Note: All transient jet data taken and the full code for the 1-D jet model are detailed in papers [1] and [2] and are available for download and use with citation at the link below.

Click here for link to download portal for experimental data and 1-D model code

The performance of internal combustion engines that use direct injection strategies is strongly dependent upon the fluid mechanics associated with the fuel injection process. While a significant amount of literature has been published on the behavior of free shear flows with quasi-steady flow rates, little documentation is available on the impact of transient rate of injection on jet characteristics and the subsequent combustion process in engines.

High-speed optical measurements of unsteady liquid fuel jets under a variety of engine conditions have been acquired in the optical engine laboratory at the Engine Research Center. These simultaneous measurements included shadowgraphy of fuel vapor, Mie scattering of liquid fuel, and OH* chemiluminescence for high temperature combustion.

High-speed imaging system used for optical measurements of in-cylinder processes

Ray-tracing diagram of the paths followed by the light imaged for shadowgraphy in vapor measurements

A new adaptive edgefinding algorithm capable of dealing with non-constant backgrounds and illumination levels was developed using the Otsu [3] technique. When comparing the transient jet penetration results to 1-D estimations informed by quasi-steady scalings [4] and that allow for experimental ROI inputs but do not consider the jet breakup length [5,6], deviations are apparent.

oldmodel — Left: Comparison of experimental jet penetration data for one jet of a six-hole fuel injector injecting diesel fuel into air at a density of 22 kg/m3 with a 75 MPa orifice pressure drop to results at the same conditions from the 1-D model described in [5], for both constant ROI [4] and an experimental 75 MPa ROI. The estimated inviscid liquid column Bernoulli penetration for the experimental ROI is also shown.
Right: Mass rate-of-injection measurements at various injection pressures for a six orifice, 110 um orifice diameter, Bosch CRIN 2 fuel injector during the initial transient period. Markers represent every fourth data point. Data processed with 30 kHz low-pass filter.

However, if the initial jet penetration is estimated based on the ROI by assuming an inviscid liquid column of constant diameter (shown in yellow above), it matches the initial penetration well. Therefore, there is a period of time where a portion of the injected fuel retains its initial momentum and is relatively undisturbed by the ambient air: the distance over which this occurs is the “jet breakup length.”

ogif — Example unprocessed jet video for diesel fuel at 15 kg per cubic meter and 125 C ambient conditions, with a fuel rail pressure of 30 MPa.

lgif — Example liftoff length video with Otsu-algorithm tracked boundaries, for diesel fuel at 15 kg per cubic meter and 115 C ambient conditions, with a fuel rail pressure of 30 MPa.

The above GIFs show samples of the visible jet and chemiluminescence videos taken at each case. The liftoff length video is shown in its processed form, after the Otsu technique has been applied. The predictions of prevalent quasi-steady literature assumes early scaling with time of t¹ and long-term scaling with time of t^1/2. The three-zone scaling with time for these transient jets is shown below.

powerfit — Left: Jet penetration on a log-log scale, with curve fits included. Diesel fuel at an intake temperature of 115 C, and in-cylinder density of 15 kg/m³. “a” represents the power scaling with time.
Right: Power fits applied to all experimental conditions tested over three regions.

It is clear that although the power scaling of penetration with time decreases as the jet penetrates, the spread of values at each region is quite large. Also, the average initial scaling is greater than unity, the value predicted by the literature for quasi-steady jets. The fit quality also decreases closer to the injector tip. It is evident that in the early penetration times, a single quasi-steady relationship does not apply for transient jets, leading to the development of the 1-D model, explained here.

References

[1] Neal, N. and Rothamer, D., “Measurement and Characterization of Fully Transient Diesel Fuel Jet Processes in an Optical Engine with Production Injectors,” Experiments in Fluids, Vol. 57, No. 10, pp. 1-19, 2016.

[2] Neal, N. and Rothamer, D., “Evolving 1-D Transient Jet Modeling by Integrating Jet Breakup Physics,” International Journal of Engine Research, Vol. Submitted for publication, No., 2016.

[3] Gonzalez, R. Digital image processing. Prentice Hall, Upper Saddle River, N.J, 2008.

[4] Hiroyasu, H. and Arai, M., “Structures of Fuel Sprays in Diesel Engines,” SAE Technical Paper, 900475, 1990.

[5] Musculus, M.P.B. and Kattke, K., “Entrainment Waves in Diesel Jets,” SAE Technical Paper, Vol. 2009-01-1355, No., 2009.

[6] Pastor, J.V., Lopez, J.J., Garcia, J.M., and Pastor, J.M., “A 1D Model for the Description of Mixing-Controlled Inert Diesel Sprays,” Fuel, Vol. 87, No. 13-14, pp. 2871-2885, Oct 2008.