Bit by bit, the new rocket ship that is supposed to blast America into the second Space Age and return astronauts to the moon appears to be coming undone.

First was the discovery that it lacked sufficient power to lift astronauts in a state-of-the-art capsule into orbit. Then engineers found out that it might vibrate like a giant tuning fork, shaking its crew to death.

Now, in the latest setback to the Ares I, computer models show the ship could crash into its launch tower during liftoff.

The issue is known as "liftoff drift." Ignition of the rocket's solid-fuel motor makes it "jump" sideways on the pad, and a southeast breeze stronger than 12.7 mph would be enough to push the 309-foot-tall ship into its launch tower.

Worst case, the impact would destroy the rocket. But even if that doesn't happen, flames from the rocket would scorch the tower, leading to huge repair costs.

"We were told by a person directly involved [in looking at the problem] that as they incorporate more variables into the liftoff-drift-curve model, the worse the curve becomes," said one NASA contractor, who asked not to be named because he wasn't authorized to discuss Ares.

"I get the impression that things are quickly going from bad to worse to unrecoverable."

NASA says it can solve -- or limit -- the problem by repositioning and redesigning the launchpad.

Engineers say that would take as much as a year and cost tens of millions of unbudgeted dollars.

What happens with Ares I is crucial to the future of the U.S. manned space program -- and of Kennedy Space Center. KSC is looking at thousands of layoffs after the space shuttle is retired in 2010. Its work force won't grow again until a new rocket launches.

In addition, huge expenditures on the rocket could bankrupt the agency's moon plans and prompt a new president to halt the program, delaying America's return to space.

Work in progress

NASA officials are now looking at ways to speed up the development of Ares and are reluctant to discuss specific problems. But they insist none is insurmountable.

"There are always issues that crop up when you are developing a new rocket and many opinions about how to deal with them," said Jeff Hanley, manager of the Constellation program, which includes Ares, the first new U.S. rocket in 35 years.

"We have a lot of data and understanding of what it's going to take to build this."

Still, Ares' woes have created unprecedented rifts inside the agency.

Now several engineers are speaking out, saying Ares should be canceled because it's expensive and potentially dangerous.

"It's time for a rethink," said Jeff Finckenor, an award-winning NASA engineer who last month quit the Ares program in frustration over the way the program is being managed.

Internal documents and studies obtained by the Orlando Sentinel appear to support concerns expressed by Finckenor and others. Nonetheless, NASA's leaders maintain that Ares will be ready for launch in 2015.

"At the highest levels of the agency, there seems to be a belief that you can mandate reality, followed by a refusal to accept any information that runs counter to that mandate," said Finckenor, whose farewell letter to his colleagues denouncing NASA management was posted (without his permission) on NASAWatch.com, an independent Web site.

The Sentinel reviewed more than 800 pages of NASA documents and internal studies and interviewed more than a dozen engineers, technicians and NASA officials involved with the project. Most, fearing retribution from NASA management, spoke on condition that their names would not be used.

All agreed that, eventually, NASA would be able to get Ares I to fly. The real question, they said, is whether the agency will be able to build it on time and on budget. What's more, they said, it will never be the robust, simple rocket that NASA intended.

"If they push hard enough, yes, it will fly," said one NASA engineer working on Ares. "But there are going to be so many compromises to be able to launch it, and it will be so expensive and so behind schedule, that it may be better if didn't fly at all."

Project earned C grade

NASA had to quell near-revolts by astronauts and scientists who last month took issue during a preliminary design review of Ares I. In the end, they were cajoled into backing the review.

The review graded the rocket against 10 criteria from NASA's program-management handbook. Seven of the marks were the equivalent of a C or a D. Overall, the project earned a grade-point average of 2.1, a low C.

The reasons for the low grades included concerns that its electronics and control systems could be shaken apart on liftoff and the launch-drift issue.

Astronauts, whose prime concern is safety, are still not happy.

Leroy Chiao, a former space-station commander who retired in 2005, stays in touch with his colleagues.

"I would say that I have heard various concerns," he said. "If I were still in the corps, I'd be skeptical about when is this thing going to fly and will we be able to put all the fixes in place."

One reason the astronauts are angry, Chiao and others say, is because NASA earlier this year relaxed its own safety requirements when it realized that Ares I could not meet rigid rules demanding triple redundancy on all critical systems.

The extra systems added too much weight. So, engineers said, NASA rewrote the rules to allow managers to decide how many backup systems each component needed. At one point, according to Finckenor, NASA considered throwing out all redundant systems in the launch-abort system, the emergency escape for the astronauts in case something goes wrong on liftoff.

Griffin's baby

Ares is in many ways the brainchild of NASA Administrator Michael Griffin.

In 2004, President Bush called on NASA to retire the shuttle and design a new rocket system capable of returning humans to the moon by 2020 and Mars by 2030.

At the time, Griffin was the highly respected head of the Space Department at Johns Hopkins University's Applied Physics Laboratory and had written a scholarly paper proposing a rocket design similar to the Ares I. It was revolutionary, with a first stage created by stacking the shuttle's solid rocket boosters, a liquid-fueled second stage and a manned capsule on top.

In April 2005, Griffin was appointed NASA administrator with a mandate to get the moon program moving. Within months, he organized a study that passed over other proven rockets and chose the Ares I as safe, simple and relatively inexpensive because it used lots of parts from the shuttle.

Experts say its problems stem from changes to the original design. These modifications, such as changing the engines and making the solid rocket boosters longer, created unexpected problems, including excessive shaking and the launch drift.

In a recent interview, Griffin defended NASA's process.

"We have been doing design work and development the way it has always been done. I mean, this is among the very hardest things that human beings do," he said.

"There has never been an aerospace system developed without problems, and there likely never will be. In the end NASA has always fixed them, and we will fix them this time."