New software could help diesel engines run on alternative fuels

0

An associate professor at The Illinois Institute of Technology has developed a smart computer model that could allow diesel engines to run on alternative fuels. To take advantage of this, diesel-powered vehicles would only need to upgrade their software suite, if required.

Associate Professor Carrie Hall used a combination of machine learning (ML) and computer modeling to achieve the feat. This development is welcome in order to accelerate our transition to highly polluting fuels such as diesel.

It is hoped that this development will significantly improve the durability of large diesel-powered vehicles, such as trucks, which rely heavily on diesel due to the long distances they must travel regularly. For the time being, the complete electrification of freight vehicle fleets is not really feasible.

The software could also help some aircraft.

At this time, it is not viable to simply replace diesel with an alternative, as most engines are fuel specific. Although biodiesel is an option, it would be great if diesel engines could become truly multi-fuel.

“Since we’re focusing on a software upgrade, someone can actually install it in their vehicle without incurring a lot of extra cost,” Hall explained. “They won’t really have to change the hardware of their vehicle.”

This software upgrade could be an important stepping stone to help trucks move away from diesel fuel for good.

“It is expected that with electric vehicles being more common for passenger cars in the United States, there will be a lot of additional gasoline that will not be used. This gasoline can be used on heavier vehicles. is a strategy that is still being explored,” Hall added. “Making engines smart enough to use a wider range of fuels also opens the door to other possibilities, such as using carbon-neutral fuels. or carbon negatives.”

This could be a game-changer for heavy-duty vehicles which account for about 1/4 of all on-road fuel consumption in the United States while only accounting for about 1% of all vehicles. Improving their effectiveness should therefore become the short- and medium-term objective.

“Everything we do is about trying to get cleaner, more efficient vehicles,” says Hall.

An alternative fuel that could be considered is gasoline. However, as any diesel-powered vehicle owner knows, it’s not a good idea without adapting the engine.

The main reason for this is that diesel and gasoline react differently. Gasoline generally requires a spark to ignite, and the resulting explosion spreads evenly through the engine cylinder.

Diesel, on the other hand, tends to ignite spontaneously after being compressed in the cylinder. When trying to run gasoline in a traditional diesel engine, the cylinder may explode or not burn at all.

The model could allow the use of several fuels with a simple software update

Because of this, Hall realized that timing is everything, as engine efficiency usually relies heavily on the harmonious operation of multiple cylinders.

“If the fuel burns a little too early or too late, you don’t really get the full benefit and the efficiency is worse,” Hall explained.

To make this possible, engine management systems therefore need real-time information about when the fuel has ignited.

“The things that are actually going on inside the cylinder of the engine are really hard to measure in a cheap way,” says Hall. “So what we’re trying to do is take the information that we get from simpler, cheaper sensors that are outside of the cylinder of the engine where the combustion happens, and diagnose what’s happening at inside the engine,” she added.

And it all has to happen in a split second, all the time.

“Our models are used to provide system feedback,” says Hall. “Understanding the moment of [fuel ignition] gives us an idea of ​​how it related to something like fuel injection, which we could then tweak based on that feedback.

Currently, the kind of computational speed needed can be achieved by using machine learning techniques or by storing large tables of data. Hall, however, took a different approach.

“We tried to create models based on the underlying physics and chemistry, even when we have these very complicated processes,” says Hall. “Recently, there has been interest in the use of neural networks to model combustion. The problem is then it’s just a black box, and you don’t really understand what’s going on underneath, which is hard to control, because if you’re wrong you can have something spinning very evil.

So Hall looked for ways to simplify existing calculations and methods to speed up the process.

“We’ve tried to capture all of the underlying effects, even if it’s in a more detailed way than we know we’re really going to be able to use for real-time control, and let that be our point of reference. . Then we simplify it by using elements like neural networks in a strategic way, but we keep this overall structure in order to understand what each element means and what it actually does inside,” says Hall.

This resulted in a lighter, more adaptable model that can be adapted to different fuels with a simple update.

This is key to Hall’s research and his recent work builds on his experience working with new fuels in the past – such as fuel blends. Hall is also a member of a collaborative group that recently received $2 million from the US Department of Energy to test new applications of a low-carbon fuel called dimethyl ether.

Hall’s control model, which Illinois Tech Research Assistant Professor Michael Pamminger (Ph.D. MAE ’21) worked on as a student in Hall’s research group, is one component of a larger project aimed at understanding how to use gasoline in diesel engines and was conducted in collaboration with Argonne National Laboratory, Navistar and Caterpillar.

“We are working with these companies to try to help them understand the underlying combustion processes, but also to create tools that they can potentially integrate into their own software and then enable their next generation of engines to use these fuels and use them well,” says Hall.

Share.

Comments are closed.