Protected by the shell of its huge launch-rocket during blast-off, NASA’s Orion Multi-Purpose Crew Vehicle (MPCV) must get back to earth on its own at mission’s end. The flight plan for this next-generation craft includes a dramatic ocean splash down reminiscent of the Apollo programme that pre-dated the Space Shuttle’s smooth runway landings.
To keep capsule and crew safe under the huge re-entry and splashdown loads — temperatures exceeding 2,650°C and speeds up to 25,000mph — a 5m diameter ablative thermal protection system is secured to the MPCV’s base with a carbon graphite and titanium carrier structure. As the ablative thermal protection system of this heat shield reaches extremely high temperatures, portions of it fall away from the vehicle to remove excessive thermal energy. The remaining carrier structure has to survive the brunt of the impact when it hits the water to help keep the astronaut module intact.
With the first unmanned launch-and-return test of Orion scheduled this Autumn, NASA engineers and contractors were highly motivated to get to a final design for the MPCV that achieved ideal weight and performance targets. In late summer of 2012, NASA’s chief engineer for the Orion project, Julie Kramer, contacted the space agency’s Engineering and Safety Centre (NESC) and requested some novel ideas for how to reduce the spacecraft’s mass.
NESC’s mission is to perform value-added independent testing, analysis, and assessments of NASA's high-risk projects to ensure safety and mission success. In the zero-failure tolerance environment of spaceflight, a second look from NESC boosts confidence that results are optimum and the final concept has widespread team support.
How to take a load off
At some 1,360kg, the 'baseline' composite-and-titanium design for the wagon-wheel shaped carrier structure that supports the MPCV’s thermal protection system (See image 3, left-side bottom figure ) was one of the largest components of the crew module, and thus became a prime target for the weight-reduction exercise. Mike Kirsch, project manager and principal engineer of the NESC’s Orion Heat Shield Carrier Structure Assessment Team takes up the story:
“The goal of the Orion program is to go well beyond Earth orbit, around the moon and eventually to an asteroid or Mars. Mass obviously becomes of paramount importance on such long trips. Wouldn’t you rather carry extra water, food, oxygen and propellant? By being efficient in our structural design, we can bring along more of those consumables. And when it comes to splashdown, a lighter vehicle comes in with less energy upon impact.”
Kirsch’s team, which included technical lead Jim Jeans, president of Structural Design & Analysis, Inc. (a longtime contractor for NASA) knew what engineering software they’d apply to the heat shield design assessment program: Collier Research Corporation’s HyperSizer.
They’d used the tool extensively on the Composite Crew Module (CCM), an earlier NESC project designed to demonstrate that composites could indeed be used as a primary structure of the crew module. Jeans is currently employing HyperSizer on other aspects of the Orion vehicle, and on the James Webb Space Telescope.
The first-ever software commercialised out of NASA, HyperSizer provides stress analysis and sizing optimisation, reducing the weight of aircraft, wind turbine blades and other structures in addition to space vehicles. Whether designed with composite or metallic materials, a typical HyperSizer optimization produces weight savings of between 25 and 40 percent. According to Jeans, there was no way they could have optimised the heat shield alternatives they proposed without HyperSizer.
Structural concepts
The baseline design for the heat shield consisted of a solid laminate carbon-graphite skin secured to the capsule by a carrier structure with a spoke-like pattern of titanium I-beams in the aforementioned wagon-wheel shape. The concept is similar to the aeroshell that protected the Rover for Mars entry.
Carbon graphite designs can be tailored, in that modifications can continue to be made en route to final manufacturing. However, in this case, the result was a design that weighed more than it needed to. With an initial goal of cutting out some 360kg, the NESC team considered both material and structural modifications to the baseline. Kirsch again:
“We needed to come up with a lighter structure that could still withstand the aerodynamic pressure of the Earth’s atmosphere re-entry and support the thermal protection system so the ablative material in the heat shield could do its job. Reentry is a pretty severe load case. But even more important is when the crew module actually hits the water. That water landing is the event that drives the design of the heat shield carrier structure. Using parachutes we try to take as much energy out of it before that impact, which is a tricky, dynamic situation based on wind and wave conditions. Ideally you want the capsule to knife in, not belly-flop. The design must be robust to the wide range of possible wind and wave conditions.”
The team developed a series of analytical models to predict how the heat shield carrier structure as a whole - particularly the internal support webs - would react under a wide range of splashdown scenarios. Landing simulations were run in LS-DYNA, a transient non-linear Finite Element Analysis (FEA) solver. The dynamic landing simulations were loaded into HyperSizer, which then controlled relevant parameters (such as material thickness and location of stiffeners) within each model to optimise, and then compare, different design iterations. HyperSizer can concurrently evaluate many different combinations of the variables that influence design, so it rapidly identified those configurations that had the lowest mass, according to Kirsch.
Sharing interim results among designers and analysts was enabled by the software’s ability to display summary images showing critical load case, margins of safety or failure modes. For Jeans, this was a game changer, particularly the ability to do trades between different construction methods, with apples-to-apples comparisons. “We could investigate many different configurations and be confident that we were making the right choices,” he says.
HyperSizer enabled 40 different variations – involving steel, aluminium, stainless steel, titanium, carbon graphite, honeycomb systems, T-stiffened, I-stiffened and so on – to be rapidly studied. Within ten weeks the team had identified six candidates with minimum mass configurations that significantly exceeded the original 360kg weight reduction goal. James Ainsworth, a structural engineer for Collier Research, was the HyperSizer technical advisor to the NESC group:
"This was a complex problem that we had not previously analysed in HyperSizer. The landing simulations evaluated are similar to a car crash where a vehicle is slamming into something at a high velocity and the entire event takes place over a few milliseconds. Our software allowed the team to evaluate the stresses and strains at every single time step, and use that data for detailed sizing and final analysis in HyperSizer. That was one of the biggest technical challenges we achieved on this project, to set up a process that could import the loads from every time step and size the model to find the optimum lightweight, robust structure. To put it in perspective, each one of those landing events was about 12 gigabytes of data and we crunched through some 50 landings.”
HyperSizer V7
The kind of transient, dynamic analysis that the NESC performed on heat shield candidates is a new feature in the latest release of HyperSizer, Version 7. James Ainsworth again:
“We included new analytical methods to account for the load redistribution that happens when you have nonlinear material and geometric responses, such as material plasticity and plastic bending. HyperSizer also handles the dynamic landing events by processing thousands of time steps for each landing simulation. A benefit of working closely with a customer like NASA is that we could readily accommodate their analysis needs by enhancing our software in real-time to keep pace. As a result, in addition to the full range of FEA solvers, we can now support, and automatically optimise, a structural design based on LS-DYNA.”
Their in-depth evaluations of a variety of engineering concepts led the NESC team to consider alternative designs that incorporated load sharing with the crew module backbone, replaced the exiting wagon-wheel stringer with an H-beam configuration, or switched the composite carbon graphite skin to a titanium orthogrid skin. The titanium orthogrid version emerged as their final proposal.
NESC built a test structure to verify that they understood the physics of the dynamic impact on this final alternative. The real-world tests provided sufficient data to validate the ability of HyperSizer to model those physics with its enhanced analytical tools.
As NASA’s Orion team continued their own work on reducing the weight of the baseline design, NESC’s insights from the HyperSizer analyses informed discussions that led to a reduction of the final weight of the baseline design by 23 percent, eliminating many kilograms of unnecessary weight. “Our work provided some independent perspective on where the baseline design was in the spectrum of alternatives, and highlighted additional efficiencies that were possible,” says Kirsch.
The lighter, stronger MPCV that resulted from the NASA/NESC partnership is scheduled for launch in 2016 with the redesigned version of the heat shield.