Scientists have discovered that the builder of the Titanic struggled for years to obtain enough rivets and riveters and ultimately settled on faulty materials that doomed the ship, which sank 96 years ago today.

The builder's own archive, the two scientists say, harbors evidence of a deadly mix of low-quality rivets and lofty ambition as the builder labored to construct the world's three biggest ships at once — the Titanic and two sisters, Olympic and Britannic.

For a decade, scientists have argued that the storied liner went down fast after hitting the iceberg because the ship's builder used substandard rivets that popped their heads and let tons of icy seawater rush in. More than 1,500 people died.

When the safety of the rivets was first questioned 10 years ago, the builder ignored the accusation and said it did not have an archivist who could address the issue. Historians say that the new evidence uncovered in the archive of Harland & Wolff, in Belfast, Northern Ireland, settles the argument and finally solves the riddle of one of the most famous sinkings of all time. The company now insists that findings are deeply flawed.

Each of the great ships under construction required 3 million rivets that acted like glue to hold everything together, and the scientists in a new book say the shortages peaked during Titanic's construction.

"The board was in crisis mode," Dr. Jennifer Hooper McCarty, a team member who studied the archive, said in an interview. "It was constant stress. Every meeting it was, 'There's problems with the rivets and we need to hire more people."'

The team collected other clues from 48 Titanic rivets recovered from the wreck, modern tests, computer simulations, comparisons with century-old metals as well as careful documentation of what engineers and shipbuilders of that era considered state of the art.

The scientists say the troubles all began when the colossal plans forced Harland & Wolff to reach beyond its usual suppliers of rivet iron and include smaller forges, as disclosed in company and British government papers. Small forges tended to have less skill and experience.

Adding to the threat, the company, in buying iron for Titanic's rivets, ordered No. 3 bar, known as "best" — not No. 4, known as "best-best," the scientists found. They also discovered that shipbuilders of the day typically used No. 4 iron for anchors, chains and rivets.

So the liner, whose name was meant to be synonymous with opulence, in at least one instance relied on cheap materials.

The scientists studied 48 rivets that divers recovered over two decades from the Titanic's resting place — two miles down in the North Atlantic — and found many riddled with high concentrations of slag. A glassy residue of smelting, slag can make rivets brittle and prone to fracture.

"Some material the company bought was not rivet-quality," said Dr. Timothy Foecke, a team member at the National Institute of Standards and Technology, a federal agency in Gaithersburg, Md.

The company also faced shortages of skilled riveters, according to archive papers. McCarty said that for a half year, from late 1911 to April 1912, when Titanic set sail, the company's board addressed the shortfalls at every meeting.

For instance, on October 28, 1911, Lord William Pirrie, the company's chairman, expressed concern over the lack of riveters and called for new hiring efforts.

In their research, the scientists found that good riveting took great skill. The iron had to be heated to a precise cherry red color and beaten by the right combination of hammer blows. Mediocre work could hide problems.

"Hand riveting was tricky," said McCarty, whose doctoral thesis at Johns Hopkins University analyzed Titanic's rivets.

Steel beckoned as a solution. Shipbuilders of the day were moving from iron to steel rivets, which were stronger. And machines could install them, improving workmanship and avoiding labor problems.

The rival Cunard line, the scientists found, had switched to steel rivets years before, using them, for instance, throughout the Lusitania.

The scientists discovered that Harland & Wolff also used steel rivets — but only on Titanic's central hull, where stresses were expected to be greatest. Iron rivets were chosen for the ship's stern and bow.

And the bow, as fate would have it, is where the iceberg struck. Studies of the wreck show that six seams opened up in the ship's bow plates. And the damage, Foecke noted, "ends close to where the rivets transition from iron to steel."

The scientists argue that better rivets would have probably kept the Titanic afloat long enough for rescuers to have arrived before the icy plunge, saving hundreds of lives.

The two metallurgists make their case, and detail their archive findings, in "What Really Sank the Titanic," a new book by Citadel Press.

Reactions run from anger to admiration. James Alexander Carlisle, whose grandfather was a Titanic riveter, has bluntly denounced the rivet theory on his Web site. "NO WAY!"

For its part, Harland & Wolff, after long silence, now rejects the charge. "There was nothing wrong with the materials," Joris Minne, a company spokesman, said last week. He noted that Olympic sailed without incident for 24 years, until retirement. (The Britannic sank in 1916 after hitting a mine.)

David Livingstone, a former Harland & Wolff official, called the book's main points misleading. He said big shipyards often had to scramble. On a recent job, he noted, Harland & Wolff had to look to Romania to find welders.

And Livingstone called the slag evidence painfully circumstantial, saying no real proof linked the hull opening to bad rivets. "It's only waffle," he said of the team's arguments.

But a naval historian praised the book as solving a mystery that has baffled investigators for nearly a century.

"It's fascinating," said Tim Trower, who reviews books for the Titanic Historical Society, a private group in Indian Orchard, Mass. "This puts in the final nail in the arguments and explains why the incident was so dramatically bad."

The new disclosures, he added, cast Harland & Wolff as "responsible for the severity of the damage."

Titanic had every conceivable luxury: cafes, squash courts, a swimming pool, Turkish baths, a barbershop and three libraries.

The lavish air extended to safety. The White Star Line, in a brochure, described the ship as "designed to be unsinkable."

During her inaugural voyage, on the night of April 14, 1912, the ship hit the iceberg around 11:40 p.m. and sank in a little more than two and a half hours. Most everyone assumed the iceberg had torn a huge gash in the ship's starboard hull.

The discovery in 1985 of Titanic's resting place began many new inquiries. In 1996, an expedition found, beneath obscuring mud, not a large gash but six narrow slits where bow plates appeared to have parted.

Naval experts suspected that rivets had popped along the seams, letting seawater rush in under high pressure.

A specialist in metal fracture, Foecke got involved in 1997, analyzing two salvaged rivets. He was astonished to find about three times more slag than occurs in modern wrought iron.

In early 1998, he and a team of marine forensic experts announced their rivet findings, calling them tentative.

Foecke, in addition to working at the National Institute of Standards and Technology, also taught and lectured part time at Johns Hopkins. There he met McCarty, who became intrigued by the riddle, as did her thesis adviser.

The team acquired many rivets from salvors who pulled up hundreds of artifacts from the sunken liner. The two scientists also collected old iron of the era — including some from the Brooklyn Bridge — to make comparisons. The new work seemed to only bolster the bad-rivet theory.

In 2003, after graduating from Johns Hopkins, McCarty traveled to England and located the Harland & Wolff archives at the Public Record Office of Northern Ireland, in Belfast.

She also explored the archives of the British Board of Trade, which regulated shipping and set material standards, and of Lloyd's of London, which set shipbuilding standards. And she worked at Oxford University and obtained access to its libraries.

What emerged was a picture of a company stretched to the limit as it struggled to build the world's three biggest ships simultaneously. She also found complacency. For instance, the Board of Trade gave up testing iron for shipbuilding in 1901 because it saw iron metallurgy as a mature field, unlike the burgeoning world of steel.

McCarty said she enjoyed telling middle and high school students about the decade of rivet forensics, as well as the revelations from the British archives.

"They get really excited," she said. "That's why I love the story. People see it and get mesmerized."