Distribution of Liquid Mass in Transient Sprays Measured Using Laser-Plasma-Driven X-Ray Tomography

We report, the use of laser-plasma-driven x rays to reveal the three-dimensional (3D) structure of a highly atomizing water spray. Soft x rays approximately 5 keV are generated by means of a laser-plasma accelerator. Transmission radiography measurements are performed at different angles, by rotating a multihole injector. Using computer tomography, the local liquid volume distribution and its spatial variation are retrieved in 3D, showing up to 55% liquid fraction at the nozzle outlet, which decreases to below 7% within only 1 mm. The resolution of the liquid volume fraction is 0.5% while the spatial resolution of the radiographic images is $11.5\mu \mathrm{m}$. The x-ray source used here provides successful measurements of liquid mass distribution over a relatively large volume and is very promising for the analysis of a variety of challenging transient spray systems, e.g., the injection of liquid synthetic and biofuels used for future clean-combustion applications.

Figure 1

Illustration of the experimental setup where femtosecond soft x-ray pulses are generated using a laser plasma accelerator (left). The incident intense laser pulse (red) propagates from left to right and generates the x-ray beam (purple) after being focused into a gas cell containing 99% helium and 1% nitrogen. The enlarged inset illustrates the laser plasma x-ray generation where the background plasma density is shown in light blue and white. The emitted electrons (blue) are deflected from the x-ray beam using a strong dipole magnet and imaged on a Lanex screen to obtain the electron-beam spectra. An aluminum foil and a Kapton vacuum window stops the laser radiation while allowing the x-ray beam to exit the vacuum chamber. The GDI spray nozzle is mounted on a motorized rotation stage. The enlarged inset shows a microscope image of the six holes of the injector each defined with a number. On the right, an averaged (25 single shots) x-ray image of the spray is provided.

Figure 2

Process for obtaining radiography images of the liquid path length. First, single-shot images (a) are recorded and an averaged image over 25 recordings is deduced (b). Then a calibration curve, shown in (c), is applied. The calibration curves shown in (c) correspond to the transmission of an x-ray beam with the same spectrum as measured at the detector as a function of equivalent path length for pure water and a mixture of water with 20% KI. Finally, the liquid path length is deduced, as exemplified for the rotation angle of $0^{\circ }$ in (d), ${60}^{\circ }$ in (e), and ${120}^{\circ }$ in (f).

Figure 3

(a) 3D tomographic reconstruction of the spray where the false colors indicate the liquid volume fraction in percent. (b) Contours for every 5% liquid volume fraction of jet no.2. (c) Line out of the liquid volume fraction along the center of jet no.2.

Figure 4

Slice of the tomographic reconstruction showing the liquid volume fraction (in %) at different distances from the nozzle. Horizontal slices are shown on the left: (a) just at the exit of the nozzle (each jet is numbered), (b) at $500\mu \mathrm{m}$ below, (c) at 1 mm below, (d) at 1.5 mm below, and (e) at 2 mm below. Vertical slices are shown on the right: (f) along the center of the spray plume no.1, (g) the center of the spray plume no.2 and 6, (h)–(j) along the center of the spray plume no.3–5, respectively.

Figure 5

(a) Experimental setup for 2P LIF imaging of the spray plume no.1. (b) Example of a resulting single-shot 2P LIF image where the laser sheet enters the spray at $70\mu \mathrm{m}$ away from the orifice center. (c) Enlarged view of the near nozzle region (red dashed box) while (d) is the corresponding $30\mu \mathrm{m}$ thickness vertical slice from the tomogram.