Development of a High-Pressure, High-Temperature, Optically Accessible Continuous-Flow Vessel for Fuel-Injection Experiments

Abstract

The focus of this work was to develop a continuous-flow vessel with extensive optical access for characterization of engine-relevant fuel-injection and spray processes. The spray chamber was designed for non-reacting experiments at pressures up to 1380 kPa (200 psi) and temperatures up to 200°C. Continuous flow of inert “sweep gas” enables acquisition of large statistical data samples and thus potentially enables characterization of stochastic spray processes. A custom flange was designed to hold a common-rail diesel injector, with significant flexibility to accommodate other injectors and injector types in the future. This flexibility, combined with the continuous flow through the chamber, may enable studies of gas-turbine direct-injection spray processes in the future. Overall, the user can control and vary: injection duration, injection pressure, sweep-gas temperature, sweep-gas pressure, and sweep-gas flow rate. The user also can control frequency of replicate injections. There are four flat windows installed orthogonally on the vessel for optical access. Optical data, at present, include global spray properties such as liquid-phase fuel penetration and cone angle. These measurements are made using a high-speed spray-visualization system (up to 100 kHz) consisting of a fast-pulsed LED (light emitting diode) source and a high-speed camera. Experimental control and data acquisition have been set up and synchronized using custom LabVIEW programs. The culmination of this development effort was an initial demonstration experiment to capture high-speed spray-visualization movies of n-heptane injections to determine liquid-phase fuel penetration length (i.e., liquid length) and spray cone angle. In this initial experiment, fuel-injection pressure was ∼120 MPa (1200 bar) and the injection command-pulse duration was 800 μs. At room conditions, liquid length and nominal spray cone angle were ∼170 mm and ∼14.5°, respectively. In contrast, with air flow in the chamber at 100 psi and 100°C, liquid length was considerably shorter at ∼92 mm and spray cone angle was wider at ∼16.5°. Future experiments will include the continuation of these measurements for a wider range of conditions and fuels, extension of high-speed imaging to vapor-phase fuel penetration using schlieren imaging techniques, and detailed characterization of spray properties near the injector nozzle and near the liquid length.

Experimental measurements of n-heptane liquid penetration distance and spray cone angle for steady conditions relevant to early direct-injection low-temperature combustion in diesel engines

“…Specifications and capabilities are briefly discussed here; further details can be found in a previous article. 23 Important vessel specifications and capabilities are listed in Table 1. The CPFV was designed for non-reacting experiments at temperatures and pressures up to 200°C and 1380 kPa (200 lbf/in 2 ), respectively.…”

“…As discussed in previous articles, 23,27 a HSSV system has been designed and implemented to acquire timeresolved fuel-spray measurements. The HSSV system consists of a custom light source (a linear array of light-emitting diodes (LEDs)), a high-speed complementary metal-oxide semiconductor (CMOS) camera (Phantom v7.3; Vision Research, Inc., Wayne, NJ), and electronic devices necessary for synchronizing the light source, camera, and fuel injector.…”

“…As detailed in previous articles, 23,27 high-speed images are post-processed using a custom graphical user interface (GUI) developed in MATLAB. A fixed threshold of 3% of maximum (491 counts out of maximum possible 16,383 counts) is applied to each image to convert the intensity values to either 1 (if intensity is greater than or equal to the threshold) or 0 (if intensity is less than the threshold).…”

“…Previous work has shown that this approach works as long as proper black references are used and maximum intensity of recorded images is sufficiently high. 4,5,23,28 The 3% threshold used in the present work was determined through a detailed sensitivity study, as discussed in a previous article. 23 Example screenshots from the MATLAB GUI are shown in Figure 4.…”

“…Development and validation of the flow vessel has been described in a previous paper. 23 In the current study, liquid penetration distances and spray cone angles have been measured via high-speed spray visualization for a range of ambient temperatures, densities, and flow velocities, as well as injection durations and injection pressures. Results are compared with predictions of liquid length and cone angle based on the Siebers 3 scaling law, as well as predictions of vapor-phase penetration based on the gasjet model of Naber and Siebers.…”