Less Beans, More Flavor: The Astonishing Physics Trick Transforming Coffee

Penn researchers discovered how to make a richer cup of pour-over coffee using fewer beans by tapping into fluid dynamics.

Through creative experiments using transparent particles, lasers, and high-speed cameras, they revealed how specific pouring techniques—like using a gooseneck kettle and pouring from the right height—can maximize flavor extraction. Their findings not only improve coffee brewing but also offer insight into broader systems like erosion and water filtration.

Brewing Better Coffee with Fewer Beans

The cost of arabica beans, the main ingredient in most coffee, has surged in recent years, driven by four back-to-back seasons of poor weather. Climate change has only made things worse, putting stress on the delicate temperature range that Coffea arabica plants need to thrive. Faced with this pressure, physicists at the University of Pennsylvania asked a bold question: Can we make great coffee using fewer beans?

“There’s a lot of research on fluid mechanics, and there’s a lot of research on particles separately,” says Arnold Mathijssen, assistant professor in the School of Arts & Sciences. “Maybe this is one of the first studies where we start bringing these things together.”

Their findings, published on April 8 in Physics of Fluids, explore how fluid dynamics can be used to boost coffee extraction, so that fewer grounds still produce a flavorful cup.

A high-speed camera catches water penetrating the simulated coffee bed. By modeling how the jet interacts with the grounds, the team found the most efficient flow pattern for extracting flavor with fewer beans. Credit: Ernest Park

Maximizing Extraction with Less Coffee

“We tried finding ways where we could use less [or] as little coffee as possible and just take advantage of the fluid dynamics of the pour from a gooseneck kettle to increase the extraction that you get from the coffee grounds—while using fewer grounds,” says coauthor Ernest Park, a graduate researcher in the Mathijssen Lab.

But to study the process clearly, they had to make the invisible visible, explains coauthor Margot Young, a graduate researcher in the Mathijssen Lab.

“Coffee’s opacity makes it tricky to observe pour-over dynamics directly, so we swapped in transparent silica gel particles in a glass cone,” Park explains.

How Avalanches and Laminar Flow Help Extraction

Using a laser sheet and high-speed camera, they watched as water poured from above triggered “miniature avalanches” within the particle bed. This tumbling motion helped stir the particles, improving extraction by increasing contact between water and grounds.

A key factor in this process is laminar (smooth and nonturbulent) flow,  made possible by a gooseneck kettle, even with a gentle pour-over flow. “If you were just to use a regular water kettle, it’s a little bit hard to control where the flow goes,” says Park. “And if the flow isn’t laminar enough, it doesn’t dig up the coffee bed as well.”

The Art and Science of Pouring Height

The team discovered that when water is poured from a height, it creates a stronger mixing effect.

“When you’re brewing a cup, what gets all of that coffee taste and all of the good stuff from the grounds is contact between the grounds and the water,” explains Young. “So, the idea is to try to increase the contact between the water and the grounds overall in the pour-over.”

They found that if poured from too great a height, the water stream breaks apart into droplets, carrying air with it into the coffee cone, which can actually decrease extraction efficiency.

A high-speed camera catches water penetrating the simulated coffee bed. By modeling how the jet interacts with the grounds, the team found the most efficient flow pattern for extracting flavor with fewer beans. Credit: Ernest Park

Testing Real Coffee for Scientific Accuracy

The researchers conducted additional experiments with real coffee grounds to measure the extraction yield of total dissolved solids. Their results confirmed that the extraction of coffee can be tuned by prolonging the mixing time with slower but more effective pours that utilize avalanche dynamics.

For thicker water flow, they found that higher pours resulted in stronger coffee, confirming their observations about increased agitation with higher pour heights. When using a thinner water jet, the extraction remained consistently high across different pour heights, possibly due to the longer pouring time required to reach the target volume.

From Coffee to Waterfalls and Beyond

The study is a love letter to coffee—and it’s also a window into the team’s broader research. “We weren’t just doing this for fun,” Mathijssen says. “We had the tools from other projects and realized coffee could be a neat model system to explore deeper physical principles.”

Those principles extend well beyond the kitchen, notes Young. “This kind of fluid behavior helps us understand how water erodes rock under waterfalls or behind dams,” she says. Even wastewater treatment and filtration systems involve similar dynamics, Mathijssen adds.

Real-World Applications and Future Research

The project also reflects ongoing research in the lab, as Park is working on microscale active surfaces that use rotating magnetic fields to clean biofilms from medical devices.

Young, meanwhile, is investigating ultra-fast biological flows, using the same high-speed imaging setup to study how tiny vortices generated by lung cilia help clear pathogens.

“You can start small, like with coffee,” Mathijssen says. “And end up uncovering mechanisms that matter at environmental or industrial scales.”

Explore Further: Researchers Reveal Simple Trick To Make Your Coffee Stronger

Reference: “Pour-over coffee: Mixing by a water jet impinging on a granular bed with avalanche dynamics” by Ernest Park, Margot Young and Arnold J. T. M. Mathijssen, 8 April 2025, Physics of Fluids.
DOI: 10.1063/5.0257924

Arnold Mathijssen is an assistant professor in the Department of Physics & Astronomy in the School of Arts & Sciences at the University of Pennsylvania.

Ernest Park is a Ph.D. candidate in the School of Arts & Sciences.

Margot Young is a Ph.D. candidate in the School of Arts & Sciences.

Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.