Uncertainty and Sensitivity Analysis for Reentry Flows with Inherent and Model-Form Uncertainties

Hosder, Serhat; Bettis, Benjamin R.

doi:10.2514/1.a32102

Cited by 48 publications

(39 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To do this, polynomial chaos methods can be implemented using an intrusive or a nonintrusive approach. Although an intrusive method may appear straightforward in theory, for complex problems this process may be time consuming, expensive, and difficult to implement [11]. In contrast, the nonintrusive approach can be easily implemented to construct a surrogate model that represents a complex computational simulation, because no modification to the deterministic model is required.…”

Section: Uncertainty Quantification Methodologymentioning

confidence: 97%

“…In recent studies [11,12,14,16,17], the polynomial chaos method has been used as a means of UQ over traditional methods, such as Monte Carlo, for its computational efficiency. Polynomial chaos is a surrogate modeling technique based on the spectral representation of the uncertainty.…”

Section: Uncertainty Quantification Methodologymentioning

confidence: 99%

“…The first objective is to demonstrate the use of stochastic response surfaces based on nonintrusive polynomial chaos (NIPC) for efficient quantification of uncertainty in system performance metrics, as well as performance boundaries. Previous work by Hosder and Bettis [11] and Bettis et al [12], as well as Hosder et al [13,14], in the area of using stochastic expansions based on NIPC as a means of efficient uncertainty quantification (UQ) has been extensively investigated. These works included studies involving the propagation of both aleatory (inherent) and epistemic uncertainties through a variety of stochastic model problems.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Quantification of Margins and Uncertainties for Integrated Spacecraft Systems Models

West

Hosder

Winter

2015

Journal of Spacecraft and Rockets

Self Cite

View full text Add to dashboard Cite

The objective of this study was to introduce an efficient and accurate approach to the quantification of margins and uncertainties for integrated spacecraft systems models. In this study, stochastic expansions, based on nonintrusive polynomial chaos, were used for efficient representation of uncertainty both in design metrics and associated performance limits of a system. Additionally, procedures were outlined for analyzing systems that contain different uncertainty types between the performance metrics and performance limits. These methodologies were demonstrated on two model problems, each possessing mixed (epistemic and aleatory) uncertainty, which was propagated through the models using second-order probability. The first was a complex system model of highly nonlinear analytical functions. The second was a coupled multisystem, physics-based model for spacecraft reentry. The performance metrics consisted of two systems used to determine the maximum g load, the necessary bank angle correction, and maximum convective heat load along a reentry trajectory. Overall, the methodologies and examples of this work have detailed an efficient approach for measuring the reliability of complex spacecraft systems models, as well as the importance of quantifying margins and uncertainties for the design of reliable systems. Nomenclature c i = mass fraction of species i F = performance metric FL, FU = lower and upper performance limit h = enthalpy or altitude, km h D = enthalpy of diffusion, J∕kg h 0 = total enthalpy, J∕kg h 0 f = heat of formation, J∕k mol Le = Lewis number M LW , M UP = lower and upper performance gate margin m = mass, kg N s = number of samples N t = number of terms in a total-order polynomial chaos expansion n = number of random dimensions P = pressure, Pa Pr = Prandtl number p = order of polynomial expansion _ q = heat flux, W∕cm 2 r = orbital radius, km S = reference area, m 2 s = downrange distance, km T = temperature, K U = velocity, m∕s U F = performance metric uncertainty U FL , U FU = lower and upper performance limit uncertainty U LW , U UP = lower and upper performance gate uncertainty α = deterministic coefficient in the polynomial chaos expansion α = generic uncertain function β = confidence level γ = flight-path angle, deg ϵ = wall emissivity θ = longitude, deg μ = dynamic viscosity, kg∕m · s ξ = standard input random variable ρ = density, kg∕m 3 σ = Stefan-Boltzmann constant, 5.67 × 10 −8 W∕ m 2 -K 4 , or bank angle, deg ϕ = latitude, deg Ψ = random basis function or heading angle, deg ω = planetary body rotation rate, rad∕s Subscripts c = conduction d = diffusion e = boundary-layer edge condition r = radiation w = wall condition ∞ = freestream condition

show abstract

Section: Uncertainty Quantification Methodologymentioning

confidence: 97%

Section: Uncertainty Quantification Methodologymentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Quantification of Margins and Uncertainties for Integrated Spacecraft Systems Models

West

Hosder

Winter

2015

Journal of Spacecraft and Rockets

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recently, Eldred et al [31] have demonstrated mixed UQ using different methods like interval optimization, second-order probability [19,28,30,32] and DSTE. They investigated the use of nested sampling for mixed UQ, where each sample taken from the epistemic distributions at the outer loop results in an inner loop sampling over the aleatory probability distributions.…”

Section: Dste For Epistemic and Mixed Uncertainty Representationmentioning

confidence: 99%

“…A sampling based computational strategy for the representation of epistemic uncertainty in model predictions with evidence theory was introduced by Helton et al [15] to reduce the computational cost of crude Monte Carlo method. In the present paper, a stochastic response surface constructed using point-collocation NIPC [26][27][28][29][30] has been implemented as a response surrogate in uncertainty analysis in order to reduce the computational cost.…”

Section: Dste For Epistemic and Mixed Uncertainty Representationmentioning

confidence: 99%