CFD Scheme Development

eMLP-VC (enhanced Multi-dimensional Limiting Process for Vorticity Conservation)
High-order accurate spatial discretization scheme for rotorcraft aerodynamics and aeroacoustics 

The recent growing interest in urban air mobility (UAM) worldwide has led to the demand for physical analyses of the aerodynamic performance and aeroacoustic characteristics of electric vertical takeoff and landing (eVTOL) rotorcraft. In a UAM eVTOL rotorcraft, the smaller vortices generated from multiple propulsors interact with lifting surfaces such as wings, fuselage, and propellers in a complex manner. Therefore, to accurately predict the performance and noise of UAM eVTOL rotorcraft, vortices should be preserved with little dissipation; and their interactions must be modeled precisely. These requirements need a numerical algorithm with a refined resolution than that allowed by conventional schemes. 

A newly modified enhanced multidimensional limiting process (eMLP) for vorticity conservation (eMLP-VC) improves the original eMLP by accounting for the fact that most rotorcraft flowfields are vortex dominated and subsonic. For advanced capability in preserving vortices, the distinguishing criterion was modified through vortex profile analysis, and low-Mach-number adjustment was performed by reconstructing the interpolated primitive variables.

You can check the details of eMLP-VC in contents below.

LAI (Local-order-of-Accuracy Index)
Error assessment technique for shock-capturing schemes in discretized domain

High-order spatial schemes, which effectively treat the dissipative characteristics of computational fluid dynamics solvers, are known to assure the precise prediction of rotorcraft aeroacoustics. However, such schemes may differ in numerical performance, even if they have the same theoretical accuracy. This is because the nonlinear part in a high-order spatial scheme locally reduces the order of accuracy depending on the grid number and flow characteristics. Therefore, it is necessary to compare each scheme in terms of local accuracy and to identify which numerical characteristics are required to retain high-order accuracy. 

A local-order-of-accuracy index (LAI) is newly proposed to quantitatively measure the accuracy of a spatial scheme. The LAI represents the local accuracy of each scheme, including the effect of the grid number and local flowfield characteristics. 

You can check the details of LAI in contents below. 

ICEPAC-K (Ice Contour Estimation and Performance Analysis Code in KFLOW)
(Ongoing research) Aircraft  ice accretion solver in KFLOW solver
with Mr. Soonho Shon and Mr. Younghyo Kim (Ph.D students in SNU AVDL)

For aircraft safety, predicting and preventing aircraft icing is very important, as it has been identified as a significant contributor to aircraft accidents. Icing accretion is being studied by various institutions, including NASA Glenn, and numerical models are being developed to predict it. Seoul National University has successfully conducted various predictions for fixed-wing aircraft through its own ICEPAC-K program. Our team have been trying to expand ICEPAC-K for rotorcraft ice accretion prediction. 

You can check the details of ICEPAC-K in contents below. 

DIRK-types (Diagonally Implicit Runge-Kutta schemes)
High-order accurate temporal integration schemes for implicit solver
with Prof. Soo Hyung Park in Konkuk University

Due to the increasing spatial accuracy of CFD solvers, the level of temporal error becomes more significant than the level of spatial error, resulting in blurred physical phenomena. The DIRK-type schemes are high-order accuracy implicit schemes designed to solve these problems. Currently, researchers are examining methods for applying these schemes to existing numerical solvers efficiently and robustly.

You can check the details of DIRK-types in contents below. 

FAMUS (Fully Automated MUlti-physics Simulator)
meshless CFD solver developed by SNU, ADD, and NEXTFoam in South Korea. 

Conventional grid-based solvers have inherent limitation that the grids should be created by manually. Since the newly developed rotorcraft have various configurations with multiple propulsors and additional lifting surfaces, meshless CFD solver, FAMUS, which do not need grids, could be a powerful alternative. 

With the cooperation of SNU HRL, I've been tried to improve FAMUS solver to use in rotorcraft flowfield. Several rotor blades were used to validate FAMUS. You can check the details of the validation results in here. Also, you can check the more information about FAMUS in here.