Back
 ENG  Vol.9 No.10 , October 2017
Effect of Parallel Computing Environment on the Solution Consistency of CFD Simulations—Focused on IC Engines
Abstract: For CFD results to be useful in IC engine analysis, simulation results should be accurate and consistent. However, with wide spread use of parallel computing nowadays, it has been reported that a model would not give the same results against the same input when the parallel computing environment is changed. The effect of parallel environment on simulation results needs to be carefully investigated and understood. In this paper, the solution inconsistency of parallel CFD simulations is investigated. First, the concept of solution inconsistency on parallel computing is reviewed, followed by a systematic CFD simulations specific to IC engine applications. The solution inconsistency against the number of CPU cores was examined using a commercial CFD code CONVERGE. A test matrix was specifically designed to examine the core number effect on engine flow, spray and combustion submodels performance. It was found that the flow field simulation during the gas exchange process is the most sensitive to the number of cores among all submodels examined. An engineering solution was developed where local upwind scheme was used to control the variability, which showed good performance. The implication of the observed inconsistency was also discussed.
Cite this paper: Keum, S. , Grover Jr., R. , Gao, J. , Yang, X. and Kuo, T. (2017) Effect of Parallel Computing Environment on the Solution Consistency of CFD Simulations—Focused on IC Engines. Engineering, 9, 824-847. doi: 10.4236/eng.2017.910049.
References

[1]   Dennis, J.B., Gao, G.R. and Sarika, V. (2012) Determinacy and Repeatability of Parallel Program Schemata. Proceedings of the 2012 Data-Flow Execution Models for Extreme Scale Computing, Minneapolis, 19-23 September 2012, 1-9.
https://doi.org/10.1109/DFM.2012.10

[2]   IEEE Standard (2008) IEEE Standard for Floating-Point Arithmetic. IEEE Standard 754-2008, 29 August 2008.

[3]   Demmel, J. and Nguyen, H.D. (2013) Fast Reproducible Floating-Point Summation. 21st IEEE Symposium on Computer Arithmetic, Austin, 7-10 April 2013, 163-172.
https://doi.org/10.1109/ARITH.2013.9

[4]   Anonymity (2008) Procedure for Estimation and Reporting of Uncertainty Due to Discretization in CFD Applications. Journal of Fluid Engineering, 130.

[5]   Blackford, L.S., Cleary, A., Whaley, R.C., Demmel, J., Dhillon, I., Ren, H., Stanley, K., Dongarra, J. and Hammarling, S. (1997) Practical Experience in the Numerical Dangers of Heterogeneous Computing. ACM Transactions on Mathematical Software, 32, 133-147.
https://doi.org/10.1145/264029.264030

[6]   Villa, O., Daniel, C.-M., Gurumoorthi, V., Marques, A. and Krishnamoorthy, S. (2009) Effects of Floating-Point Non-Associativity on Numerical Computations on Massively Multithreaded Systems. Cray User Group Meeting.

[7]   Bocchino Jr., R., Adve, V.S., Adve, S.V. and Snir, M. (2009) Parallel Programming Must Be Deterministic By Default. Proceeding of HotPar’09 Proceedings of the First USENIX Conference on Hot Topics in Parallelism, Berkeley, 29-31 March 2009.

[8]   Blelloch, G.E., Fineman, J.T., Gibbons, P.B. and Shun, J. (2012) Internally Deterministic Parallel Algorithms Can Be Fast. Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming, New Orleans, 25-29 February 2012, 181-192.

[9]   Senoner, J.-M., Garcia, M., Mendez, S., Staffelbach, G., Vermorel, O. and Poinsot, T. (2008) Growth of Rounding Errors and Repetitivity of Large-Eddy Simulations. AIAA Journal, 46, 1773-1781.
https://doi.org/10.2514/1.34862

[10]   Poinsot, T., Garcia, M., Senoner, J.M., Gicquel, L., Staffelbach, G. and Vermorel, O. (2010) Numerical and Physical Instabilities in Massively Parallel LES of Reacting Flows. Journal of Scientific Computing, 49, 78-93.
https://doi.org/10.1007/s10915-010-9432-8

[11]   Staffelbach, G., Senoner, J.M., Gicquel, L. and Poinsot, T. (2008) Large Eddy Simulation of Combustion on Massively Parallel Machines. High Performance Computing for Computational Science—VECPAR, Toulouse, 24-27 June 2008, 444-464.

[12]   Schneiders, L. (2013) An Accurate Moving Boundary Formulation in Cut-Cell Methods. Journal of Scientific Computing, 235, 786-809.

[13]   Freitas, C.J. (2002) The Issue of Numerical Uncertainty. Applied Mathematical Modeling, 26, 237-248.
https://doi.org/10.1016/S0307-904X(01)00058-0

[14]   Senecal, P.K., Richards, J., Pomraning, E., Yang, T., Dai, M.Z., McDavid, R.M., Patterson, M.A., Hou, S. and Shenthaji, T. (2007) A New Parallel Cut-Cell Cartesian CFD Code for Rapid Grid Generation Applied to In-Cylinder Diesel Engine Simulations. SAE Paper No. 2007-01-0159.

[15]   Ra, Y. and Reitz, R.D. (2008) A Reduced Chemical Kinetic Model for IC Engine Combustion Simulations with Primary Reference Fuels. Combustion and Flame, 155, 713-738.
https://doi.org/10.1016/j.combustflame.2008.05.002

[16]   Oliver, T.A. and Moser, R.D. (2011) Bayesian Uncertainty Quantification Applied to RANS Turbulence Models. Journal of Physics, Conference Series, 318, Section 4.

[17]   Ferziger, J.H. and Peric, M. (2001) Computational Methods for Fluid Dynamics. Springer, Berlin.

[18]   Trefethen, L.H. (1996) Finite Difference and Spectral Methods for Ordinary and Partial Differential Equations. Unpublished Text.
http://people.maths.ox.ac.uk/trefethen/pdetext.html

[19]   Gorle, C., Emory, M., Larsson, J. and Iaccarino, G. (2012) Epistemic Uncertainty Quantification of RANS Modeling of the Flow over a Wavy Wall. Center for Turbulence Research Annual Brief.

 
 
Top