"Penetrating Vision," Feature Article, May 2003

penetrating vision

Now you see it? X-rays are probing the unknowns of fuel injector sprays.

by John DeGaspari, Associate Editor

An investigative technique using X-rays is causing engine designers to sit up and take notice. For the first time, they say, it may be possible to understand the structure and dynamics of fuel spray in the near-nozzle region of an injector. Still early in development, the technique could have a major impact on nozzle design, fuel injection parameters, and spray modeling and simulation, researchers believe.

Fuel injection systems in diesel and direct-injection gasoline engines are tightly controlled environments where the tiniest of many variables might influence engine performance. The key to cleaner-burning, more efficient engines depends on a better understanding of the interplay of many factors—timing, pressure, flow rate, and temperature among them—that influence fuel mixing and combustion dynamics.

In an experimental setup, X-rays pass through a window on one side of an injection chamber, pene-
trating the fuel plume. A detector on the other side records the emerging ener-
gy, creating a detailed pic-
ture of the fuel spray mass. (Click photo to enlarge.)

There are still plenty of unknowns inside the combustion chamber. One is a lack of quantifiable information about spray plumes close to the injection nozzle. Traditional investigative tools, involving probes or lasers, have not been effective, because light is reflected off the cloud of the droplets the way car headlights bounce off fog on a dark night.

Researchers at Argonne National Laboratory in Argonne, Ill., believe they have found a way to penetrate this shroud to reveal details about the plume's interior. They are using X-rays generated by Argonne's Advanced Photon Source. Their work, funded by the U.S. Department of Energy, has gained the interest of Robert Bosch and General Motors.

PROBING PLUMES

According to Jin Wang, a physicist at Argonne and the principal investigator of the research team, the payoff for better combustion control is higher efficiency to reduce particulates, nitrogen oxides, and other emissions from the tailpipe. Because the factors that result in these emissions are interrelated, it's important to control the whole combustion process from the beginning, Wang said.

"You have to know the structure and dynamics of sprays and how they are atomized in the cylinder before you can do a really good combustion analysis," he said.

According to Ramesh Poola, a senior project engineer at the Electro-Motive Division of General Motors Corp. in LaGrange, Ill., information for a fundamental understanding of the spray or to do advanced computer modeling is lacking. Poola is a former colleague of Jin Wang at Argonne and is one of the original developers of the X-ray technique.

Argonne is using intense X-rays generated by the Advanced Photon Source, a 1.1-kilometer-long circular electron storage ring that generates high-intensity radiation.

The X-ray image of the diesel spray at the bottom provides more detailed information about the dense core, which is obscured in the optical image above it. The detailed information about the plumeÕs interior could improve CFD models.

In the test setup, fuel is sprayed from a nozzle into a chamber. X-rays enter the chamber through a window and penetrate the spray plume. A photodiode detector opposite the X-ray source records the energy emerging from the fuel spray. Then, the information is transferred to an oscilloscope and stored on a computer.

The intensity of the X-rays that come out is compared to what went in; the difference indicates the mass of the fuel in the plume. The pattern of X-rays recorded by the detectors creates a picture of the amount of fuel at a given position of the spray at a particular moment of time.

The X-rays generated from the Advanced Photon Source allow the taking of time-resolved measurements in nanosecond intervals. Researchers used the single wavelengths from a broad spectrum to calculate the mass of the spray very precisely. Wang said that the team has been able to detect droplets as small as four nanograms.

Argonne's research team is collaborating with scientists at Cornell University to improve the X-ray detection technique to make it less labor-intensive, Wang said. The Argonne team is using an ultrafast X-ray camera to collect the spray images more efficiently. It is also improving the physical test chamber with a transparent window for the X-rays that can withstand higher pressures and higher temperatures, similar to an actual diesel cylinder. Wang said that the lab has a long-term goal of building a test chamber in which both flow and combustion experiments could be performed. The lab's present focus is only on fuel spray flow.

WHY THE INTEREST?

Many applications that involve dense sprays with high flow rates and pressures could benefit from the detailed information, said Scott Parrish, a senior research engineer at General Motors R&D and Planning in Warren, Mich. Parrish, who has chaired the instrumentation committee of the Institute for Liquid Atomization and Spray Systems, a group that studies liquid sprays, said that the need for better diagnostic techniques to obtain measurements in dense sprays is a perennial topic at meetings. He suggested that the information could be of interest to applications such as gas turbines, and agri- cultural and industrial sprays, as well as diesel and gasoline direct-injection engines.

Even very subtle changes in the geometry of a hole in a diesel injection nozzle may result in a difference in performance, although any differences in the mass of the droplets or atomization of the fuel spray might not be evident with conventional testing techniques.

Robert Bosch, a supplier of fuel injection nozzles, has contributed equipment to the research team. Johannes Schaller, an engineer who became familiar with the X-ray technique at Argonne when he worked on diesel engines for Bosch's Research and Development Center, said the method is unique because it is non-intrusive but gives never-before-available information on the structure of the spray. A mechanical probe—a more conventional tool—doesn't provide information about the local fuel mass fraction, he said.

Nozzle designs could benefit from more detailed X-ray image of the gasoline spray on the right, compared to the optical image on the left, according to one fuel injection system manufacturer.

And while fuel sprays in direct-injection gasoline engines are not as dense, they are still hard to quantify. Stefan Arndt, a senior researcher at Bosch's R&D Center in Stuttgart, Germany, said the X-ray view could be useful in optimizing nozzles for direct-injection gasoline engines, which could lower fuel consumption and emissions.

The research is still in its early stages. It has provided better understanding about fuel flow, but has not yet resulted in any changes to nozzle design, Schaller said. So far, tests have been limited to lower backpressures and single-hole symmetrical nozzles that are easier to observe, he said.

Roger Krieger, a group manager in the Powertrain Systems Research Lab, is responsible for the compression-
ignition engines group at GM R&D and Planning. He said that GM's strategy for developing engine technology is to integrate computer modeling and experiment. He said the X-ray experiments reveal quantifiable information about the void fraction—or vapor component—inside the fuel spray plume.

BETTER SPRAY MOLDING

The X-ray technique could help to provide numerical simulations in the near-nozzle region as well as experimental validation for those simulations, Schaller said.

It may also be useful in Bosch's long-term goal of comprehensive simulation of fuel flow from nozzle to tailpipe. Near term, the technique could provide a better understanding of phenomena such as cavitation inside and outside the nozzle, he said.

Arndt said that quantitative information about fuel sprays would help to improve computer models as they evolve and become more complicated. This could save time in developing new injector designs, he said.

GM's Krieger said that new information about fuel spray a few millimeters from the nozzle could be compared to current computer models to determine if they should be revised.

Time-resolved X-ray images of a fuel spray show shockwaves, evidently produced by the liquid penetrating into the gas. Researchers say the phenomenon could affect fuel atomization and combustion.

GM plans to set up experiments comparing sprays from nozzles with different geometries of nozzle holes. It is working with Argonne as well as with the Engine Research Center at the University of Wisconsin in Madison to define new experiments that would be closer to fuel sprays that actually occur in the engine environment.

General Motors' work with the Engine Research Center is focused on both diesel and direct-injection gasoline engines. According to Rolf Reitz, the Engine Research Center's director, quantifiable information could provide new information about what form the fuel takes when it exits the fuel injector. For example, is it fragmented or continuous and, if it is continuous, what kind of breakup mechanism is involved, he asked. More quantifiable information about the spray could cause some theories about the fuel's atomization to be revisited, he said.

There are questions of how well test conditions mimic a true engine environment. For example, researchers have infused the chamber with high-molecular-weight gas, sulfur hexafluoride, to simulate the dense ambient atmosphere in a diesel engine. Krieger observed that the heavy gas goes only partway toward mimicking actual chamber densities at the time of ignition. "It's in the right direction, but is a long way from the real densities."

Results should be interpreted cautiously because techniques intended to mimic engine conditions might also change other parameters, Reitz said. Introduction of heavy gas to replicate high-pressure diesel conditions could influence factors such as the sound speed. Consequently, a test might show effects associated with supersonic flow that wouldn't occur in an actual engine. Advances in X-ray imaging techniques and instrumentation could help overcome these hurdles, said Wang, the Argonne physicist.

The X-ray experimentation at Argonne has already led to intriguing discoveries. One is shockwaves, evidently produced by the liquid penetrating into the gas. The discovery matters because shockwaves would lead to very discontinuous gradients and pressures that could disrupt the liquid and influence atomization, Reitz said.

Not everyone is sure what to make of shockwaves. Stefan Arndt of Bosch said that it is too early to say. Ramesh Poola at GM pointed out that the waves may be tied to the conditions of the experiments and not translate to the actual diesel engine environment, after all. Wang said he has seen pictures of shockwaves in real engine conditions at Bosch's R&D Center in Stuttgart.