Airflow and Particle Deposition in a Dry Powder Inhaler

A CFD Simulation

J. Milenkovic

, A. H. Alexopoulos

and C. Kiparissides

CPERI, CERTH, 6th km Harilaou-Thermi rd., Thermi, Greece

Department of Chemical Engineering, Aristotle University of Thessaloniki, Thessaloniki, Greece

Keywords: Dry Powder Inhaler, Turbuhaler, CFD, Particle, Deposition.

Abstract: In this work the steady-state flow in a commercial dry powder inhaler device (i.e., Turbuhaler) is described.

The DPI geometry is constructed in a CAD/CAM environment (i.e., CATIA v5) and then imported into

GAMBIT where the geometry is discretized into a computational grid. The Navier-Stokes equations are

solved using FLUENT (v6.3) considering different flow models, i.e., laminar, k-ε, k-ε RNG, and k-ω SST.

Particle motion and deposition are described using an Eulerian-fluid/Lagrangian-particle approach. Flow

and particle deposition for a range of mouthpiece pressure drops (i.e., 800-8800Pa), as well as particle sizes

corresponding to single particles and aggregates (i.e., 0.5-20μm) are examined. The total volumetric outflow

rate, the overall particle deposition as well as the particle deposition sites in the DPI are determined. The

transitional k-ω SST model for turbulent flow was found to produce results most similar to a reference

Large Eddy Simulation solution as well as experimental results for the pressure drop in the DPI. Realistic

particle deposition results could only be obtained by considering a nonideal sticking coefficient

corresponding to a critical capture velocity of 2.7m/s. Overall, the simulation results are found to agree well

with available experimental data for volumetric flow and particle deposition.

1 INTRODUCTION

Dry Powder Inhalers “DPI”s are one of the principle

means of delivering pharmaceuticals due to their

ease of use and cost-effectiveness. The main

function of a DPI device is the adequate dispersion

and delivery of particles. Initially the particles are in

the form of a loose powder which, under the action

of airflow is broken up and dispersed as particle

aggregates which are then further broken up into

fine particles (Ashurst et al. 2000); (Newman and

Busse, 2002); (Tobyn et al., 2004); (Islam et al.,

2008); (Alagusundaram et al., 2010). Powder

properties, e.g., cohesion, charge, size, and size

distribution, influence powder dispersion and the

breakage of particle agglomerates (French et al.,

1996); (Zeng et al., 2000); (Finlay, 2001); (Newman

and Busse, 2002); (Chan, 2006).

One of the common problems with DPIs is the

loss of powder/drug due to deposition within the

device. In order to provide the maximum drug dose

per inhalation and to ensure minimal dose-to-dose

variation it is necessary to minimize the drug losses

due to internal deposition. It is also desired to have

good control over the dispersibility of the powder,

release of drug (when attached to powder particles),

and breakup of agglomerates in order to achieve the

desired particle/agglomerate size distributions at the

DPI mouthpiece outflow (Alagusundaram et al.,

2010). Consequently, if the underlying processes are

better understood one can achieve the desired

outflow particle distribution which will conceivably

minimize oropharyngeal losses and also permit

better targeting for drug delivery in the respiratory

tract.

Due to the complex and transient flow structures

observed in most commercial DPIs as well as the

dynamic powder breakup and dispersion processes

only a small number of Computational Fluid

Dynamics “CFD” investigations have been

conducted (Schuler et al., 1999); (Ligotke, 2002).

Systematic computational studies have led to a better

understanding of the function of DPI devices. For

example, Coates et al. (2004, 2005, 2006) studied

the Aerolizer DPI in detail including the effects of

air-intake, mouthpiece, and internal grid which led

to improvements in the design and function of the

DPI. Recently, the discrete element method, DEM,

250

Milenkovic J., H. Alexopoulos A. and Kiparissides C..

Airﬂow and Particle Deposition in a Dry Powder Inhaler - A CFD Simulation.

DOI: 10.5220/0004058102500259

In Proceedings of the 2nd International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH-2012),

pages 250-259

ISBN: 978-989-8565-20-4

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

coupled to continuous phase-models has been

implemented to describe the powder dispersion

process within the inhaler (Tong et al., 2010);

(Calvert et al., 2011). From the current state-of-the-

art it is clear that the proper description of the

agglomerate strength as well as the

particle/agglomerate interaction with the inhaler

walls are key processes that determine the final

dispersion and size distribution of pharmaceutical

powders (Adi et al., 2011).

The Turbuhaler (AstraZeneca) is a multidose dry

powder inhaler that is widely used to deliver a

number of drugs (typically for asthma), e.g.,

terbutaline sulphate, (as Bricanyl), or budesonide (as

Pulmicort), to the upper respiratory tract (Wetterlin,

1988). Each dose is initially in the form of loosely

packed particle agglomerates, ~10-20μm in size,

which are released into a mixing/dispersion

chamber, where they are broken up into particles,

~1μm in size, which are then directed to the

inhalation channel of the device (Tsima et al., 1994;

Wetterlin, 1988). The proper function of the

Turbuhaler is dependent on the dynamic volumetric

flow as well as the peak inspiratory flow rate

attained during inhalation, the amount of particles

lost due to deposition within the device, and the

adequate dispersion and breakup of the powder

agglomerates in the airflow exiting the mouthpiece.

Recent experimental investigations have provided

detailed information on particle capture as well as

the percent and size distribution of escaped particles

in the outlet flow (de Koning et al., 2001); (Hoe et

al., 2009); (Abdelrahim, 2010).

In this work the steady airflow in a Turbuhaler

DPI is determined by CFD simulations and particle

motion as well as deposition is determined by

Eulerian-fluid/Lagrangian-particle simulations. In

what follows the DPI geometry, the discretization

procedure, and the CFD simulations are described in

detail. Next the results for steady-state airflow are

presented follow by the results for particle

deposition. Finally, the computational results are

compared to available experimental data.

2 RESULTS

The Turbuhaler DPI geometry was constructed in a

CAD/CAM environment (i.e., CATIA v5 R19) and

then imported into GAMBIT (v2.1) where a series of

computational grids were constructed consisting of 2

–

tetrahedral cells with a maximum

skewness of 0.85 (Figures 1, 2, and 3). The

computational grids were originally refined in

regions where large gradients of flow were expected.

Figure 1: Turbuhaler dry powder inhaler.

Figure 2: Turbuhaler dry powder inhaler CAD geometry.

Figure 3: Turbuhaler Dry Powder Inhaler Computational

Grid (1 10

tetrahedral cells).

AirflowandParticleDepositioninaDryPowderInhaler-ACFDSimulation

251

Further refinement was conducted within

FLUENT based on actual velocity gradients

observed in initial solutions.

The Navier-Stokes equations for airflow were

solved using the commercial CFD software (i.e.,

FLUENT v6.3). The SIMPLEC scheme was

employed to describe pressure-velocity coupling.

Second order discretization was used for pressure

and third order MUSCL for momentum and

turbulent variables. Convergence of CFD

simulations was assumed when the residuals were <

-4

. Zero gauge pressure boundary conditions were

employed at all the inflows, i.e., two powder loaded

cylinders (see bottom of Figure 3a) and four extra air

inlets in the DPI dispersion chamber (see Figure 1).

Different steady state airflows were simulated by

imposing a wide range of pressure drops at the

mouthpiece outflow ranging from 800 to 8800Pa

corresponding to volumetric flow rates of 20 to 70

l/min. Steady-state airflow can be considered an

approximation to dynamic inhalations where the

flow rate has approached the peak inspiratory value.

Eulerian-fluid/Lagrangian-particle simulations of

particle motion and deposition were conducted for

particles between 0.5-20μm in size encompassing

the single particle and particle agglomerate size

ranges of typical pharmaceutical powders employed

in the Turbuhaler. Particles were assumed to be

released instantaneously at t = 0 and uniformly from

a surface located immediately upstream from the

powder storage site. Powder dispersion was assumed

to occur instantaneously after which no further

breakage occurred. Consequently, particles in

motion were taken to be constant in size. Upon

collision with the inhaler walls particles either

deposited or reflected. No collision-induced

breakage was examined in this work. The capture

efficiency of particles with the inhaler walls was

assumed to be either equal to one or a function of the

velocity magnitude.

2.1 Simulations of Airflow in the

Turbuhaler DPI

According to the range of volumetric airflows

examined in this work, e.g. Q = 20 - 70l/min, the

local Reynolds numbers, Re = Q ρ / μ A

1/2

, where ρ

and μ are the density and the viscosity of air and A is

the cross-sectional area, ranged from 130-16,000.

Consequently the transitional SST k-ω model was

employed to describe the transitional turbulent flows

encountered in the DPI.

Computational grids, varying between 2 10

and

2 10

tetrahedral elements, were employed to test for

convergence. The 1 10

grid was found to provide

essentially identical results as the 2 10

grid and was

used for the results presented in this paper. It should

be noted that the computational grid was extended

from the mouthpiece by 20mm in order to minimize

recirculation effects at the outflow surface and to

improve convergence behaviour.

Figure 4: Velocity magnitude in the Turbuhaler DPI

(mouthpiece pressure drop ΔP = 800Pa).

Figure 5: Tangential velocity in the Turbuhaler DPI (ΔP =

800Pa).

In Figures 4-6 the velocity magnitudes as well as

the tangential and radial velocities are displayed

along an axial (i.e., zx) plane and several planes

normal to the z-axis (i.e., xy sections). As can be

observed, the airflow in the DPI device is found to

be laminar in the inhalation channel with two jet

flows emanating from the powder storage cylinders.

SIMULTECH2012-2ndInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

252

Figure 6: Radial velocity in the Turbuhaler DPI

(ΔP = 800Pa).

Figure 7: Mouthpiece velocity vectors (ΔP = 800Pa).

Figure 8: Mouthpiece tangential velocity contours

(ΔP = 800Pa).

In the dispersion chamber the flow is

characterized by large eddies and secondary flows.

In the helical region significant tangential flows

develop and persist about halfway up the

mouthpiece extension. The tangential motion

induced by the helical airway in the mouthpiece is

significant reaching 83% of the maximum velocity

magnitude. It should be noted that the velocity

profiles observed for larger flow rates, e.g., 60 l/min,

are qualitatively similar.

The mouthpiece outflow of a DPI is very

important as it determines the dispersion and flow

behaviour of the particles in the oral cavity and the

upper respiratory tract and consequently influences

particle losses in the oral cavity and throat regions.

In Figures 7 and 8 the mouthpiece outflow for a

pressure drop of ΔP = 800Pa is shown in terms of

velocity magnitude and tangential velocity. It is

clear that the flow is strongly influenced from the

preceding helical region and that the axial and

tangential components of the velocity are

nonuniform. Moreover, the strongly localized

tangential and axial airflows at the mouthpiece cause

recirculation flows in both the tangential and axial

directions, further complicating the flow.

Large Eddy Simulations “LES” fully resolve the

large scale motion of turbulent flows thus providing

more information and accurate results compared to

Reynolds Averaged Navier-Stokes “RANS”

approaches, e.g., k-ε, k-ω. The computational burden

of LES is significant (e.g., at least an order of

magnitude more than with RANS models).

Consequently, only a single case (i.e., ΔP = 800Pa)

of steady-state flow in the Turbuhaler DPI was

simulated with LES using FLUENT.

In Figure 9 the results for the mean velocity

magnitude obtained with LES is shown. The main

flow structures are similar with the k-ω SST results

in Figure 4 but, as expected, differences can be

observed in the flow details as well as in secondary

flows.The enhanced resolution of eddies and

secondary flows with the LES is demonstrated in

Figures 10 and 11 depicting the tangential and radial

flow, respectively. Compared to the radial and

tangential flows predicted with the k-ω SST model

(Figures 6 and 7) there are many differences, e.g., in

the large eddies of the mouthpiece extension.

In Figure 12 the magnitude of the RMS velocity

fluctuations in the DPI is shown. Significant velocity

fluctuations are observed at the top of the dispersion

chamber (~6m/s) and in the mouthpiece extension

(~9m/s). The intensity of fluctuations (e.g., RMS

velocity / velocity magnitude) varies within the

device up to a value of ~50% indicating significant

local fluctuations around the mean for the length

fluctuations of the individual velocity components

AirflowandParticleDepositioninaDryPowderInhaler-ACFDSimulation

253

scales of flow resolved within the LES. The RMS

range from 1-8m/s for the axial velocity component

and 1-4m/s for the other components with different

spatial variations within the device. These results

demonstrate that the fundamental assumption of

local turbulence isotropy of the RANS models is

incorrect In Figure 13 the tangential velocities at the

outlet surface for ΔP = 1400Pa are shown. It is clear

that the tangential velocities predicted by the k-ω

SST and LES turbulence models are very similar. In

fact the k-ω SST turbulence model provided the

most similar to the LES results compared to the

other RANS turbulence models (e.g., standard k-ε,

RNG k-ε). Consequently, despite the observed

differences in secondary flows (Figures 9-11) the k-

ω SST model was employed for all the simulations

of this work

Figure 9: Velocity magnitude in the Turbuhaler DPI –

LES results (ΔP = 800Pa).

Figure 10: Tangential velocity component in the

Turbuhaler DPI – LES results (ΔP = 800Pa).

Figure 11: Radial velocity component in the Turbuhaler

DPI – LES results (ΔP = 800Pa).

Figure 12: RMS velocity magnitude in the Turbuhaler DPI

(ΔP = 800Pa).

2.2 Simulation of Particle Motion and

Deposition in the Turbuhaler DPI

Eulerian-fluid/Lagrangian-particle simulations were

performed for all the flows examined in section 2.1.

These simulations are generally valid for particle

volume fractions <10%. For effective powder

dispersion the solids volume ratio in the DPI device

is approximately 10

-2

-10

-4

depending on the location

and the flow rate. Consequently, the particle phase

was assumed to not influence the airflow.

The total particle deposition in the DPI device

was determined assuming either a 100% capture

efficiency, σ, or a capture efficiency based on a

critical velocity magnitude. The later case was

SIMULTECH2012-2ndInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

254

implemented within FLUENT using a user-defined

function for the capture efficiency based on the

normal velocity of the particle at the moment of

collision with the walls.

(a)

(b)

Figure 13: Tangential velocity component at the

mouthpiece exit (ΔP = 1400Pa). (a) LES (b) k-ω SST.

Figure 14: Overall particle deposition in the Turbuhaler.

Single-sized simulations were performed with

particles ranging from 0.5-20μm. Particle sizes 0.5-

1.5μm correspond to individual particle constituents

of the agglomerates. Agglomerate breakage and

redispersion effects were not considered. Instead, the

agglomerates in the powder storage cylinders were

assumed to break-up rapidly into their constituent

particles. Clearly, agglomerate breakage and flow

occur simultaneously and this is an area which

requires further investigation.

For micron sized particles inertial forces

dominate the deposition process and for particles

<100μm gravity can be ignored during the time-

scale of a single inhalation. In Figure 14 the total

deposition for single-sized particles ranging from

0.5-10μm, for an ideal capture efficiency, i.e., σ = 1,

and for two pressure drops, e.g., 800 and 1400Pa,

are shown. These simulation results indicate that for

a pressure drop of 800Pa the deposition of 0.5-1μm

particles is 19-24% but that of agglomerates 5-10μm

is 90-100%. The predicted total particle deposition

in the DPI increases with volumetric flow to large,

and unrealistic, values (Figure 14). Smaller

deposition values can be obtained by considering

less than 100% particle capture efficiency. Other

mechanisms such as agglomerate breakage dynamics

and/or redispersion of deposited agglomerates could

also result in smaller values of particle deposition.

0.0

0.1

0.2

0.3

0.4

Absolute Deposition

2μm

5μm

1μm

(b)

0 102030405060

Axial Position, mm

0.0

0.2

0.4

0.6

0.8

Fractional Deposition

2μm

5μm

1μm

(a)

Figure 15: Particle deposition. (a) Fractional cumulative

deposition, (b) Local deposition (ΔP = 800Pa).

In Figure 15 the axial fractional cumulative

deposition distribution and the local fractional

deposition for ΔP = 800Pa are shown. The results

indicate significant differences in the deposition

patterns with particle size with most deposition

occurring in the dispersion chamber and the helical

region.

AirflowandParticleDepositioninaDryPowderInhaler-ACFDSimulation

255

Figure 16: Particle Deposition – Effect of Pressure drop.

(a) ΔP = 800Pa, (b) ΔP = 5400Pa. D = 1μm. σ = 1.

The spatial distribution of particles deposited on

the DPI walls was visualized using Tecplot. In

Figure 16 particle depositions for two pressure

drops, i.e., 800 and 5400Pa, are shown. It is clear

that the larger pressure drop results in increased

velocities and total particle deposition but also

significantly different particle deposition patterns.

The increased deposition for large pressure drops in

the helical region is caused by the increased

tangential flow in this region.

In Figure 17 the effect of particle size on the

distribution of deposited particles in the DPI device

is shown. Comparing particle sizes of 1 (see Figure

16a), 2 and 5μm (Figure 17) significant differences

in the total deposition as well as the deposition

distribution are observed. The significant particle

deposition that occurs in the mouthpiece region

(which includes the helical region) is actually a

common problem in many commercial DPI devices

where about half the internal deposition occurs (de

Koning et al., 2001).

The results of Figures 16 and 17 can be used to

optimize the design of the DPI. For example, the

helical region of the Turbuhaler could be redesigned

so that smaller radial and tangential velocities

develop leading to decreased particle collisions in

this region.

Figure 17: Particle Deposition – Effect of Particle Size. (a)

D = 2 μm, (b) D = 5 μm. ΔP = 800Pa. σ = 1.

0 5 10 15 20 25 30

Diameter, ì

0.0

0.1

0.2

0.3

Fractional Volume

Fractional Number

Figure 18: Fractional particle number and volume

distrbution. Inset photo 120x80μm.

Figure 18 displays the particle size distribution

of freely flowing powder containing Budesonide

(Pulmicort). The peak in the number distribution is

at D

=2.2μm while for the volume distribution it is

at 4.5μm. It was found that a Rosin Rammler

distribution, f(D), with a shape parameter value of n

= 1 and a mean diameter of D

= 2.2μm, i.e.,

()

D/D

eD/1)D(f

−

(1)

is a good approximation to the distribution depicted

in Figure 18. The injected, escaped and deposited

fractional volume distributions for ΔP = 800Pa are

SIMULTECH2012-2ndInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

256

provided in Figure 19. It is observed that, due to the

size-dependent deposition efficiency, the particle

distribution exiting the device is significantly

different than the injected particle distribution. The

shape of the injected particle size distribution affects

the total number of particles deposited in the DPI

device due to the different number of large particles

which deposit to a larger degree (Table 1).

246

Diameter, ì

0.0

0.1

0.2

0.3

0.4

Fractional Volume

Injected:

Deposited:

Escaped:

Figure 19: Particle Deposition. (a) Fractional Cumulative

Deposition, (b) Local.Deposition (ΔP = 800Pa).

Table 1: Particle deposition. Effect of injected particle

distribution. (RR Log = Rosin Rammler logarithmic)

Number

Injected

Number

Deposited

Single-size 248 195 78.6

RR Log 2480 2059 83.0

RR Log 4712 3970 84.2

RR Linear 18848 17981 95.4

2.3 Comparison to Experimental Data

Figure 20: Volumetric flow in the Turbuhaler.

The computational results of this work were

compared to the experimental results of de Koning

et al (2001) and Abdelrahim (2010) for the

Turbuhaler in terms of flow and particle deposition.

In Figure 20 the predicted steady-state

volumetric flows are plotted against the outlet

pressure drop applied at the mouthpiece. Both

laminar and k-ω SST models for flow are examined.

It is clear that both models agree very well with the

experimental data for all flow rates with the k-ω SST

model being slightly more accurate.Ιn this work the

capture efficiency is related to a critical normal

velocity, v

, above which particles reflect (assuming

no dissipation of momentum). Τhe developed by

Brach and Dunn (1992). According to this model the

critical normal velocity is.

7/10

⎟

⎠

⎞

⎜

⎝

⎛

(2)

where D is the particle diameter and the effective

stiffness parameter E is given by

()

5/2

2/3

kk5

51.0E

⎟

⎠

⎞

⎜

⎝

⎛

+π

(3)

and ks kp are determined by:

ν−

and

ν−

(4)

where ν

and ν

and E

are the Poisson’s ratio

and Young’s modulus of the surface and particle,

respectively.

In the case of lactose particles (ν

=0.4 and

=1.0GPa) colliding with polystyrene surfaces

(ν

=0.35 and E

=4.1GPa) the critical velocity was

determined to be v

= 2.7m/s.

In Figure 21 the total, dispersion chamber, and

mouthpiece particle depositions for 1400Pa (or 30

l/min) are compared to the experimental data of de

Koning et al. (2001). This critical velocity value

results in an overall capture efficiency of ~42.5%,

the mouthpiece, dispersion chamber, and total

particle deposition results for Q = 30 l/min are in

good agreement to the experimental data. It should

be noted that a 100% capture efficiency leads to very

large total deposition values, i.e., 75%, for this flow

rate (see Figure 14) and even larger for larger flow

rates, e.g., Q>30 l/min.

In Figure 22 the predicted total particle

deposition are compared to the experimental data of

de Koning et al (2001) and Abdelrahim (2010) for

flowrates Q = 30, 40, 50, 60 and 70 l/min and for

two different inspired volumes, i.e., 2 and 4l

AirflowandParticleDepositioninaDryPowderInhaler-ACFDSimulation

257

(Abdelrahim, 2010). For a critical velocity of v

2.7m/s and a particle diameter of D = 2μm the

agreement with the experimental data is good

considering the different experimental conditions

(e.g., dynamic inhalation vs. steady state

simulations) and the simplicity of the particle

deposition model (e.g., velocity cut-off capture

efficiency and single-size size distribution).

Different values of v

are also shown to provide an

indication of the sensitivity of particle deposition on

the value of v

Figure 21: Regional particle deposition in the Turbuhaler.

Q = 30 l/min. v

= 2.7m/s.

30 40 50 60 70

Flow Rate, l/min

% Total Deposition

: CFD Results

: de Koning et al. (2001)

: Abdelrahim (2010) 2l

: Abdelrahim (2010) 4l

2.7m/s

= 5.0m/s

1.0m/s

Figure 22: Total particle deposition in the Turbuhaler. D =

2 μm. V

= 2.7 m/s. Comparison between experimental

results of de Koning et al. (2001), Abdelrahim (2010) and

computational CFD results.

3 CONCLUSIONS

This work has demonstrated the use of CFD to

determine the complicated airflow as well as particle

motion and deposition in the Turbuhaler DPI. As the

flow was either locally laminar or transitionally

turbulent the transitional SST k-ω model for

turbulence was employed. LES results revealed

some differences in the large eddies and secondary

flows but were otherwise closest to the k-ω SST

results. The simulations revealed complicated flows

with intense recirculation patterns in the dispersion

chamber and strong tangential flows in the helical

region of the mouthpiece.

Particle deposition was found to depend on size

and flow rate and occurred predominantly in the

dispersion chamber and the mouthpiece. The

computational solutions were compared to

experimental data for volumetric flow and regional

deposition of de Koning et al. (2001) and good

agreement was observed for volumetric flow.

Particle deposition data were in agreement to

experimental data only for capture efficiencies less

than 100%. A simple collision model by Brach and

Dunn (1992) was employed to determine the critical

velocity for particle capture, i.e., v

=2.7m/s, which

was found to produce total particle depositions

similar to the experimental values of de Koning et al

(2001) and Abdelrahim (2010).

Future work will involve the simulation of

dynamic inhalations and will elaborate on the

particle collision model. The particle collision model

can be extended by including the effects of particle

properties (e.g., size, shape, and charge), surface

properties (e.g., roughness, charge), as well as

humidity.

REFERENCES

Abdelrahim, M. E., (2010) Emitted dose and lung

deposition of inhaled terbutaline from Turbuhaler at

different conditions, Respiratory Medicine, 104, 682-

689.

Alagusundaram M., N. Deepthi, S. Ramkanth, S.

Angalaparameswari, T. S. M. Saleem, K.

Gnanaprakash, V. S. Thiruvengadarajan, C.

Madhusudhana, C. M. Alagusundaram et al., (2010)

Dry Powder Inhalers - An Overview, Int. J. Res.

Pharm. Sci., 1(1), 34-42.

Ashurst I., A. Malton, D. Prime and B. Sumby, (2000)

Latest advances in the development of dry powder

inhalers, PSTT, 3(7), 246-256.

Brach RM, Dunn P.F (1992). A Mathematical Model of

Impact and Adhesion of Microspheres. Aerosol Sci.

Technol., 23, 51-71.

Calvert G., A. Hassanpour, M. Ghadiri, (2011)

Mechanistic analysis and computer simulation of the

aerodynamic dispersion of loose aggregates, Chemical

Engineering Research and Design, 89, 519–525.

Chan H-K, (2006) Dry powder aerosol drug delivery –

Opportunities for colloid and surface scientists,

Colloids and Surfaces A: Physicochem. Eng. Aspects,

284-285, 50-55.

SIMULTECH2012-2ndInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

258

Coates M. S., D. F. Fletcher, H-K. Chan and J. A. Raper,

(2004) Effect of Design on the Performance of a Dry

Powder Inhaler Using Computational Fluid Dynamics.

Part 1: Grid Structure and Mouthpiece Length, J. of

Pharmaceutical Sciences, 93, 2863–2876.

Coates M. S., H-K Chan, D. F. Fletcher and J. A. Raper,

(2005) Influence of Air Flow on the Performance of a

Dry Powder Inhaler Using Computational and

Experimental Analyses, Pharmaceutical Research,

22(9), 923-932.

Coates M. S., H-K. Chan, D. F. Fletcher and J. A. Raper,

(2006) Effect of Design on the Performance of a Dry

Powder Inhaler Using Computational Fluid Dynamics.

Part 2: Air Inlet Size, J. of Pharmaceutical Sciences,

95(6), 1382-1392.

de Koning J. P., M. R. Visser, G. A. Oelen, A. H. de Boer,

ThW van der Mark, P. M. J. Coenegracht, ThFJ

Tromp and H. W. Frijlink, (2001) Effect of Peak

Inspiratory Flow and Flow Increase Rate on In Vitro

Drug Deposition from Four Dry Powder Inhaler

Devices, in Dry Powder Inhalation: Technical and

Physiological Aspects, Prescribing and Use, Thesis,

Rijksuniversiteit Groningen, Ch. 6., 83-94.

Finlay W., (2001) The Mechanics of Inhaled

Pharmaceutical Aerosols. An Introduction, Academic

Press, London.

French D. L., D. A. Edwards and R. W. Niven, (1996) The

Influence of Formulation on Emission Deaggregation

and Deposition of Dry Powders for Inhalation, J.

Aerosol Sci., 27(5), 769-783.

Hoe S., D. Traini, H-K. Chan, P. M. Young, (2009)

Measuring charge and mass distributions in dry

powder inhalers using the electrical Next Generation

Impactor (eNGI), European J. of Pharmaceutical

Science, 38, 88-94.

Islam N. and E. Gladki, (2008) Dry powder inhalers

(DPIs) - A review of device reliability and innovation,

Int. J. of Pharmaceutics, 360, 1-11.

Ligotke M. W., (2002) Development and characterization

of a dry powder inhaler. In: Dalby RN, Byron PR,

Peart J, Farr SJ, editors. Respiratory drug delivery

VIII, vol. I. Tucson, AZ: Serentec Press Inc., 419–422.

Newman S. P. and W.W. Busse, (2002) Evolution of dry

powder inhaler design, formulation, and performance,

Respir Med., 96(5), 293-304.

Olsson et al.: Patent Application Publication. Pub. No. US

2007/0107721 A1. Pub. Date: May, 17, 2007.

Schuler C., A. Bakshi, D. Tuttle, A. Smith, S. Paboojian,

H. Snyder, D. Rasmussen, A. Clark, (1999) Inhale’s

dry-powder pulmonary drug delivery system:

Challenges to current modeling of gas-solid flows. In:

Proceedings of FEDSM99: 3rd ASME/JSME joint

fluids engineering conference and 1999 ASME Fluids

Engineering summer meeting, FEDSM99-7895.

Tobyn M., J. N. Staniforth, D. Morton, Q. Harmer and M.

E. Newton, (2004) Active and intelligent inhaler

device development, Int. J. of Pharmaceutics 277, 31–

37.

Tong Z. B., R. Y. Yang, K. W. Chu, A. B. Yu, S. Adi, H-

K. Chan, (2010) Numerical study of the effects of

particle size and polydispersity on the agglomerate

dispersion in a cyclonic flow, Chemical Engineering

Journal 164, 432–441.

Tsima M. P., G. P. Martin, C. Marriott, D. Gardenton and

M. Yianneskis, (1994) Drug delivery to the respiratory

tract using dry powder inhaler, Int. J. of

Pharmaceutics, 101, 1-13.

Wetterlin K., (1988) Turbuhaler: A New Powder Inhaler

for Administration of Drugs to the Airways.

Pharmaceutical Research, 5(8), 506-508.

Zeng X-M, G. P. Martin, C. Marriott and J. Pritchard,

(2000) The influence of carrier morphology on drug

delivery by dry powder inhalers, Int. J. of

Pharmaceutics, 200, 93–106.

AirflowandParticleDepositioninaDryPowderInhaler-ACFDSimulation

259