Houston -- The shuttle Columbia probably was pierced in a way that allowed super- hot gases to flood its wheel well and left wing as it re-entered Earth's atmosphere, investigators said Thursday.

In their first significant finding, the accident review board said it was not possible for excessive heat from a missing tile alone to create the dramatic temperature increases registered by internal sensors in Columbia's wheel well minutes before it disintegrated over east Texas on Feb. 1.

NASA engineers told the board that the temperature increase probably resulted from the wing being breached, though it is still unclear how or exactly when it occurred.

Also on Thursday, NASA said remains from all of Columbia's crew had been identified at Dover Air Force Base in Delaware.

NASA has said it is examining whether insulating foam that broke off an external fuel tank during launch damaged the delicate tiles that protect the shuttle during its fiery re-entry to Earth, or whether debris in space from defunct spacecraft or a meteoroid could have struck the shuttle at high speed. Engineers continue to study what could have led to penetration of the wing.

On Thursday, NASA outlined a dramatic, multi-layered chronology of the shuttle's final minutes that picks up as the craft hit the first appreciable atmosphere, about 395,000 feet above the Pacific, northwest of Hawaii, about 5: 44 a.m. PST. About seven minutes later, the vehicle entered the peak heating phase, when temperatures climb above 2,000 degrees. Less than 1 1/2 minutes later, flight controllers in Houston saw the first unusual heat reading, in the left landing gear system in the wheel well.

In less than a minute, a temperature sensor on a wheel well strut and four temperature sensors near the back of the left wing failed. Shortly afterward, Jeff Kling, a specialist in mechanical systems at NASA's Houston Mission Control, sounded an alert.

Just past 5:53 a.m., as Columbia streaked over Sonoma County at 23 times the speed of sound, another rear elevon hydraulic line temperature sensor failed. And seconds later, as the shuttle passed west of Sacramento, a sensor on the left main brake line, near the landing gear door, showed an unusual temperature increase that continued until the shuttle lost contact with the ground.

Within the next minute, another brake line temperature sensor showed unusual heating and the onboard computer directed the orbiter's wing flaps to move to correct an increasing drag on the shuttle's left side.

About 5:54 a.m., as the spacecraft approached the Nevada border, two sensors on the fuselage above the left wing showed excessive heating.

Additional unusual temperature readings popped up as the shuttle sped across Nevada at almost 22 times the speed of sound. At the same time, something on the structure was causing increasing drag.

Over Arizona, landing gear sensors registered leaps in temperature, and sensors on the bottom and then the top of the left wing failed. About 5:56 a.m. , the shuttle started the first of four planned s-turns, or banks, to bleed off energy and slow down enough to land in Florida.

Northwest of Albuquerque, flight controllers saw strange readings from a left outboard landing gear tire pressure sensor, and seconds later from a second tire pressure sensor in the same area. With the shuttle at about 216, 000 feet, traveling a little more than 20 times the speed of sound, they saw the beginning of "sharp" elevon movements as the flight control systems fought to maintain the shuttle's crucial aerodynamic orientation.

In the next seconds, the left main landing gear tire pressures and temperature readings disappeared, and the left inboard wheel registered a decrease in temperature. At 5:58 and 39 seconds, the onboard computer system sounded an alarm alerting the crew to the loss of tire pressure readings.

About 5:59 a.m., telemetry from one sensor indicated Columbia's left main landing gear was down and locked. However, two sensors continued to show that the gear was still locked in the stowed position.

The accident investigation board Thursday reported NASA's conclusion that the gear had not deployed, and that the single sensor reading indicating otherwise was faulty. The drag on the shuttle would have been "far, far greater" than what the telemetry showed at this point, NASA spokesman James Hartsfield said.

But the drag was increasing rapidly because of whatever was happening to the left wing, the telemetry showed.

Flight Director Leroy Cain asked the mechanical systems team if the problem was in the instruments.

"Those are also off-scale low," Kling said.

Shuttle pilot Col. Rick Husband tried again to talk to Mission Control: "Roger, uh, buh . . ."

Seconds after this final, abortive communication, two of the shuttle's right-side jets fired to help keep the craft in the right flight orientation, and the elevon motions grew, the left one moving upward by 8.11 degrees.