ALBUQUERQUE, N.M. — From the lessons of the 2003 Space Shuttle Columbia disaster to today’s routine commercial returns, heat shields—formally known as thermal protection systems (TPS)—remain essential for surviving extreme heat and friction during atmospheric reentry and hypersonic flight.
A team at Sandia National Laboratories has developed new methods to rapidly evaluate TPS materials for hypersonic vehicles. Over three years, engineers blended computer modeling, laboratory experiments, and flight testing to understand how heat shields behave under intense temperatures and pressures and to predict performance much faster than traditional approaches.
Project lead Justin Wagner said the effort began with a practical need: “Predict the response of heat shields more rapidly to assist Department of Defense customers.” The goal, he added, is to reduce the number of materials that require full qualification while improving understanding of those that do.
The team studied materials ranging from common graphite to advanced carbon-based and ceramic composites. Hundreds of samples were produced by a materials science team led by Bernadette Hernandez‑Sanchez, with contributions from Oak Ridge National Laboratory.
Because full flight conditions can’t be perfectly replicated on Earth, researchers used targeted experiments to mimic key elements:
- An inductively coupled plasma torch—with plasma hotter than the surface of the sun—to observe how small samples ablate (burn away) under extreme heat.
- Sandia’s National Solar Thermal Test Facility, concentrating sunlight to generate high temperatures for larger slabs.
- A hypersonic shock tunnel to simulate brief bursts of Mach 10 flow, capturing the aerodynamics and thermal loads present in hypersonic flight.
Results were compared against advanced ablation models built by collaborators at the University of Minnesota Twin Cities, with added materials data from University of Colorado Boulder, University of Illinois Urbana–Champaign, and Kratos Inc.
Using the experimental data, a modeling team led by Scott Roberts developed a full-physics model of material properties, aerodynamics, and heat transfer. A second team led by Jon Murray trained a reduced-order model (ROM)—a compressed representation that captures the most important features—using machine learning.
- The ROM was about 90% accurate compared to the full-physics model for similar missions and designs.
- It runs thousands of times faster: seconds on a desktop versus days on a supercomputer, enabling rapid design iteration and mission assessments.
- The team is working toward automated retraining, so updates in full-model material properties flow seamlessly into the ROM.
To verify real-world performance, the team flew heat-shield samples on two suborbital rocket launches through the Multi‑Service Advanced Capability Hypersonics Test Bed. Samples—ranging from coin-sized to 4‑inch wedges—were outfitted with temperature sensors. Additional instruments included:
- Optical emission spectrometer on the first flight.
- Laser absorption spectroscopy (with partners Purdue University and PSE Technology) on the second.
Next up: a 2026 test flight sponsored by the Air Force Research Laboratory via the Prometheus program, featuring a new tile with multiple material samples mounted on the nose of a reentry capsule. If recovered post‑flight, researchers plan to measure ablation depth and analyze residual chemistry to further validate the models.
“Flight gets you everything,” said mechanical engineer Katya Casper, noting that while ground tests capture pieces of the environment, in-flight data confirms how materials behave under the full spectrum of hypersonic conditions.
