Search NASAโŒ• Search

Engineering topics

Hall, Doyle T.

Publications and source records attributed to Hall, Doyle T..

12 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS2019

Determining Appropriate Risk Remediation Thresholds from Empirical Conjunction Data Using Survival Probability Methods

Satellites sometimes maneuver before conjunctions to remediate the risk of an on-orbit collision. Many missions use probability of collision (P(sub c)) thresholds to decide when such maneuvers should be performed. These thresholds tend to be conservative because of policies that require satellites survive their lifetimes without collision with high confidence (e.g., 99.9%). This study presents a semi-empirical method to estimate remediation P(sub c) thresholds that satisfy such lifetime risk requirements. The formulation combines survival probability analysis with empirical conjunction histories to estimate remediation thresholds as a function of satellite size, remaining on-orbit duration, lifetime collision probability limit, collision consequence, and other parameters.

Risk remediation thresholdsโ†—
NASA NTRS2019

Implementation Recommendations and Usage Boundaries for the Two-Dimensional Probability of Collision Calculation

The two-dimensional (2D) probability of collision (๐‘ƒ๐‘) estimation method relies on several assumptions that must be satisfied for accurate results. Monte Carlo analysis of ~44,000 conjunctions indicates that 2D-๐‘ƒ(sub ๐‘) pro-vides accurate estimates for most typical conjunctions, but occasionally underestimates ๐‘ƒ(sub ๐‘) significantly, indicating an assumption violation. A test to detect large-amplitude underestimation inaccuracies can be based on how much โ€œoffset-from-TCAโ€ 2D-๐‘ƒ(sub ๐‘) values vary during a well-defined time interval bracketing closest approach. The test successfully detects all large-amplitude 2D-๐‘ƒ(sub ๐‘) underestimations found to date, but with a high false-alarm rate. The analysis also provides implementation recommendations and usage boundaries for the 2D-๐‘ƒ(sub ๐‘) method.

Hall, Doyle T.โ†—
NASA NTRS2019

Implementation Recommendations and Usage Boundaries for the Two-Dimensional Probability of Collision Calculation

The two-dimensional (2D) probability of collision (๐‘ƒ๐‘) estimation method relies on several assumptions that must be satisfied for accurate results. Monte Carlo analysis of ~44,000 conjunctions indicates that 2D-๐‘ƒ(sub ๐‘) pro-vides accurate estimates for most typical conjunctions, but occasionally underestimates ๐‘ƒ(sub ๐‘) significantly, indicating an assumption violation. A test to detect large-amplitude underestimation inaccuracies can be based on how much โ€œoffset-from-TCAโ€ 2D-๐‘ƒ(sub ๐‘) values vary during a well-defined time interval bracketing closest approach. The test successfully detects all large-amplitude 2D-๐‘ƒ(sub ๐‘) underestimations found to date, but with a high false-alarm rate. The analysis also provides implementation recommendations and usage boundaries for the 2D-๐‘ƒ(sub ๐‘) method.

Hall, Doyle T.โ†—
NASA NTRS2018

High Fidelity Collision Probabilities Estimated Using Brute Force Monte Carlo Simulations

The NASA Conjunction Assessment Risk Analysis team has implemented new software to estimate the probability of collision (P (sub c)) for Earth-orbiting satellites. The algorithm employs a brute force Monte Carlo (BFMC) method which differs from most other methods because it uses orbital states and covariances propagated from their orbit determination epoch times using the full set of Astrodynamics Support Workstation higher order theory models, including the High Accuracy Satellite Drag Model. This paper de-scribes the BFMC algorithm, presents comparisons of BFMC P (sub c) estimates to those calculated using other methods, and discusses the implications for conjunction risk assessment.

Probabilitiesโ†—
NASA NTRS2017

Remediating Non-Positive Definite State Covariances for Collision Probability Estimation

The NASA Conjunction Assessment Risk Analysis team estimates the probability of collision (Pc) for a set of Earth-orbiting satellites. The Pc estimation software processes satellite position+velocity states and their associated covariance matri-ces. On occasion, the software encounters non-positive definite (NPD) state co-variances, which can adversely affect or prevent the Pc estimation process. Inter-polation inaccuracies appear to account for the majority of such covariances, alt-hough other mechanisms contribute also. This paper investigates the origin of NPD state covariance matrices, three different methods for remediating these co-variances when and if necessary, and the associated effects on the Pc estimation process.

Hall, Doyle T.โ†—
NASA NTRS2017

Relevance of the American Statistical Society's Warning on P-Values for Conjunction Assessment

On March 7, 2016, the American Statistical Association (ASA) issued an online editorial paper on the "context, process, and purpose of p-values." According to the paper, "the statement articulates in non-technical terms a few select principles that could improve the conductor interpretation of quantitative science, according to widespread consensus in the statistical community." This "consensus" statement was accompanied by 21 "commentaries" expressing a diversity of opinions among the panel of experts ASA convened. In the present work, we express our view that the ASA p-value warning has relevance to the space object Conjunction Assessment (CA) community.

p-valuesโ†—
NASA NTRS2017

Time Dependence of Collision Probabilities During Satellite Conjunctions

The NASA Conjunction Assessment Risk Analysis (CARA) team has recently implemented updated software to calculate the probability of collision (P (sub c)) for Earth-orbiting satellites. The algorithm can employ complex dynamical models for orbital motion, and account for the effects of non-linear trajectories as well as both position and velocity uncertainties. This โ€œ3D P (sub c)โ€ method entails computing a 3-dimensional numerical integral for each estimated probability. Our analysis indicates that the 3D method provides several new insights over the traditional โ€œ2D P (sub c)โ€ method, even when approximating the orbital motion using the relatively simple Keplerian two-body dynamical model. First, the formulation provides the means to estimate variations in the time derivative of the collision probability, or the probability rate, R (sub c). For close-proximity satellites, such as those orbiting in formations or clusters, R (sub c) variations can show multiple peaks that repeat or blend with one another, providing insight into the ongoing temporal distribution of risk. For single, isolated conjunctions, R (sub c) analysis provides the means to identify and bound the times of peak collision risk. Additionally, analysis of multiple actual archived conjunctions demonstrates that the commonly used โ€œ2D P (sub c)โ€ approximation can occasionally provide inaccurate estimates. These include cases in which the 2D method yields negligibly small probabilities (e.g., P (sub c)) is greater than 10 (sup -10)), but the 3D estimates are sufficiently large to prompt increased monitoring or collision mitigation (e.g., P (sub c) is greater than or equal to 10 (sup -5)). Finally, the archive analysis indicates that a relatively efficient calculation can be used to identify which conjunctions will have negligibly small probabilities. This small-P (sub c) screening test can significantly speed the overall risk analysis computation for large numbers of conjunctions.

Hall, Doyle T.โ†—
NASA NTRS2013

LEGEND, a LEO-to-GEO Environment Debris Model

LEGEND (LEO-to-GEO Environment Debris model) is a three-dimensional orbital debris evolutionary model that is capable of simulating the historical and future debris populations in the near-Earth environment. The historical component in LEGEND adopts a deterministic approach to mimic the known historical populations. Launched rocket bodies, spacecraft, and mission-related debris (rings, bolts, etc.) are added to the simulated environment. Known historical breakup events are reproduced, and fragments down to 1 mm in size are created. The LEGEND future projection component adopts a Monte Carlo approach and uses an innovative pair-wise collision probability evaluation algorithm to simulate the future breakups and the growth of the debris populations. This algorithm is based on a new "random sampling in time" approach that preserves characteristics of the traditional approach and captures the rapidly changing nature of the orbital debris environment. LEGEND is a Fortran 90-based numerical simulation program. It operates in a UNIX/Linux environment.

Liou, Jer Chyiโ†—
NASA NTRS1995

Determining large-scale heliospheric structure using ultraviolet resonance line observations

Currently the Pioneer 10 and Voyager 1 and 2 spacecraft are beyond the orbit of Pluto, traveling outward from the Sun. Each is capable of detecting ultraviolet radiation resonantly scattered from hydrogen and helium atoms in the heliosphere and local interstellar medium. These observations are particularly well suited for the investigation of the large-scale heliospheric H and He distributions because the Voyager spacecraft are heading upstream, into the direction of local interstellar flow, whereas Pioneer 10 is heading downstream. Observations of the brightest resonance line, H Lyman-alpha, reveals that beyond about 20 AU from the Sun, upstream intensities decrease less quickly as a function of solar distance than downstream intensities. This implies that the heliospheric H distributions in the upstream and downstream directions are significantly different. Heliospheric H atoms originate in the local interstellar flow, and must penetrate through the heliospheric interface, where they are subject to charge exchange collisions with solar wind and interstellar protons. Models indicate that this process is probably responsible for the upstream/downstream difference in H Lyman-alpha. In addition, a recent spectroscopic determination of the H atom velocity distribution in the inner heliosphere implies a significant deceleration in the bulk flow speed of the heliospheric hydrogen gas relative to the helium flow, an effect that is also likely due to H-p charge exchange occurring in the upstream heliospheric interface region. In this presentation, recent heliospheric resonance line observations and their interpretations will be reviewed, focusing on their sensitivity to large-scale heliospheric structure.

Hall, Doyle T.โ†—
NASA NTRS1993

Upper limit on Titan's atmospheric argon abundance

An analysis is conducted on the Voyager 1 UV Spectrometer solar-occultation data and a Titan flyby spectrum of the north polar region dayglow, in order to infer the tropopausal Ar mixing ratio's upper limit as a function of the CH4 mixing ratio, f(CH4). The mole-fraction upper limit of tropopausal Ar mixing ratio ranges from 0.01 to 0.1 at f(CH4) of 0.026 to as low as 0.08 at f(CH4) of 0.05. Since the best fits to the solar occultation data require f(CH4) of more than 0.26, the Ar mixing ratio must be lower than 0.1.

Strobel, Darrell F.โ†—