Pith. sign in

REVIEW 3 major objections 4 minor 26 references

Improved Directional State Transition Tensors for Accurate Aerocapture Performance Analysis

T0 review · 3 major / 4 minor · reviewed 2026-08-03 · deepseek-v4-flash

Pith's one-line read Aerocapture perturbation propagation is cheaper and more accurate when state transition tensors are aligned with higher-order Cauchy-Green tensor eigenpairs instead of linear ones.

desk verdict A genuinely useful extension of DSTT basis selection, but the validation never gets close to the nonlinear regime where the paper claims it wins. read the letter →

arxiv 2512.12475 v2 pith:EH27W3CF submitted 2025-12-13 eess.SY cs.SY

classification eess.SYcs.SY
keywords aerocapturestatetransitiontensorsdirectionalhigher-orderCauchy-Greentensoreigenpairsuncertaintypropagationquantityofinterestnonconservativedynamics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that state transition tensors, higher-order Taylor maps of the trajectory flow, become practical for aerocapture when they are directionalized along the dominant nonlinear stretching direction of the dynamics rather than the usual linear stretching direction. It derives augmented higher-order Cauchy-Green tensors, including selective versions for chosen state subsets and quantity-of-interest versions for functions like energy or apoapsis radius, and uses their dominant tensor eigenpairs to build reduced-dimension DSTTs. The paper shows that a single such nonlinear direction produces more accurate perturbation propagation than DSTTs built from up to six linear stretching directions, at lower computational cost. This matters because aerocapture dynamics are dominated by nonconservative aerodynamic forces whose strength varies wildly, making cheap and accurate nonlinear propagation desirable for onboard guidance and uncertainty quantification.

What carries the argument

The carrying mechanism is the tensor eigenpair of an augmented higher-order Cauchy-Green tensor. From the STTs one forms tensors that expand the squared norm of the final state, or of a selected subset for sCGTs, or of a quantity-of-interest map for qCGTs, in powers of the initial perturbation. The dominant z-eigenvector of the symmetrized third- and fourth-order tensors is the initial perturbation direction that maximizes the final nonlinear response; these vectors become the rows of the rotation matrices that directionalize the second- and third-order DSTTs. Because each DSTT order can use a different rotation matrix, the alignment can track the nonlinear stretching direction appropriate t

What would settle it

Propagate the same 10,000 perturbed aerocapture trajectories with realistic entry dispersions, such as kilometer-level position errors, tens of meters per second velocity errors, and larger density perturbations, and compare the DSTT directionalized by the nominal HOCGT eigenvector against full STTs and Monte Carlo integration: if the DSTT error grows disproportionately or a multi-direction basis becomes clearly better, the single-direction assumption fails.

Watch

Extended reading notes

Core claim

The central claim is that, for aerocapture, the direction used to reduce state transition tensors must come from the nonlinear dynamics, not the linearized dynamics. The paper constructs higher-order Cauchy-Green tensors from the STTs and computes their dominant tensor eigenpairs, which identify initial perturbation directions that produce the largest final state change through the nonlinear dynamics. These directions are used to build DSTTs with one latent dimension, and the paper shows that these single-direction DSTTs match or beat DSTTs built from the top six eigenvectors of the second-order CGT, at a fraction of the cost. For quantities of interest, it derives selective sCGTs and quanti

Load-bearing premise

The load-bearing premise is that the dominant nonlinear stretching direction, computed offline from the nominal trajectory via a higher-order Cauchy-Green tensor eigenpair, remains the dominant direction across the entire flight and for the actual perturbation sizes; the paper only verifies this with very small perturbations and reports that the dominant eigenvalue switches modes near peak dynamic pressure.

Editorial extensions

If this is right

  • DSTTs with a single latent dimension reduce perturbation propagation to matrix-vector multiplication rather than tensor contraction, making higher-order nonlinear propagation computationally cheap enough for onboard use.
  • For aerocapture, directionalizing along third- and fourth-order HOCGT directions preserves more of the original STT than adding additional second-order linear directions, so accuracy per integrated equation is improved.
  • Quantity-of-interest DSTTs allow specific performance metrics such as terminal energy to be propagated with substantially lower error than generic DSTTs or linear covariance methods.
  • When a quantity of interest has discontinuous or poorly conditioned partials, as with apoapsis radius, building the DSTT from a selective sCGT over the constituent states is more accurate than a qCGT built from the function itself.
  • The augmented-HOCGT construction is dynamics-general, so the same basis-selection method can be applied to other nonconservative, nonlinear flight regimes beyond aerocapture.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper validates with very small perturbations, near machine precision for third-order terms, so the dominance of the chosen HOCGT direction at realistic aerocapture dispersions remains untested; a perturbation-magnitude sweep would determine whether the single-direction assumption holds.
  • Because the dominant third-order HOCGT eigenvalue switches between two modes around 200 and 400 seconds, a piecewise basis or a small multi-direction basis may be needed for robust performance; the paper notes the switch but does not explore such adaptations.
  • The same tensor-eigenpair reasoning could be used to choose directions for uncertainty propagation in other phases of planetary flight, such as entry or powered descent, where nonconservative forces also dominate.
  • The qCGT failure for apoapsis suggests that checking the conditioning of the quantity-of-interest partials before constructing a qCGT could serve as a practical screening criterion.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper extends directional state transition tensors (DSTTs) to aerocapture by proposing new ways to choose the reduced-dimension basis: directions derived from tensor eigenpairs of higher-order Cauchy-Green tensors (HOCGTs), and two new augmented constructions—selective CGTs (sCGTs) for chosen state subsets and quantity-of-interest CGTs (qCGTs) for scalar outputs such as specific energy and apoapsis radius. The authors first show, via a decomposed CGT analysis, that the maximal linear-stretching direction is time-varying for aerocapture because of aerodynamic forces. They then construct DSTTs using maximal HOCGT/sCGT/qCGT eigendirections and compare against traditional DSTTs built from second-order CGT eigenvectors in a Uranus aerocapture application. The central claims are that the new DSTTs reduce computational cost while maintaining or improving accuracy in nonlinear perturbation propagation, and that the second-order CGT direction is inappropriate for higher-order DSTT directionalization in aerocapture.

Significance. The theoretical framework is a plausible and useful extension of the Jenson-Scheeres tensor-eigenpair results and Boone-McMahon DSTT theory. The decomposed CGT analysis in Section IV.A gives insight into why the linear stretching direction changes during aerocapture, and the sCGT/qCGT constructions are natural generalizations that other researchers may adopt. The paper is transparent about the SS-HOPM caveats and honestly reports the negative result for the apoapsis-radius qCGT. However, the quantitative support for the headline claim is weak: the Monte Carlo tests are run at perturbation scales far below realistic aerocapture dispersions, and the robustness of the precomputed single-direction basis under realistic perturbations is not tested. If the method is revalidated at appropriate perturbation magnitudes, the contribution would be significant for onboard nonlinear uncertainty propagation.

major comments (3)
  1. [Section V.B, Table 2, Figs. 9–10] The validation is performed at perturbation scales that are not representative of aerocapture dispersions and that place the test in the linear regime. The text says the Monte Carlo distribution is '1e−14 in each coordinate in nondimensional units'; Table 2 lists sigma_r = 2.56 m, which is roughly 1e-7 if the Lu nondimensionalization uses a planetary-radius length scale. The two statements differ by seven orders of magnitude, and either value is far below realistic entry-corridor dispersions (flight-path angle errors of 0.01–0.1 deg correspond to ~2e-4–2e-3 rad). Since Fig. 9 shows second-order STT errors at machine precision for perturbation magnitude 1e-6, the MC results in Fig. 10 cannot discriminate between nonlinear propagation methods. The abstract and conclusions claim that HOCGT-directionalized DSTTs 'outperform traditional DSTTs in nonlinear perturbation propagation'; this claim
  2. [Section IV.B–IV.D, Figs. 4–5] The selection of a single maximal HOCGT eigendirection is not shown to be robust. Figures 4 and 5 show the dominant third-order HOCGT eigenvalue switching between two modes near 200 and 400 s, with corresponding discontinuities in the maximal eigenvector. The DSTTs in Section V use one direction per order computed offline from (t_f, t_0), and no experiment varies the initial perturbation magnitude or direction across the ensemble to assess whether the neglected orthogonal components eventually dominate the truncation error. Because the method's premise is that a single precomputed direction captures the dominant higher-order stretching, the lack of a robustness test is a gap between the analysis and the central accuracy claim. Add a perturbation-magnitude sweep and/or a comparison of single-direction versus multi-direction DSTTs at realistic dispersions.
  3. [Section V.A.1, Eq. (52)] The normalized Frobenius-norm error measures how well a DSTT reconstructs the very STT from which it was built (via Eqs. (16)–(17) and the HOCGTs in Eqs. (29)–(30)). It is therefore a projection-style approximation metric, not an independent accuracy measure; rankings based on it are partly self-referential. The paper should state this limitation explicitly and rely on the numerical-integration comparisons (Figs. 9–11) for external validation. As written, the text in Section V.A.1 draws strong conclusions from Eq. (52) alone, for example that using a higher-order direction 'improves DSTT approximation accuracy'—this is only a statement about reconstruction fidelity in the chosen norm, not a demonstration of propagation accuracy at realistic perturbation sizes.
minor comments (4)
  1. [Section IV.C.1, Eqs. (34)–(35) and Eq. (45)] Equations (34) and (35) contain repeated dummy indices that violate the Einstein summation convention as printed (S_i,j appears twice with the same j in one term), and Eq. (45) has a κ4λ4 index mismatch. If the intended contractions use distinct dummy indices, the derivation is sound; otherwise the definitions are incorrect. Please correct the index notation.
  2. [Section V.B, text near '1e−14'] The phrase 'chosen to be1e−14' is missing a space. More importantly, the relationship between the nondimensional sigma stated in the text and the dimensional sigma in Table 2 should be explained explicitly, since the current text appears inconsistent by a factor of about 1e7.
  3. [Section V.A.2, Fig. 9] The caption and text state that each initial perturbation is scaled to nondimensional magnitude 10^-6, but the horizontal axis description is slightly confusing ('per perturbation magnitude along the R direction'). Clarify that the magnitude is the total perturbation norm and that the angle κ varies the direction.
  4. [Section VI / Abstract] The claim of 'significantly reduce computational cost' is not supported by any runtime or operation-count comparison. The reduction in the number of DSTT terms is clear from the formalism, but the offline cost of computing full STTs and SS-HOPM eigenpairs is not quantified. Please include at least a complexity table or representative runtime for the methods compared.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the improved DSTT directions are validated against independent numerical integration, and no prediction reduces by construction to a fitted input.

full rationale

The derivation chain is not circular in the sense defined here. The HOCGT/sCGT/qCGT directions are functions of the STTs (Eqs. 28-30, 33-35, 43-48), and the DSTTs are rotations/contractions of those same STTs (Eqs. 16-17); therefore the normalized Frobenius-norm comparison in Eq. 52 and Fig. 7 is a self-consistency/reconstruction check rather than an external prediction. The paper does not rest its central accuracy claim on that metric alone: perturbation propagation and QoI performance are checked against numerical integration of the full nonlinear dynamics (Figs. 6, 9-11), including 10,000-sample Monte Carlo propagation of terminal apoapsis radius and energy (Sec. V.B). These are external benchmarks not constructed from the DSTT basis, so the claim that HOCGT/qCGT-directionalized DSTTs outperform traditional DSTTs is not forced by construction. The self-citations [12,13] are used to motivate the premise that second-order CGT directions fail for aerocapture, but the paper independently re-establishes that premise with decomposed CGT eigenvector analysis and HOCGT eigenpair comparisons (Figs. 2-5, 7). The qCGT/DSTT energy result is a designed consequence of tailoring the basis to the quantity of interest, not a fitted parameter renamed as a prediction. The small perturbation scales used in the Monte Carlo (Sec. V.B) and Fig. 9 are a legitimate scope/robustness limitation for the nonlinear-regime claim, but that is an external-validity concern, not circularity.

Assumptions & free parameters 7 free parameters · 7 assumptions · 0 invented entities

The method itself introduces no fitted constants: the HOCGT eigenpairs are computed from the STTs, and the STTs are integrated from the stated dynamics. The free parameters that matter are inherited modeling inputs: the exponential-atmosphere constants (fitted to UranusGRAM in Ref. [21] with an incompletely specified 'density perturbation scale of two'), the ad hoc zeta_ref=20 normalization, the application-driven selection of states for sCGTs, the SS-HOPM restart hyperparameters, and the test perturbation scale (stated inconsistently). The r_a-qCGT further assumes that apoapsis radius is differentiable as a function of final state; the authors themselves show this assumption is violated by the hyperbolic-to-elliptic discontinuity in semi-major axis, which they give as the reason the r_a-qDSTT underperforms.

free parameters (7)
  • Atmosphere scale height H = 54.72 km
    Exponential atmosphere model rho = rho0 exp((h0-(r-Rp))/H). Constant fitted to UranusGRAM data (Ref. [21]); a central input to the lift/drag accelerations that drive the DSTT behavior.
  • Atmosphere reference density rho0 = 6.40e-3 kg/m^3
    Second fitted constant of the exponential atmosphere, paired with H.
  • Density perturbation scale = two (multiplicative, ambiguous)
    Section II: atmosphere constants fit to UranusGRAM 'with a density perturbation scale of two'; this scale parameter is not defined and its effect on the trajectory is not quantified.
  • ln(rho) normalization factor zeta_ref = 20
    Section II: ad hoc normalization of the density state to improve numerical conditioning of the variational equations.
  • sCGT selected states = [r, V, gamma]
    Section IV.C.1: selection of position, velocity, flight path angle for the selective HOCGT is application-driven, with no optimality criterion.
  • SS-HOPM restarts and dedup threshold = 100 initial guesses; cos(10^-3 rad)
    Section IV.B: hyperparameters chosen to mitigate the lack of global convergence guarantee in SS-HOPM.
  • MC test perturbation sigma = 1e-14 nondim (text) / ~1e-7 nondim (Table 2)
    Section V.B: test dispersion size controls whether nonlinear terms matter; text and table disagree by ~7 orders of magnitude.
assumptions (7)
  • domain assumption Exponential atmosphere model rho = rho0 exp((h0 - (r - Rp))/H) with constant H=54.72 km and rho0=6.40e-3 kg/m^3 describes the Uranian atmosphere over the entry corridor.
    Section II; the zeta=ln(rho) state and the aerodynamic accelerations L and D are computed from this model, so the entire STT/DSTT analysis inherits it.
  • domain assumption 3DOF point-mass dynamics with J2 gravity, spherical planet, and constant rotation rate Omega adequately represent aerocapture.
    Section II, Eqs. (1)-(9); excludes atmosphere winds and higher-order gravity terms.
  • domain assumption The Taylor expansion of the solution flow converges and third-order STT truncation captures the nonlinear perturbation propagation of interest.
    Section III; used in every STT/DSTT propagation (Eqs. 15, 19). Only validated at near-linear perturbation scales in this paper (Section V.B).
  • standard math SS-HOPM with 100 restarts and cos(10^-3 rad) dedup finds the maximum z-eigenpair of the (augmented) HOCGT.
    Section IV.B; SS-HOPM (Ref. [24]) only guarantees an eigenvalue; the paper checks the winner against six orthogonal directions (Fig. 6), not a global certificate.
  • domain assumption The quantity-of-interest function q(x_z) has well-defined partial derivatives up to third order on the intervals used.
    Eqs. (36)-(48); the authors state this fails for apoapsis radius because semi-major axis is discontinuous at the hyperbolic-to-elliptic transition (Section V.B.1), which is their explanation for the r_a-qDSTT failure.
  • domain assumption The dominant nonlinear-stretching direction computed from the nominal trajectory is stable across the initial-perturbation ensemble.
    Needed for offline DSTT construction (Sections IV.D, V.A); Fig. 5 shows mode-switching, and no sensitivity to perturbation magnitude is reported.
  • standard math STTs exhibit the flow composition property, so a set of STTs over [t0, tf] maps perturbations between any sub-intervals.
    Section III; used to construct per-interval DSTTs from single-trajectory integrations; basis of Refs. [2,3].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improved Directional State Transition Tensors for Accurate Aerocapture Performance Analysis." pith.science (2026). https://pith.science/paper/EH27W3CF

@misc{pith2026251212475,
  author       = {Pith},
  title        = {Pith review of: Improved Directional State Transition Tensors for Accurate Aerocapture Performance Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EH27W3CF}},
  note         = {Machine review of arXiv:2512.12475}
}
read the original abstract

Aerocapture is particularly challenging for semi-analytical propagation because the dynamics are dominated by nonconservative forces whose magnitudes vary significantly throughout the trajectory. State transition tensors (STTs), higher-order Taylor series expansions of the solution flow, have been widely used as a computationally efficient semi-analytical propagation method for orbital scenarios, but have not previously been applied to aerocapture. However, computing higher-order STTs requires integrating exponentially many equations as the state dimension increases. Directional state transition tensors (DSTTs) mitigate this cost by projecting the state into a reduced-dimension basis. This work develops novel dynamics analysis techniques to identify effective bases for this reduction, including augmented higher-order Cauchy Green tensors tailored to quantities of interest such as apoapsis radius. Results show that DSTTs constructed along these bases significantly reduce computational cost while maintaining accuracy in predicted apoapsis radius and terminal energy. In particular, certain of these DSTTs outperform traditional DSTTs in nonlinear perturbation propagation for key state subsets and quantities of interest. These results establish STTs and DSTTs as practical tools for aerocapture performance analysis to enable robust guidance and navigation.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 5 canonical work pages

  1. [1]

    Uranus Flagship-class Orbiter and Probe Using Aerocapture,

    Dutta, S., Shellabarger, E., Scoggins, J. B., Gomez-Delrio, A., Lugo, R., Deshmukh, R., Tackett, B., Williams, J., Johnson, B., Matz, D., Geiser, J., Morgan, J., Restrepo, R., and Mages, D., “Uranus Flagship-class Orbiter and Probe Using Aerocapture,” Proceedings of the AIAA SCITECH 2024 Forum, AIAA, Orlando, FL, 2024. https://doi.org/10.2514/6.2024-0714

  2. [2]

    Nonlinear Semi-Analytic Methods for Trajectory Estimation,

    Park, R. S., and Scheeres, D. J., “Nonlinear Semi-Analytic Methods for Trajectory Estimation,”Journal of Guidance, Control, and Dynamics, Vol. 30, No. 6, 2007, pp. 1668–1676. https://doi.org/10.2514/1.29106

  3. [3]

    Nonlinear Mapping of Gaussian Statistics: Theory and Applications to Spacecraft Trajectory Design,

    Park, R. S., and Scheeres, D. J., “Nonlinear Mapping of Gaussian Statistics: Theory and Applications to Spacecraft Trajectory Design,”Journal of Guidance, Control, and Dynamics, Vol. 29, No. 6, 2006, pp. 1367–1375. https://doi.org/10.2514/1.20177

  4. [4]

    Directional State Transition Tensors for Capturing Dominant Nonlinear Effects in Orbital Dynamics,

    Boone, S., and McMahon, J., “Directional State Transition Tensors for Capturing Dominant Nonlinear Effects in Orbital Dynamics,”JournalofGuidance,Control,andDynamics,Vol.46,No.3,2023,pp.431–442. https://doi.org/10.2514/1.G006910

  5. [5]

    Linear Covariance Analysis of Entry and Aerocapture Trajectories in an Uncertain Atmosphere,

    Ridderhof, J., Albert, S. W., Tsiotras, P., and Schaub, H., “Linear Covariance Analysis of Entry and Aerocapture Trajectories in an Uncertain Atmosphere,”Proceedings of the AIAA SCITECH 2022 Forum, AIAA, San Diego, CA & Virtual, 2022. https://doi.org/10.2514/6.2022-1216

  6. [6]

    End-to-End Mars Aerocapture Analysis Using Linear Covariance Techniques 27 and Robust Trajectory Optimization,

    Joshi, J., Woffinden, D., and Putnam, Z. R., “End-to-End Mars Aerocapture Analysis Using Linear Covariance Techniques 27 and Robust Trajectory Optimization,”Proceedings of the 2022 AAS/AIAA Astrodynamics Specialists Conference, AIAA/AAS, Charlotte, NC, 2022. AAS 22-678

  7. [7]

    Two-Stage Polynomial Chaos Expansion: An Extension Of Uncertainty Quantification TechniquesForMulti-ModalDistributionsinAerocapture,

    Grace, M. J., and McMahon, J. W., “Two-Stage Polynomial Chaos Expansion: An Extension Of Uncertainty Quantification TechniquesForMulti-ModalDistributionsinAerocapture,”ProceedingsoftheAIAASCITECH2022Forum, AIAA,SanDiego, CA & Virtual, 2022. https://doi.org/10.2514/6.2022-1769

  8. [8]

    Finite-Dimensional Density Representation for Aerocapture Uncertainty Quantification,

    Albert, S. W., Doostan, A., and Schaub, H., “Finite-Dimensional Density Representation for Aerocapture Uncertainty Quantification,”Proceedings of the AIAA SCITECH 2021 Forum, AIAA, Virtual, 2021. https://doi.org/10.2514/6.2021-0932

Show all 26 references
  1. [9]

    ApplicationOfGaussianMixtureModelsForNonlinearUncertaintyPropagationDuringMartian Aerocapture,

    Grace,M.,andMcMahon,J.,“ApplicationOfGaussianMixtureModelsForNonlinearUncertaintyPropagationDuringMartian Aerocapture,”Proceedings of the 2020 AAS/AIAA Astrodynamics Specialist Conference, AAS, Virtual, 2020. AAS 20-485

  2. [10]

    Applications of Induced Tensor Norms to Guidance Navigation and Control,

    Kulik, J., Ruth, M., Orton-Urbina, C., and Savransky, D., “Applications of Induced Tensor Norms to Guidance Navigation and Control,”Journal of Guidance, Control, and Dynamics, Vol. 48, No. 10, 2025, pp. 2180–2198. https://doi.org/10.2514/1. G009054

  3. [11]

    Nonlinearity and Uncertainty Informed Moment-Matching Gaussian Mixture Splitting,

    Kulik, J., and LeGrand, K. A., “Nonlinearity and Uncertainty Informed Moment-Matching Gaussian Mixture Splitting,”arXiv preprint, 2024. https://doi.org/10.48550/arXiv.2412.00343

  4. [12]

    Efficient Higher-Order Analytical Covariance Analysis For Aerocapture,

    Calkins, G. E., Woffinden, D., and McMahon, J., “Efficient Higher-Order Analytical Covariance Analysis For Aerocapture,” Proceedings of the 2024 AAS/AIAA Astrodynamics Specialists Conference, AIAA/AAS, Broomfield, CO, 2024. AAS 24-241

  5. [13]

    Dynamics-Informed Higher-Order Uncertainty Quantification for Aerocapture,

    Calkins, G. E., McMahon, J. W., and Woffinden, D. C., “Dynamics-Informed Higher-Order Uncertainty Quantification for Aerocapture,”Proceedings of the 2025 AIAA/AAS Spaceflight Mechanics Meeting, AIAA/AAS, Kaua’i, HI, 2025. AAS 25-317

  6. [14]

    Efficient Nonlinear Spacecraft Navigation using Directional State Transition Tensors,

    Boodram, O., Boone, S., and McMahon, J., “Efficient Nonlinear Spacecraft Navigation using Directional State Transition Tensors,”Proceedings of the 2022 AAS/AIAA Astrodynamics Specialists Conference, AIAA/AAS, Charlotte, NC, 2022. AAS 22-670

  7. [15]

    Time-Varying Directional State Transition Tensor for Orbit Uncertainty Propagation,

    Zhou, X., Armellin, R., Qiao, D., and Li, X., “Time-Varying Directional State Transition Tensor for Orbit Uncertainty Propagation,”arXiv preprint, 2024. https://doi.org/10.48550/arXiv.2412.07060

  8. [16]

    Efficient Orbit Determination Using Measurement-Directional State Transition Tensor,

    Zhou, X., Qiao, D., Malcolm, M., and Li, X., “Efficient Orbit Determination Using Measurement-Directional State Transition Tensor,”IEEE Transactions on Aerospace and Electronic Systems, Vol. 61, No. 3, 2025, pp. 8014–8027. https://doi.org/10.1109/TAES.2025.3527410

  9. [17]

    SemianalyticalMeasuresofNonlinearityBasedonTensorEigenpairs,

    Jenson, E.L., andScheeres, D.J., “SemianalyticalMeasuresofNonlinearityBasedonTensorEigenpairs,”JournalofGuidance, Control, and Dynamics, Vol. 46, No. 4, 2023, pp. 638–653. https://doi.org/10.2514/1.G006760

  10. [18]

    X., Busemann, A., and Culp, R.,Hypersonic and planetary entry flight mechanics, The University of Michigan Press, Ann Arbor, MI, 1980

    Vihn, N. X., Busemann, A., and Culp, R.,Hypersonic and planetary entry flight mechanics, The University of Michigan Press, Ann Arbor, MI, 1980. 28

  11. [19]

    Optimal Aerocapture Guidance,

    Lu, P., Cerimele, C. J., Tigges, M. A., and Matz, D. A., “Optimal Aerocapture Guidance,”Journal of Guidance, Control, and Dynamics, Vol. 38, No. 4, 2015, pp. 553–565. https://doi.org/10.2514/1.G000713

  12. [20]

    Performance Analysis of Aerocapture Systems for Uranus Orbiter,

    Deshmukh, R., Dutta, S., Lugo, R., Restrepo, R., Mages, D., Johnson, B., Matz, D., Geiser, J., Scoggins, J. B., Shellabarger, E., Gomez-Delrio, A., and Williams, J., “Performance Analysis of Aerocapture Systems for Uranus Orbiter,”Proceedings of the AIAA SCITECH 2024 Forum, AI...

  13. [21]

    UranusGlobalReferenceAtmosphericModel(Uranus-GRAM): User Guide,

    Justh,H.L.,Cianciolo,A.D.,Hoffman,J.,andAllen,G.A.,Jr.,“UranusGlobalReferenceAtmosphericModel(Uranus-GRAM): User Guide,” NASA TM-2010017250, Jun. 2021

  14. [22]

    Entry Guidance: A Unified Method,

    Lu, P., “Entry Guidance: A Unified Method,”Journal of Guidance, Control, and Dynamics, Vol. 37, No. 3, 2014, pp. 713–728

  15. [23]

    Bounding nonlinear stretching about spacecraft trajectories using tensor eigenpairs,

    Jenson, E. L., and Scheeres, D. J., “Bounding nonlinear stretching about spacecraft trajectories using tensor eigenpairs,”Acta Astronautica, Vol. 214, 2024, pp. 159–166. https://doi.org/10.1016/j.actaastro.2023.10.013

  16. [24]

    Shifted Power Method for Computing Tensor Eigenpairs,

    Kolda, T. G., and Mayo, J. R., “Shifted Power Method for Computing Tensor Eigenpairs,”SIAM Journal on Matrix Analysis and Applications, Vol. 32, No. 4, 2011, pp. 1095–1124. https://doi.org/10.1137/100801482

  17. [25]

    Algorithm 862: MATLAB tensor classes for fast algorithm prototyping,

    Bader, B. W., and Kolda, T. G., “Algorithm 862: MATLAB tensor classes for fast algorithm prototyping,”ACM Transactions on Mathematical Software, Vol. 32, No. 4, 2006, pp. 635–653. https://doi.org/10.1145/1186785.1186794

  18. [26]

    Tensor Toolbox for MATLAB,

    Bader, B. W., and Kolda, T. G., “Tensor Toolbox for MATLAB,” https://www.tensortoolbox.org, Oct. 2025. Version 3.7. 29

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.