NASA Explores Data-Driven Approach to Risk-Informed NDE for Spaceflight Hardware
Nondestructive evaluation (NDE) is central to fracture control for spaceflight hardware, but it also adds cost and schedule pressure. This has led to questions about whether NDE requirements can be selectively reduced, or “descoped,” without increasing risk.
NASA’s Technical Standard NASA-STD-5019A requires a combined approach of design, analysis, testing, NDE, and tracking of fracture-critical parts to ensure damage tolerance. Within this framework, flaws smaller than NDE detection limits are assumed to exist, and components must still demonstrate they can survive their required service life. NDE primarily serves to screen out larger flaws that could lead to failure.
The NASA Engineering and Safety Center (NESC) examined whether a probabilistic approach could support NDE descope decisions while maintaining an equivalent risk posture, particularly for single-wrought materials. Traditional damage tolerance is treated deterministically, with fixed detection thresholds and pass/fail outcomes, but actual risk depends on the likelihood that flaws exist, are missed, and propagate to failure.
The standard approach ensures safety by overlapping two conditions: critical flaws are unlikely below a certain size, and detectable flaws are unlikely to be missed. Removing NDE eliminates one of these safeguards, making risk dependent on the probability that a critical flaw exists.
To estimate this risk, the NESC used historical inspection data as a proxy. A dataset of 33,630 bolt-hole inspections over three years identified six crack-like indications. Using a conservative statistical bound and assuming any undetected critical flaw leads to failure, the resulting structural reliability was approximately 0.9996, expressed as 3.4 “nines.”
Sensitivity studies showed that increasing sample size provides limited gains, with achievable reliability ultimately constrained by observed defect rates. At similar rejection rates, a minimum of about 5,000 inspections is needed to avoid significant increases in risk relative to current human spaceflight standards.
The method relies on stable, time-invariant processes so that historical data remain predictive, and it requires careful justification when combining datasets across parts. It is intended to inform—not replace—existing fracture control practices and would require review by NASA’s Fracture Control Board and Technical Authority.
For the NDE community, this work reflects a shift toward data-driven, risk-informed inspection strategies. While NDE remains a critical safeguard, the study shows that, in limited cases with sufficient data and process control, reduced inspection scope may be evaluated without substantially increasing risk.
NASA/Aubrey Gemignani
NASA/Sam Lott
A free download with more details about NESC’s method for estimating risk when reducing NDE is available at https://nasa.gov/wp-content/uploads/2026/03/techup2025-pg57-58-innov-tech-nde-descoping.pdf?emrc=360746.
