Searching stochastic method of the submersible electric pump flow part optimal shape

Abstract

BACKGROUND: Modern approaches based on numerical optimization methods are used in development of pumps in order to achieve the high level of efficiency. These methods with use of parametrical models of flow sections of pumps help to find either global or local extremum of an objective function and to obtain the optimal geometric shape of the flow section. The study object is a dewatering electric submersible pump with cantilever working wheel arrangement used in shipbuilding industry.

AIM: Improvement of energy indicators of the flow section of the high-speed electric submersible pump at the nominal operation mode by means of calculation with the use of the stochastic algorithm of search of the global extremum of an objective function that is pump hydraulic efficiency.

METHODS: Search of the optimal shape if the flow section was conducted with the use of mathematical modelling of three-dimensional flow of a viscous fluid in the computational region of the studied object. The parametrical model of the flow section includes nine geometric variables (four variables for a working wheel, five variables for an outlet element) and their variation ranges. The chosen algorithm of search of the global extremum of the objective function is the Latin hypercube method which ensure uniform distribution of geometric variables in the search space. According to the compiled matrix of experiments with various combinations of parameters, direct fluid dynamics simulations of flow in the flow section were performed.

RESULTS: Simulation results showed that it was managed to improve the pump hydraulic efficiency by 7% in comparison with the initial flow section. According to the velocity field analysis, energy indicators improvement was achieved mainly by increasing the width of channels between an electric motor and a housing, leading to velocities decreasing and lowering the level of hydraulic losses in the outlet element of the flow section of the pump.

CONCLUSONS: The Latin hypercube method proved its effectiveness in search of the global extremum of the pump hydraulic efficiency. The obtained result can be used in further as the initial point in search of the local extremum with direct methods in the area around the found global extremum of the objective function.

Keywords

electric submersible pump / mathematical model / numerical optimization / the Latin hypercube method / global extremum / objective function / turbulence model

References

[1] Gorbatov DA, Zharkovsky AA, Adrianov AV. Consideration of various options for guiding influence on hydraulic losses in the elements of the flow part of a submersible electric pump. In: Hydraulic, hydraulic drives and hydropneumatic automation. Materials of the XXV International Scientific and Technical Conference dedicated to the 75th anniversary of the Department of Fluid Mechanics and Hydraulic Machines named after. V.S. Kvyatkovsky” Research University “MPEI”. Moscow: MEI; 2021:60–65. (In Russ). EDN: PYVRDU

[2] Горбатов Д.А., Жарковский А.А., Адрианов А.В. Влияние различных вариантов направляющего аппарата на гидравлические потери в элементах проточной части погружного электронасоса. В кн.: Гидравлические машины, гидроприводы и гидропневмоавтоматика. Материалы XXV Международной научно-технической конференции, посвященной 75-летию кафедры «Гидромеханики и гидравлических машин им. В.С. Квятковского» Национального исследовательского университета «МЭИ». Москва: МЭИ, 2021. С. 60–65. EDN: PYVRDU

[3] Galdin DN, Kretinin AV, Pechkurov SV. Optimization of the profile of a spatial impeller of a centrifugal pump using a parameterized model of the flow path and an artificial neural network. Nasosy. Turbiny. Sistemy. 2021;3(40):22–31. (In Russ). EDN: EVAYEI

[4] Галдин Д.Н., Кретинин А.В., Печкуров С.В. Оптимизация профиля пространственного рабочего колеса центробежного насоса с использованием параметризованной модели проточной части и искусственной нейронной сети // Насосы. Турбины. Системы. 2021. № 3 (40). С. 22–31. EDN: EVAYEI

[5] Chaburko PS, Lomakin VO, Kuleshova MS, Baulin MN. Comprehensive optimization of the flow part of a sealed pump using the LP-tau search method. Nasosy. Turbiny. Sistemy. 2016;1(18):55–61. (In Russ). EDN: WDEQVX

[6] Чабурко П.С., Ломакин В.О., Кулешова М.С., Баулин М.Н. Комплексная оптимизация проточной части герметичного насоса методом ЛП-тау поиска // Насосы. Турбины. Системы. 2016. № 1 (18). С. 55–61. EDN: WDEQVX

[7] Svoboda DG, Ivanov EA, Zharkovsky AA, Shchutsky SYu. Optimization of the flow path of an axial pump using direct methods. Izvestiya vysshikh uchebnykh zavedeniy. Mashinostroenie. 2022;12(753):116–123. (In Russ). EDN: LGLYSX. doi: 10.18698/0536-1044-2022-12-116-123

[8] Свобода Д.Г., Иванов Е.А., Жарковский А.А., Щуцкий С.Ю. Оптимизация проточной части осевого насоса с использованием прямых методов // Известия высших учебных заведений. Машиностроение. 2022. № 12 (753). С. 116–123. EDN: LGLYSX. doi: 10.18698/0536-1044-2022-12-116-123

[9] Svoboda D.G., Ivanov E.A., Zharkovsky A.A., Shchutsky S.Yu. Optimization of the flow path of an axial pump using a response surface. Izvestiya vysshikh uchebnykh zavedeniy. Mashinostroenie. 2022;8(749):74–83. (In Russ). EDN: IIUJUA. doi: 10.18698/0536-1044-2022-8-74-83

[10] Свобода Д.Г., Иванов Е.А., Жарковский А.А., Щуцкий С.Ю. Оптимизация проточной части осевого насоса с использованием поверхности отклика // Известия высших учебных заведений. Машиностроение. 2022. № 8 (749). С. 74–83. EDN: IIUJUA. doi: 10.18698/0536-1044-2022-8-74-83

[11] Valyukhov SG, Galdin DN, Korotov VV, Rusin VA. Using approximation models to optimize the profile of a centrifugal pump impeller. Nasosy. Turbiny. Sistemy. 2020;2(35):58–65. (In Russ). EDN: RRUDUX

[12] Валюхов С.Г., Галдин Д.Н., Коротов В.В., Русин В.А. Использование аппроксимационных моделей для выполнения оптимизации профиля рабочего колеса центробежного насоса // Насосы. Турбины. Системы. 2020. № 2 (35). С. 58–65. EDN: RRUDUX

[13] Garbaruk AV, Strelets MKh, Travin AK, Shur ML. Modern approaches to turbulence modeling. Sankt Petersburg: Izd-vo Politekhn. un-ta, 2016. (In Russ).

[14] Гарбарук А.В., Стрелец М.Х., Травин А.К., Шур М.Л. Современные подходы к моделированию турбулентности. СПб.: Изд-во Политехн. ун-та, 2016.

[15] Gulich JF. Centrifugal Pumps. Berlin Heidelberg: Springer-Verlag, 2010. doi: 10.1007/978-3-642-12824-0

[16] Gulich J.F. Centrifugal Pumps. Berlin Heidelberg: Springer-Verlag, 2010. doi: 10.1007/978-3-642-12824-0

RIGHTS & PERMISSIONS

Eco-Vector