The Physics of the Perfect Pour Over

New York Times

23-04-2025

More than a billion cups of coffee are consumed daily: French-press, espresso, cold brew, whatever it takes.
Arnold Mathijssen, a physicist at the University of Pennsylvania, is partial to pour-over coffee, which involves manually pouring hot water over ground beans and filtering it into a pot or mug below. Surely, he figured, applying the principles of fluid dynamics to the process could make it even better.
With two students of similar mind, Dr. Mathijssen began studying how to optimize the pour in a pour over. Their science-backed advice: Pour high, slow and with a steady stream of water. This ensures the greatest extraction from minimal grounds, enhancing the coffee's flavor without added beans or cost.
The findings, published this month in the journal Physics of Fluids, highlight how processes that unfold in the kitchen — from making foie gras to whipping up a plate of cacio e pepe — can inspire new scientific directions. In turn, science can enhance the art of cuisine.
'Kitchen science starts off with a relatively low entry barrier,' Dr. Mathijssen said. 'But it's more than just cute. Sometimes fundamental things can come out of it.'
Dr. Mathijssen primarily studies the physics of biological flows, such as the way bacteria swim upstream in blood vessels. But when he lost access to his lab during the Covid-19 shutdown, he started playing with his food — literally. He shook up bottles of whiskey, tested the stickiness of pasta and slid coins down slopes made of whipped cream and honey. The interest culminated in a 77-page review, structured like a menu, of the physics involved in making a meal.
'It got totally out of hand,' Dr. Mathijssen said. 'You just realize science is everywhere.'
Dr. Mathijssen has since returned to the lab, but the passion for kitchen physics has stuck. The coffee study was partly inspired by a scientist in his group who kept detailed notes about pour-over brews prepared in the lab each day. The notes included information about where the beans had come from, the extraction time and the brew's flavor profile.
Ernest Park, a graduate student in the lab, designed a formal experiment. Using silica gel beads in a glass cone, the scientists simulated the action of water being poured over coffee grounds from different heights, recording the dynamics of the system with a high-speed camera.
Then they brewed pots of real coffee, pouring from a gooseneck kettle, at varying heights. The resulting liquid was allowed to evaporate in an oven until all that remained were the coffee particles extracted from the grounds.
They found that more coffee particles remained when they had poured slowly, which increased the time the water was in contact with the grounds. Holding the kettle higher helped with the mixing, preventing the water from draining along the sides, between the grounds and the filter.
This type of flow caused what the researchers described as an avalanche effect. The water eroded the center of the pile of coffee grounds, thus suspending some of the grains, which settled and built up on the sides. Eventually, the sides collapsed inward and the process started again. This increased the flavor extracted from the coffee grounds, but only as long as the water was allowed to flow continuously.
'Your jet of water coming out should look like a smooth column all the way down,' said Margot Young, a graduate student — and former barista — involved in the study. 'If you see it starting to break up, or you can see droplets, then you have to pour from lower down.'
The scientists conducted informal taste tests, although these did not make it into the final publication. 'Taste-wise, it's very subjective,' Mr. Park said. 'So we always suggest that you try it yourself.'
Mr. Park noted that the study examined only water poured into the center of the coffee grounds, although future experiments could explore other techniques, like making swirls or spirals.
Scientific phenomena observed in the kitchen typically have analogues outside its walls. The dynamics between a jet of hot water and a bed of coffee grains, for instance, are similar to the erosion of land that can occur around waterfalls and dams. A stirred pot of soup assumes the same shape as the liquid mirrors of some telescopes. Observations of soap bubbles by Agnes Pockels, a 19th-century German homemaker, gave rise to the field of surface science and laid the groundwork for nanotechnology.
In 2022, Dr. Mathijssen helped assemble an array of studies, produced by scientists around the world, into a collection called Kitchen Flows. He is now helping compile a second collection, which so far consists of more than 30 studies, including insights into the behavior of an egg yolk, the sloshing of a bottle of beer and the most efficient way to boil pasta.
Dr. Mathijssen also plans to continue exploring the many paths to perfect coffee, such as the physics behind the formation of the milk and espresso layers in a latte. 'I want to do some more work in this direction,' he said. 'And then maybe also something about cold brews.'