Tomographic PIV 계측에 있어서의 영향인자 분석

Abstract

본 논문은 Tomographic PIV계산에 있어서 속도벡터의 오차에 영향을 미치는 인자를 분석한다. 영향인자로 카메라교정과 복셀 영상 재구성으로 선정하였으며, MLOS(multiplicative line of sight), MART (multiplicative algebraic reconstruction technique), MART(multiplicative algebraic reconstruction technique), SMART(simultaneous multiplicative algebraic reconstruction technique) 알고리듬을 이용하여 영상을 재구성하였다. 카메라 교정방법으로는 10-parameter법과 변환 행렬법을 선정하여 정량적 분석을 하였으며, 그 결과 10-parameter법이 높은 성능을 보였다.

링와(ring vortex) 형태의 유동장의 가상영상을 이용하여 복셀 영상을 비교적 계산속도가 빠른 MLOS방법으로 재구성한 다음, Tomo-PIV계산을 수행한 결과 링(ring)형태의 유동패턴을 재구성할 수 있었다.

구축한 10-파라미터법, MLOS기반의 Tomo-PIV계산방식으로 사각형 몰수체 후류에 대한 실험 계산한 결과 2차 와(secondary vortex)를 재현할 수 있었다. | In this paper, the influences factors in tomographic PIV calculations are quantitatively investigated.

The error propagation coming from the camera calibration has been investigated, and the error propagation coming from the reconstruction of voxel images, in which MLOS (multiplicative line of sight), MART (multiplicative algebraic reconstruction technique), SMART (simultaneous multiplicative algebraic reconstruction technique) algorithms are adopted, has been investigated.

The influences of the camera calibration method (10-parameter and transfer matrix) to the Tomo-PIV results have been also quantitatively investigated. It has been verified that the 10-parameter method has shown better performance than transfer matrix.

Based on MLOS reconstruction algorithm, the calculation results of Tomo-PIV showed shorter calculation time, and it has been adopted to the calculation of a ring vortex, from which it has been validated that the original flow patterns of the ring vortex has been reconstructed.

Further, the constructed Tomo-PIV system has been adopted to an experimental flow images of the wake of a square bluff body, from which the secondary flow structures of the wake have been clearly captured.