 Mechanical Engineering  Research Publications
Mechanical Engineering  Research Publications
Permanent URI for this collection
47 results
Filters
Reset filtersSettings
Statistics
Citations
Search Results
Now showing
1  10 of 47

ItemSimultaneous microPIV measurements and realtime control trapping in a crossslot channelAkbaridoust, F ; Philip, J ; Hill, DRA ; Marusic, I (Springer, 20181201)Here we report novel microPIV measurements around micronsized objects that are trapped at the centre of a stagnation point flow generated in a crossslow microchannel using realtime control. The method enables one to obtain accurate velocity and strain rate fields around the trapped objects under straining flows. In previous works, it has been assumed that the flow field measured in the absence of the object is the one experienced by the object in the stagnation point flow. However, the results reveal that this need not be the case and typically the strain rates experienced by the objects are higher. Therefore, simultaneously measuring the flow field around a trapped object is needed to accurately estimate the undisturbed strain rate (away from the trapped object). By combining the microPIV measurements with an analytical solution by Jeffery (Proc R Soc Lond A 102(715):161–179, 1922), we are able to estimate the velocity and strain rate around the trapped object, thus providing a potential fluidic method for characterising mechanical properties of micronsized materials, which are important in biological and other applications.

ItemA framework to develop datadriven turbulence models for flows with organised unsteadinessLav, C ; Sandberg, RD ; Philip, J (Elsevier, 20190415)Turbulence modelling development has received a boost in recent years through assimilation of machine learning methods and increasing availability of highfidelity datasets. This paper presents an approach that develops turbulence models for flows exhibiting organised unsteadiness. The novel framework consists of three parts. First, using triple decomposition, the highfidelity data is split into organised motion and stochastic turbulence. A datadriven approach is then used to develop a closure only for the stochastic part of turbulence. Finally, unsteady calculations are conducted, which resolve the organised structures and model the unresolved turbulence using the developed bespoke turbulence closure. A case study of a wake with vortex shedding behind a normal flat plate, at a Reynolds number of 2,000, based on plate height and freestream velocity, is considered to demonstrate the method. The approach shows significant improvement in mean velocity and Reynold stress profiles compared with standard turbulence models.

ItemA New DataDriven Turbulence Model Framework for Unsteady Flows Applied to WallJet and WallWake FlowsLav, C ; Philip, J ; Sandberg, R (The American Society of Mechanical Engineers, 20191105)The unsteady flow prediction for turbomachinery applications relies heavily on unsteady RANS (URANS). For flows that exhibit vortex shedding, such as the walljet/wake flows considered in this study, URANS is unable to predict the correct momentum mixing with sufficient accuracy. We suggest a novel framework to improve that prediction, whereby the deterministic scales associated with vortex shedding are resolved while the stochastic scales of pure turbulence are modelled. The framework first separates the stochastic from the deterministic length scales and then develops a bespoke turbulence closure for the stochastic scales using a datadriven machinelearning algorithm. The novelty of the method lies in the use of machinelearning to develop closures tailored to URANS calculations. For the walljet/wake flow, three different mass flow ratios (0.86, 1.07 and 1.26) have been considered and a highfidelity dataset of the idealised geometry is utilised for the sake of model development. This study serves as an a priori analysis, where the closures obtained from the machinelearning algorithm are evaluated before their implementation in URANS. The analysis looks at the impact of using all length scales versus the stochastic scales for closure development, and the impact of the extent of the spatial domain for developing the closure. It is found that a twolayer approach, using bespoke trained models for the near wall and the jet/wake regions, produce the best results. Finally, the generalisability of the developed closures is also evaluated by applying a given closure developed using a particular mass flow ratio to the other cases.

ItemSpatial averaging effects on the streamwise and wallnormal velocity measurements in a wallbounded turbulence using a crosswire probeBaidya, R ; Philip, J ; Hutchins, N ; Monty, JP ; Marusic, I (IOP Publishing, 20190801)The spatial averaging effects due to a crosswire probe on the measured turbulence statistics in a wallbounded flow are investigated using a combined approach of direct numerical simulation data, theoretical methods and experiments. In particular, the wire length (l), spacing ( ) and angle ( ) of a crosswire probe configured to measure the streamwise and wallnormal velocities are systematically varied to isolate effects of each parameter. The measured streamwise velocity from a crosswire probe is found to be an average of the filtered velocities sensed by the two wires. Thus, in general, an increase in the sensor dimensions when normalised by viscous units leads to an attenuated variance for the streamwise velocity ( ), resulting from a larger contribution to the spatial averaging process from poorly correlated velocities. In contrast, the variance for the wallnormal velocity ( ) can be amplified, and this is shown to be the result of an additional contributing term (compared to ) due to differences in the filtered wirenormal velocity between the two wires. This additional term leads to a spurious wallnormal velocity signal, resulting in an amplified variance recorded by the crosswire probe. Compared to the streamwise and wallnormal velocity variances, the Reynolds shear stress ( ) perhaps surprisingly shows less variation when l, and are varied. The robustness of Reynolds shear stress to the finite sensor size is due to two effects: (i) Reynolds shear stress is devoid of energetic contributions from the nearisotropic fine scales unlike the and statistics, hence crosswire probe dimensions are typically sufficiently small in terms of viscous unit to adequately capture the statistics for a range of l and investigated; (ii) the dependency arises due to cross terms between the filtered velocities from two wires, however, it turns out that these terms cancel one another in the case of Reynolds shear stress, but not for the and statistics. We note that this does not, however, suggest that is easier to measure accurately than the normal stresses; on the contrary, in a companion paper (Baidya et al 2019 Meas. Sci. Technol. 30 085301) we show that measurements are more prone to errors due to uncertainty in probe geometry and calibration procedure.

ItemSensitivity of turbulent stresses in boundary layers to crosswire probe uncertainties in the geometry and calibration procedureBaidya, R ; Philip, J ; Hutchins, N ; Monty, JP ; Marusic, I (IOP Publishing, 20190801)The sensitivity of measured turbulent stresses to uncertainties in the probe geometry and calibration procedure is investigated for a crosswire probe in a turbulent boundary layer using direct numerical simulation data. The errors investigated are guided by experiments, and to replicate the full experimental procedure, the crosswire calibration procedure is simulated to generate a voltagetovelocity mapping function, which is then utilised to calculate the measured velocity from simulated crosswire voltages. We show that wire misalignment can lead to an incorrect mean wallnormal velocity and Reynolds shear stress in the nearwall region due to the presence of shear. Furthermore, we find that misalignment in the wire orientation cannot be fully accounted for through the calibration procedure, presumably due to increased sensitivity to an outofplane velocity component. This has strong implications if using a generic commercial crosswire probe, since inclining these probes to gain access to the nearwall region can lead to a large error (up to 10%) in turbulent stresses and these errors can manifest in the log region and beyond to half the boundary layer thickness. For uncertainties introduced during the calibration procedure, the Reynolds shear stress is observed to exhibit an elevated sensitivity compared with other turbulent stresses. This is consistent with empirical observations where the repeatability in the Reynolds shear stress is found to be the poorest.

ItemKinematics of local entrainment and detrainment in a turbulent jetMistry, D ; Philip, J ; Dawson, JR (Cambridge University Press (CUP), 20190725)In this paper we investigate the continuous, local exchange of fluid elements as they are entrained and detrained across the turbulent/nonturbulent interface (TNTI) in a high Reynolds number axisymmetric jet. To elucidate characteristic kinematic features of local entrainment and detrainment processes, simultaneous highspeed particle image velocimetry and planar laserinduced fluorescence measurements were undertaken. Using an interfacetracking technique, we evaluate and analyse the conditional dependence of local entrainment velocity in a frame of reference moving with the TNTI in terms of the interface geometry and the local flow field. We find that the local entrainment velocity is intermittent with a characteristic length scale of the order of the Taylor microscale and that the contribution to the net entrainment rate arises from the imbalance between local entrainment and detrainment rates that occurs with a ratio of two parts of entrainment to one part detrainment. On average, an increase in local entrainment is correlated with excursions of the TNTI towards jet centreline into regions of higher streamwise momentum, convex surface curvature facing the turbulent side of the jet and along the leading edges of the interface. In contrast, detrainment is correlated with excursions of the TNTI away from the jet centreline into regions of lower streamwise momentum, concave surface curvature and along the trailing edge. We find that strong entrainment is characterised by a local counterflow velocity field in the frame of reference moving with the TNTI which enhances the transport of rotational and irrotational fluid elements. On the other hand, detrainment is characterised by locally uniform flow fields with the local fluid velocity on either side of the TNTI advecting in the same direction. These local flow patterns and the strength of entrainment or detrainment rates are also observed to be strongly influenced by the presence and relative strength of vortical structures which are of the order of the Taylor microscale that populate the turbulent region along the jet boundary.

ItemPressure and spanwise velocity fluctuations in turbulent channel flows: Logarithmic behavior of moments and coherent structuresMehrez, A ; Philip, J ; Yamamoto, Y ; Tsuji, Y (American Physical Society, 20190402)We study the logarithmic behavior of the pressure variance (p+2) from the datasets obtained from direct numerical simulations of turbulent channel flow for friction Reynolds number Reτ up to 4000. The higherorder moments of p were found to follow logarithmic behaviors at the same distances from the wall where (p+2) shows its log profile. The same results have been confirmed for the spanwise velocity fluctuations w at the same Reynolds numbers, with both p and w following a superGaussian behavior. The minimum Reynolds number for (p+2) and (w+2) log profiles to appear is Reτ≈500, where flow structures O(h) or less were found to significantly contribute to these profiles. The configuration of the hairpin eddy structures obtained from the conditional sampling at different wallnormal locations showed a strong link between p and w fluctuations. Positive pressure fluctuations are located between the legs of the hairpin eddy, while the negative pressure fluctuations are consistent with the head part of the hairpin eddy. Positive and negative spanwise velocity fluctuations are strongly positioned with the legs of the hairpin eddy, consistent with the counterrotating motion resulting from the eddy legs. The structures were also found to be geometrically selfsimilar such that their length and their width increase linearly with the distance from the wall.

ItemReproducing AS/NZS terraintype wind profiles in a shortfetch windtunnelKevin, K ; Philip, J ; Monty, J ; Klewicki, J (AWES, 2018)

ItemExperimental Study of Turbulent NonTurbulent Interface in a Planar Mixing Layer using Kinetic Energy CriteriaBalamurugan, G ; Rodda, A ; Philip, J ; Mandal, AC (FMFP, 2018)Using kinetic energy (KE) criteria, the Turbulent NonTurbulent Interface (TNTI) in a planar mixing layer has been identified and studied in this paper. The mixing layer is generated in a simple way by blocking the top half of the wind tunnel test section with a mesh of suitable solidity. High resolution particle image velocimetry (PIV) is used to acquire the velocity data. The threshold for identifying the interface is obtained using the area algorithm suggested by [2]. The mean and rms profiles of velocity and the spanwise vorticity with respect to the TNTI show finite jump at the interface location. The fractal dimension of the KE interface based on the box counting method is found to be ∼1.2.

ItemImprovement In Unsteady Wake Prediction Through Machine Learning Based Rans Model TrainingLav, C ; Sandberg, R ; Philip, J (ISUAAAT15, 2018)The inability of RANS to correctly capture profile wake dynamics and decay prevents the accurate prediction of unsteady losses and poses additional challenges to the aeromechanic verification of both turbines and compressors. This paper addresses this problem by introducing a novel technique applied to improve wake prediction through a RANS based calculation. In particular, for the first time, a datadriven approach is used to develop bespoke turbulence closures to be used in the context of unsteady RANS (URANS) calculations. The new closure is obtained from an evolutionary machine learning algorithm. The algorithm requires a highfidelity dataset which, in this study, is provided through a DNS for the wake downstream of a normal flat plate with a Reynolds number = 2,000, based on the freestream velocity and plate height. URANS are conducted with the developed closure and the mean flow statistics are compared with the highfidelity data. The results from the developed closure show excellent agreement with the reference data. Furthermore, the generalisability of the developed closure is evaluated by considering other flat plate wake data, which differ in both the aspect ratio and the Reynolds number from the case used in this study to develop the closures. The results once again show remarkable improvement compared with the standard URANS.