Multi-Objective Optimization of a High Specific Speed Centrifugal Volute Pump Using 3D Inverse Design Coupled With CFD Simulations

This paper presents three different multi-objective optimization strategies for a high specific speed centrifugal volute pump design. The objectives of the optimization consist of maximizing the efficiency and minimizing the cavitation while maintaining the Euler head. The first two optimization strategies use a 3D inverse design method to parametrize the blade geometry. Both meridional shape and 3D blade geometry is changed during the optimization. In the first approach Design of Experiment method is used and the efficiency computed from CFD computations, while cavitation is evaluated by using minimum pressure on blade surface predicted by 3D inverse design method. The design matrix is then used to create a surrogate model where optimization is run to find the best tradeoff between cavitation and efficiency. This optimized geometry is manufactured and tested and is found to be 3.9% more efficient than the baseline with little cavitation at high flow. In the second approach the 3D inverse design method output is used to compute the efficiency and cavitation parameters and this leads to considerable reduction to the computational time. The resulting optimized geometry is found to be similar to the more computationally expensive solution based on 3D CFD results. In order to compare the inverse design based optimization to the conventional optimization an equivalent optimization is carried out by parametrizing the blade angle and meridional shape. Two different approaches are used for conventional optimization one in which the blade angle at TE is not constrained and one in which blade angles are constrained. In both cases larger variation in head is obtained when compared with the inverse design approach. This makes it impossible to create an accurate surrogate model. Furthermore, the efficiency levels in the conventional optimization is generally lower than the inverse design based optimization.

Effects of the Circulating Jet Flow on the Suction Flow Field and Cavitation in a Canned Motor Pump

Cavitation is a general phenomenon in centrifugal pumps. When the inlet pressure near the leading edge of the blade is lower than the saturated pressure, cavitation would develop in the impeller. As cavitation occurs, the pump head will drop rapidly and the pump efficiency will decrease. In addition, severe vibration and noise will be induced. Cavitation performance is considered as an important factor in many industrial applications, and affected by various conditions. The canned motor pump is a special type of non-seal centrifugal pump. The pump and motor are integrated. In order to cool the motor and lubricate the bearing during the operation, a portion of fluid, called the circulating flow, is withdrawn from the impeller outlet, and then flows along the cooling circle within the motor. Finally, the circulating fluid moves through the hollow shaft and merges with the main suction flow near the impeller inlet, which can be defined as the circulating jet flow. The jet flow will alter the uniform velocity distribution at the impeller inlet as its direction is opposed to the main suction flow. Consequently, it is expected that the cavitation performance of the pump will drop drastically. It is necessary to analyze the effect of the jet at the pump inlet on the cavitation performance. In this paper, in order to illustrate the jet flow on the pump performance, a numerical simulation method is applied to depict the fluid flow field and cavitation performance of a canned motor pump. For the turbulent model, the standard k-ε turbulent model is adopted. To capture the cavitation performance of the pump, the Zwart-Gerber-Belamri cavitation model was used to investigate the steady cavitation flow through the entire flow channel. It can be seen from the numerical results that the internal jet flow formed by the coolant circulation has a significant effect on cavitation performance. At the pump inlet, the velocity field is divided into three regions: the internal jet flow region, the main-stream region, and the backflow region. The internal jet presents a typical submerged jet structure and its existence results in the non-uniform inlet flow distribution. For the jet flow, it extends to the pump inlet and exhibits an asymmetric characteristic. The static pressure near the impeller inlet with the internal jet is drastically reduced compared to the case without the internal jet structure, and a local low-pressure region occurs around the outlet of the jet nozzle. The cavitation performance of the pump with the internal jet drops obviously. At the off-design condition, the cavitation performance of the pump is seriously degraded. From quantitative data, it indicates that the NPSH3 increases by more than 1.51 times compared with that of the original impeller under design condition. The cavitation inception occurs on the suction side of the leading edge of the impeller near the hub, and then cavitation also occurs near the outlet of the jet nozzle. Finally, the cavitation occurs in the transition region between the internal jet and the main-stream flow regions. So, it is believed that the deterioration of cavitation performance is caused by the combined effect of the non-uniform flow distribution and the cavitation at the internal jet region.

Operational Performance and Locally Resolved Outflow of Brush Seals

As a result of the superior leakage efficiency of brush seals compared to conventional labyrinth seals, compliant contacting filament seals are used to increase the efficiency of jet engines as well as stationary gas and steam turbines. The widespread application of brush seals at different and varying pressure differences combined with variable contacting velocities at the rotor surface requires a profound understanding of the influences of different design parameters on the operational leakage performance. In order to systematically investigate the impact of different design parameters on sealing performance, a new cold air test rig was developed. The new test rig with rotating shaft enables hot-wire anemometry measurements downstream of the seals. These measurements provide insight into the locally resolved flow structure in addition to the integral leakage measurements. For the investigations, one welded and five different clamped brush seals at rotational speeds up to 3000rpm and pressure differences across the seals up to 500kPa are considered. Therefore, the influence of two different designs on the flow through the bristles is presented. For the clamped brush seals, variations of the front and backing plate are investigated. Additionally, the effects of bristle diameter and three different axial inclinations of the bristle pack on the sealing efficiency are shown. Furthermore, initial wear development during the first 30 to 60 hours of brush seal operation at varying experimental conditions is presented and linked to the design parameters. Consequently, the effects of major design aspects on the operational performance of brush seals are examined and presented.

Effects of Trailing Edge Position of Splitter Blade on the Pressure Pulsation in a Low Specific Centrifugal Pump

A combination of experimental and numerical simulation was carried out to analyze influence of trailing edge position of splitter blade on the pressure fluctuation in low specific pumps with and without splitter blades under different flow rates. Performance experiments and PIV tests were performed to verify the results of numerical calculation. Several monitor points were placed in the calculation model pump to collect the pressure fluctuation signals, which were processed by Fast Fourier Transform to obtain the frequency results for further analysis. Besides, turbulence intensity and relative velocity distribution were also analyzed in regions of impeller and volute. The results showed that compared with prototype without splitter blade and the splitter blade schemes, when the trailing edge of splitter blade deviates to the suction side of main blade, the maximum pressure pulsation amplitudes are the lowest at different monitoring points of model pump. And the variation of pressure pulsation amplitude in this scheme is relatively stable with the change of flow rates compared with other schemes. Furthermore, the splitter blade scheme with an appropriate trailing edge position has the lowest average turbulence intensity and optimal relative velocity distribution in main flow passage component. Therefore, this paper proposes a reference scheme of the trailing edge position of the splitter blade to effectively decrease predominate pressure pulsation amplitude.

The Effects of Multiphase, Multi-Viscosity Fluids on Axial Thrust and Cooling Flow Performance of a Canned Motor Pump

A canned motor pump for technology demonstration in subsea oil and gas systems was tested to provide data to understand the design, modeling and performance characteristics of a canned motor pump operating in multiphase flow. This paper discusses the impact of multiphase flow and fluid viscosity on axial thrust and motor cooling flow characteristics. The technology demonstrator is a two-stage, low specific speed (Ns∼550) centrifugal pump designed to deliver 140 gpm at 600 feet of head at 3930 rpm. The 61 hp canned motor is cooled by a small portion of the pump discharge fluid, which is drawn downward through the motor by pump out vanes on the hub of the second stage impeller. The multiphase test loop is equipped for both water and light oil operation in low pressure and ambient temperature conditions. Testing occurred over a range of conditions to simulate varying fluid properties and operating scenarios. Shaft rotational speed varied between 2000 and 4250 rpm with pump liquid flow rates from 25 to 250 gpm. These operating scenarios were repeated for both water and light oil (∼2 cP) with multiphase flow ranging from 0–20% gas volume fraction (GVF) using injected air. Testing results indicate a detectable impact from the different fluids and GVF’s tested, which can be related to features such as the second stage impeller pump-out vane and regions within the motor cavities. In water-air tests, increasing GVF led to the following: motor input power reduced by 5%; axial thrust increased by 100%; motor cooling fluid temperature rise increased by 100%; and pressure rise in the second stage pump out vanes reduced by 30% - directly impacting motor cooling flow rate, temperature rise, and axial thrust. In the oil-air tests, multiphase flow showed similar tendencies with reduced magnitude. Notably, the effects due to air injection do not appear at GVF below 15% with oil-air mixtures, unlike water-air tests which demonstrated effects across all GVFs. The test results provide insight into the behavior of variable viscosity, multiphase flow in the canned motor pump cooling passages, as driven by the second stage impeller pump out vanes. These observed characteristics can be used to design flow control features and evaluate operational impacts, while the performance data obtained can be used to assess the behavior of flow models for this application.

Inter-Blade Vortex Characteristics With the Blockage Effects of Runner Blade in a Francis Hydro Turbine Model

This study presents the numerical analysis on the inter-blade vortex characteristics along with the blockage effects of runner blade in a Francis hydro turbine model with various flow rate conditions. The turbine model showed different flow characteristics in the runner blade passages according to operating conditions, and inter-blade vortex was observed at lower flow rate conditions. This inter-blade vortex can lead to performance reduction, vibration, and instability for smooth operation of turbine systems. The previous study on blockage effects on various runner blade thickness, showed its influence on hydraulic performance and internal flow characteristics at low flow rate conditions. Therefore, the inter-blade vortex characteristics can be altered with the blockage effects at low flow rate conditions in a Francis hydro-turbine. For investigating the internal flow and unsteady pressure characteristics, three-dimensional steady and unsteady Reynolds-averaged Navier-Stokes calculations are performed. These inter-blade vortices were captured at the leading and trailing edges close to the runner hub. These vortex regions showed flow separation and stagnation flow while blockage effects contributed for decreasing the inter-blade vortex at low flow rate conditions.

Experimental Investigation on Cavitation Performance of Torque Converter Using Transparent Model

In this study, we experimentally investigated the influence of the amount of dissolved air in working fluid and the rotation speed ratio of turbine to pump elements on cavitation phenomenon in automotive torque converter. In order to directly observe the cavitation phenomenon, transparent model was used. The applied charge pressure was varied to change the significance of cavitation. The pump and turbine torques were simultaneously measured to clarify the relation between torque performance and cavitation phenomenon. As a result, the cavitation region was found to depend on the speed ratio; cavitation occurred on the suction side of turbine blades at low speed ratios while in the pump region at high speed ratios. The effect of the amount of dissolved air was significant, which enhanced the growth of cavitation bubbles through the deposition of dissolved air. In such cases, with the further decrease of charge pressure, a large number of gaseous cavitation bubbles appeared in the whole flow passage. The torque performance was deteriorated at this stage.

Numerical Method for Studying Bearing Gap Pressure Wave Development and Subsequent Performance Mapping of Externally Pressurized Gas Journal Bearings

In previous work, numerical methods were developed to determine the pressure waves (pressure distribution) in the bearing gap of round externally pressurized gas bearings (EPB’s) that were pressurized through porous liners (PL bearings) or through liners with rows of feedholes (FH bearings). When integrated and differentiated these pressure portraits yield the net hydrodynamic force (FH) between the shaft and the bushing and the mass flow rates through the bearing gap. These results successfully replicated force-deflection curves and mass flow rate data for experimentally tested prototype FH and PL bearings over a wide range of mass flow constriction and clearances. Subsequently the numerical study was expanded to a broader design space of clearance and mass flow compensation. Also, a bearing performance mapping method of mapping the normalized bearing load over the clearance-eccentric deflection plane was developed for different levels of mass compensation. These performance maps produced a very interesting result as they indicated certain areas in the design space of FH bearings where static instability (negative stiffness) would be encountered. This static instability was not observed in the experimental data but is noted in references as known to occur in practice. Because this numerical method is based on the development of pressure wave portraits, the FH pressure wave could then be “dissected” in the areas of the onset of static instability which gave much insight as to the possible causes of static instability. This initial work, then, was perhaps the first to predict where in design space static instability would occur and yield some insight via examination of the corresponding pressure waves as to the cause. The numeric techniques developed, however are in no way limited to non-rotating bearings but are extensible to rotating bearings. The method is also easily extensible to examination of any configuration of feedholes or orifices. Nor is it limited to parallel deflections but can yield results for unbalanced loads. The method is also not limited to round bearings but can be applied to any cross-section configuration of bearing gap cross section such as a 3 lobed bearing or a slotted 3 lobed bearing. Examination of the resulting pressure wave development patterns for different scenarios can be examined to garner insight as to the causes of differing performance that can be applied to alterations towards optimization. Thus sharing in detail the developed numerical method underlying these studies seems worthwhile.

Effect of Rotating Speed on Performance of Electrical Submersible Pump

High speed rotating pump is the current trend in pump’s development and application, which has the advantages of compact size and energy-saving features. The electrical submersible pump, typically called an ESP, is an efficient and reliable artificial-lift method for lifting moderate to high volumes of fluids from wellbores, which have been wildly used for oil or groundwater extraction. To verify the similarity of pump performance under different rotating speeds, a typical ESP is selected as the model pump. By employing the numerical simulation and performance testing methods, the external performance characteristics and internal flow fields under different rotating speeds of the pump are studied. The entire computational domain is established by two stages ESP, and then meshed with the high-quality structured grid based on the Q-type and Y-type block topology. Grid sensitivity analysis is carried out to determine the appropriate mesh density for mesh independent solution. SST k-ω turbulence model with standard wall function in conjunction with Reynolds-Averaged Navier-Stokes (RANS) equations is used to solve the steady flow field. The results show that the increase in the rotating speed could increase the ESP’s head significantly. ESP’s external characteristics under different speeds meet the similar conversion rule quite well. In addition, the flow field distributions in the main flow components of the pump have great similarity at different rotating speeds. The experimental test results for a prototype show good agreement with the simulation results, including the pump’s head, efficiency and axial force. This paper provides a data set for further understanding of the effects of rotating speeds on ESP’s performance and inner flow fields.

Flow Visualization Study of Jet and Bucket Interactions in Traditional and Hooped Pelton Runner

This paper aims to study the flow pattern in and around a bucket of a Traditional and a Hooped Pelton runner at single injector operation and illustrates different stages of jet interaction. High speed photography is used to study the flow pattern, keeping the camera in different positions relative to the jet and to the bucket. It is concluded from the results that the flow visualization study, provides exceptional observations with an absolute frame of reference to mark the bucket duty period of a single-jet Pelton runner. The small scale models display erosion damages at the bucket lips, this indicated that the high pressure occur in the early stage of interaction. This fact is substantiated by the present flow visualization studies for the first time. The uncertainty of the free surface outflow within the Pelton turbine bucket establishes good documentation. The results are helpful to know the interaction between the jet and bucket of Pelton turbine.