Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index; none of the 56 sources its claims cite is checked for support yet.

Recent advances in modelling of frost formation for mechanical systems

Open access
Tao, Yong · 2025 · Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences

Abstract

The physics underlying frost and ice formation has been extensively studied over the past few decades, with significant contributions to our understanding of this phenomenon. These insights have primarily been applied to engineering systems with refrigerated surfaces, such as refrigerators, freezers and heat pumps of various sizes. Despite considerable progress, the dynamic and complex mechanisms governing frost and ice formation remain an active area of research, as competing factors continue to challenge predictive accuracy. The increasing interest from stakeholders in reducing energy consumption and carbon footprints in mechanical systems further underscores the importance of advancing modelling and simulation capabilities in this domain. This article critically reviews recent advances in frost and ice formation modelling, with a focus on identifying key research gaps and opportunities for further exploration. This article is part of the theme issue ‘Heat and mass transfer in frost and ice’.

Overview

CoChat Summary

TL;DR

This review surveys three decades of frost-growth modelling for mechanical systems, covering physics-based, data-driven, and geometry/surface-treatment approaches. It finds that despite progress, models remain condition-specific, frost density prediction lags thickness prediction, and generic validated models are still lacking.

Key findings

  • The review identifies persistent gaps: lack of generic widely applicable models/correlations, need for real-time validation of frost thickness and density, uneven frosting at ultra-low temperatures, and later-stage mechanisms; it states that Song & Dang’s 2018 gaps remain valid.
  • Physics-based modelling advanced from Tao et al.’s two-period porous-media model, which introduced an effective mass diffusivity up to seven times the molecular diffusivity of water vapour in air, to air–frost full-domain CFD models that remain sensitive to mesh and empirical correlations.
  • Data-driven models report high accuracy within limited ranges: Tahavvor’s ANN/fractal model agreed within 2% for frost density and 5% for frost thickness; Moradkhani et al. used 249 thickness and 54 density data points (80/20 train/test) and reported AAREs of 1.06% for thickness and 0.83% for density, with >99% of predictions within ±10%.
  • Surface and geometry studies show effects such as about 10% shorter defrosting with superhydrophobic VMFE coatings, and CFRP fin frost thickness increases of 71.5% and 75.9% when surface temperature decreased from −10°C to −20°C at ambient 5°C and 10°C; generally frost thickness increases with higher air velocity, higher relative humidity, and lower cold surface temperature, while frost density increases with higher wall temperature and air velocity, with relative-humidity impact debatable.

Methods

  • Design: critical/narrative review of frost-growth modelling methodologies for mechanical systems over the last three decades, organized into physics-based models, data-driven models, and geometric/surface characterizations.
  • Data or sample: reviews representative experimental, analytical, numerical CFD, and machine-learning studies and correlations from the 1940s onward; no new experiments or primary data are collected.
  • Main analysis: qualitative synthesis and comparison of model formulations, validation ranges, reported errors, and research gaps; no systematic review protocol or meta-analysis is described.

Limitations

  • Paper-stated: reviewed models lack generic applicability; real-time validation of frost thickness/density is needed; data are limited at ultra-low temperatures and later frost stages; high-fidelity air–frost models face moving-boundary/mushy-zone tracking, mesh sensitivity, and extensive computational time; initial frost thickness and density are often guessed and adjusted.
  • Paper-stated: data-driven models are limited to available experimental conditions and ranges, e.g. Moradkhani et al. caveat that agreement is limited to forced-convection data, and Abbas et al. report a preliminary correlation capturing 82% of data within a 30% error margin.
  • Evident from the text: as a review, it is selective and focuses on mechanical-system applications and basic configurations, explicitly referring readers to other reviews for many experimental works; no systematic search or quantitative synthesis is reported.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1098/rsta.2024.0369

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1. Two-period model predicting frost growth [1]: (a) porous-media model describing thermal and mass diffusion within the frost layer during the full growth period; (b) ice column model for initial growth period; (c) transition time between two periods based on simplified empirical observations: moist air temperature, T∞ , humidity ratio, W∞ , velocity, U∞ , heat and mass transfer coefficients, h, hm, cold plate temperature, Tc, ice column temperature, Tβ , ice column diameter, d, frost thickness, δ f , frost density, ρ f, frost effective thermal conductivity, keff and effective mass diffusivity, Deff . Reprinted from [10], with permission from Elsevier.
Figure 1. Two-period model predicting frost growth [1]: (a) porous-media model describing thermal and mass diffusion within the frost layer during the full growth period; (b) ice column model for initial growth period; (c) transition time between two periods based on simplified empirical observations: moist air temperature, T∞ , humidity ratio, W∞ , velocity, U∞ , heat and mass transfer coefficients, h, hm, cold plate temperature, Tc, ice column temperature, Tβ , ice column diameter, d, frost thickness, δ f , frost density, ρ f, frost effective thermal conductivity, keff and effective mass diffusivity, Deff . Reprinted from [10], with permission from Elsevier.CC BY · PubMed Central
Figure 2. Three-dimensional visualization of the frost layer at an instant in time with contours of temperature subject to turbulent flow [16,25]. The frost layer thickness varies along x and y. At point D on the frost surface, the thickness is given by Sf^(∗).The plate temperature is held constant at Tp^(∗), but the frost surface temperature varies in x and y. As shown in panel (b), the temperature increases linearly from the plate temperature to the frost surface temperature. The frost density also varies along x and y. However, along the z-direction, only an average frost density value is considered (i.e. no variation along z). Reprinted from [32], with permission from Elsevier.
Figure 2. Three-dimensional visualization of the frost layer at an instant in time with contours of temperature subject to turbulent flow [16,25]. The frost layer thickness varies along x and y. At point D on the frost surface, the thickness is given by Sf^(∗).The plate temperature is held constant at Tp^(∗), but the frost surface temperature varies in x and y. As shown in panel (b), the temperature increases linearly from the plate temperature to the frost surface temperature. The frost density also varies along x and y. However, along the z-direction, only an average frost density value is considered (i.e. no variation along z). Reprinted from [32], with permission from Elsevier.CC BY · PubMed Central
Figure 3. Frost growth on two parallel cold plates forming a channel airflow is modelled as in a symmetric domain where the airflow cross-section is subject to a finite height and an adiabatic thermal boundary condition at the top [42]. Reprinted from [3], with permission from Elsevier.
Figure 3. Frost growth on two parallel cold plates forming a channel airflow is modelled as in a symmetric domain where the airflow cross-section is subject to a finite height and an adiabatic thermal boundary condition at the top [42]. Reprinted from [3], with permission from Elsevier.CC BY · PubMed Central
Figure 4. Definition of (a) a square fin and (b) an annular fin, both subject to a cross flow of moist air [44]. Reprinted from [2], with permission from Elsevier.
Figure 4. Definition of (a) a square fin and (b) an annular fin, both subject to a cross flow of moist air [44]. Reprinted from [2], with permission from Elsevier.CC BY · PubMed Central
Figure 5. (a) A representative morphology of a micropillar patterned substrate via SEM and a sessile DI water droplet placed on such a substrate, indicating superhydrophobic characteristic [48,53]. (b) Definition of pitch L. Reprinted from [53], with permission from Elsevier.
Figure 5. (a) A representative morphology of a micropillar patterned substrate via SEM and a sessile DI water droplet placed on such a substrate, indicating superhydrophobic characteristic [48,53]. (b) Definition of pitch L. Reprinted from [53], with permission from Elsevier.CC BY · PubMed Central
Figure 6. Three surfaces with different contact angles [49]. Reprinted from Applied Thermal Engineering [54], with permission from Elsevier.
Figure 6. Three surfaces with different contact angles [49]. Reprinted from Applied Thermal Engineering [54], with permission from Elsevier.CC BY · PubMed Central
Figure 7. Frost growth on an AgI stripe patterned surface with (a) three different spacings, and (b) with a comparison between a flat surface and a patterned surface (400 micron metre spacing), showing a horizontal growth trend on strip patterned surface [50]. Reprinted from International Journal of Heat and Mass Transfer [55], with permission from Elsevier.
Figure 7. Frost growth on an AgI stripe patterned surface with (a) three different spacings, and (b) with a comparison between a flat surface and a patterned surface (400 micron metre spacing), showing a horizontal growth trend on strip patterned surface [50]. Reprinted from International Journal of Heat and Mass Transfer [55], with permission from Elsevier.CC BY · PubMed Central

Verification

Claims the paper makes

0 Valid · 8 Not valid · 40 Not checked · 44 Own results and interpretations

22 more claims

Create a free account to see all 92 claims and to verify the 56 references still open.

See all claims — free account
  • Probably not a substantial claim. 57
  • It may state two separate claims as one. 20

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 56 references

Research works printed but not linked: 2

  1. Mao Y, Besant R, Rezkallah K. 1992 Measurement and correlations of frost properties with airflow over a flat plate. ASHRAE Trans. Symp 98, 65–78.
  2. Mao Y, Besant R. 2005 Measurement and correlations of frost properties with laminar airflow at room temperature over a flat plate. ASHRAE Trans. Symp 111, 739–745.

Cited by

See all 6

Metadata

Authors
Tao, Y.
Affiliations
Mechanical Engineering, Cleveland State University
Venue
Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
Published
2025-07-17
Volume
383
Issue
2301
Pages
20240369-20240369
ISSN
  • 1364-503X (ISSN-L)
  • 1471-2962
OpenAlex
W4412488992
Cited by
6
References
56
Citations by year
2026: 6
FWCI
0.28
Type
journal article
Language
en
Open access
Licence
Locations
  • Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • EngagedScholarship @ Cleveland State University (Cleveland State University) (submittedVersion, cc-by)
Topics
  • Physical Sciences › Engineering › Mechanical Engineering › Refrigeration and Air Conditioning Technologies
  • Physical Sciences › Engineering › Aerospace Engineering › Icing and De-icing Technologies
  • Physical Sciences › Chemical Engineering › Filtration and Separation › Freezing and Crystallization Processes
Keywords
frost formation, ice formation, heat pumps, refrigerators, freezers, energy consumption, carbon footprint, heat and mass transfer, prediction accuracy

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-18, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-18 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Recent advances in modelling of frost formation for mechanical systems. https://app.cochat.ai/papers/10.1098/rsta.2024.0369
The paper
Tao, Y. (2025). Recent advances in modelling of frost formation for mechanical systems. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences. https://doi.org/10.1098/rsta.2024.0369

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.