Pith. sign in

REVIEW 3 major objections 3 minor 27 references

Female and Combined Male-Female Injury Risk Functions for the Anterior Pelvis Under Frontal Lap Belt Loading Conditions

T0 review · 3 major / 3 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper reports the first female iliac wing fracture tolerance measurements under frontal lap belt loading and derives injury risk functions in which age, not sex, is the significant predictor.

desk verdict First female iliac wing fracture data under lap belt loading, with a transparent but uncompensated clustering problem; the dataset is worth having, but the sex and age conclusions need a cluster-level reanalysis. read the letter →

arxiv 2501.08911 v1 pith:GZZXELNG submitted 2025-01-15 q-bio.TO physics.bio-phphysics.med-ph

Connor Hanggi , Joon Seok Kong , James Caldwell , Bronislaw Gepner , Martin Östling , Jason R. Kerrigan This is my paper · ORCID
classification q-bio.TOphysics.bio-phphysics.med-ph MSC 62N0162P10
keywords SurvivalanalysisPMHStestinginjuryriskfunctionfracturetolerancefemalepelvisiliacwinglapbeltloadingWeibullmodel
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 to provide the first direct measurements of female iliac wing fracture tolerance under frontal lap belt loading, along with injury risk functions built from those measurements. Twenty female iliac wings from ten pelvises were tested in the same component-level belt-loading setup previously used for male wings; fourteen fractured with known forces and six were right-censored. Fitting Weibull survival models to the female data and to combined male-female data, the authors find that age is the only statistically significant covariate, while sex, belt position, height, weight, and pelvis size are not. If these results hold, restraint designers can use the new risk functions to estimate female pelvis fracture risk from measured belt forces, and the absence of a sex effect suggests a single age-based risk curve may serve both sexes.

What carries the argument

The load-bearing object is the parametric Weibull survival model, which treats the force at fracture as a time-to-event outcome and incorporates right-censored observations (tests where fracture did not occur or occurred only after boundary conditions changed). The model yields injury risk functions of the form $1 - e^{-(F \cdot e^{-\theta})^{k}}$ with shape $k$ and intercept $\theta$, optionally shifted by covariates, and models are compared by AICc. This machinery is what lets the authors pool exact and censored failures from only 20 female wings and then combine them with 22 male wings to test sex.

What would settle it

Re-fit the female and combined Weibull survival models with a shared frailty or cluster-robust variance that accounts for left-right correlation within each pelvis; if the age covariate's p-value rises above 0.05 or the sex covariate becomes statistically significant in the combined data, the paper's central conclusions would not survive.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that female iliac wing fracture tolerance under frontal lap belt loading can be measured at the component level and follows the same Weibull survival form used for male data. The female base model is $\Pr(\text{Fracture} \mid \text{Force}) = 1 - \exp\left(-(\text{Force} \cdot e^{-8.762})^{1.754}\right)$, and in the combined male-female data age is the only significant covariate ($p = 0.003$), whereas sex is not ($p = 0.567$). The authors interpret this as evidence that iliac wing fracture tolerance does not differ between sexes once force is accounted for, and that age-related bone quality, not body size or belt position, dominates the variation in fracture force.

Load-bearing premise

The paper's statistical conclusions assume that all twenty iliac wing observations are statistically independent, even though pairs of wings come from the same ten pelvises and left-right fracture forces are correlated, leaving the effective sample size somewhere between ten and twenty.

Editorial extensions

If this is right

  • The female risk function can be used to convert lap-belt-relevant forces measured in anthropomorphic test devices or human body models into a predicted probability of iliac wing fracture for female occupants.
  • Because sex was not a significant covariate in the combined data, restraint optimization may not need separate male and female pelvis risk curves, though age should be included.
  • Age's significance in female and combined data but not clearly in the male data suggests that collecting younger male specimens could sharpen the age effect.
  • The non-significance of belt position means a single risk function may apply across ASIS-centered and slightly higher belt placements.

Reading between the lines

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

  • A clustered Weibull model that treats left and right wings from the same pelvis as correlated might change the p-values; the paper's own limitation statement acknowledges the effective sample size is between 10 and 20, so the age and sex conclusions are less secure than the nominal p-values suggest.
  • If local bone microstructure (cortical thickness and trabecular density near the ASIS/AIIS) is collected as planned, it may replace age as the dominant covariate, since age is likely a proxy for bone quality.
  • The lack of a sex effect at the component level does not contradict field findings that women have higher pelvis injury risk; exposure, body size, and belt fit may still drive the difference.
  • Future restraint designs for reclined automated-driving seating, where lap belt loads are higher, could use these curves to set force limits that protect older female occupants.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. This manuscript reports the first experimental data on female iliac wing fracture tolerance under frontal lap belt loading. Twenty female iliac wings from ten pelvises were loaded with belt webbing; fourteen fractured at known forces (exact observations) and six were right-censored. Weibull survival models were fit to the female data and to combined male-female data from Moreau et al., with univariate covariates (age, sex, belt position, belt angle, anthropometry). The main quantitative results are the risk functions in Table 3, including the female base model Pr(Fracture|Force) = 1 - exp(-(Force*exp(-8.762))^1.754), and the claim that age is the only statistically significant covariate (female p=0.007, combined p=0.003) while sex is not (p=0.567).

Significance. If the statistical concerns are resolved, the data and risk functions are a significant contribution: they are, to the authors' knowledge, the first female component-level iliac wing tolerance data under belt loading, they replicate a published male test protocol, and they provide empirically fitted Weibull risk functions with explicit handling of censored observations. The paper also makes a falsifiable prediction (the risk curves in Figure 3 and Table 3) that can be used in HBMs and ATDs. The main inferential conclusions, however—that age matters and sex does not—rest on an independence assumption that the authors themselves flag as violated.

major comments (3)
  1. [Limitations (p. 14-15), Table 3, Table A1] The manuscript states that the effective sample size is between ten and twenty because of left-right correlation, yet all Weibull models assume mutual independence. This assumption directly affects every p-value in Table A1 and every AICc comparison in Table 3. For example, the combined age CI [-0.032, -0.007] and the sex CI [-0.540, 0.296] in Table 3 are computed under n=42 independent observations, whereas sex and age are pelvic-level covariates with only 21 independent pelvises (10 female, 11 male). Positive within-pelvis correlation, which Figure A6 suggests, would make these intervals wider and could change the significance of age and the model-selection results. Please reanalyze with a cluster-robust variance estimator, a shared frailty model, or a cluster bootstrap, and report the resulting p-values, CIs, and AICc (or an equivalent small-sample criterion computed at the pelvis level).
  2. [Results, Table A1; Discussion] The conclusion that age is 'the only covariate' and that sex is not significant is based on a series of univariate models, with no multivariable model that includes age and sex together. Since female and male samples differ in age range (23-84 vs 50-77 years), the univariate sex effect may be confounded by age (and vice versa). Please fit a combined model with age and sex (and possibly their interaction) to separate the two effects, or clearly state that the data cannot distinguish them.
  3. [Table 3] The absence of a statistically significant sex effect (p=0.567) is a low-power null result, not evidence of equivalence: the sex coefficient CI [-0.540, 0.296] in the combined model spans a wide range that includes clinically meaningful differences in fracture force (roughly a factor of exp(0.54)=1.7 on the rate scale). With only 10 female and 11 male pelvises, the study is underpowered to detect all but very large sex effects. Please report a power or precision analysis and phrase the conclusion as 'no sex effect was detected in this sample' rather than 'sex is not beneficial' as a covariate.
minor comments (3)
  1. [Methods, Data Analysis] The phrase 'double censored data' is imprecise because the data include exact and right-censored observations only; 'right- and exact-censored' or 'doubly censored' would be clearer if both types are intended.
  2. [References] Reference [16] appears to contain a typo: 'Biochemical Data' should likely read 'Biomechanical Data'.
  3. [Results, Survival Models] The manuscript does not report any check of the Weibull proportional-hazards or shape assumption; a brief assessment (e.g., log-cumulative hazard plots or a likelihood-ratio test against a more flexible model) would strengthen the choice of distribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the risk functions are openly fitted to measured fracture data, and the cited male dataset and methods are used as external inputs, not as assumptions that contain the female or combined results.

full rationale

The paper's derivation chain is empirical rather than definitional. Twenty female iliac wings were tested, forces at fracture were recorded as exact or right-censored observations, and Weibull survival models were fit to those data; the Methods state that 'Injury risk functions were created by fitting Weibull survival models to data that integrated censored and exact failure observations.' The resulting Table 3 risk functions are therefore fitted outputs, not predictions derived from assumptions that already contain the answer. The age and sex coefficients are regression estimates from observed data, and neither covariate is defined in terms of fracture force, so there is no self-definitional loop. The paper's reliance on Moreau et al. [14] supplies the male comparison dataset, the experimental fixture concept, and the choice of Weibull survival modeling; that is external data and method transfer, not a self-citation chain that forces the female or combined conclusions. No uniqueness theorem is invoked to forbid alternative models, and no ansatz is smuggled in via citation to produce the risk curves. The paper's own limitation—that 'the effective sample size of the study is somewhere between ten and twenty due to the correlation seen between the left and right sides of the same pelvis' and that the models 'assume all observations are mutually independent'—is a substantive statistical validity concern that could affect p-values, confidence intervals, and AICc comparisons, but it is not circularity. The central claims are openly fitted to the collected data and are falsifiable against the reported observations; therefore the circularity score is 0.

Assumptions & free parameters 16 free parameters · 5 assumptions · 0 invented entities

The models are empirical fits; every number in Table 3 is a fitted parameter. The main unfunded assumptions are statistical independence, non-informative censoring, and transferability of the component-level force to real belt loads. No new physical entities are introduced.

free parameters (16)
  • Female base model Weibull shape = 1.754
    Fitted to female fracture data (n=20), Table 3.
  • Female base model Weibull intercept = 8.762
    Fitted to female fracture data (n=20), Table 3.
  • Female position model Weibull shape = 1.743
    Fitted to female fracture data (n=20), Table 3.
  • Female position model Weibull intercept = 8.676
    Fitted to female fracture data (n=20), Table 3.
  • Female position covariate coefficient = 0.216, not significant
    Fitted to female fracture data (n=20), Table 3.
  • Female age model Weibull shape = 2.121
    Fitted to female fracture data (n=20), Table 3.
  • Female age model Weibull intercept = 9.577
    Fitted to female fracture data (n=20), Table 3.
  • Female age covariate coefficient = -0.017 per year, significant
    Fitted to female fracture data (n=20), Table 3.
  • Combined base model Weibull shape = 1.682
    Fitted to combined male-female data (n=42), Table 3.
  • Combined base model Weibull intercept = 8.698
    Fitted to combined male-female data (n=42), Table 3.
  • Combined sex model Weibull shape = 1.674
    Fitted to combined male-female data (n=42), Table 3.
  • Combined sex model Weibull intercept = 8.764
    Fitted to combined male-female data (n=42), Table 3.
  • Combined sex covariate coefficient = -0.122, not significant
    Fitted to combined male-female data (n=42), Table 3.
  • Combined age model Weibull shape = 1.877
    Fitted to combined male-female data (n=42), Table 3.
  • Combined age model Weibull intercept = 9.810
    Fitted to combined male-female data (n=42), Table 3.
  • Combined age covariate coefficient = -0.020 per year, significant
    Fitted to combined male-female data (n=42), Table 3.
assumptions (5)
  • domain assumption Weibull survival model correctly characterizes force-at-fracture distribution with right censoring.
    The parametric form is assumed following Moreau et al. and McMurry and Poplin, not derived from mechanism.
  • domain assumption Censoring is non-informative; belt webbing tears and displacement limits are unrelated to latent fracture force.
    Four of six censored observations were due to belt tears, which may depend on ASIS geometry, so this assumption is load-bearing.
  • domain assumption Left and right iliac wings are independent observations.
    The paper acknowledges this is likely false and may bias the models; all p-values inherit this assumption.
  • domain assumption Measured loadcell force equals the force applied to the iliac wing at fracture under the simplified belt loading.
    The component test isolates belt force; conversion to real belt force relies on the transfer function in Richardson et al. [15].
  • domain assumption Specimen sex and age records are accurate and the female sample represents the female population.
    Ten pelvises with ages 23 to 84 years form a small, possibly age-skewed sample.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Female and Combined Male-Female Injury Risk Functions for the Anterior Pelvis Under Frontal Lap Belt Loading Conditions." pith.science (2026). https://pith.science/paper/GZZXELNG

@misc{pith2026250108911,
  author       = {Pith},
  title        = {Pith review of: Female and Combined Male-Female Injury Risk Functions for the Anterior Pelvis Under Frontal Lap Belt Loading Conditions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GZZXELNG}},
  note         = {Machine review of arXiv:2501.08911}
}
read the original abstract

Purpose: Iliac wing fractures due to lap belt loading have been observed in laboratory settings for 50 years and recent data suggest they are also occurring in the field. Automated driving systems (ADS) and other occupant compartment advancements are expected to offer enhanced flexibility in seating orientation, which could place a greater reliance on the seatbelt to restrain occupants. Such changes may increase seatbelt loads and create new challenges in successfully restraining occupants and mitigating injury to areas such as the pelvis. Injury criteria exist for component-level male iliac wing fractures resulting from frontal lap belt loading, but not for females. Methods: This study explored female iliac wing fracture tolerance in the same loading environment as a previous study that explored the fracture tolerance of isolated male iliac wings. Male and female fracture data were combined to evaluate the effect of sex. Injury risk functions were created by fitting Weibull survival models to data that integrated censored and exact failure observations. Results: Twenty female iliac wings were tested; fourteen of them sustained fracture with known failure forces (exact), but the remaining six wings either (1) did not fracture, or (2) fractured after an event that changed the boundary conditions (right censored). The fracture tolerance of the tested specimens ranged widely (1134 - 8759 N) and averaged 4240 N (SD 2516 N). Conclusion: Female data and combined male-female data were analyzed. Age was the only covariate investigated in this study that had a statistically significant effect and improved the predictive performance of the models.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 22 canonical work pages

  1. [1]

    Seating positions and activities in highly automated cars – a qualitative study of future automated driving scenarios,

    S. Jorlöv, K. Bohman, and A. Larsson, “Seating positions and activities in highly automated cars – a qualitative study of future automated driving scenarios,” presented at the IRCOBI, Antwerp, Belgium, 2017

    work page 2017

  2. [3]

    Occupant activities and sitting positions in automated vehicles in China and Sweden,

    M. Östling and A. Larsson, “Occupant activities and sitting positions in automated vehicles in China and Sweden,” presented at the 26th International Technical Conference on the Enhanced Safety of Vehicles (ESV), Eindhoven, Netherlands., 2019

    work page 2019

  3. [4]

    Effect of legroom proportions and individual factors for sitting with crossed legs: implications on the interior design of automated driving vehicles,

    M. J. Stanglmeier, F. Schulte, G. Schauberger, R. J. Bichler, A. Schwirtz, and F. K. Paternoster, “Effect of legroom proportions and individual factors for sitting with crossed legs: implications on the interior design of automated driving vehicles,” Ergonomics, vol. 64, no. 11, pp. 1393–1404, Nov. 2021, doi: 10.1080/00140139.2021.1933201

    arXiv 2021

  4. [5]

    Challenges for Occupant Safety in Highly Automated Vehicles across Various Anthropometries,

    B. Gepner et al., “Challenges for Occupant Safety in Highly Automated Vehicles across Various Anthropometries,” presented at the 26th International Technical Conference on the Enhanced Safety of Vehicles (ESV), Eindhoven, Netherlands., 2019

    work page 2019

  5. [6]

    Pelvis kinematics and injuries of reclined occupants in frontal impacts,

    R. Richardson et al., “Pelvis kinematics and injuries of reclined occupants in frontal impacts,” presented at the IRCOBI, Munich, Germany, 2020

    work page 2020

  6. [7]

    Comparison of Pelvis Kinematics and Injuries in Reclined Frontal Impact Sled Tests Between Mid-Size Adult Male and Female PMHS,

    J. Shin, “Comparison of Pelvis Kinematics and Injuries in Reclined Frontal Impact Sled Tests Between Mid-Size Adult Male and Female PMHS,” presented at the International Research Council on the Biomechanics of Injury (IRCOBI), Cambridge, UK, 2023

    work page 2023

  7. [8]

    Kinematics and Dynamics of the Pelvis in the Process of Submarining using PMHS Sled Tests,

    C. Luet, X. Trosseille, P. Drazétic, P . Potier, and G. Vallancien, “Kinematics and Dynamics of the Pelvis in the Process of Submarining using PMHS Sled Tests,” presented at the 56th Stapp Car Crash Conference, Oct. 2012, pp. 2012-22–0011. doi: 10.4271/2012-22-0011

    work page doi:10.4271/2012-22-0011 2012

  8. [9]

    Reference PMHS Sled Tests to Assess Submarining,

    J. Uriot et al., “Reference PMHS Sled Tests to Assess Submarining,” presented at the 59th Stapp Car Crash Conference, Nov. 2015, pp. 2015-22–0008. doi: 10.4271/2015-22-0008

    work page doi:10.4271/2015-22-0008 2015

Show all 27 references
  1. [10]

    Reference PMHS Sled Tests to Assess Submarining of the Small Female,

    X. Trosseille et al., “Reference PMHS Sled Tests to Assess Submarining of the Small Female,” presented at the 62nd Stapp Car Crash Conference, Nov. 2018, pp. 2018-22–0003. doi: 10.4271/2018-22-0003

    2018 doi

  2. [11]

    Thorax of 3-Point Belt Wearers During a Crash (Experiments with Cadavers),

    A. Fayon, C. Tarriere, G. Walfisch, C. Got, and A. Patel, “Thorax of 3-Point Belt Wearers During a Crash (Experiments with Cadavers),” presented at the 19th Stapp Car Crash Conference (1975), Feb. 1975, p. 751148. doi: 10.4271/751148

    1975 doi

  3. [12]

    Results of 49 Cadaver Tests Simulating Frontal Collision of Front Seat Passengers,

    G. Schmidt, D. Kallieris, J. Barz, and R. Mattern, “Results of 49 Cadaver Tests Simulating Frontal Collision of Front Seat Passengers,” presented at the 18th Stapp Car Crash Conference (1974), Feb. 1974, p. 741182. doi: 10.4271/741182

    1974 doi

  4. [13]

    A Methodology to Replicate Lap Belt Loading Conditions from a Sled Impact Test in a Non-Impact Dynamic Environment on Whole-Body Postmortem Human Subjects,

    D. Moreau et al., “A Methodology to Replicate Lap Belt Loading Conditions from a Sled Impact Test in a Non-Impact Dynamic Environment on Whole-Body Postmortem Human Subjects,” presented at the IRCOBI, Online, 2021. 18

    2021

  5. [14]

    Development of an Injury Risk Function for the Anterior Pelvis Under Frontal Lap Belt Loading Conditions,

    D. Moreau et al., “Development of an Injury Risk Function for the Anterior Pelvis Under Frontal Lap Belt Loading Conditions,” Ann. Biomed. Eng., vol. 51, no. 9, pp. 1942–1949, Sep. 2023, doi: 10.1007/s10439-023-03244-8

    1942 doi

  6. [15]

    Evaluation of lap belt-pelvis load transfer in frontal impact simulations using finite element occupant models,

    R. E. Richardson, B. Gepner, J. R. Kerrigan, and J. L. Forman, “Evaluation of lap belt-pelvis load transfer in frontal impact simulations using finite element occupant models,” Traffic Inj. Prev., pp. 1–9, Sep. 2024, doi: 10.1080/15389588.2024.2381084

    2024

  7. [16]

    Data Censoring and Parametric Distribution Assignment in the Development of Injury Risk Functions from Biochemical Data,

    R. W. Kent and J. R. Funk, “Data Censoring and Parametric Distribution Assignment in the Development of Injury Risk Functions from Biochemical Data,” presented at the SAE 2004 World Congress & Exhibition, Mar. 2004, pp. 2004-01–

    2004

  8. [17]

    Statistical Considerations in the Development of Injury Risk Functions,

    T. L. McMurry and G. S. Poplin, “Statistical Considerations in the Development of Injury Risk Functions,” Traffic Inj. Prev., vol. 16, no. 6, pp. 618–626, Aug. 2015, doi: 10.1080/15389588.2014.991820

    2015

  9. [18]

    Statistical Simulations to Evaluate the Methods of the Construction of Injury Risk Curves,

    A. Petitjean and X. Trosseille, “Statistical Simulations to Evaluate the Methods of the Construction of Injury Risk Curves,” presented at the 55th Stapp Car Crash Conference, Nov. 2011, pp. 2011-22–0015. doi: 10.4271/2011-22-0015

    2011 doi

  10. [19]

    Allometry and Sexual Dimorphism in the Human Pelvis,

    B. Fischer and P. Mitteroecker, “Allometry and Sexual Dimorphism in the Human Pelvis,” Anat. Rec., vol. 300, no. 4, pp. 698–705, Apr. 2017, doi: 10.1002/ar.23549

    2017 doi

  11. [20]

    Automobile injury trends in the contemporary fleet: Belted occupants in frontal collisions,

    J. Forman et al., “Automobile injury trends in the contemporary fleet: Belted occupants in frontal collisions,” Traffic Inj. Prev., vol. 20, no. 6, pp. 607–612, Aug. 2019, doi: 10.1080/15389588.2019.1630825

    2019

  12. [21]

    Vulnerability of Female Drivers Involved in Motor Vehicle Crashes: An Analysis of US Population at Risk,

    D. Bose, M. Segui-Gomez, ScD, and J. R. Crandall, “Vulnerability of Female Drivers Involved in Motor Vehicle Crashes: An Analysis of US Population at Risk,” Am. J. Public Health, vol. 101, no. 12, pp. 2368–2373, Dec. 2011, doi: 10.2105/AJPH.2011.300275

    2011

  13. [22]

    Injury vulnerability and effectiveness of occupant protection technologies for older occupants and women,

    C. J. Kahane, “Injury vulnerability and effectiveness of occupant protection technologies for older occupants and women,” National Highway Traffic Safety Administration, Washington, DC, DOT HS 811 766, 2013

    2013

  14. [23]

    Comparing the effects of age, BMI and gender on severe injury (AIS 3+) in motor-vehicle crashes,

    P. M. Carter, C. A. C. Flannagan, M. P. Reed, R. M. Cunningham, and J. D. Rupp, “Comparing the effects of age, BMI and gender on severe injury (AIS 3+) in motor-vehicle crashes,” Accid. Anal. Prev., vol. 72, pp. 146–160, Nov. 2014, doi: 10.1016/j.aap.2014.05.024

    2014 doi

  15. [24]

    Regression and time series model selection in small samples,

    C. M. Hurvich and C.-L. Tsai, “Regression and time series model selection in small samples,” Biometrika, vol. 76, no. 2, pp. 297–307, 1989, doi: 10.1093/biomet/76.2.297

    1989 doi

  16. [25]

    Therneau, A package for survival analysis in R

    T. Therneau, A package for survival analysis in R. R package version 3.2-13. [Online]. Available: https://CRAN.R-project.org/package=survival

  17. [26]

    The Compressive Behavior of Bone as a Two-Phase Porous Structure,

    D. Carter and W. Hayes, “The Compressive Behavior of Bone as a Two-Phase Porous Structure,” J. Bone Jt. Surg., vol. 59, no. 7, pp. 954–962, 1977

    1977

  18. [27]

    Multimodel Inference: Understanding AIC and BIC in Model Selection,

    K. P. Burnham and D. R. Anderson, “Multimodel Inference: Understanding AIC and BIC in Model Selection,” Sociol. Methods Res., vol. 33, no. 2, pp. 261–304, Nov. 2004, doi: 10.1177/0049124104268644. 19 Supplementary Material for Hanggi C, Kong J, Caldwell J, Gepner B, Ӧstling M,...

    2025 doi

  19. [317]

    doi: 10.4271/2004-01-0317

    2004 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.