Abstract
Gradient-based optimisation using adjoints is an increasingly common approach for industrial flow applications. For cases where the flow is largely unsteady however, the adjoint method is still not widely used, in particular because of its prohibitive computational cost and memory footprint. Several methods have been proposed to reduce the peak memory usage, such as checkpointing schemes or checkpoint compression, at the price of increasing the computational cost even further. We investigate incomplete checkpointing as an alternative, which reduces memory usage at almost no extra computational cost, but instead offers a trade-off between memory footprint and the fidelity of the model. The method works by storing only selected physical time steps and using interpolation to reconstruct time steps that have not been stored. We show that this is enough to compute sufficiently accurate adjoint sensitivities for many relevant cases, and does not add significantly to the computational cost. The method works for general cases and does not require to identify periodic cycles in the flow.
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Berggren, M.: Numerical solution of a flow-control problem: vorticity reduction by dynamic boundary action. SIAM J. Sci. Comput. 19(3), 829–860 (1998)
Christakopoulos, F., Jones, D., Müller, J.D.: Pseudo-timestepping and verification for automatic differentiation derived CFD codes. Comput. Fluids 46(1), 174 – 179 (2011). https://doi.org/10.1016/j.compfluid.2011.01.039
Christianson, B.: Reverse accumulation and implicit functions. Optim. Methods Softw. 9(4), 307–322 (1998). https://doi.org/10.1080/10556789808805697
Giles, M., Pierce, N.: Adjoint equations in CFD—duality, boundary conditions and solution behaviour. Am. Inst. Aeronaut. Astronaut. (1997). https://doi.org/10.2514/6.1997-1850
Griewank, A., Walther, A.: Algorithm 799: Revolve: An implementation of checkpointing for the reverse or adjoint mode of computational differentiation. ACM Trans. Math. Softw. 26(1), 19–45 (2000). https://doi.org/10.1145/347837.347846
Gunzburger, M.: Adjoint equation-based methods for control problems in incompressible, viscous flows. In: Flow, Turbulence and Combustion, vol. 65, pp. 249–272. Kluwer Academic Publishers (2000). https://doi.org/10.1023/A:1011455900396
Hascoet, L., Pascual, V.: The Tapenade automatic differentiation tool: Principles, model, and specification. ACM Trans. Math. Softw. 39(3), 20:1–20:43 (2013). https://doi.org/10.1145/2450153.2450158
Heuveline, V., Walther, A.: Online checkpointing for parallel adjoint computation in PDEs: application to goal-oriented adaptivity and flow control. In: Euro-Par 2006 Parallel Processing, pp. 689–699. Springer (2006)
Hückelheim, J., Xu, S., Gugala, M., Müller, J.D.: Time-averaged steady vs. unsteady adjoint: a comparison for cases with mild unsteadiness. In: 53rd AIAA Aerospace Sciences Meeting. American Institute of Aeronautics and Astronautics (2015). https://doi.org/10.2514/6.2015-1953
Jameson, A.: Aerodynamic shape optimization using the adjoint method. In: Lectures at the Von Karman Institute, Brussels (2003)
Krakos, J.A., Darmofal, D.L.: Effect of small-scale output unsteadiness on adjoint-based sensitivity. AIAA J. 48(11), 2611–2623 (2010). https://doi.org/10.2514/1.J050412
Lee, B.J., Liou, M.S.: Unsteady adjoint approach for design optimization of flapping airfoils. AIAA J. 50(11), 2460–2475 (2012). https://doi.org/10.2514/1.J051663
Nadarajah, S., Jameson, A.: Optimum shape design for unsteady three-dimensional viscous flows using a nonlinear frequency-domain method. J. Aircr. 44(5), 1513–1527 (2007). https://doi.org/10.2514/1.27601
Ratanaworabhan, P., Ke, J., Burtscher, M.: Fast lossless compression of scientific floating-point data. In: Proceedings of the Data Compression Conference, DCC ’06, pp. 133–142. IEEE Computer Society, Washington, DC, USA (2006). https://doi.org/10.1109/DCC.2006.35
Rumpfkeil, M., Zingg, D.: A general framework for the optimal control of unsteady flows with applications. In: 45th AIAA Aerospace Sciences Meeting and Exhibit. American Institute of Aeronautics and Astronautics (2007). https://doi.org/10.2514/6.2007-1128
Schneider, T.M., Eckhardt, B., Yorke, J.A.: Turbulence transition and the edge of chaos in pipe flow. Phys. Rev. Lett. 99, 034, 502 (2007). https://doi.org/10.1103/PhysRevLett.99.034502
Tim Wildey, E.C.C., Shadid, J.: Adjoint based a posteriori error estimates using data compression. In: Moitinho de Almeida, C.T.J.P., Díez, P., Parés, N. (eds.) VI International Conference on Adaptive Modeling and Simulation (2013)
Wang, Q., Hu, R., Blonigan, P.: Least squares shadowing sensitivity analysis of chaotic limit cycle oscillations. J. Comput. Phys. 267(0), 210–224 (2014). https://doi.org/10.1016/j.jcp.2014.03.002
Xu, S., Jahn, W., Müller, J.D.: CAD-based shape optimisation with CFD using a discrete adjoint. Int. J. Numer. Methods Fluids 74(3), 153–168 (2014)
Xu, S., Radford, D., Meyer, M., Müller, J.D.: Stabilisation of discrete steady adjoint solvers. J. Comput. Phys. 299, 175–195 (2015)
Acknowledgements
This project has received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no [317006].
This research utilised Queen Mary’s MidPlus computational facilities, supported by QMUL Research-IT and funded by EPSRC grant EP/K000128/1.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer International Publishing AG
About this chapter
Cite this chapter
Hückelheim, J.C., Müller, JD. (2019). Checkpointing with Time Gaps for Unsteady Adjoint CFD. In: Minisci, E., Vasile, M., Periaux, J., Gauger, N., Giannakoglou, K., Quagliarella, D. (eds) Advances in Evolutionary and Deterministic Methods for Design, Optimization and Control in Engineering and Sciences. Computational Methods in Applied Sciences, vol 48. Springer, Cham. https://doi.org/10.1007/978-3-319-89988-6_7
Download citation
DOI: https://doi.org/10.1007/978-3-319-89988-6_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-89986-2
Online ISBN: 978-3-319-89988-6
eBook Packages: EngineeringEngineering (R0)