Kazuyuki Sugimura

Multi-Objective Design Exploration of a Centrifugal Impeller Accompanied With a Vaned Diffuser

The problem of designing a centrifugal blower was explored using multi-objective genetic algorithm and data mining techniques. Blade-to-blade regions of an impeller and a diffuser were modeled and time-averaged non-uniform inflow to the diffuser was considered. The design objectives were blower efficiency and uniformity of the inflow to the diffuser. The impeller’s shape was represented by NURBS curves and then optimized. The obtained non-dominated solutions showed a trade-off relationship and the design variables controlling the trade-off were found to be related to the dimensions of the vane-less diffuser and the load balance of the impeller. We also applied Decision Tree Analysis and Rough Set Theory to reveal design rules, which led to good performance. Although the design rules derived from the optimization result and data mining results partly agreed with each other, we also clarified that there were some differences due to the characteristics of the data mining methods we used.© 2007 ASME

Kriging-model-based multi-objective robust optimization and trade-off-rule mining using association rule with aspiration vector

A new design method called MORDE (multi-objective robust design exploration), which conducts both a multi-objective robust optimization and data mining for analyzing trade-offs, is proposed. For the robust optimization, probabilistic representation of design parameters is incorporated into a multi-objective genetic algorithm. The means and standard deviations of responses of evaluation functions to uncertainties in design variables are evaluated by descriptive Latin hypercube sampling using Kriging surrogate models. To extract trade-off control rules further, a new approach, which combines the association rule with an “aspiration vector,” is proposed. MORDE is then applied to an industrial design problem concerning a centrifugal fan. Taking dimensional uncertainty into account, MORDE then optimized the means and standard deviations of the resulting distributions of fan efficiency and turbulent noise level. The advantages of MORDE over traditional approaches are shown to be the diversity of the solutions and the quantitative controllability of the trade-off balance among multiple objective functions.

Multi-objective optimization and design rule mining for an aerodynamically efficient and stable centrifugal impeller with a vaned diffuser

This article presents a combined use of multi-objective optimization and quantitative design rule mining methods to improve the aerodynamic efficiency and stability of a centrifugal impeller with a vaned diffuser. A time-averaged but spatially distributed flow is considered at the mixing plane to evaluate the flow uniformity, which affects aerodynamic stability. First, the impellers shape has been optimized using a multi-objective genetic algorithm to improve efficiency and flow uniformity. It was found that the trade-off among non-dominated solutions can be controlled by the vane-less diffusers dimensions and the aerodynamic load distribution. A compromise solution has been experimentally shown to improve both objectives. Second, decision tree analysis and rough set theory have been applied to extract design rules for improving each objective. Although the design rules derived from both methods are consistent with regard to the main effects of design variables, some differences are found regarding the interaction effects.

Aerodynamic Shape Optimization and Knowledge Mining of Centrifugal Fans Using Simulated Annealing Coupled With a Neural Network

An aerodynamic shape optimization method suitable for “inexpensive” centrifugal impellers and diffusers has been developed. The shapes are parameterized using non-uniform rational B-spline curves with special attention being paid to the blade’s edge profiles. A hybrid algorithm combining simulated annealing and a neural network is employed for collaborative optimization. The simulated annealing and neural network take turns in controlling the optimization processes, not only for maximizing the efficiency of global exploration, but also for minimizing the risks of automation failures or of reaching an incorrect optimum. A statistical analysis was also conducted using the neural network to extract design knowledge. By applying the proposed method to a centrifugal impeller and diffuser design problem, we obtained innovative shapes for the leading edge of the impeller and the trailing edge of the diffuser. Important design parameters related to the new shapes were identified through the design space analysis.Copyright

Performance Improvement of a Return Channel in a Multistage Centrifugal Compressor Using Multiobjective Optimization

Multistage centrifugal compressors are widely used in applications in the oil and gas fields where compressors are operated for long periods, and hence their reliability is very important. For its operation to be cost effective, a compressor is required to exhibit high efficiency and a wide operating range. To improve the aerodynamic performance of the centrifugal compressor, many investigations on impellers and diffusers have been conducted [1–5]. Previous investigations suggested that higher efficiency can be achieved by improving the blade loading distribution of the impeller in the case of high and medium flow coefficients. Other papers suggested that wedge-type impellers applied in the case of a low flow-coefficient region also provide higher efficiency. Previous investigations on diffusers reported that half-guide vane-type diffusers provided high efficiency in the case of high flow coefficients. The velocity in the return channel is considerably lower than that in the impeller and diffuser. Therefore, the affect of total pressure loss in the return channel on the overall performance at the design flow rate is relatively small. However, it has been confirmed that the residual swirl flow at the outlet of the return channel leads to insufficient head rise in the next impeller stage [6–8]. Therefore, optimization of the return channel is necessary to minimize the loss in its passage and the residual swirl flow at its outlet. Hildebrandt [9] optimized the return vane and return bend separately by using a multiobjective optimization method to minimize the loss coefficient. Aalburg et al. [10,11] optimized the return channel using the design of experiment (DOE) by decreasing the diffuser outlet-to-inlet radius ratio while maintaining the return channel performance. Aalburg et al. also experimentally confirmed that by using a diffuser with a smaller outlet-to-inlet radius ratio, the efficiency and power could be further decreased. However, in the above studies, the residual swirl flow was not considered in the objective functions used in the optimization. Therefore, in this study a multiobjective optimization based on a genetic algorithm was performed to determine the optimum shape of the return channel to minimize the loss in the passage and the residual swirl flow at the outlet of the return channel. The improved performance of the optimum return channel was experimentally confirmed using a two-stage test compressor.

Multi-Objective Optimization of Pin-Fin Heatsinks

The multi-objective optimization of pin-fin heatsinks using a Kriging approximation model is presented based on systematic experimental results. Thermal resistance and pressure drop are the objective functions in this study. Pareto solutions to the objective functions are illustrated. We derived the design rules for the diameter, height, and pitches for the uniform staggered arrays of pin-fin heatsinks by correlating the objective functions with design variables. We also analyzed the contribution of all design variables to the thermal resistance as well as the pressure drop. We found that both the thermal resistance and the pressure drop are the most sensitive to the ratio of transverse pitch to pin-fin diameter.Copyright

Multi-Objective Robust Design Optimization and Knowledge Mining of a Centrifugal Fan That Takes Dimensional Uncertainty Into Account

A new design approach named MORDE (m ulti-o bjective r obust d esign e xploration), in which multi-objective robust optimization techniques and data mining techniques are combined, is proposed in this paper. We first developed a widely applicable design framework for multi-objective robust optimization. In this framework, probabilistic representation of design variables are introduced and Kriging models are used to approximate relations between design variables with uncertainty and multiple design objectives. A multi-objective genetic algorithm optimizes the mean and standard deviation of the responses. We then applied the framework to the real-world design problem of a centrifugal fan used in a washer-dryer. Taking dimensional uncertainty into account, we optimized the means and standard deviations of the resulting distributions of fan efficiency and turbulent noise level. Steady Reynolds-averaged Navier Stokes simulations were used to build Kriging models that approximate these objective functions. With the obtained non-dominated solutions, we demonstrated how to analyze features of solutions and select design candidates. We also attempted to acquire design knowledge by applying several data mining techniques. Self-organizing map was used to visualize and reuse the high dimensional design data. Decision tree analysis and rough set theory were used to extract design rules to improve the product’s performance. We also discussed differences in types of rules, which were extracted by both methods.Copyright

Optimizing a Suction Channel to Improve Performance of a Centrifugal Compressor Stage

Design parameters for a suction channel of process centrifugal compressors were investigated, and an optimizing method to improve efficiency by using the new design parameters was proposed. Both pressure loss and circumferential flow distortion in the suction channel were evaluated by using computational fluid dynamics (CFD). The main dimensions, which had a large influence on pressure loss and circumferential flow distortion, were identified by using design of experiments (DOE). Next, the passage sectional area ratios Ac /Ae , Ae /As , and Ac /As were found to be the dominant design parameters for the pressure loss and circumferential flow distortion, where Ac , Ae and As are passage sectional areas for the casing upstream side, casing entrance and impeller eye, respectively. Then the shape of the suction channel was optimized using Ac /Ae , Ae /As , and Ac /As . Finally, to evaluate the improvement effect of optimizing the values of Ac /Ae , Ae /As , and Ac /As on compressor stage performance, a base suction channel and an optimized type of suction channel were manufactured and tested. The design suction flow coefficient was 0.1 and the peripheral Mach number was 0.78. Test results showed that the optimized suction channel achieved 3.8% higher stage efficiency than the base one while maintaining the overall operating range from surge to choke. The method for optimizing suction channels by using the three described design parameters was concluded to be very effective for improving the stage efficiency.Copyright

Proposal and Experimental Verification of Design Guidelines for Centrifugal Compressor Impellers With Curvilinear Element Blades to Improve Compressor Performance

This study numerically and experimentally examines the effects of applying curvilinear element blades to fully-shrouded centrifugal impellers on the performance of the centrifugal compressor stages. The design suction coefficient of the target impellers was 0.073. Our previous study confirmed that the application of curvilinear element blades could improve the stage efficiency of similar types of centrifugal compressors. However, a detailed explanation of the relation between the stall margin and the application of the curvilinear element blades remains to be given. The purpose of this study is to investigate the effects of using these blades on the impeller flow field and the stall margin in further detail.The curvilinear element blades we developed for centrifugal turbomachinery were defined by the coordinate transformations between a revolutionary flow-coordinate system and a cylindrical coordinate system. All the blade sections in the transferred cylindrical coordinate system were moved and stacked spanwise in accordance with the given “lean profile,” which meant the spanwise distribution profile of movement of the blade sections, to form a new leaned blade surface. The effects of the curvilinear element blades on the impeller flowfield were investigated by conducting numerical simulations using this method. We next considered the optimum design guidelines for impellers with curvilinear element blades. Then we designed a new impeller using these design guidelines and evaluated the performance improvement of a new compressor stage by conducting numerical simulations.As mentioned in several papers, we numerically confirmed that curvilinear element blades with a negative tangential lean profile improved the velocity distribution and stage efficiency because they help to suppress the secondary flows in the impeller. The negative tangential lean mentioned in this paper represents the lean profile in which the blade hub end leans forward in the direction of the impeller rotation compared to the blade shroud end. At the same time, we also found that the stall margin of these impellers deteriorated due to the increase in relative velocity deceleration near the suction surface of the shroud in the forward part of the impeller. Therefore, we propose new design guidelines for impellers with the curvilinear element blades by applying a negative tangential lean to line element blades in which the blade loading of the shroud side in the forward part of the impeller is reduced. We confirmed from the numerical simulation results that the performance of the new compressor stage improved compared to that in the corresponding conventional one.The new design guidelines for the curvilinear element blades were experimentally verified by comparing the performance of the new compressor stage with the corresponding conventional one. The measured efficiency of the new compressor stage was 2.4 % higher than that of the conventional stage with the stall margin kept comparable. A comparison of the measured velocity distributions at the impeller exit showed that the velocity distribution of the new impeller was much more uniform than that of the conventional one.© 2014 ASME

Conjugate Heat Transfer Analysis in an Actual Gas Turbine Rotor Blade in Comparison With Pyrometer Data

Conjugate Heat Transfer (CHT) was analyzed in a first stage rotor blade in an actual gas turbine. The main objectives of this research were to simulate and validate improvements to the accuracy of predicting temperature on the surfaces of rotor blades in a gas turbine and compare these with experimental results.This simulation was carried out under similar conditions to those during gas turbine operation. Computational grids were generated based on CAD data obtained from the rotor blades with fully resolved rib turbulators and pin fins for both fluid and solid domains during CHT analysis. A tetrahedral mesh with prism layers was used and the y+ of the first mesh adjacent to the wall was kept at less than 1.0 over the whole surface. Thermal barrier coating was modeled by adding thermal resistance at the fluid-solid interfaces. Inlet boundary conditions for the external- and internal-gas-flow regions were defined based on one-dimensional analysis and measured results. Steady Reynolds-averaged Navier-Stokes simulation was carried out using the Shear Stress Transport (SST) turbulence model. The simulated results were compared with measured data obtained from a pyrometer and thermocouple.The temperature distributions predicted from CHT analysis agreed with those obtained from an experiment near the leading edge of the rotor blades. However, the temperature distribution at the center of the pressure side had a difference of 50 K with that obtained from the experimental data. The heat transfer coefficients on the surfaces of the blades were almost equal to those on the pressure side. Thus, we considered that the internal cooling flows contributed more to temperature distributions on the surfaces of the blades rather than the external gas flows. The main stream in the internal cooling flow passages leaned toward one side of the walls and the temperatures on this side became lower than those obtained from the experimental results. Therefore, we suspect CHT analysis underestimated the mixing effect generated by the rib turbulators. It is important to solve the complex flow phenomena in internal cooling passages to better predict the accuracy of temperature distributions on the surfaces of blades.Copyright