Detection of Water and Ice on Bridge Structures by AC Impedance and Dielectric Relaxation Spectroscopy (Phase III) - FY10 NATSRL

Principal Investigator:

John Evans, Professor, UMD-Chemistry and Biochemistry

Project Summary:

In the interest of improved public safety, researchers are developing low-cost sensing systems to monitor surface conditions on bridge and road decks. These sensing systems are based on the measurement of the impedance response of a sensor in contact with or close proximity to ice, water, or aqueous solutions of deicing chemicals. Both a. c. impedance and dielectric relaxation (DR) spectroscopies were considered, and researchers have chosen to focus on the latter technology since it can be implemented with less expensive electronic components. Determination of surface condition is accomplished by utilizing commercially available electronics (normally employed for soil moisture measurement) to carry out DR by time domain reflectometry. The reflected waveforms captured by this technique are sensitive to the dielectric properties of material in contact with the passive metal electrodes of the sensor. As such, with careful design of the electrode geometries and configurations, the physical state (frozen vs. liquid) of precipitation and deicing chemicals on the deck or road surfaces can be determined by sophisticated mathematical analysis of the captured waveforms.

Researchers have completed the evaluation of more than 40 prototype sensors of varying electrode configuration and geometry. The design of ruggedized versions of the best of these will be finalized prior to their deployment (in late summer of 2010) at two remote sites, one at the NATSRL I35 research facility and the other on a Mn/DOT bridge in District One.

Software to process the captured waveforms, transmitted from a remote system to a central processing facility via cellular data modem has also been developed. In addition to archiving sensor responses, automated processing of current data at this central location can now be carried out for multiple installations. While the analysis involves sophisticated chemometric and statistical treatment of the data, the output variable is a single, simple Boolean (two state) value which indicates either safe or unsafe conditions (ice, frost, snow, or frozen electrolyte solutions) at the monitored sites. It is anticipated that this information can then be used to trigger signage to alert motorists to slippery conditions, to activate automated deicing systems, or to notify maintenance personnel of the immediate need for the application of deicing chemicals at the site.

Sponsor:

Project Details: