Deterministic lateral displacement (DLD) in the high

Reynolds number regime

Dipta Saha
Roll: 0419102049
Department of Mechanical Engineering
Bangladesh University of Engineering and Technology (BUET)
Dhaka-1000, Bangladesh

References:
[1]Huang, L.R., Cox, E.C., Austin, R.H. and Sturm, J.C., 2004. Continuous particle separation through deterministic lateral displacement. Science, 304(5673), pp.987-990.
[2] Dincau, B.M., Aghilinejad, A., Chen, X., Moon, S.Y. and Kim, J.H., 2018. Vortex-free high-Reynolds deterministic lateral displacement (DLD) via airfoil pillars. Microfluidics and
Nanofluidics, 22(12), p.137.
[3] Dincau, B.M., Aghilinejad, A., Hammersley, T., Chen, X. and Kim, J.H., 2018. Deterministic lateral displacement (DLD) in the high Reynolds number regime: high-throughput and dynamic
separation characteristics. Microfluidics and Nanofluidics, 22(6), p.59.

1
 Introduction

Microfluidic Particle Sorting

A lab-on-a-chip (LOC) is a device that integrates one or several laboratory functions on a single integrated circuit.

Microfluidic Particle Sorter is a subset of lab on chip devices.

Blood analysis: Counting and Sorting cell types

 CD-4 for HIV
 RBC / WBC / Platelets
 Circulating Tumor Cells (CTC)

Environment Analysis:
 Bacteria sizing
 Other living organisms

Bangladesh University of Engineering and Technology 2

 Introduction

Deterministic lateral Displacement

Microfuidic DLD uses tilted pillar arrays that generate a fluid
bifurcation and a unique number of streamlines between the
gaps.

These devices use an array of micro pillars to influence

particle displacement according to hydrodynamic forces that
emerge under flow conditions.

 Highly efficient method.

 Continuous and high throughput.
 Fine resolution, separation resolution 12%.

Bangladesh University of Engineering and Technology 3

 Introduction

Deterministic lateral Displacement

 Critical diameter,
Center to center distance between
obstacles, Row shifting distance with
respect to previous row, . Gap
width, d

Modes of particle propagation

• Zigzag mode (Particles that are

smaller than the critical diameter)

• Displacement mode (a particle Fig:

with a radius larger than the
critical diameter)
(A) Geometric parameters defining the obstacle matrix.
(B) Three fluid stream do not mix as they flow through the matrix. Small particle
stays at the same lane.
(C) A particle radius larger than the lane 1 follows a streamline passing through
particle’s center.

Bangladesh University of Engineering and Technology 4

 Methods

Numerical Simulation Parameters

 Geometry:
• 2D geometries
• Circular and airfoil shaped posts

Working platform:
• COMSOL Multiphysics
• Finite element based software

Boundary Condition:
• Inlet constant flow velocity (0.1<Re<75)
• Outlet was fixed at atmospheric pressure
• Walls are no slip condition

Solver Model: 1 2
• Continuity equation
• Navier-stokes equation
• Particle tracing model

Working Fluid Property:

• Density, (a) Cylindrical Shape (b) Airfoil Shape
• Viscosity,
Bangladesh University of Engineering and Technology 5
 Results

Velocity Field within DLD array

(a) Laminar flow, no apparent
vortices or separation.

(b) Vortices begin to appear in

the wake of the pillars.

(c) Vortices have grown

significantly.

(d) Vortices size is nearly

identical with ‘c ’.

Fig: Variation of velocity field with increasing Reynolds number

Bangladesh University of Engineering and Technology 6
 Results

Trajectory of identical particles

Fig: (a) Particle follow zigzag mode, yielding no net displacement, (b) the same particle shows bumped mode, resulting
lateral displacement

By evaluating streamline evolution and vortex effects in this analysis, it is clear that the traditional models
for approximating Dc are insufficient for high-Re flows.

Bangladesh University of Engineering and Technology 7

 Results

Trajectory of identical particles

Experimental observation:

While a 10 micrometer particle shows a zigzag mode at Re=6, it

shows a bumped mode at higher Re.

Transition from left to right represents a particle transition from

zigzag mode to bumped mode

Fig: x axis represent outlet channel number and y axis

represent percentage of particles
Bangladesh University of Engineering and Technology 8
 Results

Velocity Field within DLD array

Here, we can clearly see
that the pillars with no angle
of attack do not develop
vortices up to Re = 100.

Fig: Numerically generated streamlines for different flow condition

Bangladesh University of Engineering and Technology 9

 Results

Variation of Dp /Dc,I with Re

(a) The neutral configuration

yields clear zig-zag and
bumped modes, with a very
thin transition region.
(b) The negative AoA configuration
yields a wider transition region.
Both configurations predict a
similarly decreasing Dc with
increasing Re

Fig : Comparison of Dc shift for two different airfoil pillar configuration

Bangladesh University of Engineering and Technology 10

Conclusions
• For Cylindrical shaped pillars array, the micro vortices emerges at higher Re.

• The micro vortices changes the critical diameter of the array that’s why separation performance can vary.

• Micro pillars with airfoil shape with zero angle of attack does not create any micro vortices at higher
Reynolds number.

• Micro vortices arise when the angle of attach increase.

• Airfoil shaped pillar with zero angle of attack shows high resolution separation performance at higher Re
number over other process.

Bangladesh University of Engineering and Technology 11

Thank you

Bangladesh University of Engineering and Technology 12