In 1936, a dim, unremarkable star in the constellation Orion began to brighten. Within a matter of months it had grown roughly a hundred times more luminous, and when it peaked it was blazing at a level astronomers had only ever seen in dying stars. Yet FU Orionis, as it came to be known, was not dying. It was being born. The strange eruption baffled observers for decades and eventually gave its name to an entire class of young, violently flickering stars — the FUors — whose behavior has remained one of the more stubborn puzzles in the study of how stars and planets take shape.
On the afternoon of October 6, 2026, that puzzle comes to Sharp Laboratory on the University of Delaware's Newark campus, when Dr. Rajika Kuruwita, a research fellow at the Australian National University, delivers an astronomy seminar titled "Understanding stellar multiplicity in embedded environments." The talk is free and open to the public, running from 4:00 to 5:00 p.m.
Kuruwita, who goes by Reggie, works at Mount Stromlo Observatory in Australia's capital territory, where her research centers on star formation, multiplicity, stellar interactions, and three-dimensional magnetohydrodynamical simulations — the art of recreating, on a supercomputer, the churning clouds of gas and dust inside which stars are assembled. Her central subject is a fact that tends to surprise people who picture stars as solitary suns: most stars are not born alone. They arrive in pairs, triples, and larger clusters, and the interactions among those siblings, and with the surrounding natal environment, shape everything that follows.
That framing is the heart of her seminar. Stars still embedded in their birth clouds, she argues, do not form in isolation. They jostle against their companions and against the gas that surrounds them, and the result is a chaotic adolescence: sudden accretion bursts, warped and deformed protoplanetary discs, and the occasional outright ejection of a star from its family. Many of the stars that appear single today, she notes, likely had siblings once — companions that were stripped away or flung off before the system settled down.
Kuruwita's path to these questions runs through some of the more exotic corners of stellar dynamics. Her Master of Research examined the common envelope phase, a brief and violent stage in the lives of close binary stars in which one star's expanding envelope engulfs its companion, dragging the two into a death spiral. Her simulations, run with the hydrodynamics code ENZO, tested whether material falling back onto the pair could tighten their orbit further — a mechanism that earlier models had struggled to reproduce. Her doctoral work turned to the formation of binary systems themselves, using magnetohydrodynamical simulations with the FLASH code to study how binarity affects the outflows that young stars launch, and what all of it means for the prospects of planet formation around double stars.
More recently her attention has moved toward a statistical question: how do the observed numbers of multiple-star systems arise? Observations of young stellar systems show a distinctive bimodal distribution — companion frequency peaks at two different separations — and the long-standing explanation has been that binaries form through two separate channels, one involving the fragmentation of dense cores and the other involving the fragmentation of the discs around individual stars. In a 2022 paper, Kuruwita and collaborators examined the contribution of each channel, work that underpins the multiplicity statistics she will compare against her cluster simulations in Newark.
It is the second half of her talk, though, that carries the most dramatic hook. Kuruwita has been zooming into individual stars inside her simulated star clusters to attack the FU Orionis problem directly, and the answer her team found is not the one most astronomers expected. In a paper submitted in September 2026, she and her coauthors report that FU Orionis may not be bursting at all — at least not in the way the textbooks describe.
The conventional picture, developed over decades by researchers including Lee Hartmann and Scott Kenyon, holds that an FUor flare is a sudden dump of mass from a protoplanetary disc onto a young, low-mass star. The trouble is that models built around an isolated disc run out of material quickly, producing bursts far shorter than the century-long eruption astronomers actually observe at FU Orionis. Kuruwita's approach was to start much further out, at the scale of the interstellar medium itself, where protostars are still actively fed by long streams of gas. Her team ran an unbiased search through a star cluster formation simulation resolved at 25 astronomical units — the distance from the Sun to Uranus — hunting for FUor candidates. Most of the candidates turned out to be low-mass secondary stars in highly eccentric binary orbits, exactly like FU Orionis. Zooming in further, to a resolution of 0.79 astronomical units, they found something striking: as the star swings toward its companion at the closest point of its orbit, its orbital speed turns supersonic, stripping away the gas around it and choking off accretion. Then, just past that closest approach, the star slows, and gas rushes back in through a process known as Bondi-Hoyle-Lyttleton accretion, piling material onto the star in a sudden surge. The tails of gas left behind by this process look, in the simulation, like the accretion streamers astronomers observe in real star-forming regions.
If the mechanism is right, it reframes the FUor phenomenon: not a disc dumping its contents onto a star, but a binary pair's orbital rhythm alternately starving and then gorging one of its members. It also carries a prediction — that FU Orionis, after its long bright spell, has quietly returned to its former, dimmer accretion rate. That claim is testable, and the coming years of observation will decide whether the eccentric-binary picture holds.
For a general audience, the appeal of the seminar lies in that arc: a mystery that began when Franklin Roosevelt was in the White House, pursued through some of the largest computer simulations ever built of star birth, and still unsettled enough that a visiting researcher can stand up in a lecture hall and argue that the famous 1936 eruption was something other than what it seemed. The talk is aimed at a broad audience, and the questions it raises — why stars come in pairs, why some flare, why our own Sun ended up alone — are as much about where we came from as about the distant clouds of Orion.