This article was submitted to Cryospheric Sciences, a section of the journal Frontiers in Earth Science
This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
There is a complex coupling relationship between the airflow and snow cover. In a period of hours or even days, the airflow will cause the redistribution of snow, and the redistribution of snow will cause the airflow to change. This study develops a dynamic mesh technology applied in snow drifting simulation through a realtime dynamic mesh update to depict the snow surface evolution process under longperiod snow drifting, and a solver application named driftScalarDyFoam based on OpenFOAM is implemented. This solver divides the longperiod snow drifting process into several stages, in each of which a snow transport equation is applied to predict the spatial distribution of snow, and finally, the snow surface evolves according to the erosion–deposition model. This method that we have proposed has been validated for several measured cases, including snow distribution on a flat roof and snow distribution around a building.
Aeolian transport, which is typified by snow drifting, manifests as the interaction between wind and a large number of tiny snow particles. In cold regions, a related phenomenon may cause various types of problems such as equipment failure, traffic interruption, and structural collapse (
For snowdrift related to buildings, the SteadyRANS CFD simulation based on the Eulerian approach was proven to provide sufficiently accurate results with a low computational cost (
However, the previous research studies are more inclined to use commercial CFD software (e.g., Ansys Fluent), which is unfavorable for researchers who want to reproduce or improve this method. OpenFOAM (or Opensource Field Operation And Manipulation) (
The continuity and Navier–Stokes equations, used as governing equations for incompressible airflow, are expressed as follows:
The snow transport in a suspension layer is calculated by using a convection–diffusion equation (
Snow particle saltation can be divided into four subprocesses including the aerodynamic entrainment, particle trajectories, particlebed collisions, and wind modification (
However,
The establishment of the erosion and deposition model can reflect the evolution of snow distribution during snow drifting, which was first proposed by
When erosion occurs, Fick’s laws of diffusion (
DriftScalarDyFoam is an incompressible separate solver considering the pseudotransient assumption of longperiod snow drifting and the dynamic evolution of snowdrift, implemented within the OpenFOAM framework, and parallelized with MPI. Its source code is available and released under the GNU General Public License (GPL).
The redistribution of the snow surface can be described as a firstorder timedependent ordinary differential equation (ODE) for snow mass exchange
An explicit Euler method is adopted to discretize the continuous time, which is expressed as
Since
Step 1) Prepare a computational domain and initialize fields.
Step 2) Enter a new global time step and start the subcycle in this global time step.
Step3) Update all fields with the SIMPLE algorithm as a steadystate scalar transport with oneway coupling, in which the equations of pressure
Step4) Calculate the mass exchange rate
Step5)Adjust the snow surface according to the mass exchange rate
Flowchart of the pseudotransient scheme for longperiod evolution of snow drifting.
In
In this study, a more practical method is applied, in which the inverse distance between the boundary point and its neighbor face center is used as the weight to help implement the interpolation from face value to node value. This method is similar to the RBF interpolation method, but the inverse proportional function we used does not meet the Gaussian distribution. After the interpolation, we specify a vector field
The snow distribution on a flat roof is a typical research case for snow drifting around buildings (
The calculation model used in this study is the prototype described by
Overview of parameters used in simulation of snow drifting in case I.
Variable  Physical meaning  Value  Relevant literature 

Building height  8 m 


Reference velocity at roof height 
4.84 m/s 


Snow bulk density  150 kg/m^{3} 


Duration time of snow drifting  5 × 10^{5}s  N/A 

Angle of repose  50° 


Falling velocity of snow  0.2 m/s 


Threshold friction velocity  0.2 m/s 


Initial depth of snow cover  80 cm 


Aerodynamic roughness height  4 mm 

Computation conditions used in simulation of snow drifting in case I.
Computation domain  16 
Grid discretization  Minimum grid width is 0.02 H 
Turbulence model  Standard kepsilon (SKE) 
Inflow boundary  Profile of velocity and turbulence intensity is set according to the form proposed by 
Downstream boundary  Zero gradient for outflow and fixed value (0) for reverse flow (inletOutlet) 
Upper boundary  Zero gradient condition is used for all variables, 
Snow distribution on the roof using different numbers of calculation stages.
An interesting phenomenon is that the time required for erosion to reach a stable time in the front region is much shorter than that that in the rear region (
Time evolution of the snow distribution on a 3D flat roof using 100 stages (Δ
Variation of friction velocity using a total of 100 stages (Δ
The comparison with the wind tunnel test indicates that our simulation results are within the expected range (
Snow depth at
In regions of high snowfall and strong wind in winter, snowdrift is formed around buildings due to longperiod snow drifting, which may cause difficulties for vehicular traffic and pedestrians (
The model and simulation parameters of Case II are similar to Case I (
Overview of parameters used in simulation of snow drifting in case II.
Variable  Physical meaning  Value  Relevant literature 

Building height  1 m 


Reference velocity at roof height 
3.0 m/s  N/A 

Snow bulk density  150 kg/m^{3} 


Duration time of snow drifting  1 × 10^{4}s  N/A 

Falling velocity of snow  0.2 m/s 


Threshold friction velocity  0.15 m/s 


Initial depth of snow cover  0 cm 


Aerodynamic roughness height  0.2 mm 

Computation conditions used in simulation of snow drifting in case II.
Computation domain  16 
Grid discretization  Minimum grid width is 0.02H 
Turbulence model  Standard kepsilon (SKE) 
Inflow boundary  Profile of velocity and turbulence intensity is set according to the form proposed by 
Inflow boundary (snow)  Fixed value of 0.005 kg/m^{3} 
Downstream boundary  Zero gradient for outflow and fixed value (0) for reverse flow (inletOutlet) 
Upper boundary  Zero gradient condition is used for all variables, 
Snow distribution around the cube at
Comparison of normalized snow depths obtained from previous CFD simulation (
In the comparison of snow depth at
Velocity vectors and contours around the cube.
This study mainly introduces the solver driftScalarDyFoam developed based on OpenFOAM, which refers to the classic snow transport model and erosion/deposition formula, and incorporates the multistage model to achieve simulations with longterm duration or complex meteorological conditions.
Through the prediction of snow distribution on flat roofs and around buildings, driftScalarDyFoam has been shown to be robust and reliable in its results, which depicts its versatility in the research and production of snow engineering. Compared with previous numerical studies (
However, the limitation of the theoretical model hinders the accuracy of this solver, including the oneway coupling between snow and airflow and turbulence effect on snow particles. In addition, the application of this model in mountainous regions (
The original contributions presented in the study are included in the article/
XC: conceptualization of this study, methodology, software, validation, formal analysis, data curation, writing—original draft, and visualization. ZY: project administration and funding acquisition.
This project is jointly supported by the National Natural Science Foundation of China (Nos. 51378428 and 52178506).
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
The Supplementary Material for this article can be found online at: