Bibliography Glossary Geophysical
About Engineering Applications Geophysical Methods Glossary Bibliography
Traffic Vibrations

Many of the problems caused by traffic vibrations are more related to human perception and harmful effects than actual physical damage. These are often addressed in Environmental Impact Statements in which future effects are predicted and evaluated and compared to recommended threshold criteria. They are outlined in the FTA Guidance Manual Transit Noise and Vibration Impact Assessment. Although this reference focuses primarily on Rail Transit issues, the concepts and procedures are usually applicable to roads, highways, and bridges.

Traffic vibrations are addressed either during the planning stage, prior to construction when engineers anticipate potential problems, or during the operating stage, usually in response to a complaint. Engineers are not generally concerned about the impact of vibrations directly on the highway structure, with the notable exception of bridges.

Human Perception

Of all vibration effects, human perception is the most difficult to quantify, since it varies according to many independent factors, such as the following:

There have been many attempts recorded in literature to label human sensitivities, and descriptions have ranged from barely perceptible to annoying, and the units used have included acceleration, velocity, and displacement. There are now international conventions for measuring and evaluating these vibrations, and for most highway applications, the more simple evaluations suffice. People are similar enough that if a vibration is characterized in literature as "annoying," most people will find it annoying.

Procedures for measurement

The FTA Manual (Ch. 11) offers general guidelines. Griffin (1990, p.453) describes methods for evaluating whole-body exposure to vibrations that are basically the same as the ANSI Standards. These procedures can be quite complicated and are, in most cases, unnecessary unless litigation is anticipated. Moreover, some of the procedural aspects are outdated by modern digital instrumentation. One important difference is that the ANSI Guidelines require measurement of acceleration, but the FTA guideline suggests velocity. There is sufficient research to support both, but perhaps the more modern data are in acceleration. Included in this section is an example generalized procedure entitled Ambient Ground Vibration Measurement borrowed from GEOVision that can be helpful. Bruel & Kjaer have published an on-line web manual that is a primer on vibration measurement. A generalized approach that will work in most applications is presented below. Table 10 defines the strength of perception of various magnitudes of vibration.

Table 10. Approximate strength of perception for various magnitudes of weighted vibration. (M.J. Griffin, 1990, p. 262)

Semantic label r.m.s. weighted acceleration (m/sec) Semantic label
Very strong perception 0.315
Strong perception
Very clear perception 0.08
Clear perception
Perception probable 0.020
Perception improbable

It should be noted that human perception sensitivity is most sensitive below 2 Hz horizontally, between 5 and 15Hz vertically, and decreases at higher frequencies. It cannot be overemphasized that merely measuring the vibration levels and interpreting them in terms of human perception only partly addresses the problem. There are many factors that contribute to vibrations as perceived by humans, and cannot be addressed using only geophysics.

The most important axis of measurement for human perception vibrations is vertical, since people are generally more sensitive to vertical motions than horizontal motions. The general Ambient Ground Vibration Measurement methodology below should also be followed. However, there are some key differences:

  1. Measurements should be made using an accelerometer.
  2. Triaxial measurements are best, but the vertical axis is the minimum required.
  3. Minimum resolution should be 0.001m/sec2 (0.039in/sec2).
  4. Digital recordings should be made either directly in the frequency domain (such as with a spectrum analyzer), or anticipating analysis in the frequency domain.
  5. Sufficient data should be collected for at least 20 averages to achieve some statistical benefit.
  6. The bandwidth of interest is generally 1 to 100 Hz; the sampling rate should be at least 250 Hz.
  7. Data should be recorded at the location and time period the subject claims is the "worst."
  8. Careful consideration should be given to the location of the complaint, the hours during which the vibration is perceived, the duration, the frequency, and whether the vibration events are associated with sound. Each factor can contribute to weighting factors that could elevate or reduce the evaluated motion. For example, the same vibration measured in a factory setting will be much more tolerable than in a doctor's office.

Harmful Effects

Having addressed harmful effects on people, this section discusses harmful effects on critical processes. It is not the role of this web manual to discuss all the possible processes that can be affected by vibrations, but it is important to have some idea of the types of processes that can be affected. This will allow engineers to understand why some types of vibration measurement programs are necessary, and when they should be used.

Usually, critical processes involve imaging (both medical and graphical), medical procedures, delicate manufacturing, or assembly operations. Some examples of these include the following:

In most cases, hospitals and manufacturing plants will evaluate and design their facilities to mitigate the effects of vibrations. To do this, they measure the existing ambient conditions and design foundations or, under worst-case conditions, full floating foundations. These are expensive, but do an excellent job. In summary, if the highway is already in existence, facilities will make appropriate accommodations.

Sometimes, however, it will be necessary to route a highway or build a bridge near such a facility, and the potential harmful effects must be evaluated and addressed. The best approach, as outlined in the FTA Manual (1995, ch. 11), is as follows:

  1. Obtain from the facility or equipment manufacturer the threshold of tolerance for vibration. Make sure this is the threshold at the input to the foundation, and not at the input to the machine or instrument itself. Some analysis may be required to make this conversion.
  2. Evaluate the vibration propagation/attenuation characteristics of the soils between the proposed highway or bridge, and the critical process.
  3. Predict what the vibration amplitude and frequency characteristics will be emanating from the source.
  4. Combine 1, 2, and 3 to predict and compare what the motions will be at the equipment foundation.

The general measurement procedures outlined under Human Perception above will be equally useful here.

There are no known cases in which traffic vibrations have actually caused structural damage. However, in such a case, the approach and procedures outlined under Construction Vibrations below will apply.

Example Procedure for Ambient Ground Vibration Measurements
(Used with Permission from GEOVision Geophysical Services)

Accurate Characterization of ground vibrations is often needed for siting of vibration-sensitive facilities or for diagnosis of existing vibration-related problems. Due to the small amplitudes of these vibrations, special instrumentation and special care is needed to obtain meaningful data.

The objective of ground vibration measurements is to accurately obtain site vibration data in terms of time histories, spectra (Power Spectral Density (PSD) or response), or statistical parameters; obtain data over time to determine the statistical variability of ground vibrations; and obtain data with sufficient quality to assist in identification of the major contributors to the site vibrations.

The objective of this procedure is to provide a formal framework for a skilled, experienced engineer to accomplish high-quality, consistent ground vibration measurements. This procedure covers equipment, measurements, and basic data analysis, but does not include advanced analyses or reporting.


  1. Kinemetrics ES-T Episensor Broadband Accelerometer or equivalent (figure 221). All three channels (X, Y, and Z) or any subset may be used for a specific project. Multiple sensors are good if available.
  2. Portable digital recorder with very-low-noise preamplifier and digital resolution of a least 16 bits. The Kinemetrics K2 Seismograph or SSR-1 Digital Recorder are specified in this procedure, but others may be used if equivalent or better.
  3. Laptop computer with recorder control software (Kinemetrics Quicktalk, Quicklook, and PSD) and communications cable.
  4. Power/signal cable from recorder to sensor.
  5. Batteries to operate system.
  6. Solar panel (if required) for long-term unattended measurements.
  7. GPS antenna (if needed) for precision timing.
  8. Field enclosure (if needed) for environmental protection.
  9. Tape measure for accurate location of measurements. DGPS may be used if available.

High resolution sensor for ambient vibration monitoring.  (Kinemetrics)

Figure 221. High resolution sensor for ambient vibration monitoring. (Kinemetrics)

The measurement site should be selected to be representative of "average" site conditions. Measurement site or general site conditions that could affect vibrations (proximity of roads, mechanical equipment, construction, etc.) should be carefully web manualed in a field notebook.

If the site is under construction, the construction activities should be web manualed (type, location, durations) in a field notebook. Unless construction vibrations are the subject of the investigation, efforts should be made to avoid measurements during times of construction activities.

Other environmental conditions that could affect vibrations should be web manualed in a field notebook. These may include wind, temperature, humidity, rain, noise, or others.

The sensor(s) and recorder(s) should have been calibrated within 12 months of the vibration measurements. A copy of these calibration data should accompany the instrumentation. Calibration constants important to the measurements (sensor sensitivities, sensor natural frequencies, and recorder gains) should be recorded in the field notebook.

Immediately prior to the field measurements, the sensor-recorder sensitivity should be checked by tilting each sensor through its full-scale range (90 degrees for a 1g accelerometer, 14.5 degrees for a ¼g accelerometer), and the output measured using the voltmeter function on the recorder. These data should be recorded in the field notebook. Although not a calibration, results of this sensitivity check should be within ±5% of the formal calibration data.

Measurement Procedure

The general field procedures (for 5 minute/hour survey) are as follows:

  1. Install the recording station. This should be at least 3 m from the sensor. For multiple-day measurements, provide adequate power and environmental protection.
  2. Connect the sensor to the recorder and turn on the recorder. Connect to the computer and turn on computer.
  3. Set up recorder parameters. Include the setting of sensor calibration data where appropriate. web manual recorder parameters in field notebook. Depending on the recorder, download the parameter file and save it to diskette for the project file.
  4. Check sensor zero settings. If greater than 500 mv, manually adjust sensor zero.
  5. Record a functional test. Download file. View the file with QL16 or other appropriate file viewer. Make sure response is normal.
  6. Mark location of each measurement/Run numbers on the Site Map.
  7. Proceed with measurement. Manually record the first event (typically 5 minutes) using the keyboard trigger. Download the file and check with QUICKLOOK.
  8. Set up the recorder for event recording. For 7-day ambient tests, set instrument to record 5 minutes of data every hour. Start acquisition.
  9. Wait on site until the next measurement period. Check the recorder to verify that it has correctly performed the automatic data collection. If not, troubleshoot and repeat
  10. Leave the site if desired. For multiple-day measurement programs, site inspection once per day is required (twice per day is recommended). Data should be downloaded and memory cleared if more than 75% of available recorder memory is used.
  11. Site conditions should be noted for each site visit. If possible, ask site personnel about unusual conditions that may have occurred while the engineer was not present. Examples would be the passage of large, heavy vehicles.
  12. Daily, or upon completion of the measurement period, download data to the PC and check all data using QUICKLOOK.
  13. If there is more than one site, repeat these steps for each additional site.

Required Field Records

Much of the above information will be automatically recorded in the seismograph header at the time of recording (gains, filtering, date and time) and need not be recorded on the paper log:

  1. Site Map showing to scale the locations of measurements referenced to physical objects.
  2. Diskettes with backup copies of data on hard disk, labeled with measurement designation, record ID numbers, date, and tester name.

Basic Analysis and Interpretation

Following completion of fieldwork, the recorded digital records are processed by computer and analyzed by an experienced engineer to produce plots and tables of transient motion (time histories) and steady-state motion (converted from PSD's).

Again, the specific procedure varies according to the project requirements. The following steps are for projects using the K2 recorder only.

  1. Retrieve K2 files and transfer to PC for analysis. NOTE: K2 file names are randomly assigned by the recorder, and will differ from Measurement ID numbers. Save the original *.evt files, and then copy to new files with logical, project-based names.
  2. Run QUICKLOOK on all *.evt files. Visually review all time-series data for quality. Note any unusual data aspects in the field notebook.
  3. Using the print command, print the QUICKLOOK screen for each data file. This will act as a "road map" for future data analyses. Mark (in ink) any notes or observations on the paper plots.
  4. Run Kinemetrics' PSD program. For each channel, calculate and plot the PSD. Use the default window, overlap, and gains.
  5. Using the print command, print the PSD plot for each channel. Mark any notes or observations on the paper plots.
  6. Export the PSD data for each channel to ASCII using the "print to file" function.
  7. After each of the data channels is completed, import the ASCII PSD data to an EXCEL spreadsheet. Make a single plot containing all channels of PSD data. Print this plot and include with the field files.


A "Field File" should be made for each distinct measurement project. If a project contains distinct sites, separate field files shall be made.

The field file should contain:

Although the scope of this procedure does not include reporting, the material contained in the field file is appropriate for inclusion as an appendix or in the main body of a project report.