vortex breakdown with angle of attack of 23.6◦ in the region near to
the sting tip. Iso-contours of vorticity along the span at different locations over the chord before and
after critical vortex breakdown angle is plotted.
5. Conclusion
Thus, transonic ELBM can be used to analyse vortical systems and shock structures of flow over a
delta wing without the use of any explicit turbulence models. The current solver has captured the
pressure losses, secondary vortex structures and the presence of terminating and cross-flow shocks
over the wing planform.
References
[1] Hanumantharayappa, M. N., Thantanapally, C., Namburi, M., Kumaran, V., and Ansumali, S. (2021).
LES/DNS of flow past T106 LPT cascade using a higher-order LB model. In AIAA Propulsion and Energy
2021 Forum (p. 3485).
[2] Chu, Julio. Experimental surface pressure data obtained on 65 delta wing across Reynolds number and
Mach number ranges. Vol. 3. National Aeronautics and Space Administration, Langley Research Center,
1996.
[3] Donohoe, S. R. and Bannink, W. J., "Surface Reflective Visualisations of Shock-wave/Vortex Interactions
Above a Delta Wing," AIAA Journal , Vol. 35, No. 10, 1997, pp. 1568-1573
[4] Escudier, M., "Vortex Breakdown: Observations and Explanations," Progress in Aerospace Sciences,
Vol. 25, 1988, pp. 189-229.
[5] Hoeijmakers, H. W. M., "Modelling and Numerical Simulation of Vortex Flow in Aerodynamics" AGARD
Conference Proceedings "Vortex Flow Aerodynamics", AGARD-CP-494, July 1991, pp. 1.1-1.46.
[6] Deléry, J. M., "Physics of Vortical Flows", Journal of Aircraft, Vol. 29, No. 5, 1992, pp. 856-876.
[7] Deléry, J. M., "Aspects of Vortex Breakdown," Progress in Aerospace Sciences, Vol. 30, No. 1, 1994, pp.
1-59
[8] Gursul, I., "Recent Developments in Delta Wing Aerodynamics," Aeronautical Journal, 2004, pp. 437-452
[9] Schiavetta, Lucy and Boelens, Okko and Fritz, Willy. Analysis of transonic flow on a slender delta wing
using CFD 24th AIAA Applied Aerodynamics Conference, p 3171 , 2006
[10] Kandil, O. A., Kandil, H. A., and Liu, C. H., “Shock-Vortex Interaction over a 65-degree Delta Wing in
Transonic Flow,” AIAA 24th Fluid Dynamics Conference, AIAA-93-2973, July 1993
[11] Erickson, G., Schreiner, J., and Rogers, L., “Multiple Vortex and Shock Interactions at Subsonic,Transonic
and Supersonic Speeds,” AIAA-90-3023, 1990
[12] Elsenaar, A., Hjemberg, L., Bütefisch, K. A., and Bannink, W. J., “The International Vortex Flow Ex-
periment,” Validation of Computational Fluid Dynamics - AGARD-CP-437 Volume 1 , AGARD, 1988, pp.
9.1–9.23.
[13] Erickson, G. E. and Rogers, L. W., “Experimental Study of the Vortex Flow Behaviour on a Generic
Fighter Wing at Subsonic and Transonic Speeds,” 19th AIAA Fluid Dynamics, Plasma Dynamics and
Lasers Conference, AIAA-87-1262 , June 1987.
[14] Namburi, M., Krithivasan, S., and Ansumali, S., "Crystallographic lattice Boltzmann method," Scientific
reports, Vol. 6, No. 1,2016, pp. 1-10
6
Transonic Flow over Delta Wing using ELBM
[15] Kolluru, P. K., Atif, M., Namburi, M., and Ansumali, S., "Lattice Boltzmann model for weakly compressible
flows" PhysicalReview E, Vol. 101, No. 1, 2020, p. 013309
[16] Atif, M., Namburi, M., and Ansumali, S., "Higher-order lattice Boltzmann model for thermohydrodynam-
ics," Physical ReviewE, Vol. 98, No. 5, 2018, p. 053311
[17] Elsenaar, A. and Hoeijmakers, H. W. M., "An Experimental Study of the Flow over a Sharp-Edged Delta
Wing at Subsonic and Transonic Speeds," AGARD Conference Proceedings "Vortex Flow Aerodynamics"
, AGARD-CP-494, July 1991, pp. 15.1-15.19.
[18] Atif, M., Kolluru, P. K., Thantanapally, C., and Ansumali, S. (2017). Essentially entropic lattice Boltzmann
model. Physical review letters, 119(24), 240602.
[19] Krithivasan, Siddharth, Siddhant Wahal, and Santosh Ansumali. "Diffused bounce-back condition and
refill algorithm for the lattice Boltzmann method." Physical Review E , 89.3 (2014): 033313.
[20] Houtman, E. M., Bannink, W. J. ."Experimental and numerical investigation of the vortex flow over a delta
wing at transonic speed." AGARD, Vortex Flow Aerodynamics 11 p(SEE N 92-12996 04-02). , (1991)
6. Copyright Statement
The authors confirm that they, and/or their company or organization, hold copyright on all of the
original material included in this paper. The authors also confirm that they have obtained permission,
from the copyright holder of any third party material included in this paper, to publish it as part of their
paper. The authors confirm that they give permission, or have obtained permission from the copyright
holder of this paper, for the publication and distribution of this paper as part of the ICAS proceedings
or as individual off-prints from the proceedings.
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: