Post on 03-Oct-2021
Purdue University Purdue University
Purdue e-Pubs Purdue e-Pubs
Publications of the Ray W. Herrick Laboratories School of Mechanical Engineering
7-2020
Noise Source Identification and Noise Directivity Analysis of Noise Source Identification and Noise Directivity Analysis of
Bladeless Fans by Combined CFD and CAA Method Bladeless Fans by Combined CFD and CAA Method
Ang Li Purdue University, li2467@purdue.edu
Jun Chen Purdue University
Yangfan Liu Purdue University
J Stuart Bolton Purdue University, bolton@purdue.edu
Patricia Davies Purdue University
Follow this and additional works at: https://docs.lib.purdue.edu/herrick
Li, Ang; Chen, Jun; Liu, Yangfan; Bolton, J Stuart; and Davies, Patricia, "Noise Source Identification and Noise Directivity Analysis of Bladeless Fans by Combined CFD and CAA Method" (2020). Publications of the Ray W. Herrick Laboratories. Paper 232. https://docs.lib.purdue.edu/herrick/232
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for additional information.
Influence of Geometric Parameters on Aerodynamic
and Aeroacoustic Performance of Bladeless Fans
Ang Li, Jun Chen, Yangfan Liu,
Stuart Bolton, Patricia Davies
Ray W. Herrick Laboratories
Purdue University
West Lafayette, 47907, USA
August 1st, 2019
AJKFLUIDS 2019-5220 07.28-08.01 San Francisco, CA, USA
OUTLINE
⚫ Background & Motivation
⚫ Methodology
➢ Experiments
➢ Numerical Simulations
⚫ Results
⚫ Conclusions
2AJKFLUIDS 2019-522007.28 - 08.01, 2019, San Francisco, CA, USA
BACKGROUND: FANS IN INDUSTRY
Vehicle Engine CPU Radiator Fan
HVAC Cooling Fan
Applications
– Cooling system
– Ventilation
– Thermal comfort
Features
– High flow rate
– Low noise level
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 3
Wind channel
airfoil
... ,,.,.,.~ impeller
Step2-0utlet ai,·
" r
St~pS- lnM & Outlet ai,·
r Base
Step3- Suctioned air Step4- Eotraimuent
©
BLADELESS FAN
• Bladeless fans launched by Dyson
Working mechanism of the bladeless fan
Advantages Jafari et al. (2015)
– The produced wind is softer and more uniform.
– Flow rate at downstream is larger.
– No visible rotating blade is safer for children.
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 4
AN EXAMPLE OF THE BLADELESS FAN
– UK, Dyson. “Dyson Cool Fans - Air Multiplier Technology Explained - Official Video.” YouTube, YouTube, 5 Mar. 2014, www.youtube.com/watch?v=bUJ-X1rsKV4.
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 5
PREVIOUS WORK
Flow field structure outside the bladeless fan over the center plane
Li et al. (2016) Li et al. (2014)
Curve 1 Prototype Curve 4
Pressure (Pa)
Curve 2 Curve 3 Curve 5
Pressure distribution near the slit with different Coanda surface curvatures
Jafari et al. (2015) Jafari et al. (2016) Jafari et al. (2016)
. Power Level (d B) Surface Acoustic
38 30 ll IJ 5
(a)
"' :; 0::
"' 0
i.:
" s = 0 > " .:: =
0
2500
OT=2mm OT=3mm
....... ~60;;-- 18100 10 20 4° Flow Rate (L/s) Inlet Volume
400 350 300 250 200 150 100 50 0
-50 -100 -150
Streamlines outside the bladeless fan over the center plane
Sound power contour Effect of the outlet thickness and outlet angle on volume flow rate at downstream
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 6
OBJECTIVE & RESEARCH STRATEGY
▪ Characterize the aerodynamic and aeroacoustic performances of the
bladeless fan by a combined 3D numerical and experimental study
▪ Investigate the influence of geometric parameters of the wind channel
on bladeless fan’s performance
707.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
Bladeless fan Prototype
Experiments Numerical Simulations
c
H
Influence of the geometric parameters of the wind channel
x0
METHODOLOGY: VELOCITY MEASUREMENT AT FAR FIELD
Measurement in Herrick PBE Lab
– Measurement position: 837
– Measurement duration at each position: 30s
– Sampling rate: 1s
– Accuracy: ±(2%+0.03m/s of indicated values)
X velocity contour
3D ultrasonic anemometer 0.8m
@x = 1.5m
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
yz
x
x = 1.5m
X velocity component(m/s)
3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
8
--
• • -+ -+ -+
METHODOLOGY: SOUND PRESSURE MEASUREMENT AT RECEIVERS
Measurement in Herrick anechoic chamber
x = 1.5m
Receiver 1
z = 0.8m
z
x = 0.1m
Receiver 2
B&K intensity probe
Sound pressure level @ two receivers
x
70 cu a.. ::i.
0 -- Receiver 1 N 60 II -- Receiver 2 -~
50 Cl.
Ill "O 40 -Q) > Q) 30 Q) "-::J Cf) 20 Cf) Q) "-Cl.
"O 10 C ::J 0
Cf) 0 0 1000 2000 3000 4000 5000
Frequency (Hz)
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 9
2m
Mass flow rate ---
PROTOTYPE OF THE BLADELESS FAN
H
x0
c
d=2mm, H=3cm, c=12cm, x0/c=10%
Cross-section of the wind channel
Qinlet
= 0.11kg/s
Computational domain
Temperature: 25℃~ 30℃ Reynolds number at the intake and the slit: 𝐑𝐞𝐢𝐧𝐭𝐚𝐤𝐞 = 𝟔𝟕𝟐𝟑,𝐑𝐞𝐬𝐥𝐢𝐭 = 𝟑𝟎𝟎𝟎~𝟒𝟏𝟎𝟎
Blocks at slit
10cm
Wind channel
Simulation Set-up
– The number of grids: 6,320,000
– Steady RANS: 𝒌 − 𝜺 model
– LES: Smagorinsky-Lilly model
– Time step: 𝟏 × 𝟏𝟎−𝟒 s
– Flow solver: SIMPLE
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 10
~· ,--,◄ .. .,...___ :
I I I I I I I I I I
~
x = 1.5m
x y
z z
centerline
2.0
1.5
N
= 9,680,000
- '1 2 3
W ise velocity (m/s) Stream
4
MESH GENERATION OF THE BLADELESS FAN
– Mesh independency test
• Mesh for the computational domain • Mesh for the cross-section of the wind channel
𝒚+ = 𝟎. 𝟖𝟓
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 11
METHODOLOGY: DATA POSTPROCESSING
Power spectral density of sound
pressure
Sound pressure
level
Acoustics Analysis
(Solver: Matlab)
Aeroacoustic
results
Unsteady Flow
Simulation + BC & IC
(Solver: OpenFoam)
'( , )p tx
(FW-H analogy)
Aeroadynamic results
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 12
X Velocity
5 4.8 4.6 4.4 4.2 4 3.8 3.6 3.4 3.2 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
RESULTS: INSTANTANEOUS FLOW FIELD
z = 0.8m ⚫ X velocity over Z planes
⚫ LES, t = 5s to 15s
z = 0.55m z = 0.3m
1307.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
Mean X Velocity: o 0.2 0.4 o.6 0.0 1 1.2 1.4 1.6 1.0 2 2.2 2.4 2.6 2.0 3 3.2 3.4
Ill
-· I I ...... I I ...... I I I ...... I I I
~
1.4
1.2
1.0
Io.a N
0.6
-Exp. ------ RANS
· - · LES
~
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
X velocity (m/s)
@ X=1.5 m
LES, time-averaged RANS Exp.
(t = 4s to 15s)
x = 1.5m
x ycenterline
1407.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
@ z = 0.8m
@ the center plane of the bladeless fan
z z
z
AERODYNAMIC CHARACTERISTICS: MEAN FLOW
r
' \
\ \
3 2 1 0
-1 -2 -3 -4 -5 -6 -7 -8 -9 -10
AERODYNAMIC CHARACTERISTICS: MEAN PRESSURE
z = 0.8m ⚫ Pressure over Z planes
⚫ LES, averaged from t=4s to 15s
Pressure (Pa)
z = 0.55m z = 0.3m
1507.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
~
X = 0.1m x= 0.5m x= 1.5m
-- die = 1.25% -- die= 1.67% -- die= 2.08% -- die= 2.50%
X ---►
0.4
0.2
E ___. 0.0 >-
-0.2
-0.4
U/ Uinlet 0.0 0.4 0.8 1.2
0 2 3 ux (m/s)
1.6
4
3.34
1.67
(.) 0.00 >=
-1.67
EFFECT OF THE SLIT WIDTH
LES, time averaged for t = 4s to 15s, @ x =1.5 m
Slit Width
c = 12cm
d/c = 1.25% d/c = 1.67% (baseline) d/c = 2.08% d/c = 2.50%
@ z = 0.8m
1607.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
(!! ~
~
~ I
I I I I I :
I f x=0.1m
X velocity component(m/s) o 0.2 o.4 o.6 0.0 1 1.2 1.4 1.6 1.0 2 2.2 2.4 2.6 2.0 3 3.2 3.4
--Hie= 16.7% -- Hie= 25.0%
x= 0.5m x= 1.5m -- Hie= 33.3% --Hie= 41 .7%
X
u / u inlet 0~ 0.4 0.8 1.2
0.4
0.2
_s 0.0
>--0 .2 +---+--
-0.4
0 2 3
u x (m/s)
1.6
4
3.34
1.67
(.) 0.00 >=
-1.67
-3.34
EFFECT OF THE CROSS-SECTION HEIGHT
LES, time averaged for t = 4s to 15s, @ x =1.5 mH = 2cm
H = 3cm
H = 4cm
H = 5cm H/c = 16.7% H/c = 25.0% (baseline) H/c = 33.3% H/c = 41.7%
c = 12cm
@ z = 0.8m
1707.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
X velocity component(m/s) o 0.2 o.4 o.6 0.0 1 1.2 1.4 1.6 1.0 2 2.2 2.4 2.6 2.0 3 3.2 3.4
--x0/c=5% U/ Uinlet --x0/c=1 0% 0.0 0.4 0.8 1.2 1.6
--x0/c=15% ' X = 1.5m 0.4 ' 3.34 ,-X = 0.1m x= 0.5m --x0/c=20%
0.2 1.67 _,,,,. ,....._ E '- (..) - 0.0 \\ 0.00 ;;::: >- ,.,··
X -0.2 -1.67
---►
-0.4 -3.34
0 2 3 4
ux (m/s)
EFFECT OF THE SLIT LOCATION
LES, time averaged for t = 4s to 15s, @ x =1.5 m
𝒙𝟎/c = 5% 𝒙𝟎/c = 10% (baseline) 𝒙𝟎/c = 15% 𝒙𝟎/c = 20%
c = 12cm
@ z = 0.8m
1807.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
x0 = 0.6cm
x0 = 1.2cm
x0 = 1.8cm
x0 = 2.4cm
~
x=0.1m
X velocity component(m/s) o 0.2 o.4 o.6 0.0 1 1.2 1.4 1.6 1.0 2 2.2 2.4 2.6 2.0 3 3.2 3.4
- NACA0015 - Baseline
x= 0.5m x=15m - CLARK YM-15 - EPPLER478 0.4
0.2
-----S 0.0 >-
X ---►
-0.2
-0.4
0.0
I " I ·,
U/ Uinlet 0.4 0.8 1 2
·- . ----
1.6
t
· , , -...: !,..__
+ , - \ ::,. ✓. - : ::--·
3.34
1.67
(.) 0 .00 >=
---· ---- -1 .67 .,, I
----+------+------< -3.34
0 2 3 4
ux (m/s)
EFFECT OF THE PROFILE OF CROSS-SECTION
NACA 0015 LES, time averaged for t = 4s to 15s, @ x =1.5 m
CLARK YM-15
EPPLER 478
Baseline NACA0015 CLARK YM-15 EPPLER 478 Baseline
@ z = 0.8m
1907.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
' ' ' -
- i Qtotal
5
14 M-1T -e- M2 4
.---.. .---.. + ~ 12 .._, ~
~ 3~
::f 10 ~-•T! 2
8 +--:,~,-,---,-~,--..----,--~-~_J 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.~
die(%)
5
+ 14 ~
4
~12 0
~ -:-----· 3 ~
10 /I ~ 2 I
8 1 5 10 15 20
x0/c (%)
.---.. ~ 0 .._, ~N
,----,----,--------.--- 5
14 :_~ -,,.-------+-J---
- - M2 4
.---.. ~ 12 0 .._,
~ 10 +t 2
8+---,----~----_J 1 20 30 40
Hie(%)
5 14 - M1
12 - M2 4
.---.. ~ 8 0 .._,
3 .---.. ~ 0
~ 6 2 ? 4
1 2
0 CLARK YM-15 EPPLER 478 Baseline
Profile of cross-section
RATIO OF MASS FLOW RATE
Effect of slit width Effect of cross-section height
Effect of slit location
𝑄𝑡𝑜𝑡𝑎𝑙 𝑀1 =
𝑄𝑖𝑛𝑙𝑒𝑡
𝑄𝑏𝑎𝑐𝑘 𝑀2 =
𝑄𝑖𝑛𝑙𝑒𝑡
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
Effect of profile of the cross-section
20
~ 70 ro a.. ::t
0 60 N .. • • II
I!! 50 a.
co :s, 40 .. Q) > ~ 30 ~ :::J .. rJ) 20 rJ) Q)
c.. "O 10 C :::J 0 0 Cl)
~ 70 ro a.. ::t
0 60 N
II
J 50
co :s, 40 Q) > ~ 30 Q) .... :::J rJ) 20 rJ) Q) .... a.
"O 10 C :::J 0
Cl)
0
Experimental result Numerical result
- - Boundary of 95%-confidence interval
1000 2000 3000 4000 5000
Frequency(Hz)
-- Experimental result -- Numerical result
- - Boundary of 95%-confidence interval
1000 2000 3000 4000
Frequency (Hz)
5000
AEROACOUSTIC CHARACTERISTICS
x = 1.5m
Receiver 1
z = 0.8m
z
x
x = 0.1m
Receiver 2
Sound pressure level @ x = 0.1m
@ x = 1.5m
– Acoustic model: FW – H model
– Density: 𝟏. 𝟐𝟐𝟓𝐤𝐠/𝐦𝟑
– Sound speed: 𝟑𝟒𝟎𝐦/𝐬 – Reference acoustic pressure: 𝟐𝟎𝛍𝐏𝐚 – Noise source: Bladeless fan
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 21
~ I /
~ =====================~
•
ro 30 a.. :::1.
0 <;'c_ 20
!"
~ 10 ~ :2, 0 ai ~ -10 Q)
~ -20 (/)
~-30 "O
§ -40 0
Cf)
ro 30 a..
:::1.
~ 20 IL
!"
~ 10 <( [D
:2, 0 ai ~ -10 l!? ~ -20 (/)
l!? ~-30 C ~
0
-- dlc=1 .25% --d/c=1.67% -- d/c=2.08% -- d/c=2.50%
1000 2000 3000 4000 5000
Frequency (Hz)
--+------4-----< -- x0 /c=5% --x0 /c=10%
--x0 /c=15%
--x0 /c=20%
~-40+------,----+----+---~--o 1000 2000 3000 4000 5000
Frequency (Hz)
H/c=16.7% -- H/c=25.0%
H/c=33.3% H/c=41 .7%
1000 2000 3000 4000 5000
Frequency (Hz)
ro ~ 20 - :;_"rj~~~:ia;;:~ 7-----t---o N
I~ 10 0.
'
~ -10 ~
~ -20 (/) (/)
~ -30 "O C
-Baseline - NACA0015 - CLARK YM-15 - EPPLER478
i5 -40+------~-----~--(/) 0 1000 2000 3000 4000 5000
Frequency (Hz)
Receiver
z = 0.8m
z
x
EFFECT OF THE GEOMETRIC PARAMETERS ON AEROACOUSTIC PERFORMANCE
H
x0
c
Effect of slit width Effect of cross-section height
d=2mm, H=3cm, c=12cm, x0/c=10%
x = 1.5m Effect of slit location Effect of profile of the cross-section
2207.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220
CONCLUSIONS
— When the wind produced by the bladeless fan becomes more
powerful, the aerodynamic noise is louder.
— With the decrease of the slit width, the wind strength becomes
more powerful. The generated noise increases at the same time.
— The bladeless fan with the cross-section of 4cm has the best
aerodynamic performance, but the generated noise is the
loudest.
— With the slit moves away from the leading edge, both wind
strength and noise level increase.
— The profile of the cross-section affect the shape of the influence
zone, but has insignificant effect on outflow mass flow rate and
the generated noise.
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 23
ONGOING EFFORT
— Investigate the performance of the bladeless fan prototype with
the impeller in the base
— Identify the main noise source
— Analyze the noise directivity of the bladeless fan
— Come up with a general criteria to evaluate the aerodynamic
performance of the bladeless fan (i.e. strength, uniformity and
steadiness of the wind)
— Propose a criteria to evaluate the compromise between the
aerodynamic and aeroacoustic performance
— Apply the results to optimize the design of the new-generation
bladeless fan.
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 24
Glidea
ACKNOWLEDGEMENT
• Thanks to the financial support and professional feedback
provided by Midea Global Innovation Center
07.28 - 08.01, 2019, San Francisco, CA, USA AJKFLUIDS 2019-5220 25