WASHINGTON, D.C.—It’s no secret that many of America’s bridges are in sad shape. One of the reasons is the corrosion of the steel rebar used to strengthen concrete structures. Deluged with de-icing salt and acidic compounds from industrial production, the steel weakens—potentially leading to disasters.

But changing the composition of the concrete can help reinforce the steel, according to Michael Loy, 17, one of 40 finalists in the 2009 Intel Science Talent Search here in Washington, D.C. this week for the final rounds of judging. (Read all of our coverage in this In-Depth Report.)

Loy, a student at Oregon Episcopal School, discovered that adding a certain mix of sodium borate to the concrete kept the steel from corroding under extreme conditions far better than a control mixture—to the point of being “still shiny, like it’s brand new,” he says. The steel also turned out to be 10 to 20 percent stronger in tests. He also had a  unique way of presenting some of the data—as a comic [see photo above].

WASHINGTON, D.C. (March 10, 2009)—Stem cells have long been touted as potential cures or treatments for a variety of ailments from paralysis to Parkinson's disease. After all, these cells (found in bone marrow, for instance, and also in human embryos, making their use a subject of much controversy) can potentially turn into a wide variety of cells with specific functions. Or, they can throw off proteins—such as growth factors—that help other cells grow. Stem cells injected into a paralyzed patient’s spine, for instance, might help regenerate nerve tissue.

But while the flexibility of these cells is the key to their usefulness, they might not be quite as flexible as we once thought, says Julia Dory Ransohoff, 17, one of 40 finalists in the Intel Science Talent Search here in Washington this week for the competition’s final rounds. (We've been on hand to live-Twitter and to profile a few projects, everything from Splenda in drinking water to whether parents should discuss their drinking with their kids to cellulosic ethanol.)

WASHINGTON, D.C.—The excitement is building to a crescendo here as 40 top high school scientists wait for tonight's gala finale of Intel's Science Talent Search. We've been on hand to live-Twitter and to profile a few projects, everything from Splenda in drinking water to whether parents should discuss their drinking with their kids to cellulosic ethanol.

Here are some more of the highlights:

WASHINGTON, D.C.—When gas prices were sky high, lots of people talked about ethanol as a fuel of the future. In particular, many investors placed their hopes in cellulosic ethanol. Such ethanol is made from the non-edible parts of corn, such as the stalk and leaves, or from non-corn sources such as certain kinds of grasses.* Unlike the ethanol available today, it does not require the edible parts—a requirement that has raised concerns of pitting fuel versus food.

But at least as important, no one has yet figured out a way to make cellulosic ethanol for anything approaching the price of gas—or how to make much of it at all. “Right now it’s just big diseconomies of scale,” says Aditya Rajagopalan, 17, a student at Choate Rosemary Hall in Connecticut.

WASHINGTON, D.C.—People like sucralose—the artificial sweetener marketed as Splenda—because the human body can’t break it down and use it. That means the substance has almost no calories and makes it a popular ingredient in everything from cookies to diet sodas. Unfortunately, it turns out that modern wastewater treatment methods don’t break down Splenda either.

That, according to Smitha Ramakrishna, 17, one of 40 finalists in the 2009 Intel Science Talent Search who’ve gathered in Washington, DC, for the final judging rounds this week, means that the sweetener can accumulate in the water supply after people excrete it, potentially harming fish and other living things.

WASHINGTON, D.C.—If you drank as a teenager, do not tell your kids about it. That’s the lesson from Chelsea Lynn Jurman’s study of teen drinking behavior—the only social science project among the 40 finalists in the Intel Science Talent Search going through the final judging rounds here this week.

“The perception kids create becomes the reality,” says Jurman, 17, a senior at Rosyln High School in New York—and if they perceive that you drank and turned out okay, they figure they will, too.

ScientificAmerican.com is on hand to speak with several of the finalists in this prestigious national competition that is the modern incarnation of the old Westinghouse Science Talent Search, which began in the 1940s. The 40 finalists get scholarships ranging from $5,000 to the top prize of $100,000. You can read about past finalists in our "Where Are They Now" series, and follow along as we live-Twitter from Washington.

Students from New York City and its suburbs again took an outsized number of finalist slots in the annual Intel Science Talent Search, according to results released today.

Started in 1942 as the Westinghouse Science Talent Search, the contest identifies 40 top high school science students, based largely on independent research projects. Winners (who can earn up to $100,000) will be chosen once the finalists gather in Washington DC for interviews in March.

Longtime Intel/Westinghouse fixtures Stuyvesant High School and the Bronx High School of Science in the city had two finalists apiece, but Ward Melville in Long Island also produced two finalists. Three other New York state schools (John Jay High School, Smithtown High School West, and Roslyn High School) also chipped in to produce nine finalists for the Empire State. Only California, with five finalists, came close.

Yes, this addictive game that has you rotating shapes to fit into a grid until you go mad may actually be making you smarter. Past studies have found that when people play Tetris initially, they use more glucose, suggesting that their brains are working hard. Over time, though, glucose usage returns to normal. In other words, the brain has to work less hard the more it practices. It becomes primed to solve related problems quickly.

So the sisters, who grew up building increasingly complex structures with Legos, decided to develop a software program that would could react quickly to commands, and take up a minimum of space on the robot mini-computer. They also wanted a high degree of control over steering and the like. So they built what they called a "generic" robotic software program that could execute programs rapidly, do quick downloads, and leverage the three motors the FIRST Lego League allowed them.