Perspectives on Returning the Space Shuttle to Flight
Technology Today talks with a panel of SwRI experts in mechanical and materials engineering, ballistics, computational fluid dynamics and impact modeling regarding the Institute’s role in NASA’s space shuttle return-to-flight program.
TT: Describe what you did in relation to the return-to-flight program.
Grosch: I got a call at lunchtime two days after the [Columbia] accident, which happened on a Saturday. Justin Kerr from Johnson Space Center asked me, “Can you launch a large piece of foam at velocities approaching 900 feet per second?” I said I don’t know if we can, nobody else has ever done it before, but we’ll try. By Thursday afternoon, three days after we had the initial call, we had built a barrel, set up the instrumentation, built a surrogate projectile, and demonstrated that yes, we could achieve the goals that NASA set for us to launch this projectile at that velocity. Justin basically told me, “Don, your life is about to change. You’re about to become a very busy man.” And he was right. A lot of people, at SwRI and elsewhere, began to be very busy at that point.
Walker: We had done this hypervelocity work for Justin before and, in fact, the International Space Station shielding was certified here at the Institute, where we have a unique facility that can launch impactors at about 25,000 miles an hour.
Grosch: The investigation came to a fairly abrupt stop — for us, anyway — when we put that 16-inch diameter hole in Leading Edge Panel No. 8 on July 7, 2003. About three or four days later, it all started up again [with] return to flight. We began focusing on launching much smaller pieces of foam, but instead of shooting them up to 900 feet per second, to this date we’ve gone up to 2,600 feet per second with much smaller pieces of foam. That was a whole new learning curve for us. It took a lot of work, a lot of calibration shots, [but] we got a lot of data shots under our belt for return to flight. It’s close to 2,000 shots right now.
Walker: We were mostly involved in modeling. We produced analytic models for insulating foam striking thermal tiles, as well as ice striking thermal tiles and ablator materials striking thermal tiles. They essentially have three uses, these fast-running models. We have the pre-flight use, which is when they are used for some of the analysis of how risky it is to fly. Then pre-launch use, meaning once the shuttle was on the pad and, say, something like ice hits the shuttle because it falls off the gantry or from one of the fittings, do we need to go out and repair it? The other was during flight, after the orbiter went up. It was highly imaged with video cameras as well as radar, and even some cameras on planes. We also did large-scale numerical simulations using our hydrocodes. We were doing impact calculations on insulating foam, ice and ablator materials striking the thermal tiles, and we were looking at various things. We were exploring some of the extremes in the impact envelope with those models. So that was our main contribution.
Miller: We were asked to study a number of issues related to what is called the CIPAA device, which stands for Cure-In-Place Ablator Applicator, a tile repair system the astronauts could don to make repairs to the orbitor’s thermal protection tiles during EVA [extravehicular activities] — not leading-edge repairs necessarily. There are several issues associated with the repair material the CIPAA device is used to dispense, a two-part polymeric material much like bathroom silicone though a much more sophisticated formulation. When this is applied to damaged tiles, the cured product can provide thermal protection to the orbiter upon reentry through a process of controlled thermal ablation. The official designation of this special ablative material is STA-54. What was not very well understood is the bubbling behavior the material exhibits when it’s processed, packaged and stored in the flight-worthy device and then dispensed in space using the CIPAA device. There was little understanding or consensus as to why the components of STA-54 and the dispensed mixture have a propensity to bubble or evolve gas, even though the components were very carefully processed and packaged under vacuum. This bubbling phenomenon can lead to larger-than- acceptable voids in the cured product, which would be detrimental to performance of the material as an ablative heat shield. So we supported a thoroughly detailed and extensive effort to gain a better understanding of the material’s physical and chemical properties to explain the mechanism by which gas bubble formation occurs and, perhaps, provide a practical solution to the problem. In particular, what’s the source of bubble formation after processing? What are the gas diffusion characteristics of the material and how can they be measured and modeled mathematically? But most importantly, how much, gas does the material absorb, and how, during its processing? That led to other questions as well, having to do with microgravity effects in space. What happens to this material when you try to dispense it in a microgravity environment, but also under vacuum? NASA was very keen on wanting to have some level of understanding of those properties. One of the most interesting discoveries to come out of this work was that one or more of the solid constituents in STA-54 exhibited a surprisingly effective gas-to-solid-phase adsorption isotherm for air, which could explain why this material evolves gas latently. This led to a viable hypothesis for the unique bubbling behavior of the material and to the development of a semi-empirical model for predicting the diffusion of gas in the material.
Freitas: Our primary mission was to look at the two CIPAA components —there’s an ablator material that’s actually two separate materials, an A and a B part, which are stored separately and flow through tubes into a hand-gun system where they are mixed. Our job was to predict what the mixing efficiency was for these materials. What makes it complicated is that these materials were non-Newtonian, which means they can flow in unusual ways. The key was, how efficient were these static mixers. So, we made predictions of both Newtonian flow and non-Newtonian flow through the device using our computational fluid dynamics tools, and then we made estimates of the pressure drop. We showed that this current design probably has about a 50 to 60 percent mixing efficiency. I think they were doing some tests in orbit of these materials to see if it would bubble, but the gun system was not deployed operationally on the recent Discovery mission, STS-114.
Riha: We worked on one part of the impact problem. We did technical assessment of the debris transport; that is, if a piece of foam or ice comes off, where will it impact? There’s a lot of variability in this, and during the whole launch procedure as it’s going up, the flow field is going to change. So a piece of foam coming off from one point during one part of the launch may impact on one part of the wing or tile, and there’s a lot of uncertainty involved.
Part of our technical assessment was to evaluate the tools NASA was using to predict the impact of the foam or the ice. There are three parts: the probability that foam comes off; the probability that that piece of foam actually impacts a certain part of the wing or the leading edge; and the probability that damage is caused by that impact. One part of that is to look at probabilities of impact on different parts [of the shuttle] caused by foam from different parts of the external tank. NASA was working with some of their contractors and already had some tools in place for predicting the probabilities of impact on the wings.
What they came up with were essentially maximum kinetic energies, or what could cause damage on the RCC [reinforced carbon-carbon] or the tiles. So we ended up evaluating the tools they were using and their probabilistic inputs for the uncertainty on, say for example, the size of the foam that is released and the flow field parameters. We could essentially predict maximum kinetic energies that might be seen from foam liberated from different parts of the external tank and include scenarios for the three parts of the probability problem. The SwRI-developed NESSUS® program (see article on Page 12) was used and allowed quick integration of the analysis codes to predict the debris transport.
Hudak: Our work was done at the request of the NASA Engineering and Safety Center [NESC], which was established by NASA in response to the recommendations of the Columbia Accident Investigation Board following the Columbia accident. We were asked by NESC to assist with an Independent Technical Assessment of an issue related to the shuttle’s main engine. In May of 2002 cracks were discovered in a component called the flowliner inside the 17-inch diameter hydrogen fuel line that takes hydrogen from the external tanks into the space shuttle’s main engine. Our job was to integrate all the input from teammates at various NASA centers and do a probabilistic analysis. This required us to understand the physics involved and model the sources of uncertainty. Then, through simulation, we predicted what the probabilities of failure would be. Now, this takes a lot more work than a worst-case type deterministic analysis, but what it gives you is a quantitative measure of reliability. We showed that with proper inspections between flights, the probability of failure of the flowliner could be maintained below the probability of failure of the overall mission. This was the basis for a return-to-flight rationale.
Riha: Another interesting part of that is, there were 37 slots [in the flowliner], and in each slot there were eight crack location possibilities that had to be evaluated and worked into the probabilistic analysis. It’s very complex, but the probabilistic system was set up and analyzed in a short amount of time using the NESSUS software, allowing time to focus on the technical issues. We defined scenarios for nondestructive evaluation capability for flaw detection to achieve a given probability of failure.
TT: The project team received an award for this, didn’t it?
Hudak: Yes, the team got a Group Achievement Award from Ralph Roe, who is director of the NESC, so it was nice to see that NASA and NESC appreciated the effort. I should emphasize that we were part of a very large team with all the NASA centers, although we ended up being the integrator of a lot of information. One of the challenges in dealing with all these people was figuring out who had what knowledge, and getting that from them.
TT: Why did NASA include the Institute in the return-to-flight program?
Walker: The reason we were approached to do the follow-up work on return-to-flight was because of the work we had done for the Columbia investigation. We did a really good job, both experimentally and modeling-wise, and that’s what led to the follow-up work for the return-to-flight.
Grosch: During the investigation we made a lot of modifications to our [compressed gas] gun system. We learned to shoot the foam — large pieces of foam that had never been shot before — and actually, to this date, there still haven’t been any large pieces of foam shot anywhere but Southwest Research Institute.
Freitas: Those are the specific reasonings, but the general reasons are that we have done good work for NASA for years, and so we are an organization that they feel comfortable coming to.
Hudak: I think that in our case several members of the NESC knew about our long-standing reputation in probabilistic analysis and life prediction. In fact, some of the tools that we used were initially developed for NASA 15 years ago.
Miller: It helps, too, to have a multidisciplinary organization. You tend to dabble in all of the NASA offices. Our work really emanated from Johnson Space Center. We collaborated with KSC [Kennedy Space Center] as well as the NASA contractors. So, I think it’s important that we continue to have this sort of multidisciplinary approach and be able to collaborate with all of the NASA offices around the country.
TT: What obstacles has SwRI faced with this program?
Walker: The time constraint. And even though it’s been two and a half years since we lost Columbia, on the impact end it’s been, as Don pointed out, almost continuous.
Hudak: We proposed a nine-month study, and when we went to the first meeting we were told it had to be done in half that time. We met that schedule. It was very difficult, but also rewarding in the end — people just pitched in, in spite of the fact they were very busy to begin with. People really came together. That was the experience working with all the NASA people, too, at the centers. The team dynamic was probably one of the more rewarding things that came out of this experience.
Miller: Certainly the time constraint, but our project started off as a small, time- and-materials-type project, less than $20,000. After defining the problem and proposing a hypothesis as to what the source of the problem was, that very rapidly grew to a substantial program, but nonetheless with similar time constraints.
TT: Does a probabilistic analysis approach help to measure risk, or reduce risk, on future flights?
Riha: I think it will be both. They have to quantify the risk first, and there’s a lot to it.
Hudak: There were no available statistics on many of these kinds of rare events that we’ve talked about today, so NASA didn’t know what number to input into their system analysis. The approach here is to calculate the probabilities from first principles, or at least sound engineering. As David mentioned, I think the big benefit for mitigating risk is knowing what is contributing to your risk. You obviously want to put your resources into reducing the uncertainty that is controlling the risk. So you can use this approach as a research planning tool as well, to make your investments where you’re going to get the biggest return. And the return is measured in terms of lowering the risk.
TT: How did this experience affect SwRI’s culture?
Walker: I think one of our strengths was our approach, which we brought to the investigation and the return to flight. We were able to bring our expertise, which involved combining the experimental work with the large-scale numerical simulations and the analytical modeling work, and tying those three together.
Hudak: The general approach that we used in the flowliner project we’ve used for problem-solving for many other industries. We had to tune it to NASA’s needs, and this was probably one of the more complex problems that we’ve addressed thus far. However, the process of modeling the uncertainty and coupling it with deterministic models is something we’re doing for military and commercial aircraft, land-use vehicles, biomechanics of bone and implants, and deepwater offshore structures. Thus, instead of changing our culture, I believe return-to-flight analyses like these are helping to change NASA’s way of thinking. They’re now starting to think more in terms of a fully probabilistic analysis, not just gathering statistics but doing mechanics-based analysis to predict the probabilities. One of the parting comments I had from the fellow that led our independent technical assessment from NASA was that he wanted to apply this to other problems that they had.
Grosch: I think the fact that our range technicians worked with other range technicians and I worked with other experimental engineers — a lot of different people — I’m sure there are some things I’ve taken away, maybe changed some procedures to make them a little better.
TT: How do you feel about the relationship between the Institute’s role in return-to-flight and its place in history?
Hudak: Every once in a while a project comes along that makes one happy that one went into science and engineering. I think everybody who worked on our team certainly felt that way about the flowliner project. That’s why they put in a lot of extra effort to get the job done on the schedule that we needed to meet. In a nutshell, I think we all felt good about the contribution, and we’d sign up to do it again, even though it does affect our lives in little ways and not-so-little ways.
Riha: I feel really rewarded working on a program of national significance like this. And one with technology challenges. That’s why most of us work here.
Miller: For me personally, I used to live in the Space Center area and I watched many launches, including Apollo. I’ve always been fascinated by the science and engineering that goes into launching a vehicle. Even when you go back to those days in the ‘70s, to think of what technologies were available, and yet be able to launch a vehicle of the size of the Saturn rocket, for example. If it weren’t for Apollo and if it weren’t for the shuttle, I don’t think scientists and engineers would be as stimulated and challenged to be innovative in designing and building and solving problems of this magnitude. You almost need to be challenged in that respect to innovate and develop technologies that are of benefit to mankind.
Freitas: On a personal level, I’d actually met Dr. [Kalpana] Chawla* before she was an astronaut. She worked at a company at which I served on the board of directors. And so to be involved, a little bit, to help resolve some issues related to that event was nice.
Comments about this article? Contact Joe Fohn, Technology Today editor, at (210) 522-4630 or firstname.lastname@example.org.
Published in the Summer 2005 issue of Technology Today®, published by Southwest Research Institute. For more information, contact Joe Fohn.