Quantitative metrics for comparison of in-cylinder velocity fields using particle image velocimetry

Abstract

The in-cylinder flow field plays a key role in determining the combustion performance of internal combustion engines (ICEs) and it is critically important to validate numerical simulations of the flow field by comparison to experimental measurements using techniques such as particle image velocimetry (PIV). With the current trend for high-speed diagnostics, methods for quantitative comparison of vector fields are required which can provide robust spatially averaged results, without inspection of individual flow fields. The quality of match between vector fields, when quantified using current metrics such as the relevance index (RI), can be overly sensitive to the alignment of regions of low velocity such as the tumble vortex centre. This work presents complementary metrics, weighted using a function of the local velocity, for robust quantification of the alignment and magnitude differences between vector fields, the weighted relevance index (WRI) and the weighted magnitude index (WMI). These metrics are also normalized and combined in the combined magnitude and relevance index (CMRI). PIV measurements taken up to every 2 crank angle degrees within the tumble plane of a motored, optically accessible ICE are used to demonstrate the motivation for development and the application of the WRI, WMI, and CMRI metrics. The metrics are used to determine the number of cycles required to provide a representative mean flow field and to identify single cycles of interest. Variability of the flow field is quantified using the metrics and shows high variability in the region of the spark plug near typical ignition timings.

Manifold reduction techniques for the comparison of crank angle-resolved particle image velocimetry (PIV) data and Reynolds-averaged Navier-Stokes (RANS) simulations in a spark ignition direct injection (SIDI) engine

“…This undesired sensitivity in the low speed regions may lead to misleading conclusions for the validity of simulation results when compared with the experimental data. Therefore, in this study new metrics developed by the authors in previous work 34 : the Weighted Relevance Index (WRI) and Weighted Magnitude Index (WMI), are used to quantify the differences between velocity fields in terms of direction and magnitude, respectively.…”

Multi-Plane PIV Measurements in a Gasoline Direct Injection Engine

“…For details of the WRI, WMI and CMRI metrics and their application to analysis of flow fields the reader is directed to [22], [23]. A brief summary is provided here for convenience.…”

“…However, the RI can be overly sensitive to the alignment of regions of low velocity such as the tumble vortex center, where even small differences in the locations of the vortex center between the modelled and measured data can result in very low values of the RI. Recent work ( [22], [23]) has used a weighted function of the local velocity to provide a Weighted Relevance Index (WRI). The same type of approach has been applied to comparison of the magnitudes with a Weighted Magnitude Index (WMI).…”

On the use of particle image velocimetry (PIV) data for the validation of Reynolds averaged Navier-Stokes (RANS) simulations during the intake process of a spark ignition direct injection (SIDI) engine

“…19 Multiple studies have been focusing on the validation of in-cylinder flow fields modelled by RANS simulations. [20][21][22][23][24] As the RANS model only provides an averaged flow field at a certain engine phase (crank angle), the PIV measurements in multiple engine cycles need to be averaged into a single field in order to enable a comparison. One common approach is to use the phase-averaged ensemble mean (averaged flow field at fixed crank angle), assuming the engine operation is a cyclostationary process.…”