3D flow visualization

 

 

 

3D PIV measurements of flow in a nasal cavity with geometry acquisition

 

-  3D PIV measurements of flow in a nasal cavity model which is constructed using transparent silicone. Experiments were performed using refractive index-matched working fluid (mixture of glycerol and water).

 

Index matching tests for a model filled with (a) 100 % water, (b) 100 % glycerol, and (c) a refractive index-matched working fluid

 

Schematic diagram of the experimental apparatus

-  Procedure

 

Geometry acquisition: 3D shape of the nasal cavity model was reconstructed from captured PIV images.

 

Schematics of the particle triangulation and pairing procedure

 

(a) Results of the morphological operations applied to the particle accumulation results. (b) Crosssectional images of the corresponding geometry at the y - z plane

Results

-  Flow field inside the acquired geometry of the nasal cavity

 

Coronal cross-sectional images of the measured flow field: a velocity magnitude, and b turbulent kinetic energy

 

 

Coronal cross-sectional contours of the velocity magnitude and turbulent kinetic energy

-  Flow pattern analysis inside the nasal cavity

 

(a) Vector field at the cross-section, (b) stream lines and stagnation points, and (c) combined images of (b) and (c) near the middle turbinate

 

(a) Sagittal view of the nasal cavity and the streamlines inside the inferior meatus. The contours indicate (b) the velocity magnitude, (c) u, (d) v, and (e) w

 

 

 

3D PIV measurement of flow around an arbitrarily moving body

 

-  Simultaneously monitoring 3D fluid flows and the structure of an arbitrarily moving surface embedded in the flow.

 

Schematic diagram of the experimental setup associated with the 2D and 3D PIV.

(a) 2D PIV and the corresponding PIV image. (b) 3D PIV and a volumetric PIV image

 

-  Procedure

 

 

 

-  Surface shape and tomographic reconstruction volume

      From obtained surface structure, 3D reconstruction volume can be optimized.    

   

 

 

Particle volume reconstruction associated with the surface structure.

 

 

Results

 

-  Flow field around an eccentric rotating cylinder

 

Instantaneous flow field around the reconstructed cylinder

 

-  Flow field around a flapping flag

 

Time-varying flow field and the flag motion at Y = 0 for one flapping period

 

 
 
 
 
 








Tomo-PIV Measurement of Flow around an Arbitrarily Moving Body with Surface Reconstruction


S.Im, Y.J.Jeon and H.J.Sung, "Tomo-PIV Measurement of Flow around an Arbitrarily Moving Body with Surface Reconstruction", Experiments in Fluids, Vol. 56, Number 2, 2015

 

A three-dimensional surface of an arbitrarily moving body in a flow field  was reconstructed using the DAISY descript-xor and epipolar geometry constraints. The surface shape of a moving body was reconstructed with tomographic PIV flow measurement. Experimental images were captured using the tomographic PIV system, which consisted of four high-speed cameras and a laser. The originally captured images, which contained the shape of the arbitrary moving body and the tracer particles, were separated into the particle and surface images using a Gaussian smoothing filter. The weak contrast of the surface images was enhanced using a local histogram equalization method. The histogram-equalized surface images were used to reconstruct the surface shape of the moving body. The surface reconstruction method required a sufficiently detailed surface pattern to obtain the intensity gradient profile of the local descript-xor. The separated particle images were used to reconstruct the particle volume intensity via tomographic reconstruction approaches. Voxels behind the reconstructed body surface were neglected during the tomographic reconstruction and velocity calculation. The three-dimensional three-component flow vectors were calculated based on the cross-correlation functions between the reconstructed particle volumes. Three-dimensional experiments that modeled the flows around a flapping flag, a rotating cylinder, and a flapping robot fish tail were conducted to validate the present technique.

 

Figure 1 Gaussian convolved orientation maps of the histogram-equalized surface image