|
Article information
2022 , Volume 27, ¹ 6, p.45-57
Zgoda I.N., Semenov A.A.
High performance computation of thin shell constructions with the use of parallel computations and GPUs
Purpose. The Ritz method often used for calculation of thin-shell structures allows performing a transition from a variational problem to the problem of multidimensional function minimization. The main disadvantage of this method is the nonlinear increase of computational complexity with the increase of terms amount in approximation functions. This fact makes it difficult or impossible to obtain results in cases where high solution accuracy is required or when a shell structure with complex geometry is investigated. At the same time, studies aimed at improving computing performance of the Ritz method, especially in problems of shell computer modelling, are rare. The purpose of this work is to develop algorithms for high-performance calculation of the stress-strain state (SSS) of thin-walled shell structures using the Ritz method. Methodology. Analysis of the current state in the field of computer modelling for shell structures was performed. Shell SSS calculation schemes are studied. Algorithms are proposed to improve computer modelling performance by using different mathematical model properties, multi-core CPUs and graphics accelerators. Findings. Described algorithms were implemented in shell SSS modelling software previously developed by the authors. Tests of the performance have shown that they improve the calculation time by several orders compared with the non-optimized version of the software. Originality/Value. New effective algorithms have been developed for modelling the SSS of thin- walled shell structures using the Ritz method. These algorithms can be applied not only to the calculation of shells, but to any other application of Ritz method to the problem of functional minimization.
[full text]
Keywords: shell constructions, parallel computations, GPU computations, Python, Ritz method
doi: 10.25743/ICT.2022.27.6.005
Author(s):
Zgoda Iurii Nikolaevich
Position: Student
Office: Saint Petersburg State University of Architecture and Civil Engineering
Address: 190005, Russia, St-Petersburg, 2nd Krasnoarmeyskaya st., 4
Phone Office: (812) 575-05-49
E-mail: yurii.zgoda@mail.ru
SPIN-code: 1887-3538Semenov Alexey Aleksandrovich
PhD. , Associate Professor
Position: Head of Chair
Office: Saint Petersburg State University of Architecture and Civil Engineering
Address: 190005, Russia, St-Petersburg, 2nd Krasnoarmeyskaya st., 4
Phone Office: (812) 575-05-49
E-mail: sw.semenov@gmail.com
SPIN-code: 9057-9882 References:
1. Ledenev V.V., Hudyakov A.V. Obolochechnye konstruktsii v stroitelstve [Shell constructions in civil engineering]. Tambov: FGBOU VO “TGTU”; 2016: 272. (In Russ.)
2. Raeesi A., Zohrehheydariha J., Das S. Failure analysis of steel silos subject to wind load. Engineering Failure Analysis. 2017; (79):749–761. DOI:10.1016/j.engfailanal.2017.04.031.
3. Lee S.E., Sahin S., Rigo P., Park M., Paik J.K. Ultimate strength of cylindrical shells with cutouts. Ships and Offshore Structures. 2017; 12(S1):153–173. DOI:10.1080/17445302.2016.1271592.
4. Godoy L.A. Buckling of vertical oil storage steel tanks: review of static buckling studies. Thin-Walled Structures. 2016; (103):1–21. DOI:10.1016/j.tws.2016.01.026.
5. Stupishin L., Nikitin K., Kolesnikov A. Numerical research orthotropic geometrically nonlinear shell stability using the mixed finite element method. IOP Conference Series: Materials Science and Engineering. 2017; (201):012019. DOI:10.1088/1757-899X/201/1/012019.
6. Paccola R.R., Sampaio M.S., Coda H.B. Continuous stress distribution following transverse direction for FEM orthotropic laminated plates and shells. Applied Mathematical Modelling. 2016; 40(15–16):7382–7409. DOI:10.1016/j.apm.2016.03.005.
7. Solovei N.A., Krivenko O.P., Malygina O.A. Finite element models for the analysis of nonlinear deformation of shells stepwise-variable thickness with holes, channels and cavities. Magazine of Civil Engineering. 2015; 1(53):56–69. DOI:10.5862/MCE.53.6.
8. Mellouli H., Jrad H., Wali M., Dammak F. Meshless implementation of arbitrary 3D-shell structures based on a modified first order shear deformation theory. Computers & Mathematics with Applications. 2019; 77(1):34–49. DOI:10.1016/j.camwa.2018.09.010.
9. Faghih Shojaeri M., Ansari R. Variational differential quadrature: a technique to simplify numerical analysis of structures. Applied Mathematical Modeling. 2017; (49):705–738. DOI:10.1016/j.apm.2017.02.052.
10. Deng J., Guasch O., Maxit L., Zheng L. Vibration of cylindrical shells with embedded annular acoustic black holes using the Rayleigh-Ritz method with Gaussian basis functions. Mechanical Systems and Signal Processing. 2021: (150):107225. DOI:10.1016/j.ymssp.2020.107225.
11. Wang J., Li Z.L., Yu W. Structural similitude for the geometric nonlinear buckling of stiffened orthotropic shallow spherical shells by energy approach. Thin-Walled Structures. 2019; (138):430–457. DOI:10.1016/j.tws.2018.02.006.
12. Chai Y., Song Z., Li F. Investigations on the aerothermoelastic properties of composite laminated cylindrical shells with elastic boundaries in supersonic airflow based on the Rayleigh – Ritz method. Aerospace Science and Technology. 2018; (82–83):534–544. DOI:10.1016/j.ast.2018.09.040.
13. Li H., Pang F., Gao C., Huo R. Jacobi – Ritz method for dynamic analysis of laminated composite shallow shells with general elastic restraints. Composite Structures. 2020; (242):112091. DOI:10.1016/j.compstruct.2020.112091.
14. Fazzolari F.A., Carrera E. Refined hierarchical kinematics quasi-3D Ritz models for free vibration analysis of doubly curved FGM shells and sandwich shells with FGM core. Journal of Sound and Vibration. 2014; 333(5):1485–1508. DOI:10.1016/j.jsv.2013.10.030.
15. Karpov V.V. Prochnost’ i ustoychivost’ podkreplennykh obolochek vrashcheniya. V 2 ch. Ch. 2. Vychislitel’nyy eksperiment pri staticheskom mekhanicheskom vozdeystvii [Strength and stability ofthe reinforced shells of rotation. In 2 parts. Pt 2. Computational experiment with static mechanical
impact]. Ìoscow: Fizmatlit; 2010: 248.
16. Weinan E., Yu B. The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems. Communications in Mathematics and Statistics. 2018; 6(1):1–12. DOI:10.48550/arXiv.1710.00211.
17. Semenov A.A., Zgoda Yu.N. Visualization of the stress — strain state of shell structures using virtual and augmented reality technologies. Proceedings of the 30th International Conference on Computer Graphics and Machine Vision (GraphiCon 2020). Pt 2. 2020: 13-1–13-12. Available at:
http://ceur-ws.org/Vol-2744/paper13.pdf.
18. Karpov V.V. Models of the shells having ribs, reinforcement plates and cutouts. International Journal of Solids and Structures. 2018; (146):117–135. DOI:10.1016/j.ijsolstr.2018.03.024.
19. Nickolls J., Buck I., Garland M., Skadron K. Scalable parallel programming with CUDA: is CUDA the parallel programming model that application developers have been waiting for? Queue. 2008; 6(2):40–53. DOI:10.1145/1365490.1365500.
20. Virtanen P., Gommers R., Oliphant T.E. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python. Nature Methods. Nature Publishing Group; 2020; 17(3):261–272. DOI:10.1038/s41592-019-0686-2.
21. Lam S.K., Pitrou A., Seibert S. Numba: a LLVM-based Python JIT compiler. Proceedings of the Second Workshop on the LLVM Compiler Infrastructure in HPC. Association for Computing Machinery; 2015: 1–6. DOI:10.1145/2833157.2833162.
Bibliography link:
Zgoda I.N., Semenov A.A. High performance computation of thin shell constructions with the use of parallel computations and GPUs // Computational technologies. 2022. V. 27. ¹ 6. P. 45-57
|
