Non-Traditional Noise and Vibration Mitigation … Noise and Vibration Mitigation Strategies ... •A study at BART show rail dampers reduce noise over time ... R² = 0.96 86.0 88.0

Non-Traditional Noise and Vibration Mitigation Strategies

Christopher Layman, Ph.D.Shannon McKennaJudy Rochat, Ph.D.

ATS Consulting

Pasadena, CA

• We can take into account the condition of the rail and vehicles operating on the system– Maintenance (rail grinding and wheel truing)

– Low-noise vehicles

• We can capitalize on elements already included in the project:

– Extending the width of the ballast to reduce noise

– Vibration reduction due to noise barrier foundation

Non-Traditional Mitigation Measures

• There could be noise and vibration reduction measures already included in the project that are not being accounted for.

Key Presentation Take-Aways

• Noise and vibration levels can rise by 10-15 dB for severely corrugated rail or wheels

• Acoustic rail grinding and wheel truing can be used to reduce noise levels

Maintenance as a Mitigation Measure

• Mitigation recommendations will depend on the reference noise level used in the analysis

• Important to document conditions of the existing system, including rail roughness

Maintenance as a Mitigation Measure

System Lmax*, dBA

FTA Reference Level 81

Sacramento 82

San Diego 75

Maintenance was recommended

as a mitigation measure

Less mitigation is required* Lmax for a 2-car train at 50 ft, 40 mph on

ballast-and-tie track

• Maintenance could also be used as a vibration mitigation measure

Maintenance as a Mitigation Measure

Up to 15 dB difference in Force Density Level before and after re-profilingSource: Wilson, Ihrig & AssociatesNorthgate Link Extension Final Design

About 10 dB difference in vibration level before and after grindingSource: Ben Lawrence, proceedings of Acoustics 2004

• Some systems, like San Diego Trolley, have low-noise vehicles

• If you are procuring a vehicle, you can request a low-noise vehicle. Features may include:

– Wheel skirts

– Under-car absorption

– Others?

Specify a Low-Noise Vehicle

• Ballast is a sound absorbing material

• Can use a simple reflection model to determine the region of influence affected by ballast

Extending Ballast

Region of Influence

Source

Absorptive ground

• Can measure effective flow resistivity to quantify absorption

• Can model ground absorption in TNM, CadnaA, or SoundPlan

Extending Ballast: Predicting Effects

Width of ballast

beyond ties (ft)

Estimated extent

of influence*

Reduction*

3 52 ft ~1 dB

5 64 ft ~1.5 dB

10 94 ft ~2 dB

*Assumes receiver is 5 ft above ground level

Extending Ballast: Example Project

Source

Extra Ballast

Existing noise (dBA)

Allowable increase due to

project (dB)

Predicted noise (dBA)

Predicted increase (dB)

Predicted increase including ballast

effect (dB)

58 2.4 61.8 3.8 2.3*

*Assumes 1.5 dB decrease due to ballast

Ballast with Berm

• Ballast with small berm next to tracks can provide 10 dB reduction (Attenborough, Inter-Noise 2005)

Note: 38 cm = 1.2 ft, 110 cm = 3.6 ft, 530 cm = 17.4 ft

• Train vibrations measured on BART at the Walnut Creek/Pleasant Hill region.

• Wayside vibration measured with (2012) and without (2005) sound wall in place.

Vibration Mitigation with Sound Walls

B&T Track on

Embankment

• Measured Train Vibration at Two Similar Locations

Difference in Measured Vibration

-10.0

-5.0

0.0

5.0

10.0

15.0

20.0

6.3 8 10 12.5 16 20 25 31.5 40 50 63 80 100 125 160

dB

1/3 Octave Band Center Frequency, Hz

Site 3/5

Site 4/6

Sound Wall Insertion Loss

Site 4/5

Site 3/6

Stiff Foundations Block Waves[…usually]

What Have Others Found?

Jet Grouting Wall

cement–bentonite

mixtures

Track

El Realengo, Spain

Measured Insertion Loss (Red Curve)

Soil Dynamics and Earthquake Engineering 77 (2015) 238–253

50 ft

10m 14m

24m 32m

• Modeling Approach

– 3D Finite Element – Solid Mechanics

– Ignores the track and moving vehicle

– Random point sources are located at axle locations of a typical 3-car consists.

– Frequency domain – linear analysis

Can We Simulate Our Measurements?

Modeling Geometry and Details

PML (grey regions)

Soil

Concrete

Sound Wall

Soil

Probe

This entire face

is Symmetry

Vertical Insertion Loss

20*log10(|u_zref|/|u_zwall|)

• Moderately Stiff Soil

• Geometry and mesh scales with

wavelength

• Hysteretic damping throughout

• Given measurement locations and

frequencies of interest, model is limited

to f <= 70 Hz

Insertion Loss Results

Data

Model

(115ft)

• Model qualitatively agrees with measurements.

• Agrees better with farther measurement(?)

• Does embankment area have softer soil?

• Is there a coincidence frequency effect?

Spatial Patterns of Vertical Displacement

Wall+Embankment

Embankment

Sources

50 Hz

• We don’t see significant amplification.

• General trend: moderate distances from

track have highest IL

Sensitivity To Wall and Embankment Geometry

• Reducing Embankment Height

• Embankment has little effect on IL

• Reducing Foundation Depth

• Stronger dependence on foundation depth

• Take into account the noise and vibration conditions of the existing system during environmental assessments

• Identify any noise and vibration reduction measures already included in the project:

– Extending the width of ballast could provide a noise benefit

– Properly designed sound walls may reduce vibration impacts significantly. However, the optimal range appears narrow.

Conclusions

THANK YOU!

END

Modeling Parameters

• Noise and vibration can be mitigated at the:

– Source

– Path

– Receiver

Background

Noise propagation through air:

Vibration propagation through soil:

• Rail dampers are tuned to absorb specific vibration frequencies which reduces the amount of noise radiated by the rail

• Measurements at SacRT show reduction in wayside noise levels by 2-3 dB

• Rail dampers can be used where sound walls are not feasible

Rail Dampers

• A study at BART show rail dampers reduce noise over time and limit corrugation growth

• Rail dampers are easy to add to existing track

Rail Dampers

y = 1.49x + 88.2

y = 0.25x + 86.5

R² = 0.96

86.0

88.0

90.0

92.0

94.0

96.0

98.0

100.0

102.0

104.0

0 10 20 30

12

th O

Ct.

Co

rru

ga

tio

n (

dB

A)

Months after Grinding

Growth of Corrugation Noise around

450 Hz

No Rail Dampers

Rail Dampers

Linear (No Rail

Dampers)

Linear (Rail Dampers)

Date Difference*

Apr 2013 5.8

Oct 2013 2.7

Apr 2014 1.7

Sept 2014 3.0

Apr 2015 3.3

Oct 2015 2.1

* Difference in noise level for a track section

without dampers minus a section with dampers