Integral Length Scales and Time Scales of Turbulence in an Optical Spark-Ignition Engine

Abstract

In-cylinder air flow structures are known to play a major role in mixture preparation and flame development in spark-ignition engines. In this paper both LDV and PIV measurements were undertaken in an optical spark-ignition at 1500 RPM, 0.5 bar inlet plenum pressure. One of the primary PIV planes was vertical, cutting through the centrally located spark plug (tumble plane) inside the pentroof at ignition timing. The other plane was horizontal inside the pentroof 1 mm below the spark plug. LDV was conducted 1 mm below the spark plug on a line from inlet to exhaust but also on a lower line 14 mm below the spark plug. In-cylinder PIV data at specific crank angles in the intake and compression strokes were also analysed on the central tumble plane and on a horizontal plane 14 mm below the spark plug. The combination of both techniques allowed high spatial and temporal resolution as the two data sets complemented each other to provide details of mean flow and turbulence characteristics on different levels, aiming ultimately for quantification of integral time scales and length scales. LDV cycle-resolved analysis distinguished between the classic approach of using the time integral of the autocorrelation function to obtain the integral time scale and a high-frequency cut-off analysis to obtain high-and low-frequency fluctuations about an in-cycle mean.3

Flame front analysis of ethanol, butanol, iso-octane and gasoline in a spark-ignition engine using laser tomography and integral length scale measurements

“…without having injected any fuel in the intake stroke) for both vertical and horizontal planes. well as of LDV data [54,55], showed that w was slightly lower than the mean of u and v (4-5%) which effectively translated to an overestimation of the TKE by ~20% on the PIV tumble plane when using the average of u and v as an estimation for w. This should be borne in mind when making comparisons of data in the area of intersection between the two planes hereafter to avoid confusion.…”

“…To bring this into context, the integral timescale in this engine measured by LDV at ignition timing was of the order 8-16 °CA (depending on velocity component) [54,55].…”

“…The integral timescales were also calculated using the autocorrelation coefficient (by time averaging) and available crank-angle resolved LDV airflow data from the same engine at the same conditions for further interpretation [55]. It is noted here that the PIV measurements campaign in this engine was much larger and what implied from the data set used in the current paper; several parameters were studied, e.g.…”

“…This extended campaign and analysis solidified confidence to the conclusions of the subset of PIV data used in the current study. The complete set of data can be useful to the interested reader for broader interpretation of in-cylinder flow phenomena [54,55,62].…”

“…Selection of turbulence intensity values u is not always straightforward for such an exercise in real engines. Statistical analysis of in-cylinder flow velocities by both LDV and PIV methods in the engine under study showed u values of the order 3.0 m/s on the basis of typical ensemble averaging techniques that lead to 'mean' and 'RMS' values [54,55]. However, methods to distinguish in-cycle bulk velocities from high-frequency turbulence values were also applied to the LDV data of the current engine according to [71] and these led to u values of ~1.5 m/s [55].…”