Search

This paper explores the suitability of Artificial Neural Networks (ANNs) as an enabler of Design Automation in the turbomachinery industry. Specifically, the paper provides 1) a preliminary estimation of the effectiveness of ANNs to define values for design variables of reciprocating compressors (RC) and 2) a comparison of ANNs performance with traditional and more computationally demanding methods like CFD. A tailored ANN trained on a dataset composed by 350+ Baker Hughes' RC automatically assigns values to 8 geometrical variables belonging to multiple parts of the RC in order to satisfy two target conditions linked to their thermodynamic performance. The results highlight that the ANN-assigned parameters return an optimal solution for RC also when the target values do not belong to the training dataset. Their predictive capacity for RC thermodynamic performance, with respect to CFD, are comparable (i.e. less than 2% in terms of calculated absorbed power) and the approach enables a significant gain in terms of computational time (i.e. 2 minutes vs 10 hours). Future perspectives of this work may involve the integration of this tool in an advanced DA method to lead Design Engineers (DEs) during the whole design process. [ABSTRACT FROM AUTHOR]

Computational Fluid Dynamics is widely used for the analysis and the design of turbomachinery blade rows. A well established method is the application of semi-unstructured meshes, that uses a combination of structured meshes in the radial direction and unstructured meshes in the axial as well as the tangential direction. This takes advantage of the approximately two dimensional flow field through the blade rows, whereby a fine radial discretization, excepting the near wall region, is not necessary. Otherwise, it is possible to discretize particular regions, e.g. the leading and trailing edge regions, in the axial and tangential direction without generating unnecessary nodes in the far field. The meshing approach is based on the projection of a two dimensional unstructured mesh defined at a reference surface. Once, the two dimensional mesh is generated the projection is achieved by transfinite interpolation from the reference surface to further radial surfaces using a structured mesh. Due to the modeling of geometrical features, especially fillets, advanced methods for the generation of structured meshes and mesh smoothing algorithms are required. The paper presents two different approaches for the generation of an appropriate structured mesh. The first is based on the solution of elliptic partial differential equations. The second approach is based on the split of the domain into fourteen appropriately arranged blocks. Furthermore, two smoothing methods for two dimensional unstructured meshes, a constrained Laplace smoothing and an optimization based approach, are presented. Regarding a more realistic representation of the geometry, methods for the modeling of cavities, variable clearance sizes and fillets are presented. Finally, a comparison of the smoothing techniques applied to a rotor passage is presented and the influence of chosen geometrical features on the flow solution is evaluated. [ABSTRACT FROM AUTHOR]

The accurate control of the gap between static and rotating components is vital to preserve the mechanical integrity and ensure a correct functioning of any rotating machinery. Moreover, tip leakage above the airfoil tip results in relevant aerodynamic losses. One way to measure and to monitor blade tip gaps is by the so-called Blade Tip Clearance (BTC) technique. Another fundamental phenomenon to control in the turbomachines is the vibration of the blades. For more than half a century, this has been performed by installing strain gauges on the blades and using telemetry to transmit the signals. The Blade Tip Timing (BTT) technique, (i.e. measuring the blade time of arrival from the casing at different angular locations with proximity sensors) is currently being adopted by all manufacturers as a replacement for the classical strain gauge technique because of its non-intrusive character. This paper presents a novel magnetoresistive sensor for blade tip timing and blade tip clearance systems, which offers high temporal and high spatial resolution simultaneously. The sensing element adopted is a Wheatstone bridge of Permalloy elements. The principle of the sensor is based on the variation of magnetic field at the passage of ferromagnetic objects. Two different configurations have been realized, a digital and an analogue sensor. Measurements of tip clearance have been performed in an high speed compressor and the calibration curve is reported. Measurements of blade vibration have been carried out in a dedicated calibration bench; results are presented and discussed. The magnetoresistive sensor is characterized by high repeatability, low manufacturing costs and measurement accuracy in line with the main probes used in turbomachinery testing. The novel sensor has great potential and is capable of fulfilling the requirements for a simultaneous BTC and BTT measurement system. [ABSTRACT FROM AUTHOR]

This project is to design and fabricate the Tesla turbine with gas as working fluid and looking at the potential of installation a Tesla turbine onto the car in order to utilize those energy which originally generated from vehicle's engine and loss to environment due to drag force, and convert partial back to electrical energy and store within battery which hopefully can be used in hybrid vehicles or electrical vehicles for small energy consumption applications. The optimum design is determined based on the study parameters such as distance between disks and total contact area between working fluid and disks. In addition, the Tesla turbine's torque is analyzed using Computational Fluid Dynamics (CFD) Software and found that the best distance between disks is 0.001 m while the most suitable contact area between working fluid and disks is 0.099 m2. At the end of the project, it can be observed that the torque has increased by 29.6 % while the head loss has a significant drop of 26.9 %. [ABSTRACT FROM PUBLISHER]