CFD Online Logo CFD Online URL
Home > Jobs > Job Record #15688

CFD Jobs Database - Job Record #15688

Job Record #15688
TitleHigh-Fidelity Aerostructural Optimization
CategoryPhD Studentship
LocationFrance, Paris
InternationalYes, international applications are welcome
Closure DateFriday, May 31, 2019

Aerostructural optimization is a keystone process to concurrently improve
aerodynamic performance and reduce the structural mass of an aircraft. The
increasing raise of composite materials in modern aircraft structures noticeably
increased structural flexibility. In this context, fluid-structure interactions
are to be considered when designing such highly-flexible structures.
Optimization based on tightly integrated high-fidelity aerostructural analysis
is particularly attractive because many of the objectives and constraints
relevant to aircraft design include both aerodynamic and structural functionals.
To improve the design by considering a large number of parameters,
gradient-based multi-disciplinary design optimization (MDO) techniques are
required and now effective and widely used [1], [2]. However, gradient-based MDO
is efficient if the computation of gradients is fast and accurate. This means
that efficient solution strategies are required for the coupled problem driving
the nonlinear multi-disciplinary analysis (MDA) and for the coupled linear
sensitivity analysis for gradient computation purpose.
Two major approaches have been addressed in the literature: the partitioned
approach and more recently the monolithic approach. The partitioned approach
(also referred to as staggered method) has been used extensively in past
research [3], [4], [5] (including at ONERA [6], [7], [8]). It is easier to
implement, as it allows reusing existing routines from monodisciplinary solvers.
However, this approach may lack of robustness when considering strong
fluid-structure interactions and requires advanced relaxation techniques which
then often lead to large computational costs. On the other hand, the monolithic
approach offers a promising alternative which aims at solving the coupled system
directly using iterative techniques. However, other difficulties linked to the
numerical complexity of the fully coupled system arise and advanced scaling and
preconditioning strategies associated to efficient iterative solvers are
required (see [10] for an illlustration of an advanced iterative Krylov solver
applied to CFD problems). In addition, the nonlinear solution process for the
coupled MDA problem often requires a non trivial initialization step [2], [9].
Also recent demonstrations have only been performed on inviscid flows governed
by the Euler equations, which makes the current conclusions by the authors
difficult to extend to RANS simulations.

 The envisaged work plan of this PhD will be as follows: 
o Litterature review of iterative solvers and preconditioners best suited for
structured-mesh CFD problems. Specifically, preconditioners, domain
decomposition methods and advanced Krylov iterative solvers will be reviewed
(see [11] for a comprehensive review). Some demonstrative numerical experiments
will be performed with the core solver of the elsA CFD software. New promising
solution techniques will be implemented and tested in a modular Python
environment wrapping the elsA solver.
o Review of steady aeroelastic solution process and existing coupled-gradient
computational approaches for aerostructural design [8].
o Explore alternative monolithic solution strategies for the coupled nonlinear
MDA problem and the coupled tangent/adjoint linear sensitivity analysis
problems. This exploratory work will be performed in a modular Python
environment that will allow rapid and flexible prototyping of promising
preconditioners and solvers.
o Demonstrate the new fully-coupled monolithic solver efficiency on a simple
three dimensional flexible shape optimization of a standalone ONERA M6 wing and
compare with the existing staggered solver.
o Application to a full aerostructural optimization of a civil transport
aircraft composite wing box (typically the NASA Common Research Model
[1] Z. Zhang, S. Khosravi, and D. W. Zingg, "High-Fidelity Aerostructural
Optimization with Integrated Geometry Parameterization and Mesh Movement,"
Structural and Multidisciplinary Optimization, vol. 55, pp. 1217-1235, 2017. 
[2] G. K. W. Kenway, G. J. Kennedy, and J. R. R. A. Martins, "Scalable Parallel
Approach for High-Fidelity Steady-State Aeroelastic Analysis and Adjoint
Derivative Computations," AIAA Journal, vol. 52, pp. 935-951, 2014. 
[3] J. J. Reuther, J. J. Alonso, J. R. R. A. Martins, and S. C. Smith, "A
Coupled Aero-Structural Optimization Method for Complete Aircraft
Configurations," in AIAA Paper 99-0187, 1999. 
[4] K. Maute, M. Nikbay, and C. Farhat, "Coupled Analytical Sensitivity Analysis
and Optimization of Three-Dimensional Nonlinear Aeroelastic Systems," AIAA
Journal, vol. 39, pp. 2051-2061, 2001. [5] J. R. R. A. Martins, "A
Coupled-adjoint Method for High-fidelity Aero-structural Optimization," Stanford
University, Ph.D. dissertation, 2002. 
[6] M. Marcelet, J. Peter, and G. Carrier, "Sensitivity Analysis of a Strongly
Coupled System Using the Discrete Direct and Adjoint Approach," Revue Européenne
de mécanique numérique, vol. 17, pp. 1077-1106, 2008. 
[7] C. Blondeau, T. Achard, P. Girodroux-Lavigne, and R. Ohayon, "Recent
Achievements towards Aero-Structure Gradient Computation using High-Fidelity
CFD-CSM in the Onera elsA Software," in International Forum on Aeroelasticity
and Structural Dynamics, IFASD 2015, Saint Petersburg, Russia, 2015. 
[8] T. Achard, C. Blondeau, and R. Ohayon, "High-Fidelity Aerostructural
Gradient Computation Techniques with Application to a Realistic Wing Sizing,"
AIAA Journal, Vol. 56, No. 11 (2018), pp. 4487-4499, doi:10.2514/1.J056736. 
[9] Zimi, J. Zhang and David, W. Zingg, "Efficient Monolithic Solution Algorithm
for High-Fidelity Aerostructural Analysis and Optimization," AIAA Journal, vol.
56, no. 3, pp. 1251-1265, 2018. 
[10] X. Pinel and M. Montagnac, "Block Krylov Methods to Solve Adjoint Problems
in Aerodynamic Design Optimization," AIAA Journal, vol. 51, pp. 2183-2191, 2013.
[11] Y. Saad, “Iterative methods for sparse linear systems”, Second Edition,
SIAM, 2003.

Contact Information:
Please mention the CFD Jobs Database, record #15688 when responding to this ad.
NameBlondeau Christophe
Email ApplicationYes
29 Avenue de la Division Leclerc
Record Data:
Last Modified17:28:07, Saturday, February 23, 2019

[Tell a Friend About this Job Advertisement]

Go to top Go to top