Results 1 - 10 of 112
Results 1 - 10 of 112. Search took: 0.022 seconds
|Sort by: date | relevance|
[en] This study applies a novel neural-network technique, support vector regression (SVR), to forecast reliability in engine systems. The aim of this study is to examine the feasibility of SVR in systems reliability prediction by comparing it with the existing neural-network approaches and the autoregressive integrated moving average (ARIMA) model. To build an effective SVR model, SVR's parameters must be set carefully. This study proposes a novel approach, known as GA-SVR, which searches for SVR's optimal parameters using real-value genetic algorithms, and then adopts the optimal parameters to construct the SVR models. A real reliability data for 40 suits of turbochargers were employed as the data set. The experimental results demonstrate that SVR outperforms the existing neural-network approaches and the traditional ARIMA models based on the normalized root mean square error and mean absolute percentage error
[en] The article discusses performance of a power module in a wave-powered navigational buoy developed by National Institute of Ocean Technology. The power module of the buoy consists of an impulse turbine and a generator. Both unidirectional and bidirectional impulse turbine were tested for their performance. The article describes the selection, performance assessment tests and experimentation on these power modules carried out in the laboratory and in open sea trials. It was observed that unidirectional impulse turbine gave better performance than bidirectional impulse turbine for a given range of flow coefficient and pressure drop across the turbine. The performance of the wave powered navigational buoy in the open sea trials has given confidence on its use as a product and has also led to knowledge enhancement for scaling up floating wave energy devices. (author)
[en] The design of low-solidity vaned diffusers and the effect on the performance of a turbocharger compressor is discussed. The effect of vane number and turning angle was investigated while maintaining a basic design with a leading edge angle of 70 .deg., leading and trailing edge radius ratios of 1.1 and 1.3. All results are compared with those obtained with the standard vaneless diffuser configuration and it was shown that all designs increased and shifted the pressure ratio to reduced flowrates. Despite the low-solidity configuration none of the vane designs provided a broad operating range, and the vane leading edge angle was not main factor that system went into the surge condition. The diffuser of higher trailing edge angle improved the flow range for the compressor to operate at lower flow region
[en] A technique to measure radiated BPF sound from an automotive turbocharger compressor impeller is proposed in this paper. Where there are high-level background noises in the measurement environment, it is difficult to discriminate the target component from the background. Since the effort of measuring BPF sound was taken in a room with such condition in this study, no discrete BPF peak was initially found on the sound spectrum. Taking its directionality into consideration, a microphone covered with a parabolic cone was selected and using this technique, the discrete peak of BPF was clearly observed. Since the level of measured sound was amplified due to the area-integration effect, correction was needed to obtain the real level. To do so, sound measurements with and without a parabolic cone were conducted for the fixed source and their level differences were used as correction factors. Consideration is given to the sound propagation mechanism utilizing measured BPF as well as the result of a simple model experiment. The present method is generally applicable to sound measurements conducted with a high level of background noise
[en] Highlights: • Unsteady turbine performance prediction by integrating the 1-D and meanline models. • The optimum discretization length/diameter ratio is identified. • No improvement is gained by increasing the number of rotor entries. • The predicted instantaneous mass flow and output power are analysed in detail. - Abstract: Stringent emission regulations are driving engine manufacturers to increase investment into enabling technologies to achieve better specific fuel consumption, thermal efficiency and most importantly carbon reduction. Engine downsizing is seen as a key enabler to successfully achieve all of these requirements. Boosting through turbocharging is widely regarded as one of the most promising technologies for engine downsizing. However, the wide range of engine speeds and loads requires enhanced quality of engine-turbocharger matching, compared to the conventional approach which considers only the full load condition. Thus, development of computational models capable of predicting the unsteady behaviour of a turbocharger turbine is crucial to the overall matching process. A purely one-dimensional (1D) turbine model is capable of good unsteady swallowing capacity predictions, however it has not been fully exploited to predict instantaneous turbine power. On the contrary, meanline models (zero-dimensional) are regarded as a good tool to determine turbine efficiency in steady state but they do not include any information about the pressure wave action occurring within the turbine. This paper explores an alternative methodology to predict instantaneous turbine power and swallowing capacity by integrating one-dimensional and meanline models. A single entry mixed-flow turbine is modelled using a 1D gas dynamic code to solve the unsteady flow state in the volute, consequently used as the input for a meanline model to evaluate the instantaneous turbine power. The key in the effectiveness of this methodology relies on the synchronisation of the flow information of different time scales. The model is validated against experimental data generated at Imperial College London under steady and pulsating flow conditions. Three rotational speeds (27.0, 43.0, and 53.7 rps/√(K)) and four pulse flow frequencies (20 to 80 Hz) are considered for performance validation. In addition to the turbine performance, the common level of unsteadiness is also compared based on Strouhal number evaluations. Furthermore, comparisons are made with the quasi-steady assumption in order to understand the strengths and weaknesses of the current method for effective unsteady turbine performance prediction
[en] In this study, we investigate the performance variations of an automotive turbocharger compressor with respect to the height variation of the recirculating casing treatment (RCT). We use three RCT heights, namely 1.2 mm, 1.5 mm, and 1.8 mm. We vary the compressor speed from 90,000 to 150,000 rpm, and the flow rate from 0.015 kg/s to 0.08 kg/s. The calculation results of the total pressure ratio and isentropic efficiency showed good agreement with the performance data provided by the manufacturer within a 0.7 percent error. The results showed that the RCT heights of 1.2 mm, 1.8 mm, and 1.5 mm, in that order, exhibited a more uniform pressure distribution, higher pressure ratio, and wider operational range. As the number of revolutions per minute increased, we obtained typical characteristics of a compressor map having a narrower operational range in the region of higher pressure ratio
[en] Highlights: • Diverse instability patterns at full-operating conditions. • Compressor experiences stable states, stall and surge successively at low speeds. • Stable states, mild surge, deep surge successively occurs at middle and high speeds. • Different flow instability inducements at different operating conditions. - Abstract: The paper is proposed to experimentally investigate the flow instability of centrifugal compressors with vaned diffuser, and the instability inducement and mechanism at full operating range are presented and analyzed in detail. In this paper, the steady performance and transient evolution process of the compressor are measured by steady and dynamic sensors. Experimental results show that the compressor experiences diverse flow instability patterns at different operating conditions. At low rotating speeds, the compressor undergoes stable state, stall and deep surge successively with mass flow rate decreases and the inducement of the flow instability is the wave-like instability disturbance at the impeller inlet. At middle rotating speeds, the compressor experiences stable state, mild surge and deep surge successively with mass flow rate reduces, and the inducement of the flow instability is the spike-shaped instability disturbance at the diffuser inlet. At high rotating speeds, the compressor still experiences stable state, mild surge and deep surge, whereas the flow instability is dominantly induced by the characteristic of the entire compression system. The research results of this paper promote understandings of the instability mechanism and improvements of the stability of centrifugal compressors.
[en] Highlights: • An optimization is performed for a variable geometry turbocharger (VGT) vane. • An experimental setup with different vanes design is designed. • Central composite design (CCD) is applied to reduce the cases and tests. • Results are compared to free space parameter (FSP) theory. - Abstract: In this paper, central composite design (CCD) based on Design of Experiment (DoE) is applied to obtain an optimal design of the vane geometry for a variable geometry turbine (VGT). The design is tested at four different pressure ratios (1.25, 1.5, 1.75 and 2.0) on a Garrett GT1541V turbocharger. Seventeen different cases for the inlet guide vanes are proposed. All cases, each having a unique combination of vane height, thickness, length and angle, has been produced via 3D printing. The goal of this study is to ascertain how vane geometry impacts turbine efficiency, so as to arrive at the ideal configuration for this specific turbine for the investigated range of operating conditions. As a main outcome, the results demonstrate that the applied vane angle has the strongest impact on the turbine efficiency, with smaller angles yielding the most favorable results. After CCD analysis, an optimized design for the vanes geometry with 76.31% efficiency (averagely in all pressures) is proposed. As a final step, all cases are analyzed from a free space parameter (FSP) perspective, with the theoretically optimal design (e.g., FSP < 5) corresponding nicely to the best experimental results.
[en] The detailed flow characteristics of three high-pressure-ratio mixed-flow turbines were investigated under both steady and pulsating flow conditions. Two rotors featured a constant inlet blade angle, one with 12 blades and the second with 10. The third rotor was shorter and had a nominally constant incidence angle. The rotors find application on an automotive high-speed large commercial diesel turbocharger. The steady flow entering and exiting the blades has been quantified by a laser Doppler velocimetry system. The measurements were performed at a plane 3.0-mm ahead of the rotor leading edge and 9.5-mm downstream the rotor trailing edge. The turbine test conditions corresponded to the peak efficiency point at two rotational speeds, 29,400 and 41,300-rpm. The results were resolved in a blade-to-blade sense to examine fully the nature of the flow at turbocharger representative conditions. A correlation between the combined effects of incidence and exit flow angle with the isentropic efficiency has been verified. Regarding pulsating flow, the velocity data and their corresponding instantaneous velocity triangles were resolved in a blade-to-blade sense to understand better the complex phenomenon. The results highlighted the potential of a nominally constant incidence design to absorb better the inadequacy of the volute to discharge the exhaust gas uniformly along the blade leading edge. A double vortex rotating in a clockwise sense propagated on the plane normal to the meridional direction. This should be attributed to the effect of the passing blade that was acting as a blockage to the flow. The phenomenon was more pronounced near the suction and pressure surfaces of the blade, but diminished at the mid-passage region where the flow exhibited its best level of guidance. The full mixed flow turbine stage under transient conditions was modelled firstly with a 'steady' inlet and secondly with a 'pulsating' inlet boundary condition. In both cases comparison was made to experiment performance and LDV measurements. With the steady inlet boundary condition, a high level of accuracy was achieved when compared to the experimental performance and velocity field. The velocity along the leading edge showed the same discrepancy as the single passage analysis that is with the radial and axial component from mid span to the blade tip. At the trailing edge features identified in the experimental data are identified in the numerical results; the velocity field appears more 'diffused' across the plane as per the experimental data than from the single passage analysis. With the pulsating inlet boundary, the predicted velocity traces in the volute and close to the turbine lead and trailing edge show excellent agreement in both form (against time) and magnitude
[en] Highlights: • Matching of inlet and outlet distortions has a great effect on the performance. • The matching mechanism is analyzed. • The criteria for the matching is proposed. • An estimate formula for the best matching is proposed. - Abstract: The centrifugal compressor is a very common type of effective energy conversion device, which is used in a range of industrial processing equipment and automotive turbochargers. Due to the requirements and the limited installation space, centrifugal compressors are always connected with complex inlet and outlet pipe systems, including various kinds of bends, struts, and volutes with asymmetric geometries. These asymmetric components induce distortions at the compressor inlet and the outlet, which exert a strong influence on the compressor performance and flow field. This paper employs full-annulus unsteady simulations to study the matching of inlet and outlet distortions by adding a distortion model to compressor inlet and outlet. The results show that the matching does have a large influence on the compressor performance and flow field and there does exists the best matching. The best matching can neutralize the inlet and outlet distortions, keep the uniformity of flow parameters in the impeller and the diffuser, and inhibit the oscillation of mass flow in one blade passage. In addition, the peak efficiency under the best matching can be improved 0.67% and 1.16% at 80% and 100% nominal rotation speed, respectively. Finally, the matching criteria are proposed and validated, and an estimate formula is established to provide some guidance for the installation and integrated design of compression system.