A Comparison of the Flow Structures and Losses Within Vaned and Vaneless Stators for Radial Turbines

This paper details the numerical analysis of different vaned and vaneless radial inflow turbine stators. Selected results are presented from a test program carried out to determine performance differences between the radial turbines with vaned stators and vaneless volutes under the same operating conditions. A commercial computational fluid dynamics code was used to develop numerical models of each of the turbine configurations, which were validated using the experimental results. From the numerical models, areas of loss generation in the different stators were identified and compared, and the stator losses were quantified. Predictions showed the vaneless turbine stators to incur lower losses than the corresponding vaned stator at matching operating conditions, in line with the trends in measured performance.. Flow conditions at rotor inlet were studied and validated with internal static pressure measurements so as to judge the levels of circumferential nonuniformity for each stator design. In each case, the vaneless volutes were found to deliver a higher level of uniformity in the rotor inlet pressure field. [DOI: 10.1115/1.2988493]

Reynolds Number Effects on Fully-Developed Pulsed Jets Impinging on Flat Surfaces

A systematic study of the effect of the Reynolds number on the fluid dynamics and turbulence statistics of pulsed jets impinging on a flat surface is presented. It has been suggested that the influence of the Reynolds number may be somewhat different for a jet subjected to pulsation when compared to an equivalent steady jet. A comparative study of both steady and pulsating jets is presented for a Reynolds number range from Re = 4;730 to Re = 10;000. All the other factors that affect the flowfield are kept constant, which are H/d = 3, St = 0.25, and d = 30.5 mm. It was found that for the range of the Reynolds numbers tested, pulsation results in a shortening of the jet core, the centerline axial velocity component declines more rapidly, and higher values of the radial velocity component for r/d > 0.75are observed. As the Reynolds number increases, the jet spreads more rapidly, the turbulent kinetic energy and nondimensional turbulent fluctuations decrease, and the flowfield near the impinging surface changes drastically, which is evident with the development of a turbulent momentum exchange interaction away from the wall for r/d > 1.5.

Numerical Investigation of Optimal Pin Actuator Location on a Supersonic Projectile

The flowfield around a supersonic projectile using a pin actuator control method has been predicted using computational fluid dynamics. It has been predicted using both viscous and inviscid methods for a number of positions. Both methods showed that an optimal longitudinal position exists. However, the inviscid model over-predicted the lateral acceleration due to the difference in shock formation around the pin between the two approaches. The optimal location was predicted independent of solver, however the higher-fidelity solver predicted lower achievable lateral accelerations. This is due to the viscous interactions caused by the pin. The effect of projectile orientation has shown that shielding the pin leads to reduced effectiveness due to the wake of the fin enveloping the pin. When the pin is exposed to onset flow, the forces achieved are increased. There is also an increase in the achievable forces and moments with increasing Mach number.

Investigation of a ship's propeller jet using momentum decay and energy decay

Previous researchers use the velocity decay as an input to investigate the ship’s propeller jet induced scour. A researcher indicated that most of the equations used to predict the stability of various protection systems are often missing a physical background. The momentum decay and energy decay are currently proposed as an initial input for seabed scouring investigation, which are more sensible in physics. Computational fluid dynamics (CFD) and laser Doppler anemometry (LDA) experiments are used to obtain the velocity data and then transforming into momentum and energy decays. The findings proposed several exponential equations of velocity, momentum and energy decays to estimate the region exposed to the seabed scouring.

Transonic Aeroelastic Stability Predictions Under the Influence of Structural Variability

Flutter prediction as currently practiced is almost always deterministic in nature, based on a single structural model that is assumed to represent a fleet of aircraft. However, it is also recognized that there can be significant structural variability, even for different flights of the same aircraft. The safety factor used for flutter clearance is in part meant to account for this variability. Simulation tools can, however, represent the consequences of structural variability in the flutter predictions, providing extra information that could be useful in planning physical tests and assessing risk. The main problem arising for this type of calculation when using high-fidelity tools based on computational fluid dynamics is the computational cost. The current paper uses an eigenvalue-based stability method together with Euler-level aerodynamics and different methods for propagating structural variability to stability predictions. The propagation methods are Monte Carlo, perturbation, and interval analysis. The feasibility of this type of analysis is demonstrated. Results are presented for the Goland wing and a generic fighter configuration.

Transonic Aeroelastic Stability Analysis Using a Kriging-Based Schur Complement Formulation

A method is described to allow searches for transonic aeroelastic instability of realistically sized aircraft models in multidimensional parameter spaces when computational fluid dynamics are used to model the aerodynamics. Aeroelastic instability is predicted from a small nonlinear eigenvalue problem. The approximation of the computationally expensive interaction term modeling the fluid response is formulated to allow the automated and blind search for aeroelastic instability. The approximation uses a kriging interpolation of exact numerical samples covering the parameter space. The approach, demonstrated for the Goland wing and the multidisciplinary optimization transport wing, results in stability analyses over whole flight envelopes at an equivalent cost of several steady-state simulations.

Transonic aeroelastic simulation for instability searches and uncertainty analysis

In this paper the use of eigenvalue stability analysis of very large dimension aeroelastic numerical models arising from the exploitation of computational fluid dynamics is reviewed. A formulation based on a block reduction of the system Jacobian proves powerful to allow various numerical algorithms to be exploited, including frequency domain solvers, reconstruction of a term describing the fluid–structure interaction from the sparse data which incurs the main computational cost, and sampling to place the expensive samples where they are most needed. The stability formulation also allows non-deterministic analysis to be carried out very efficiently through the use of an approximate Newton solver. Finally, the system eigenvectors are exploited to produce nonlinear and parameterised reduced order models for computing limit cycle responses. The performance of the methods is illustrated with results from a number of academic and large dimension aircraft test cases.

Alternative method to evaluate discharge coefficients. Part 2: case study

The purpose of this article is to apply an alternative method whereby discharge coefficients can be estimated for the flow through a poppet valve at various lifts. Presented is the development of an operational quasi-steady flow rig. An engine cylinder head poppet valve was used as the case study. The requirement to directly measure mass flowrates using a standard conventional steady flow apparatus has been eliminated. Transient mass flowrates, pressures and temperatures of air during an inflow test for a poppet valve at various lifts were measured. Mass flowrates were also calculated from measured cylinder gas pressures and corrected for heat transfer. Using both methods to determine the mass flowrates, isentropic discharge coefficients were calculated and shown to compare within +/- 4.0 per cent of steady flow data. A computational fluid dynamics (CFD) validation of the quasi-steady flow rig is also presented.

Hydrodynamic cavitation in micro channels with channel sizes of 100 and 750 micrometers

Decreasing the constriction size and residence time in hydrodynamic cavitation is predicted to give increased hot spot temperatures at bubble collapse and increased radical formation rate. Cavitation in a 100 x 100 mu m(2) rectangular micro channel and in a circular 750 mu m diameter milli channel has been investigated with computational fluid dynamics software and with imaging and radical production experiments. No radical production has been measured in the micro channel. This is probably because there is no spherically symmetrical collapse of the gas pockets in the channel which yield high hot spot temperatures. The potassium iodide oxidation yield in the presence of chlorohydrocarbons in the milli channel of up to 60 nM min(-1) is comparable to values reported on hydrodynamic cavitation in literature, but lower than values for ultrasonic cavitation. These small constrictions can create high apparent cavitation collapse frequencies.

Experimental validation of the performance of a microreactor for the high-throughput screening of catalytic coatings

In this paper, the results of computational fluid dynamics simulations of flow, temperature, and concentration distributions used in the design of a microreactor for the high-throughput screening of catalytic coatings (Mies et al., Chem. Eng. J. 2004, 101, 225) are compared with experimental data, and good agreement is obtained in all cases. The experimental results on flow distribution were obtained from laser Doppler anemometry measurements in the range of Reynolds numbers from 6 to 113. The measured flow nonuniformity in the separate reactor compartments was below 2%. The temperature distribution was obtained from thermocouple measurements. The temperature nonuniformity between the reactor compartments was below 3 K at a maximum heat production rate of 1.3 W in ethylene oxidation at 425 degrees C over CuO/Al2O3/Al coatings. With respect to concentration gradients, a deviation from the average rate of reaction of only 2.3% was obtained at realistic process conditions in the ethylene ammoxidation process over identical Co-ZSM-5 coatings in all reactor compartments. The cross talking noise between separate compartments does not exceed 0.1% when the reactor parts have a smooth surface finish. This illustrates the importance of ultraprecision machining of surfaces in microtechnology, when interfaces cannot be avoided.

Design of a thick-walled screen for flow equalization in microstructured reactors

A systematic computational fluid dynamics (CFD) approach has been applied to design the geometry of the channels of a three-dimensional (thick-walled) screen comprising upstream and downstream sets of elongated channels positioned at an angle of 90 degrees with respect to each other. Such a geometry of the thick-wall screen can effectively drop the ratio of the maximum flow velocity to mean flow velocity below 1.005 in a downstream microstructured reactor at low Reynolds numbers. In this approach the problem of flow equalization reduces to that of flow equalization in the first and second downstream channels of the thick-walled screen. In turn, this requires flow equalization in the corresponding cross-sections of the upstream channels. The validity of the proposed design method was assessed through a case study. The effect of different design parameters on the flow non-uniformity in the downstream channels has been established. The design equation is proposed to calculate the optimum values of the screen parameters. The CFD results on flow distribution were experimentally validated by Laser Doppler Anemometry measurements in the range of Reynolds numbers from 6 to 113. The measured flow non-uniformity in the separate reactor channels was below 2%.

Header design for flow equalization in microstructured reactors

To enhance the uniformity of fluid flow distribution in microreactors, a header configuration consisting of a cone diffuser connected to a thick-walled screen has been proposed. The thick-walled screen consists of two sections: the upstream section constitutes a set of elongated parallel upstream channels and the downstream section constitutes a set of elongated parallel downstream channels positioned at an angle of 90 with respect to the upstream channels. In this approach the problem of flow, equalization reduces to that of flow equalization in the first and second downstream channels of the thick-walled screen. In turn, this requires flow equalization in the corresponding cross sections of the upstream channels. The computational fluid dynamics analysis of the fluid flow maldistribution shows that eight parallel upstream channels with a width of 300-600 pm are required per 1 cm of length for flow equalization. The length to width ratio of these channels has to be > 15. The numerical results suggest that the proposed header-configuration can effectively improve the performance of the downstream microstructured devices, decreasing the ratio of the maximum flow velocity to the mean flow velocity from 2 to 1.005 for a wide range of Reynolds numbers (0.5-10). 2006 American Institute of Chemical Engineers AlChE J, 53: 28-38, 2007.

Implementation of Menter's Transition Model on an Isolated Natural Laminar Flow Nacelle

This study evaluates the implementation of Menter's gamma-Re-theta Transition Model within the CFX12 solver for turbulent transition prediction on a natural laminar flow nacelle. Some challenges associated with this type of modeling have been identified. The computational fluid dynamics transitional flow simulation results are presented for a series of cruise cases with freestream Mach numbers ranging from 0.8 to 0.88, angles of attack from 2 to 0 degrees, and mass flow ratios from 0.60 to 0.75. These were validated with a series of wind-tunnel tests on the nacelle by comparing the predicted and experimental surface pressure distributions and transition locations. A selection of the validation cases are presented in this paper. In all cases, computational fluid dynamics simulations agreed reasonably well with the experiments. The results indicate that Menter's gamma-Re-theta Transition Model is capable of predicting laminar boundary-layer transition to turbulence on a nacelle. Nonetheless, some limitations exist in both the Menter's gamma-Re-theta Transition Model and in the implementation of the computational fluid dynamics model. The implementation of a more comprehensive experimental correlation in Menter's gamma-Re-theta Transition Model, preferably the ones from nacelle experiments, including the effects of compressibility and streamline curvature, is necessary for an accurate transitional flow simulation on a nacelle. In addition, improvements to the computational fluid dynamics model are also suggested, including the consideration of varying distributed surface roughness and an appropriate empirical correction derived from nacelle experimental transition location data.