30 years on the Road To Progressively Better Data Tire Pavement Interaction Noise and Correlation with Pavement Texture Parameters Presentation by Dr. Ricardo Burdisso 1 Lucas Spies 1 ; Sterling McBride 2 ; Ricardo Burdisso 3 ; Corina Sandu 4 and Vincent Bongioanni 5 1 [email protected]; 2 [email protected]; 3 [email protected]; 4 [email protected]; 5 [email protected]
39
Embed
Tire Pavement Interaction Noise and Correlation with ... Burdisso.pdf · • TPN is produced only by the tread pattern. • NTPN is mainly produced by the pavement (independent of
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
30 years on the Road To Progressively Better Data
Tire Pavement Interaction Noise and Correlation with Pavement
Texture ParametersPresentation by
Dr. Ricardo Burdisso
1
Lucas Spies1; Sterling McBride2; Ricardo Burdisso3; Corina Sandu4 and Vincent Bongioanni5
• Tread and Non-tread pattern noise• Relationship between pavement profile and noise
• Discussions
9
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 10
• Noise spectrogram from acceleration test (Tire 12, 45 to 65 mph).
Experiments results: Tread and non-tread pattern Noise
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 11
• Noise spectrogram from acceleration test (Tire 12, 45 to 65 mph).
Experiments results: Tread and non-tread pattern Noise
Noise component a function of vehicle speed: amplitude and frequency increases with speed.
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 12
• Noise spectrogram from acceleration test (Tire 12, 45 to 65 mph).
Experiments results: Tread and non-tread pattern Noise
Noise component with frequency content independent of speed. However, amplitude increases with speed.
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 13
• TPIN can be separated into two components: tread (TPN) and non-tread pattern (NTPN) noise
Experiments results: Tread and non-tread pattern Noise
Order tracking analysis allows to extract the tread pattern noise from the total noise signal
Optical signal used to perform order tracking analysis.
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 14
• Extraction of TPN noise:
1 revolution of the tire (window).
Experiments results: Tread and non-tread pattern NoiseOrder tracking analysis:For each window• Noise signal resampled • Compute DFT• Average DFTs (TPN in
frequency domain)• Take inverse DFT of average
DFT (TPN in time domain)• Subtract TPN signal from total
signal (NTPN in time domain)
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
65
70
75
80
85
90
A-w
eigh
ted
SP
L [d
BA
re 2
0e-6
Pa]
A-weighted Sound Pressure Level (Frequency of Interest)
• Tread and Non-tread pattern noise• Relationship between pavement profile and noise
• Discussions
25
30 years on the Road To Progressively Better Data
Discussions
26
• A large number of tire noise data was collected using an OBSI system with an optical sensor (for order tracking analysis) under multiple testing conditions.
• Pavement profile data was acquired using a scanning laser.• Tire noise was separated into two main components: tread (TPN) and non-tread-
pattern (NTPN) noise• TPN is due only the tread pattern• NTPN is mainly a function of pavement.
• The NTPN spectrum is correlated to the pavement profile spectrum only over a limited frequency range (~ 200 to 900 Hz).
30 years on the Road To Progressively Better Data
Extras
27
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 28
US460 road data – 60 mph – Tonal noise associated with the pavement distress.
• On the US460 road test section, there was a segment (~300 m) where the non-tread pattern noise component showed tonal components.
• It was speculated that the tonal noise was due to corrugation of the pavement. However, visual inspection of the pavement didn’t revealed these corrugations.
TPIN vs pavement transverse corrugation distress
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
0
0.1
0.2
0.3
0.4
A-w
eigh
ted
p2 rm
s [P
a2
]
A-weighted Power Spectrum of Sound Pressure (Frequency of Interest)
Tone section
Normal section
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 29
US460 road data – 60 mph – Tonal noise associated with the pavement distress.Spectrograms shown are for tire 20 (i.e. SRTT tire). It is important to highlight that the similar behavior was observed in all 5 tested tires.
US460 road - Tire 20 - Total Noise Spectrogram - Tone section
1 1 2 2 3 3 4 4 5
Time in seconds
1280
Freq
uenc
y (H
z)
10
15
20
25
30
35
Pow
er/fr
eque
ncy
(dB
/rad/
sam
ple)
US460 road - Tire 20 - Total Noise Spectrogram - Normal section
1 2 3 4 5 6 7 8 9
Time in seconds
1280
Freq
uenc
y (H
z)
10
15
20
25
30
35
Pow
er/fr
eque
ncy
(dB
/rad/
sam
ple)
TPIN vs pavement transverse corrugation distress
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 30
The non-tread pattern noise can be used to estimate the wavelength of the corrugated pavement.
Non- Tread pattern noise
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
0
0.1
0.2
0.3
0.4
A-w
eigh
ted
p2 rm
s [P
a2
]
A-weighted Power Spectrum of Sound Pressure (Frequency of Interest)
Tone section
Normal section
120 Hz
13.1 2.1 230120
rot tire tirecorrugation
tone
f l Hz m mmf Hz
λ − ⋅ ⋅= = ≅
: Circumference of the tire
: Rotational speed of the tire in Hz
: Frequency interval between noise tones
tire
rot tire
tone
lff
−
TPIN vs pavement transverse corrugation distress
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Wavelength (m)
25
30
35
40
Am
plitu
de (d
B) -
Ref
1m
m
Spectrum - Decibels scale
Normal section
Tone section
The corrugated pavement wavelength was confirmed from direct measurements of the pavement profile
and the computation of the spectrum.
Tone with ~ 230 mm wavelength, which matches the wavelength predicted by the noise data
Non-tread pattern TIPN can be used for efficienlty monitoring pavement distress
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 31
US460 road data – 60 mph – Tonal noise associated with the pavement distress.
Results shown are for tire 20 (i.e. SRTT tire). It is important to highlight that the similar behavior was observed in all 5 tested tires.
Total tire noise Non- Tread pattern noiseTread pattern noise
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
0
0.1
0.2
0.3
0.4
A-w
eigh
ted
p2 rm
s [P
a2
]
A-weighted Power Spectrum of Sound Pressure (Frequency of Interest)
Tone section
Normal section
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
0
0.05
0.1
0.15
0.2
A-w
eigh
ted
p2 rm
s [P
a2
]
A-weighted Power Spectrum of Sound Pressure (Frequency of Interest)
Tone section
Normal section
200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency [Hz]
0
0.1
0.2
0.3
0.4
A-w
eigh
ted
p2 rm
s [P
a2
]
A-weighted Power Spectrum of Sound Pressure (Frequency of Interest)
Tone section
Normal section
120 Hz
TPIN vs pavement transverse corrugation distress
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 32
Experiments• VT SMART road test
Test information
Number of pavements 26
Among the tested pavement there were:• 14 surface mixes asphalt sections• 8 concrete sections• 3 bridges sections• 1 Open Graded Friction Course• 1 concrete section with longitudinal grooves• 7 concrete sections with transverse grooves
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data
• There are 26 different pavements.
9/23/2018 33
Extend TPIN model to include pavement parameters.
Monthly Project Update – Confidential & Proprietary to CenTiRe
SMART Road pavement data.
S.R. Bridge
PCC 2
RR Bridge
PCC 1g
PCC 1f
PCC 1e
PCC 1d
PCC 1c
PCC 1b
PCC 1a
L1
KI
JHF
HWY. Bridge
D2
D1
C
B
A E1
E2 G
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 34
• Overall A-weighted sound pressure level for all tires (dBA)
• For the pavement tested, the tread pattern noise is not the dominant noise source.
• For a newer/smoother pavement (very limited data), the tread pattern noise component account for about 50% of the total noise (Tire 12).
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data
Non Tread Pattern Noise spectrogram - SMART road - Run # 4 @60mph
10 20 30 40 50 60 70 80 90
Time in seconds
1280
2560
Freq
uenc
y (H
z)
10
15
20
25
30
35
Pow
er/fr
eque
ncy
(dB
/rad/
sam
ple)
36
Experimental resultsSMART road data – 60 mph – Tonal noise associated with transverse grooves.
Non tread pattern noise spectrogram
HWY bridgePCC 1-a, PCC 1-c, PCC 1-e and PCC 1-gRR bridgeS.R. bridgeThe NTPN spectrogram shows tonal content at certain intervals, considered to be associated with the presence of transversal grooves on the pavement.
24
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data
Non Tread Pattern Noise spectrogram - SMART road - Run # 4 @60mph
10 20 30 40 50 60 70 80 90
Time in seconds
1280
2560
Freq
uenc
y (H
z)
10
15
20
25
30
35
Pow
er/fr
eque
ncy
(dB
/rad/
sam
ple)
37
Experimental resultsSMART road data – 60 mph – Tonal noise associated with transverse grooves.
Non tread pattern noise spectrogram
HWY bridgePCC 1-a, PCC 1-c, PCC 1-e and PCC 1-gRR bridgeS.R. bridgeThe NTPN spectrogram shows tonal content at certain intervals, considered to be associated with the presence of transversal grooves on the pavement.
13.1 2.1 201378
rot tire tiregrooves
tone
f l Hz m mmf Hz
λ − ⋅ ⋅= = ≅
24
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 38
Experimental resultsSMART road data – 60 mph – Tonal noise associated with transverse grooves.The pavement data spectrogram is also computed. The first tone appears at a similar frequency as in the NTPN spectrogram
SMART road pavement profile spectrogram
S.R
. Brid
ge PC
C 2
RR
Brid
ge
PC
C 1
g
PC
C 1
f
PC
C 1
eP
CC
1d
PC
C 1
c
PC
C 1
bP
CC
1a
L2 L1
K J I
H G F
E2
E1
HW
Y b
ridge
D2
D1 C B A
Distance in meters
0
1333
2667
4000
5333
Freq
uenc
y (H
z)
-5
0
5
10
15
20
25
Pow
er/fr
eque
ncy
(dB
/rad/
sam
ple)
30 years on the Road To Progressively Better Data30 years on the Road To Progressively Better Data 39
Experimental resultsSMART road data – 60 mph – Tonal noise associated with transverse grooves.The pavement data spectrogram is also computed. The first tone appears at a similar frequency as in the NTPN spectrogram.