Memorandum

DATE: June 1, 2018

TO: Michelle Julius, AECOM Project Manager
FROM: Herbert Singleton, Cross-Spectrum Acoustics, Inc.
SUBJECT: Calhoun Isles Condominiums Vibration Assessment

Executive Summary
This memorandum documents the methodology, process and vibration analysis conducted at the Calhoun
Isles Condominiums located at 3145-3151 Dean Court in Minneapolis, Minnesota. A segment of the
proposed Southwest Light Rail Transit (SWLRT)/Green Line extension is located next to the Calhoun
Isles Condominiums. In response to concerns that vibration from construction and transit operations
would create significant disruption to the complex, a detailed analysis of the condominium towers and
parking garage adjacent to the SWLRT alignment was performed. The measurements and analysis were
carried out by Cross-Spectrum Acoustics Inc. (CSA) under contract to AECOM.
CSA conducted a series of vibration measurements at the condominium complex during the week of April
2, 2018 to quantify (1) the existing ambient vibration conditions inside the condominium towers and
garage and (2) the vibration propagation characteristics of the building structure. The data collected from
these measurements demonstrate how the structures respond to vibration and were used to refine the
January 2017 SWLRT light rail vehicle (LRV) vibration calculations..
To characterize existing vibration conditions, CSA measured ambient vibration levels inside the structure
at various floors of Tower 1 (3151 Dean Court, the western tower), the Bridge (3145 Dean Court,
spanning between the two towers) and the Garage (located to the southwest of Tower 1). The structures
tested were selected by representative of the Calhoun Isles Condominium Association (CICA) and their
consultant, Itasca Consulting Group, Inc. Measurement locations within the structures were selected in
coordination with Itasca. The ambient vibration measurements demonstrated that existing interior
vibration levels varied from being imperceptible at quiet times to distinctly noticeable during freight train
movements and other typical interior activities such as tenants walking, using elevators, closing doors or
rolling carts. Background vibration levels were higher at the top floors of Tower 1 and the Bridge
compared with levels on the lower floors. These higher levels were likely caused by vibration from
rooftop air handling units. Existing vibration levels at the Garage were the highest of all the structures.
Existing vibration sources at the Garage included interior automobile traffic and exterior freight train
traffic.
CSA also conducted transfer mobility measurements at these three structures. Transfer mobility is the
relationship between the force generated at a source position and the resulting vibration measured at
another position. Transfer mobility describes how vibration increases or decreases as it propagates
through the ground and into the structure. The transfer mobility measurements performed at the
condominiums were used to directly measure the increase or decrease in vibration levels as vibration
waves enter the building at the foundation and travel through the structure up to higher floors.

A service of the Metropolitan Council

www.swlrt.org

6465 Wayzata Boulevard, Suite 500 • St. Louis Park, MN 55426 • Main: 612-373-3800 • Fax: 612-373-3899
Page 2 of 65

The transfer mobility measurements indicated that vibration levels at most frequencies decrease as
vibration enters and propagates upward in the building. A slight amplification at some lower frequencies
was measured on the 5th, 7th and 10th floors of the Tower 1 and at all floors of the Bridge. A low-
frequency amplification was also measured on the top two Garage floors. As noted in Federal Transit
Administration (FTA) noise and vibration impact assessment guidelines, it is common for buildings to
exhibit amplifications because of ceiling and floor resonances so these results are not unusual.
The building amplification will be used to limit construction vibration. The two most significant sources
of construction vibration likely to be used on this project, impact pile driving and vibratory rolling, are
prohibited from use near the condominiums. In addition, project specifications require the contractor to
measure vibration generated by construction equipment prior to deployment near the structure and again
when it is brought on-site. Equipment will not be used if it produces vibration levels that exceed the
project vibration limit or if it produces excessive vibration at low frequencies.
FTA guidelines include procedures for calculating interior vibration levels generated by trains using
empirical data. These calculations combine the force characteristics of the LRVs, the performance of
resilient track fasteners, the effect of the tunnel slab, soil properties, and the measured transfer mobility
of the building structures. Using FTA procedures, CSA calculated the vibration levels generated from
future light rail operations inside Tower 1, the Bridge and the Garage.
The calculations show that vibration from LRVs will be below FTA limits. Maximum vibration levels
caused by future LRV traffic will be lower than or comparable to maximum vibration levels generated by
existing freight traffic. The structural amplification measured in the Bridge and on the upper floors of
Tower 1 increase low-frequency vibration generated by LRVs. However, even with this amplification,
calculated vibration levels are well below FTA criteria, and are also lower or comparable to existing
vibration levels at those frequencies.
Finally, the volume of daily light rail operations and its effect on the Calhoun Isles Condominiums was
assessed. Based on expected light rail operating conditions, the Calhoun Isles Condominiums will be
exposed to vibration from LRVs for a total of approximately 31 minutes per day, which is equivalent to
2% of the time in a day. As part of the current assessment, ambient vibration data were used to calculate
the statistical level that represents the highest vibration levels the Calhoun Isles Condominiums is
currently exposed to over that same duration. These calculations show that exposure from 31 minutes of
LRV vibration is lower than or comparable to the existing 31-minute/2% vibration exposure. Therefore,
there will be no appreciable difference between the existing highest total daily vibration exposure and
future light rail vibration exposure.

Background
The proposed SWLRT/Green Line extension will be located in a tunnel to the west of the Calhoun Isles
Condominiums and below the Kenilworth Trail, between the Kenilworth Corridor freight tracks and the
condominium structure. As part of the Southwest Light Rail Transit Final Environmental Impact
Statement (FEIS), vibration from SWLRT LRVs operating in a subsurface tunnel were modeled at the
Calhoun Isles Condominiums structure.

The vibration model in the FEIS was based on FTA guidelines 1 and included empirically-derived data for
the LRV and soil characteristics, and an analytical model of the Kenilworth tunnel structure based on

1
Transit Noise & Vibration Impact Assessment, Federal Transit Administration, Report FTA-VA-90-1003-06, May
2006
Page 3 of 65

finite element analysis (FEA). The results of the FEA model indicate that LRVs operating in the tunnel on
track fitted with resilient fasteners do not exceed FTA vibration criteria at the Calhoun Isles
Condominiums.

Calhoun Isles residents expressed concern that the condominium structures were unique and would
respond differently to vibration than other large concrete structures. Residents requested a detailed
vibration study of the structures to determine the specific response of the structures to light rail transit
operations. In response to this request, the Metropolitan Council contracted CSA, through AECOM, to
perform transfer mobility measurements at the building. Transfer mobility is the relationship between a
vibration source and resulting vibration at another position along the ground or inside a structure. The
transfer mobility measurements performed by CSA were used to revise the FEA model and vibration
calculations previously performed as part of the FEIS. Measurements of interior ambient vibration at
Calhoun Isles were also performed to aid in contextualizing the light rail vibration calculations.

Ground-borne vibration is described as oscillating movement of the ground. This movement can be
characterized in terms of displacement, velocity or acceleration. To be consistent with FTA methods, this
memo quantifies vibration using velocity in units of root-mean-square (RMS) inches per second.
Vibration velocity levels are given in decibels (VdB) relative to 1 micro-inch per second. Ground-borne
noise is sound that is radiated from the motion of interior room surfaces in structures due to ground
vibration. Ground-borne noise levels are presented in A-weighted decibels (dBA) relative to 20
microPascals. Vibration transfer mobility amplitudes are presented in decibels relative to 1 micro-inch per
second per pound.

Measurement Equipment and Procedures

The vibration transfer mobility test procedure is shown schematically in Figure 1. First, a borehole was
drilled to the proposed top-of-rail depth of 25 feet. Then, a force was generated by hammering the top of
the borehole drill string. The force was measured using a load cell attached to the end of the drill string at
the bottom of the hole. To measure the vibration response, one accelerometer was located on the soil
outside of the structure being tested and additional accelerometers were placed inside the structure at the
building slab and at various floors above grade. The relationships between the input forces and the
vibration responses at the outdoor accelerometer and the interior accelerometer positions provide an
estimate of how vibration generated from an exterior source travels into and through the building.
This procedure was developed in the 1980s and adopted by FTA in 1995. This empirical method has
subsequently been used in hundreds of rail transit projects across the globe, including in Minneapolis. The
results from this procedure have been validated as accurate for assessing the vibration propagation
characteristics of soils and structures. The method is currently being developed into a formal standard
that will be published by the American National Standards Institute (ANSI).
Page 4 of 65

Figure 1. Transfer mobility measurement schematic (not to scale)

The vibration acceleration of the ground, floors and walls were measured with PCB 393B05 high-
sensitivity accelerometers. The PCB 393B05 accelerometers have a rated frequency response of 0.5 to
700 Hertz (Hz) and nominal sensitivities of 10 volts per unit of gravity. The exterior accelerometer was
magnetically mounted to a metal spike driven into the soil. The interior accelerometer on the building slab
was magnetically mounted to a large concrete block on the lowest level of the structure. The other interior
accelerometers were either stud-mounted to steel mounting blocks coupled to the floor with spikes,
mounted to the floor with a wax/putty compound, or stud-mounted to aluminum blocks attached to the
wall with epoxy. Photographs of the accelerometers are shown in Appendix A.
Forces were measured using a PCB 205C force sensor in a custom-built enclosure. The load cell has a
rated maximum capacity of 60,000 pounds and a frequency response of 0 to 50,000 Hertz. A Data
Translation DT9837A digital acquisition module connected to a laptop computer was used to record the
time-history signal from the load cell and each accelerometer. A PCB 401B04 transducer simulator and
Velleman HPG1MK2 signal generator were used to calibrate the signal chain. A PCB 9200D portable
shaker, with calibration traceable to the National Institute of Standards and Technology (NIST) was used
to calibrate the accelerometers prior to the measurements.
The recorded data from the accelerometers and load cell were processed in the CSA laboratory using
custom analysis software. The acceleration data were converted into velocity numerically via trapezoid
rule integration. Transfer function calculations were performed to determine the relationship between the
force generated by the falling weight and the vibration measured by each accelerometer. This relationship
is referred to as a “point source transfer mobility” (PSTM). The narrowband PSTMs were converted into
one-third octave band spectra by energy averaging the narrow band bins in each standard one-third-octave
band and adjusting for the bandpass response of each filter.
Page 5 of 65

Measurement Locations
Transfer mobility measurements were used to characterize three structures of the Calhoun Isles
Condominium complex:
• #3151 Dean Court (“Tower 1”)
• #3145 Dean Court (“Bridge”)
• Calhoun Isles Parking Garage (“Garage”)
Tower 1 is a 10-story high structure that occupies the west part of the condominium complex. The Bridge
is a 7-story suspended residential structure that spans between Tower 1 and Tower 2 (#3141 Dean Court).
The Garage is a 3-story concrete building that lies to the south west of Tower 1. Tower 1 has a basement
with a crawl space. The Garage has an underground level that extends beneath the Bridge structure. The
foundation of the two towers are supported on a mat footing and the foundation of the Garage is
supported on spread footings. These structures were selected by representatives of CICA and their
consultant, Itasca.
Transfer mobility measurements were conducted between boreholes drilled along the proposed corridor
and the three structures described above. All boreholes were drilled to a depth of approximately 25 feet,
which is the top-of-rail depth for the proposed alignment. Figure 2 shows the layout of the boreholes
relative to the Calhoun Isles structures. Boreholes A and B were used for measurements at Tower 1,
Borehole C was used for the Bridge measurement and Borehole D was used for the testing at the Garage.
Each borehole was located at the proposed location of the near-track centerline at the closest point to the
tested structure.

BOREHOLE LOCATIONS

Figure 2. Plan view of borehole locations relative to Calhoun Isles Condominium structures
Page 6 of 65

The purpose of the measurements was to quantify the reduction in vibration levels from outside to inside
the structure and the amplification or reduction of vibration level with increasing floor height. Therefore,
accelerometers were placed on different floors in each of the structures. The accelerometers on residential
floors in Tower 1 and the Bridge were placed in hallways to minimize disruption to building tenants.
During the transfer mobility measurements of the residential structure, all electrical and water systems in
the relevant structures were deactivated to reduce contamination from background vibration sources.
Vehicle traffic in the Garage and condominium driveway was halted during the transfer mobility testing
of all structures. Maps of the specific measurement positions inside the building are included in Appendix
A.

Measurement Results
Measurement results are presented in terms of the PSTM measured at each site, insertion loss derived
from PSTM data, and ambient vibration metrics.

Point Source Transfer Mobility: PSTM as a function of frequency and location are provided for each
site. The PSTM is an absolute quantity that describes the vibration propagation characteristics of the soil
between a source and receiver location as a function of frequency. For a given vibration source, a high
PSTM value indicates a higher vibration level at the receiver point, and a low value indicates a lower
vibration level. Note that a negative PSTM value means that the vibration level at the receiver location
will be very low but does not indicate attenuation of vibration.

Insertion Loss: PSTMs were used to derive outdoor-to-indoor insertion loss spectra at each of the
interior measurement positions. Insertion loss represents the difference in vibration level between the
exterior position and the interior position; a negative number represents a reduction in amplitude as
vibration travels into a structure, and a positive number represents an increase in vibration amplitude.

Ambient vibration: Ambient vibration data at each measurement position is also presented below.
During the ambient measurement period, interior mechanical and electrical systems operated normally.
Non-representative events (e.g. high levels from a person directly touching the sensor for example) were
not included in the data presented in this study. Various statistical metrics for the ambient vibration are
presented in tables 1, 2 and 3. The metrics provided in those tables include:

L eq = equivalent level which is the time-averaged level of vibration energy over the measurement
period and includes the contribution from major and minor vibration sources

L max = Maximum vibration level over the measurement period

L 2 = 2nd percentile vibration level, which is the vibration level that is exceeded 2% of the time over the
measurement period. In other words, for 2% of the time, the vibration is at the L 2 level or higher. As
an example, for a 1-hour measurement period, vibration levels would exceed the L 2 value for a total of
1.2 minutes during the period. For a 24-hour period, vibration levels would exceed the L2 value for a
total of 29 minutes during the period.

Two percent of a 24-hour day is approximately 31 minutes, which corresponds to the total amount of time
the structures will be exposed to vibration from SWLRT LRVs each day (based on anticipated operations
of 220 daily trips at a speed of 45 mph and assuming an exposure time of about eight seconds for each
three-car consist passage). The L 2 metric provides a baseline to compare current maximum ambient
Page 7 of 65

vibration levels that occur over a 31-minute period and vibration generated by LRVs over the same
period.

Tower 1/Borehole A
Borehole A was located at civil station 2781+50 of the proposed eastbound track, approximately 15 feet
from the closest façade of Tower 1. One accelerometer was located on the ground outside of unit 101.
Seven additional accelerometers were located inside the building at the following locations:
• Foundation (tunnel)
• First floor hall outside of unit 101
• Third floor hall outside of unit 303
• Fifth floor hall outside of unit 503
• Seventh floor hall outside of unit 703
• Tenth floor hall outside of unit 1003

Figure 3 shows the PSTMs collected at this position. The data show a reduction in amplitude with
increasing floor height at frequencies above 16 Hz. The difference between the outdoor “Ground” curve
and the closest interior position (“Tunnel” curve for this site) are indicative of the difference in vibration
inside and outside the building. Similar effects are shown at all the measurement sites and demonstrate
the reduction in interior vibration level due to the building mass.

Figure 3. Point source transfer mobilities for Borehole A/Tower 1 measurements

The insertion loss for Tower 1 is presented in Figure 4. The plot shows a reduction in vibration energy
with increasing floor height across most of the frequency range, as would be expected for a building of
this type. There is a small increase in vibration levels in the 12.5 and 16 Hz frequency bands at the higher
Page 8 of 65

floors. The insertion loss is relatively small at the basement tunnel and first floor compared with the
insertion loss at higher floors.

The amplification in the 12.5 Hz and 16 Hz bands are indicative of resonances at those frequencies. For
large concrete structures buildings of this type, floor resonances are typically present in the 20 to 30 Hz
range, while wood-frame structures show resonances in the 15 to 20 Hz range. The lower-frequency
resonance in Tower 1 may be due to the lighter-weight steel construction of the top three floors compared
with the heavier concrete construction of the base structure.

Table 1and Figure 5 show ambient vibration statistics and time histories of ambient vibration at the
measurement locations, respectively. Ambient vibration levels were measured between approximately
11:43am to 2:03pm and 3:08pm to 3:29pm. Background sources included heater fans in the emergency
stairwell, normal residential activity, freight train traffic, elevators and heavy vehicles moving in the
parking lot. Average background vibration velocity levels varied between 54 and 60 VdB, with a
maximum level of 80 VdB on the 10th floor. The highest indoor average vibration levels were on the 10th
floor. The increase in vibration level from 3:22pm to 3:29pm is freight train traffic which is shown in
greater detail in a later section.

Figure 4. Insertion loss for Borehole A/Tower 1 measurements

Page 9 of 65

Table 1. Ambient vibration statistics at Tower 1 (Borehole A configuration)

Location Vibration Velocity Metrics, VdB re 1µin/s
L eq L max L2
1st Floor 54 VdB 72 VdB 66 VdB
3rd Floor 55 VdB 73 VdB 67 VdB
th
5 Floor 57 VdB 77 VdB 70 VdB
7th Floor 58 VdB 76 VdB 70 VdB
10th Floor 60 VdB 80 VdB 72 VdB

Figure 5. Vibration velocity time histories at Tower 1

Page 10 of 65

Figure 5. Vibration velocity time histories at Tower 1 (cont.)

Tower 1/Borehole B
At the request of the condominium representatives, a set of measurements was performed with
accelerometers located on the walls using three accelerometers mounted in transverse positions. The
borehole was located at civil station 2781+59, approximately nine feet from the Borehole A position. The
outdoor and foundation accelerometers were located at the same positions as for the previous Tower 1
measurement. The remaining accelerometers were located in the west stairwell, mounted on aluminum
blocks affixed to stairwell walls. The configuration of the interior accelerometers was as follows:
• Stairwell wall on the 2nd floor slab: one accelerometer mounted vertically, a second accelerometer
mounted in the transverse (south) direction
• Stairwell wall on the 8th floor slab: one accelerometer mounted in the transverse (south) direction
• Stairwell wall on the 10th floor slab: one accelerometer mounted vertically, a second
accelerometer mounted in the transverse (south) direction

Figure 6 shows the measured PSTMs at these sites. As with the data collected during the Borehole A
measurements, the results indicate decreasing amplitudes with increasing floor height. The transfer
mobilities measured in the vertical direction have a greater amplitude than the results measured in the
transverse direction at the same position across most of the frequency range.
Page 11 of 65

Figure 6. Point source transfer mobilities for Borehole B/Tower 1 measurements

The insertion loss for the vertically-mounted accelerometer positions are provided in Figure 7. The
curves at these positions are similar to those shown in Figure 4. The insertion loss measured in the tunnel
shows slight differences from the curve derived from the Borehole A measurement. These differences are
likely attributable to slight variations in soil composition between the two borehole locations. The
amplification in the 12 to 16 Hz bands shown in the Figure 4 insertion loss plot are not present in Figure
7. This reflects the difference in measuring vibration on the wall which produces less movement than the
floor vibration data shown in Figure 4.

The results of the ambient vibration measurements are shown in Table 2 and Figure 8 below. Figure 8
shows the ambient vibration velocity time histories measured between 10:46am to 2:00pm and 2:34pm to
4:00pm on April 4, 2018. Ambient sources at this location included residents moving around the building,
elevator operations, personnel moving in the stairwells, door openings and closings, and freight train
traffic.
Page 12 of 65

Figure 7. Insertion loss for Borehole B/Tower 1 measurements

Table 2. Ambient vibration statistics at Tower 1 (Borehole B configuration)

Location Vibration Velocity Metrics, VdB re 1µin/s
L eq L max L2
2nd Floor Vert. 48 VdB 72 VdB 52 VdB
2nd Floor Trans. 48 VdB 71 VdB 51 VdB
8th Floor Trans. 48 VdB 70 VdB 55 VdB
10th Floor Vert. 53 VdB 76 VdB 58 VdB
10th Floor Trans. 47 VdB 71 VdB 54 VdB
Page 13 of 65

Figure 8. Vibration velocity time histories at Tower 1

Page 14 of 65

Figure 8. Vibration velocity time histories at Tower 1 (cont.)

Bridge/Borehole C
Borehole C was located at civil station 2780+85, about 115 feet from the westernmost Bridge support.
Accelerometers were located at the Bridge section of the Calhoun Isles Condominium complex that spans
between Tower 1 and Tower 2. Accelerometers were located at the following positions:
• Foundation (underground parking garage)
• Surface sidewalk (supported by the underground parking garage)
• Fourth floor hallway near units 404 and 405
• Fifth floor hallway near unit 505
• Seventh floor hallway near unit 703
• Ninth floor hallway near unit 903
• Tenth floor hallway near the Party Room

Figure 9 shows the measured PSTMs at this site. The insertion loss for the Bridge site is presented in
Figure 10.
Page 15 of 65

Figure 9. Point source transfer mobilities for Bridge measurements

Figure 10: Insertion Loss for Bridge measurements

Page 16 of 65

The transfer mobility data show a significant reduction in vibration from the sidewalk to the suspended
Bridge floors. The PSTM and resulting insertion loss at the 4th through 7th floors are comparable to each
other in amplitude, as opposed to the data measured at the parking garage and sidewalk that are
significantly different in shape and amplitude. There is approximately 8 dB of amplification in the 8 Hz
band on the 10th floor.

The results of the ambient vibration measurements are shown in Table 3 and Figure 11 below. Figure 11
shows the ambient vibration velocity time history measured between 9:30am to 1:46pm and 2:38pm to
4:00pm on April 5, 2018. Ambient sources at this location included residents moving around the building,
elevator operations, personnel moving in the stairwells, door openings and closings, and freight train
traffic. The sensor on the 9th floor failed during the measurements, so those data are not presented here.
The ambient L eq measurements inside the Bridge are slightly lower than measured in Tower 1. The
ambient average and maximum levels on the 10th floor are higher than on the lower floors. This is likely
caused by the 8 Hz amplification shown in Figure 10.

Table 3. Ambient vibration statistics at Bridge

Location Vibration Velocity Metrics, VdB re 1µin/s
L eq L max L2
4th Floor 55 VdB 77 VdB 61 VdB
5th Floor 53 VdB 77 VdB 60 VdB
7th Floor 52 VdB 76 VdB 59 VdB
10th Floor 59 VdB 79 VdB 67 VdB

Figure 11. Vibration velocity time histories at Bridge

Page 17 of 65

Figure 11. Vibration velocity time histories at Bridge (cont.)

Page 18 of 65

Garage/Borehole D
The borehole was located at civil station 2779+20. Accelerometers were placed at the following locations:
• Soil approximately three feet from west wall
• Foundation/lowest level of parking structure
• Ground level/floor 2 of parking structure
• Roof deck/floor 3 of parking structure

Figure 12 shows the measured PSTMs at the Garage and the measured insertion loss for the Garage is
given in Figure 13. The insertion loss shows a broad amplification in the frequencies between 8 Hz and
25 Hz at the two suspended floors. This is due to the design of the structure that includes uncoupled,
wide spans that are designed for movement and expansion.

Figure 12. Point source transfer mobilities for Garage measurement

Page 19 of 65

Figure 13. Insertion Loss for the Garage measurement

Table 4 shows the summary of the ambient vibration data at the Garage. Figure 14 shows ambient
vibration levels measured between 9:00am and 1:00pm at the Garage. Ambient sources included drilling
of Borehole D, automobile traffic inside the Garage, residents walking around the structure and freight
train traffic. Ambient vibration levels ranged from 59 VdB at the ground level to 67 VdB at the roof deck,
and maximum levels reach as high as 89 VdB on the top deck. The prominent source during this
measurement was the auger rig drilling Borehole D. The rig left the site at around 11:21am. Automobile
traffic in the Garage was the major internal source. Vibration levels at the upper floors of the Garage are
noticeably higher than at the foundation. The upper floors of the Garage structure are composed of
concrete slabs separated by expansion joints. They are also directly excited by automobiles driving over
the slabs. Therefore, the relatively high vibration levels are expected.

Table 4. Ambient vibration statistics at Parking Garage

Location Vibration Velocity Metrics, VdB re 1µin/s
L eq L max L2
Lower Level 59 VdB 73 VdB 66 VdB
Ground Level 64 VdB 86 VdB 72 VdB
Roof Deck 67 VdB 89 VdB 76 VdB
Page 20 of 65

Figure 14. Vibration velocity time history at the Garage

Page 21 of 65

Freight Train Traffic

During the course of the ambient measurements at each structure, vibration from freight trains moving
through the Kenilworth Corridor were logged. Freight train consists included two to five locomotives and
28 to 110 freight cars. Freight train speeds were approximately 10 mph for all events. Trains moving east
were generally fully loaded, while westbound trains were empty. The detailed vibration time histories for
freight train traffic measured at Tower 1, the Bridge and the Garage are presented in the Figures 15-17
below. Because of an equipment problem during the April 3 2:38pm freight train event, only data from
three positions were recorded.
Page 22 of 65

(a)

(b)
Figure 15. Freight train vibration time histories inside Tower 1 (a) with accelerometers in
Borehole A locations, and (b) accelerometers in Borehole B locations
Page 23 of 65

Figure 16. Freight train vibration time history inside Bridge

Figure 17. Freight train vibration time histories inside Garage

The time histories of the freight train traffic show a significant reduction in level from the outside
accelerometer position to the foundation position and the lower floors. At the higher floors of Tower 1
and the Bridge, the interior vibration levels from freight trains are greater compared to lower levels. This
is due to the smaller insertion loss at lower frequency bands combined with the substantial low frequency
vibration generated by freight trains due to their stiff suspensions combined with worn wheels and rails.

Figures 18 through 20 provide the time-averaged vibration spectra of the freight train traffic shown above.
The spectra show peaks in the 12.5 one-third octave bands which dominate the overall vibration velocity
levels on the higher floors in both Tower 1 and the Bridge. This corresponds with peaks in the PSTMs at
the upper floors.
Page 24 of 65

(a)
Figure 18. Frequency spectra of train traffic inside Tower 1 (a) with accelerometers in Borehole
A locations, and (b) accelerometers in Borehole B locations
Page 25 of 65

(b)
Figure 18. Frequency spectra of train traffic inside Tower 1 (a) with accelerometers in Borehole
A locations, and (b) accelerometers in Borehole B locations (cont.)

Figure 19. Frequency spectra of freight train traffic inside Bridge

Page 26 of 65

Figure 20. Frequency spectra of freight train traffic inside Garage

Findings
The transfer mobility and insertion loss results indicate that there is a reduction in vibration level
generated by external sources in most frequency bands of interest due to propagation through the
foundation at all locations tested. There is almost no outdoor-to-indoor reduction in the 63 Hz and 100 Hz
bands at Tower 1. There is a slight increase in vibration at frequencies in the 16 Hz band and below at the
upper floors of Tower 1 and the Bridge. There is a broad amplification in the 10 to 20 Hz region at the
upper floors of the parking garage.

Average ambient floor vibration levels varied by building location, ranging from 52 VdB at the 7th floor
of the Bridge to 67 VdB at upper level of Garage. Maximum ambient levels, ranged from 72 VdB at the
1st floor of Tower 1 to 89 VdB at the upper level of the Garage. The highest ambient levels were observed
on the top level of the Garage.

Relatively high ambient vibration levels were observed on the highest floors of Tower 1 and the Bridge
even in the absence of external surface vibration sources. This vibration appears to be generated by
rooftop heating ventilation and air conditioning (HVAC) units. Rooftop HVAC units are mounted on the
roofs of the connectors between the two towers and the Bridge, as well as a lower floor landing on Tower
2. These units do not appear to be mounted on isolation curbs and therefore may be directly vibrating the
building structure.
Page 27 of 65

Refined LRV Vibration Calculations

The building response data measured at the Calhoun Isles Condominiums were used to refine the
vibration and ground-borne noise calculations originally performed as part of the FTA environmental
process.
Ground borne vibration were modeled using FTA procedures that incorporate the force generated by rail
vehicles, the vibration propagation characteristics of the ground and the building foundation/soil
interaction. This relationship is assessed using the following equation:
L𝑉𝑉 = FDL + TMLine + CBuild
where:
L v = ground-borne vibration velocity level in decibels (VdB) referenced to 1 µinch per second

FDL = force density level in decibels referenced to 1 pound per square root of foot

TM Line = line source transfer mobility in decibels, referenced to 1 µinch per second per pound per
square root of foot

C Build = building foundation coupling adjustment

Ground-borne noise is modeled using the following equation:

LA = LV + K A-wt − 5
where
L A = A-weighted ground-borne noise level, in decibels (dBA) reference to 20 µPascals
L v = ground-borne vibration velocity level in decibels referenced to 1 µinch per second
K A-wt = A-weighting adjustment at the 1/3-octave band center frequency

The -5 dB adjustment in the above equation is based on current research 2 that demonstrates that the
equation in the 2006 version of the FTA guidance manual overestimates ground-borne noise. This
adjustment has already been incorporated into the latest version of the Federal Railroad Administration
(FRA) high-speed rail guidance manual and will be incorporated into the next version of the FTA
guidance manual which should be released later this year. Use of this adjustment is permitted by FTA as
it reflects current best practice.
The force density level of LRVs derived for the Central Corridor environmental analysis 3 was used to
represent the force from SWLRT LRVs. The FDL used in the vibration calculations for LRVs traveling
at 40 mph is shown in Figure 21. Because vibration levels generated by LRVs is proportional to LRV
45mph
speed, the force density was adjusted using the relationship 20 Log10 � � as recommended by FTA.
40mph
This adjustment is conservative as the Central Corridor analysis speed relationship is closer to
10Log to 15 Log rather than 20Log.

2
Ground-Borne Noise and Vibration in Buildings Caused by Rail Transit, Transit Cooperative Research Program
Project D-12, December 2009
3
“Vibration Measurements and Predictions for Central Corridor LRT Project,” Memorandum from Hugh
Saurenman, ATS Consulting to Kathryn O’Brien, Metropolitan Council, December 19, 2008
Page 28 of 65

The soil propagation characteristics were modeled using the line source transfer mobility (LSTM)
measured near the Calhoun Isles Condominiums for the SWLRT FEIS 4. The smoothed LSTM is shown
in Figure 22.

Figure 21. SWLRT LRV force density, 40 mph

4
“Noise and Vibration Technical Report,” Southwest LRT Project Technical Report, October 2015
Page 29 of 65

Figure 22. Line source transfer mobility used to model vibration propagation at
Calhoun Isles Condominiums

The building adjustments for each of the three structures were determined by subtracting the interior
accelerometer data at each measurement position from the exterior surface level accelerometer data.
SWLRT LRVs will operate in a tunnel structure with a 3-foot thick slab and a maximum 5-foot thick
concrete plug to form a total base that is up to 8-feet thick. This structure will attenuate the forces
generate by the LRVs which in turn will reduce ground-borne vibration levels. An estimation of the
insertion loss provided by the tunnel was generated for the previous engineering analysis 5 using finite
element analysis (FEA) methods. While this calculated insertion loss is consistent with previous field
measurements and the literature, a revised estimate of insertion loss based on Urban Mass Transportation
Administration (UMTA) 6,7 methodology (referred to as ‘UMTA Est’ below) was made based on
measured data to improve the accuracy of the projections.
Data recently collected at two North American light rail transit systems demonstrates that a concrete slab
8 to 12 inches thick beneath embedded track will provide 2 to 9 decibels of reduction in the frequency
range of 12 Hz to 200 Hz. An empirically-derived relationship published by the UMTA has indicated that

5
“Summary of Operational and Construction Vibration at the Calhoun Isles Condos,” Internal Memorandum from
Lance Meister, Cross-Spectrum Acoustics Inc, to Kelcie Campbell, SPO, January 31, 2017
6
Handbook of Urban Rail Noise and Vibration Control, Report Urban Mass Transportation Administration, Report
UMTA-MA-06-0099-82-1, October 1982
7
State-of-the-Art Review; Prediction and Control of Ground-Borne Noise and Vibration from Rail Transit Trains,
Urban Mass Transportation Administration, Report UMTA-MA-06-0049-83-4, December 1983
Page 30 of 65

increasing the thickness of the slab beyond 8 to 12 inches will provide additional reduction that is
proportional to 56 Log (thickness). These data are consistent with measured results 8 and with finite
element models of thick concrete slabs.
Figure 23 shows the measured insertion loss from a 12-inch concrete slab under embedded track (“Emb.
Slab”) measured at a light rail system in North America. The figure also shows the calculated insertion
loss for a 4-foot thick slab (“UMTA Est.”) calculated by arithmetically adding the 56Log(thickness)
UMTA relationship adjustment to the measured 12-inch concrete slab data described above using the
56Log(thickness) UMTA relationship. Finally, the figure provides the estimated vibration reduction from
a 4-foot thick concrete slab generated by the FEA model (“FEA”) used in the January 2017 engineering
analysis. The FEA-modeled insertion loss is similar to that derived using actual data adjusted using with
the empirical UMTA formula.

Figure 23. Vibration insertion loss spectra from concrete slabs in light rail tunnel

Table 5 presents the projected vibration level inside the Calhoun Isles Condominiums incorporating the
building response, tunnel structure response (based on measured slab reduction adjusted using with the
empirical UMTA method), soil properties and vehicle characteristics. While the January 2017 analysis
used finite element estimates to model the reduction from the slab, the slab adjustments incorporated into
the predicted levels for this study, shown in Table 5, are based on real-world empirical data rather than
theoretical computer modeling.

8
“Modeling the effects of structure mass in reducing ground-borne vibration and ground-borne noise in rail transit
systems,” Herbert Singleton Jr. and Scott Edwards, Presented at INTER-NOISE 2015, August 2015
Page 31 of 65

Table 5: Calculated maximum interior vibration levels from SWC LRT

Ground Borne Vibration Level (VdB re 1µin/s) Ground-Borne Noise Level (dBA re 20µPa)
Max Band Proj. FTA
Location Level† FTA Criteria Exceed? Level Criteria Exceed?
Tower 1
Tunnel 56 VdB -- -- 18 dBA -- --
1st Floor 55 VdB 72 VdB No 12 dBA 35 dBA No
3rd Floor 55 VdB 72 VdB No 4 dBA 35 dBA No
5th Floor 57 VdB 72 VdB No 3 dBA 35 dBA No
7th Floor 57 VdB 72 VdB No 5 dBA 35 dBA No
10th Floor 58 VdB 72 VdB No 0 dBA 35 dBA No
Bridge
4th Floor 61 VdB 72 VdB No 0 dBA 35 dBA No
5th Floor 60 VdB 72 VdB No 0 dBA 35 dBA No
7th Floor 61 VdB 72 VdB No 0 dBA 35 dBA No
10th Floor 67 VdB 72 VdB No 0 dBA 35 dBA No
Garage
Lower Level 58 VdB -- -- 24 dBA -- --
Ground 60 VdB -- -- 20 dBA -- --
Level
Roof Deck 64 VdB -- -- 15 dBA -- --
Projected LRT levels include vehicle, building structure and tunnel structure characteristics in addition to resilient
fasteners.
† Maximum 1/3 Octave Band Level

While the upper floors of Tower 1 and the Bridge may be susceptible to vibration in the 12 Hertz to 16
Hertz frequency range, SWLRT LRVs generate little energy in those bands as shown in Figure 21. The
LRV does produce vibration in the 8 Hz octave band which coincides with the building amplification at
the Bridge. The projections in Table 5 incorporate the vibration response of the structure at floors that
were part of this study and even with building amplification in the 8 Hz, 12 Hz and 16 Hz low frequency
bands, the projected levels from LRVs operating in a tunnel structure are well below the FTA impact
criteria.
Figure 24 shows comparisons of the projected LRV spectra with current average ambient vibration at the
10th floors of Tower 1 and the Bridge. Above 12 Hz, the projected maximum LRT spectrum is
significantly lower than the existing background vibration at those floors. At very low frequencies, the
vibration is higher than the existing average level, but lower or comparable to existing maximum levels.
Comparisons of LRT vibration to ambient levels on other floors are shown in Appendix C.
Page 32 of 65

(a)

(b)
Figure 24. Comparison of predicted light rail vehicle spectra with ambient spectra (a) Tower 1,
(b) Bridge
Page 33 of 65

Figure 25 shows projected maximum LRV vibration compared to maximum vibration levels from a
sample of existing freight train traffic at the first floor of Tower 1 and the tenth floor of the Bridge, which
are the locations in those two structures where LRV vibration will be highest. As shown in that figure,
LRV vibration will be lower than freight train vibration at Tower 1 across the frequency band. At the
Bridge, LRV vibration is significantly lower than freight train vibration at frequencies above 10 Hz.
Maximum LRV vibration in the 8 Hz band is comparable to maximum levels from existing freight trains
at this location, but levels are well below perceptual limits. Comparisons of maximum LRV vibration
levels to maximum freight train vibration levels at other locations and other freight events are provided in
Appendix D. Note that while high frequency vibration from LRV vehicles will exceed freight train
vibration levels in the Garage, LRV vibration levels in this frequency range are lower than what is
required for very sensitive laboratories, and will be lower than existing automobile traffic.
Page 34 of 65

(a)

(b)
Figure 25. Comparison of maximum light rail vehicle spectra to maximum freight train spectra
(a) Tower 1, (b) Bridge
Page 35 of 65

Finally, there has been concern expressed about the volume of LRV traffic compared to freight rail traffic
and the impact this additional volume may have on the Calhoun Isles Condominium structures and
residents. The L 2 described earlier in this report was used to compare the total daily vibration exposure
from LRV traffic to the highest existing vibration levels over the same total daily duration.
Figure 26 shows projected LRV spectra compared with the ambient L 2 spectra on the 1st floor of Tower 1
and the 7th floor of the Bridge. These are the locations in Tower 1 and the Bridge with the lowest existing
L 2 levels; the ambient vibration at these locations are at or above this L 2 level for approximately 31
minutes per day. At Tower 1, the LRT vibration spectrum is lower than then existing L 2 spectrum. At the
Bridge, the projected LRT spectrum is lower than the existing L 2 spectrum at most frequencies except for
the 8 Hz band. However, the LRT vibration level at the 8-Hz band is below FTA limits and the threshold
for human detection. Other locations inside Tower 1 and the Bridge have similar results. At the parking
garage, vibration from LRT operations is lower than the existing 31-minute L 2 level and well below
building damage thresholds. Therefore, even at a traffic level of 220 vehicles per day, LRT operations
will not affect the existing vibration environment.

Additional ambient vibration data and projected LRV vibration results are presented in Appendix C.
Page 36 of 65

(a)

(b)
Figure 26 Comparison of predicted light rail vehicle spectra with ambient L 2 (highest 31-minute
ambient level) spectrum (a) Tower 1, (b) Bridge
Page 37 of 65

Appendix A: Measurement Locations & Site Photographs

Tower 1/Borehole A

Outdoor Accelerometer
Page 38 of 65

Indoor/Foundation
Page 39 of 65

First Floor/Near Unit 101

Page 40 of 65

3rd Floor/Near Unit 303

Page 41 of 65

5th Floor/Near Unit 503

Page 42 of 65

7th Floor/Near Unit 703

Page 43 of 65

10th Floor/Near Unit 1003

Page 44 of 65

Tower 1/Borehole B

2nd Floor Slab

Page 45 of 65

8th Floor Slab

Page 46 of 65

10th Floor Slab

Page 47 of 65

Bridge/Borehole C

Parking Garage
Page 48 of 65

Sidewalk
Page 49 of 65

4th Floor
Page 50 of 65

5th Floor
Page 51 of 65

7th Floor
Page 52 of 65

9th Floor
Page 53 of 65

10th Floor
Page 54 of 65

Garage/Borehole D

Lower Level
Page 55 of 65

Ground Level/Outside
Page 56 of 65

Roof Deck
Page 57 of 65

Appendix B. PSTM Data & Coherence

Tower 1/Borehole A

Tower 1/Borehole B
Page 58 of 65

Bridge/Borehole C

Garage/Borehole D
 Page 59 of 65

Appendix C: LRV vs Ambient vibration at measurement locations

Tower 1
1st Floor 3rd Floor

5th Floor 7th Floor

10th Floor
Page 60 of 65

Bridge

4th Floor 5th Floor

7th Floor 10th Floor

 Page 61 of 65

Garage

Lower Level Ground Level

Roof Deck
 Page 62 of 65

Appendix D: LRV vs Freight Train vibration at measurement

locations

Tower 1

LRV vibration compared to April 3, 2018, 2:40pm Freight Train

 Page 63 of 65

LRV vibration compared to April 3, 2018, 3:22pm Freight Train

 Page 64 of 65

Bridge

LRV vibration compared to April 5, 2018, 1:35pm Freight Train

 Page 65 of 65

Garage
LRV vibration compared to April 6, 2018, 11:28am Freight Train