Skip to content

hyunbin-kang/MS_portfolio

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 

Repository files navigation

Portfolio of MS course

This page is the portfolio of Hyunbin Kang, a master's student at Inha University, Department of Aerospace Engineering, Aerodynamic Analysis and Design Lab.

Profile

Name : Hyunbin Kang

Major : Computational Fluid Dynamics

  • High Order Methods
  • Hypersonic Aerodynamics
  • Panel Methods

Education

  • Inha University, Department of Aerospace Engineering, Bachelor's Degree (2017-2022)
  • Inha University, Department of Aerospace Engineering, Master Course (2023-)

Email : qls1455@gmail.com(personal), qls1455@inha.edu(education)

Office : Inha University (Aerospace Campus), 36 Gaetbeol-ro, Yeonsu-gu Incheon 21999, Korea

Projects

About the projects I was involved in during my Master's Course.

1. KSAS 2023 Spring conference talk

Korean Society for Aeronautical and Space Science

alt text

Comparison of Flow Simulation over Low Reynolds Number Airfoil by Accuracy of PyFR

Intro :

  • In this study, aerodynamic investigation was conducted with Low Reynolds Number Airfoil of EFD-CFD workshop 5th case *.
  • LRN airfoil has complex flow structure, so accurate computational is needed.
  • In this study, We utilze the PyFR, a aerodynamic analysis framework using Large Eddy Simulation (LES)

Results :

Spanwise effect

  • Shorter spanwise length case has pool results of lift coefficients.

    image
    lift curve with several spanwise length

  • This result is caused by the 3-D effect of turbulent flow.

    image
    Q-criterion at angle of attack of 10 degree

Laminar Separation Bubble (LSB)

  • Low Reynolds number flow has a complex flow structure such as laminar separation, transition, reattachment etc.
  • This LSB reduce the aerodynamic performance.
  • LES can solve the LSB with high resolution.

2. ICOSAHOM 2023 conference talk

International Conference of Spactral and High Order Method

alt text

Simulating Unsteady Flow over Low-Reynolds-Number Airfoil with Separation Bubble by PyFR

Intro :

  • Based on the previous study (KSAS), parametric study is conducted for numerical and computational condition.
  • Paramters : spanwise length, order of accuracy, grid system, and precision

Results :

Non-linearity

  • LSB's movement cause the non-linear aerodynamic characteristics

    imgage
    non-linearity of lift coefficient

    image

    time-averaged x directional velocity contour

    image
    time-averaged pressure coefficient distribution

Stall structure :

  • At high angle of attack region, LSB is genarated at leading edge of airfoil.
  • Since the strong separation can't recover the pressure, full separation from LE to TE is captured.

    image
    time-averaged x directional velocity contour at high angle of attack

Discussion :

effect of Spanwise length

  • We compared the result with (0.05, 0.1, 0.25, 0.5, 1.0) times of the chord length
  • Short span is limited in including 3-D effect
  • In high AoA region, spanwise length of 0.5 times of chord length is needed at least.

effect of Order of accuracy

  • We performed the simulation with P2, P3, and P4 order with coarse, medium, and fine grid respectively
  • If the grid is fine near the surface and wake region, there are slight difference in the case of P2 order compared with P2 and P3 order.
  • The simulation with P2 coarse grid has inaccurate result. So lower order case has strong grid dependency.

effect of Grid system

  • We constructed structured grid by PointWise and unstructured grid by Gmsh.
  • There are no sigmificant difference with two grid system.
  • However, there notable difference in computation time
  • Unstructured grid's minimum edge length and minimum cell volume is 4 and 13 times smaller than structured grid. So the computation time of unstr case is 9 times larger than the structured case.

effect of precision

  • There are no siginficant difference with single and double precision(FP32 and FP64)

3. Teaching Assistant of Capstone Design

Research on developing a variable wake model for panel program

Intro :

  • Panel method is one of fast techniques of the aerodynamic analysis
  • Panel method use the virtal wake panel to predict the lift
  • When solving for high angle of attack condition, fixed wake model issues
  • So variable wake model must be developed