Sandia aids NASA in Columbia accident investigation
Sandia played a key role in helping NASA determine the cause of the space shuttle Columbia disaster. Sandia researchers’ analyses and experimental studies supported the position that foam debris shed from the fuel tank and impacting the orbiter wing during launch was the most probable cause of the wing damage that led to the breakup the Columbia.
"Sandia’s expertise in the areas of impact testing and modeling, material testing, non-continuum aerodynamics, and thermal analysis has been invaluable to our investigation teams," writes William Readdy, NASA associate administrator for space flight, in an Aug. 12 letter to Sandia. "The cooperative effort and sharing of ideas, test methods, and analytical tools have been beneficial to both our organizations." For five months, more than 30 Sandia researchers from both Sandia sites applied the Labs’ computational and experimental engineering and material sciences resources to determine if the foam impact was in fact the cause of the fate of Columbia.
"Sandia played an important role in determining the cause of the disaster," says Engineering Sciences Director Tom Bickel (9100), programmatic lead for Sandia’s efforts. "Sandia helped guide the investigation and served as an expert advisor to NASA."
Simulations and material testing work performed by Sandia staff, along with corroborating work by NASA engineers and contractors, guided large-scale testing done at Southwest Research Institute (SwRI) in San Antonio, Texas. Testing there was performed on full-scale mock-ups of parts of the wing using flight hardware from the remaining orbiter inventory and museum displays. Foam impacts on different locations of the orbiter wing leading edge and thermal protection system (TPS) tiles on the wing underside were studied to assess and demonstrate the potential damage that could have resulted during launch.
Dramatic test: 16-inch hole
Testing showed that firing foam projectiles at various locations on the wing reinforced carbon composite (RCC) leading edge panels could produce damage ranging from localized cracking of the RCC to full breakage. The most dramatic test at SwRI produced a 16-inch diameter hole in the lower half of a leading edge panel of the orbiter. Such damage was acknowledged to be catastrophic, since the hole would allow high-temperature gases to enter the left wing and melt the aluminum wing structure during reentry.
"The confirming event of the investigation was the experiment conducted using the scenario modeled by Sandia," says Tom.
Since the tests, the Columbia Accident Investigation Board (CAIB) has acknowledged the RCC leading edge foam impact scenario to be the most likely cause of the Columbia disaster. The CAIB released its initial report last week (Aug. 26) and is scheduled to issue its full report by the end of this year.
Just two days after the Feb. 1, 2003, Columbia disaster, Sandia was contacted to see how Sandia could help with the investigation. Two days after that, Tom Bickel, Carl Peterson (9100) and Basil Hassan (9115), senior aero staff and management from Engineering Sciences Center 9100, went to Johnson Space Center to determine how the Labs could assist in the investigation.
"We joined a group of engineers from other agencies that were asked to assist with the investigation," Tom says. "Sandia’s expertise in material and engineering science was a perfect fit in the investigation."
In the weeks after the accident, Basil and Carl shared being onsite full-time at NASA Johnson Space Center. They worked with NASA, Boeing, and Lockheed Martin managers and engineers from across the country to develop credible scenarios that might have led to the accident. The bulk of this work was tied to unraveling telemetry data available from the final minutes before the orbiter breakup and analyzing the locations of orbiter debris recovered across Texas following breakup. Basil and Carl also served as intermediaries to get information on the orbiter back to Sandia engineers for incorporation into the Sandia analyses.
"Because of the small amount of available information, we had a hard time getting our arms around the problem," says Basil. "After the first month the CAIB announced it wanted answers by the end of May, which seemed like very little time for an investigation of this magnitude."
In the first several months after the accident, Sandia staff worked closely with NASA engineers to perform scoping analyses to assess the credibility of postulated damage to the Columbia that could have occurred prior to reentry and could have resulted in orbiter breakup during reentry. While the foam impact was considered a potential cause, the location of the impact, size and mass of the foam, and its impact velocity remained unknowns. The initial Sandia studies focused on assessing potential aerodynamic effects and impact damage as well as trying to unravel telemetry data that indicated severe heating in the vicinity of the orbiter left wing landing gear. Along with the NASA engineers and industry contractors, Sandia analysts focused on attempting to duplicate the response of many of the sensors.
"The difficult part of that was not knowing where to start," Basil says. "More than one possible scenario could lead to the same result. All we could do was try to eliminate as many scenarios as possible and then focus on the ones that seemed the most plausible."
Fortunately, NASA located the onboard Columbia flight recorder in late March. This recorder contained a wealth of additional temperature information and other flight data that helped NASA to eliminate several accident scenarios and speed up piecing together the puzzle. From this data, it was determined that the most probable damage location was the leading edge of the left wing. This became the focus of Sandia’s participation in the study that continued heavily through July.
Serving as an independent investigative arm for NASA, Sandia staff initially evaluated a number of possible foam impact accident scenarios using the Labs’ computational capabilities. ASCI computer platforms and codes were extensively used in this work.
Initially, Sandia provided scoping analyses for foam impacting the RCC and the tiles on the underside of the wing. These computations were the first such analyses provided to NASA management. Sandia researchers indicated that foam impacting the wing underside would not have caused much damage, but impacts to the RCC could have caused severe damage to the orbiter.
"After our prediction that the foam would penetrate the RCC, we faced a very skeptical community," Tom says. "We had to convince ourselves and our colleagues that we were correct through the use of materials testing and many confirmatory simulations."
This skepticism remained until tests were conducted of foam impacting a mock-up of the orbiter’s leading edge.
Areas of concentration
Sandia’s work focused on two major areas: aerothermodynamics and impact analysis. In the aerothermodynamics area, Sandia brought significant expertise in computational fluid mechanics, rarefied gas dynamics, and material thermal response.
The researchers used a variety of internal and external computer codes to help in the analysis, including computational fluid dynamics (CFD) analyses for the orbiter at various altitudes along the trajectory, heat transfer predictions, calculations of plumes that simulated hot gas entering the wing, and material-response calculations of possible damaged wing leading edge and tile materials. The efforts of Michail Gallis (9113, noncontinuum CFD), Jeff Payne (9115, continuum CFD), and Don Potter (9115, plasma modeling) were especially critical to NASA efforts to interpret telemetry data from the final minutes leading to orbiter breakup.
Engineers in Sandia’s structural mechanics groups performed simulations of foam impacting the orbiter. As part of this effort, they developed and refined material response models for the RCC, TPS tile, and foam materials using NASA-provided data and Sandia-measured properties.
Serving as Sandia’s technical oversight, Art Ratzel (9750) says the impact analysis study became the major focus of Sandia’s efforts in the investigation after the first two months of scoping work.
"Our impact analyses centered on various aspects including foam mass, angle of the impact, and velocity," says Art. "We used our computational tools to interpret numerous scenarios and fed back this information to NASA to support the SwRI test design and diagnostics placement."
Various impact analyses were conducted on the leading edge RCC and TPS tile materials. Kenneth Gwinn and Kurt Metzinger (both 9126) worked both problems using Pronto/SPH, and David Crawford (9116) used CTH for the impact into the tile. Because of the physics of the problem, Pronto/SPH was better for the RCC leading edge impacts and CTH worked better for the tile impacts. After preliminary analyses it was determined that the potential for extensive damage that could have led to the wing failure was much greater for the RCC than for the tile, given the impact conditions provided by NASA. This led Sandia to do more detailed analyses of the RCC impact.
In addition, Sandia experimentalists played a major role in the investigation through their material characterization efforts. Wei-Yang Lu, Bonnie Antoun, and John Korellis from the Sandia/ California Materials and Engineering Sciences Center (8700) led studies on the RCC, thermal protection system (TPS) tiles, and foam impacting materials that provided data needed to populate the material response models critical to the computational studies. Moo Lee and supporting 6100 staff also supported this work using different experimental techniques that provided confirmation of foam and TPS tile material response. NASA relied extensively on this work; NASA provided Sandia with all of the RCC materials that could be made available for testing during the initial stages of the investigation. Sandia’s RCC test data was disseminated to NASA and contractor groups supporting the accident investigation, as well as to the CAIB team. It became the baseline RCC material property data used in all of the impact analyses conducted for the investigation.
Sandia has helped provide NASA with an understanding of the analyses required to perform these types of tests concerning tile and foam. The work will help NASA if a situation arises wherein it needs to assess damage to the leading edge areas and the underside tile areas. The work should also assist NASA with return-to-flight issues to determine what sort of impact the current areas can withstand and to design mitigation methods to prevent impacts that could cause damage.
Art Ratzel and Tom Bickel reflected upon the work performed in support of the Columbia accident investigation. "In looking back over the past several months, it is clear that this team approached supporting the Columbia investigation with the same commitment to provide exceptional service that Sandians have provided in supporting our National Security mission and nationally critical studies such as the Unabomber work, the TWA-800 accident, and the USS Iowa accident.
"The team assembled represented but a fraction of the Sandia staff that early-on offered their time and technical expertise to help NASA. The infrastructure and technical capabilities resident at Sandia made possible our successful support.
"The NASA engineers with whom we worked side-by-side must also be acknowledged. Under the toughest of times and overwhelming scrutiny from the media and outside investigative teams, the NASA team remained open to our needs and suggestions, and overall demonstrated excellence in engineering.
"The terrible events that resulted in our partnership cannot be forgotten, but hopefully the path forward for NASA will include future collaborations with Sandia that will benefit our national space program."