NASA is considering five options for redesigning the so-called "bi-pod" ramp system on the shuttle's external fuel tank, the area where a chunk of foam insulation broke away during Columbia's launch, hit the left wing and possibly caused a catastrophic breach. The currently favored option calls for eliminating the use of foam in the area in favor of an exposed fitting equipped with heaters to prevent pre-launch ice buildups.

Senior NASA and contractor managers and engineers met Tuesday and today in Michoud, La., where Lockheed Martin Space Systems Co. builds the external tanks, to review the bi-pod redesign options in preparation for making a final selection in the next few weeks. While most observers believe NASA will not be able to resume shuttle flights until well into 2004 at the earliest, senior agency managers say they believe launchings could resume as early as mid-to-late December. In any case, engineers plan to have the bi-pod fix in place as soon as possible, eliminating what many believe was the initiating event in the Columbia disaster.

"Clearly, there's been a lot of focus on the foam, the bi-pod foam in particular, and our redesign efforts have been ongoing in that particular activity for a couple of months already," Michael Kostelnik, NASA's deputy associate administrator for shuttle and space station, told reporters today. "We will solve the shedding problem of foam in that particular area and our technical plan for that allows us a return to flight window in December."

The external fuel tank forms the structural backbone of the shuttle "stack" for the climb to space. Solid-fuel boosters are attached on each side with massive explosive bolts while the shuttle is attached by fittings under its nose and engine compartment. The aft end of the shuttle is attached with massive fittings where 17-inch propellant lines from the external tank feed into the ship's aft engine compartment. The nose is attached by a so-called bi-pod, two cylindrical struts that meet at a single attachment plate just behind the nose landing gear doors. The struts attach to a pair of fittings on the external tank, forming a triangular bi-pod assembly.

At launch, the tank is loaded with a half-million gallons of supercold liquid oxygen and liquid hydrogen rocket fuel. From the beginning, the bi-pod strut attachment fittings have been buried in hand-sprayed foam insulation to prevent ice from forming around the aluminum assemblies. Any such ice likely would break away when subjected to the vibration of launch, posing an impact hazard to the underside of the shuttle. The foam covering the bi-pod fittings is hand carved to form an aerodynamically benign ramp allowing the smooth flow of air over the protrusions.

As it turns out, the bi-pod ramp prevented ice buildups but created another impact hazard: The foam itself.

A suitcase-size chunk of foam from Columbia's left bi-pod ramp area pulled free during launch Jan. 16 and struck the left wing's leading edge at some 500 mph. The foam strike presumably caused the breach that led to the ship's destruction during re-entry Feb. 1. Another large piece of foam broke away from the same area during an October shuttle flight and hit a booster. All told, investigators have identified at least six missions, and possibly seven, in which large pieces of foam debris fell off the bi-pod ramps of the external tanks.

It is not yet clear what causes the foam to break away. Many engineers believe some form of "cryopumping" may be involved, in which trapped air in the foam liquifies after the tank is chilled during fueling. In this scenario, heating during ascent causes the trapped liquid to warm and expand, weakening the foam in that area. But recent tests by the Columbia Accident Investigation Board indicate cryopumping alone is not the answer and that some combination of factors may be responsible.

In any case, there is little doubt the CAIB's final report, expected in late July, will include a recommendation to redesign the bi-pod attachment system to preclude foam shedding in the future. In the meantime, NASA is pressing ahead with work to do just that.

Of the five options currently on the table, the favored solution, known as option 2a, is one in which the aerodynamically shaped foam ramp is eliminated entirely in favor of an exposed fitting equipped with one or more heaters to eliminate the possibility of pre-launch ice buildups. Aluminum end covers and caps would be replaced with tougher Inconel covers. Other changes include the addition of wiring for the heaters and additional electrical power from ground-support equipment prior to liftoff.

Other options include:

• Option 1: Reduced volume ramp with a minimum 50 percent reduction in the amount of foam required; aluminum end covers replaced by super lightweight ablator material; new heater elements and wiring.
  
  
• Option 2: Partially exposed fitting with an 80 percent reduction in foam ramp volume; Inconel end covers and caps; new heater and wiring.
  
  
• Option 3: Inconel fairing over the entire bi-pod fitting, eliminating need for a foam ramp. Strip heaters would be attached to the interior of the welded fairing. This option was an early favorite.
  
  
• Option 3A: Small Inconel fairing, equipped with strip heaters, that covers the central shaft of the bi-pod fitting. No ramp foam is required.

Agency managers, with input from the Columbia Accident Investigation Board, hope to settle on a redesign option in the next few weeks.