Air Cleaner Systems

3rd EACC, July 05/06, 2007, Frankfurt
Transient Dust Load Simulation for Air Cleaner Systems
Dr. Bernhard Huurdeman, MANN+HUMMEL GMBH
Contents
Steady state flow simulation of air cleaners Modeling of pleated filter elements Simulation method for transient dust load Examples Application of the new Fluent 3-Volume Method to compact filter elements Summary
refer to protection notice ISO 16016
MANN+HUMMEL Automotive Products

Plastic components for air induction systems air intake manifolds air cleaner turbo ducts cylinder head covers Air and liquid filtration systems Crankcase ventilation
clean air side
pleated filter element
dirty air side

3
Flow Simulation of Air Cleaners

Steady state 3-D flow simulation Accurate geometry representation Calculation of pressure loss Improve geometry where possible clean air pipe clean air housing
resonator
dirty air pipe
dirty air housing incl. filter element filter element as anisotropic porous media
Simulation Results for Unloaded Filter Element

0.7
static pressure [kPa]

0.6
system pressure loss

measurements: 1.85 kPa simulation: 1.80 kPa (filter element: 0.33 kPa)
static pressure difference [kPa]
measurements
0.5
simulation
0.4
0.3
0.2
0.1
Point 6
0
fresh air pipe 1
fresh air pipe 2
fresh air pipe 3
fresh air housing
4 element 5
filter
clean air housing
clean air pipe
Point 4 Point 2 Point 1

Point 5
Point 3
Modeling of Filter Element

Exact geometry of all element pleats can not be simulated Filter element as anisotropic porous media Constant porous media coefficients dependent on pleat size and media Pressure loss dependent on viscous loss and inertia loss coefficients C1 and C2
p / b = C1 v + C 2 v
2
C1 and C2 can be determined by measurements or by simulation

6
Determination of Porosity Coefficients

2-D flow simulation of pleat geometry Result is C1 and C2 in main flow direction C1 and C2 parallel and perpendicular to element pleats can be determined by simple analytical formulas Standardized simulation using Web-application tool (EASA)
grid
velocity
filter media as isotropic porous material
inlet
main flow direction
outlet
Dust Load Measurements of Flat Filter Media

Pressure loss of flat filter media depends on dust loading This information is required for each combination of dust and filter media
45
increase in pressure loss [mbar]
40 35 30 25 20 15 10 5 0 0 5 10 15 20 25 30 added mass of dust ISO fine [g]
Cdust = mdust / Amedia

pmedia = f (Cdust )
2-D Simulation of Dust Loaded Pleats

Simulation of pressure loss of the filter element with increasing amount of dust, by increasing the filter media pressure loss. Simplification: Element pleat is loaded homogenously Result: C1 and C2 as a
total pressure loss [Pa]
600.0
EASA Web-Tool
Total pressure loss pleat

500.0
400.0
1500.0
function of separated dust
p pleat p medium
Total pressure loss medium
2500.0
7000.0

2000.0
6000.0
5000.0
4000.0
300.0
3000.0
1000.0
C1, C 2 = f (Cdust )
C1, C2 for Fluent 3.0E+09 2.5E+09
200.0
2000.0 500.0
100.0
1000.0
0.0
0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
volume flow [m/min]
volume flow [m/min]
volume flow [m/min]
300
C1 [1/m2]
1.5E+09 1.0E+09 5.0E+08 0.0E+00
150 100 50 0
C2 [1/m]
2.0E+09
C1 C2
250 200
& pelement = f (pmedia , mair )
dust ISO 16016 refer to protection notice concentration [kg/m]
Dust Load Distribution on the Filter Element using Particle Tracking for different Particle Diameter
1 m 3 m 10 m
4.50 4.00
30 m
ISO fine dust is modeled using four different particle sizes
3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 0.100
PTI fine dust
volume [%]
1.000
10.000 particle size [m]
100.000
1 m 3 m 10 m 30 m particle size: mass vol %: 20 % 32 % 24 % 24 %

10
3-D Simulation Method for Element Loading
The element surface is divided into approx. 15x15 rectangles. Then, cuboids are defined between inlet and outlet face. Amount of dust in each cuboid is computed using a particle tracking method Porous media coefficients are determined from the interpolated local amount of mass of dust
C1 [1/m2]
C1, C2 for Fluent 3.0E+09 2.5E+09 2.0E+09 1.5E+09 1.0E+09 5.0E+08 0.0E+00 300
rectangle cuboid
C1 C2
for every 10th time step. Then update of the porous media coefficients See also SAE paper 2006-01-1316
150 100 50 0
dust concentration [kg/m]
C1 , C2 = f (Cdust )
11
C2 [1/m]
Transient Simulation with particle tracking
250 200
3-D Simulation of Dust Loaded Element

dust distribution in the filter element with 10x15 cuboids (here: 400 tets per cuboid). interpolated dust distribution
cuboid
C1 interpolated distribution of C1 and C2 (good convergence behaviour)
C2
12
Example: Dust Distribution at End of Simulation
1m
3 m
10 m
30 m
13
Example: Animation of dust and velocity

velocity normal to element accumulated dust
20 18
pressure loss of filter element and air cleaner
pressure loss [mbar]
16 14 12 10 8 6 4 2 0 0
filter element air cleaner
50
100
150
200
250
300
dust load per filter surface area [g/m]
14
Air Cleaner Simulation with Different Types of Dust

Cumulative Volume Three different types of dust with mean
volume sum [%]
100 90 80 70 60 50 40 30 20 10 0 1 10 100
PTI fine SF500 4 m SF300 10 m W6 40 m
particle diameter of 4, 10, and 40 m Measurements of flat filter media and of complete air cleaner
Measured Pressure loss on flat filter media

increase in pressure loss 4 dust
particle size [m]
Simulation Grid
10 dust
40 dust dust concentration
15
Air Cleaner with Different Types of Dust

mass of dust
small particle size large particle size
half loaded
fully loaded
16

small particle size
large particle size
17

25
Good correlation of pressure

20 pressure increase [mbar]
loss increase, but strongly dependent on dust modeling. Velocity near air mass meter helps to evaluate the influence of the dust load during element life time no load
Test 4mu Test 10 mu Test 40 mu Simulation 4 mu Simulation 10 mu Simulation 40 mu
4 m
15
10 m
10
40 m
0 0 20 40 60 80 100 120 140 dust concentration [g/m]
half load
full load Here, a nearly dust load independent velocity distribution shows an uncritical influence of the dust load on the air mass meter signal.
18
Filter Element Modeling for Compact Elements

Instead of pleats, compact filter elements consist of long alternately closed channels The assumption, that the dust will distribute homogenously along the channel, is here not applicable anymore. The usage of an isothermal and dust size dependent version of the Fluent 3-Volume model for DPF is assumed to be an attractive approach. three overlapping volumes with equal grids dirty air clean air dirty air porous media clean air
V1
source terms
V2
solid
Using source terms, the dirty and clean air mass is exchanged between the two volumes.
19
Velocity Distribution and Dust Load with the 3-Volume Model

velocity distribution dirty air volume velocity distribution clean air volume filter element
7% dust load
dust distribution
scale 5 - 15 % 100 % dust load
scale 90 -110 %
20
Summary
A flow simulation method for air cleaner systems has been presented which takes dust loading into account during the lifetime of a filter element. When compared to measurements, simulation produced reasonably accurate dust distribution in filter element surface. The simulation method will be useful to improve air cleaner design regarding the quality of the air mass meter signal during element life time. Furthermore, the dust hold capacity of air cleaner systems can be compared, which will help to find the most suitable design variant. The new 3-volume method developed by Fluent is a promising approach especially for compact filter elements. The possibility to simulate an inhomogeneous dust distribution along the main filter direction, may also improve the dust load simulation method for pleated filter elements.
21

Air Cleaner Systems

Uploaded by

Copyright:

Available Formats

Air Cleaner Systems

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Air Cleaner Systems

Uploaded by

Copyright:

Available Formats

3rd EACC, July 05/06, 2007, Frankfurt

Transient Dust Load Simulation for Air Cleaner Systems

Dr. Bernhard Huurdeman, MANN+HUMMEL GMBH

refer to protection notice ISO 16016

MANN+HUMMEL Automotive Products

clean air side

pleated filter element

dirty air side

refer to protection notice ISO 16016

Flow Simulation of Air Cleaners

dirty air pipe

refer to protection notice ISO 16016

Simulation Results for Unloaded Filter Element

static pressure [kPa]

system pressure loss

static pressure difference [kPa]

fresh air pipe 1

fresh air pipe 2

fresh air pipe 3

fresh air housing

clean air housing

clean air pipe

Point 4 Point 2 Point 1

Modeling of Filter Element

C1 and C2 can be determined by measurements or by simulation

Determination of Porosity Coefficients

main flow direction

refer to protection notice ISO 16016

Dust Load Measurements of Flat Filter Media

increase in pressure loss [mbar]

40 35 30 25 20 15 10 5 0 0 5 10 15 20 25 30 added mass of dust ISO fine [g]

Cdust = mdust / Amedia

refer to protection notice ISO 16016

2-D Simulation of Dust Loaded Pleats

Total pressure loss pleat

total pressure loss [Pa]

total pressure loss [Pa]

function of separated dust

Total pressure loss pleat

Total pressure loss medium

Total pressure loss pleat

Total pressure loss medium

volume flow [m/min]

volume flow [m/min]

volume flow [m/min]

1.5E+09 1.0E+09 5.0E+08 0.0E+00

& pelement = f (pmedia , mair )

dust ISO 16016 refer to protection notice concentration [kg/m]

ISO fine dust is modeled using four different particle sizes

3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 0.100

PTI fine dust

10.000 particle size [m]

1 m 3 m 10 m 30 m particle size: mass vol %: 20 % 32 % 24 % 24 %

3-D Simulation Method for Element Loading

dust concentration [kg/m]

Transient Simulation with particle tracking

3-D Simulation of Dust Loaded Element

C1 interpolated distribution of C1 and C2 (good convergence behaviour)

refer to protection notice ISO 16016

Example: Dust Distribution at End of Simulation