NORTHWESTERN UNIVERSITY
Infrastructure Technology Institute
TEA-21 Year 5
Semi-Annual Summary Research Progress Report

2133 Sheridan Road
Evanston, Illinois 60208
Phone: (847) 491-8165
Fax: (847) 467-2056

dschulz@northwestern.edu
www.iti.northwestern.edu

NONDESTRUCTIVE TESTING AND EVALUATION OF BRIDGES

Principal Investigator: David Prine, Institute Chief Research Engineer

Michigan Street Bridge

The Michigan Street Bridge is a rolling bascule type moveable bridge located in downtown Sturgeon Bay Wisconsin (Figure 1). This 1930 vintage structure is one of only two active bridges providing access to the Northern Door County peninsula. The bridge serves as a vital physical and economic link to the city. Wisconsin Department of Transportation (WisDOT) inspections from 1994 to the present have documented the ongoing deterioration of the structure.

Figure 1. Rack and Pinion Support Structure,
Michigan Street Bridge, Sturgeon Bay, WI

The Institute has been continuously monitoring the condition of the bridge for WisDOT since 1995. This system with further 2003 upgrades is now operating with autonomous data acquisition. We are now providing user specified notification (email, pager, fax, etc) and complex data reduction techniques to the Michigan Street website. During winter of 2004 ITI engineers performed "house cleaning" operations during an on-site visit. Some effort was also spent on further software improvements which include improved digital filtering to suppress noise spikes in the data to allow improved trend monitoring. The monitoring system continues to operate reliably and provide condition data in a user friendly format.

Computer-Aided Design Educational Assistance

Interest in CAD continues to be a high priority item. During this reporting period we organized a demonstration of Autodesk Inventor for thirty students and four faculty members (Figure 2). Two engineers demonstrated the software and gave away free copies of it.

Figure 2. Autodesk Inventor Demonstration

Surveying workshop

Northwestern's ASCE student chapter and the Infrastructure Technology Institute co-sponsored a workshop on surveying. In February Dr. Kurt Bauer explained the fundamentals of surveying and mapping to a group of students. A second workshop was held on Saturday, May 22, 2004. This workshop concentrated on instrumentation and surveying techniques. Professional surveyors Fred Campeau and Joerg Feldbinder lectured the group on the essential elements of surveying and then took the group outside and surveyed a building site (Figures 3-4). One group ran a level circuit and the other group used a total station to measure distance and angles. Every person was able to take a turn at setting up each instrument and taking shots. In the end, we were able to close the circuit horizontally and vertically. There was a discussion on errors and how to correct for them.

Figure 3. Professional surveyor Fred Campeau explains surveying instrumentation

Figure 4. Northwestern student gets hands on training in surveying

Monitoring of excavation at Ford Engineering Design Center construction site

ITI engineers assisted Professor R. Finno on a project to monitor soil movement around the excavation site for the new Ford Engineering Design Center. The assistance consisted of installing sensors in and around the excavation and in the adjacent McCormick Engineering and Applied Science building (Tech). Additionally, ITI engineers installed the instrumentation that allows remote continuous monitoring of the movement. In Figure 5 below a Leica total survey station (left) and a web TV camera (right) are shown mounted on the roof at Tech. The total survey station is automated and it is taking "survey shots" from the roof of Tech and recording the data on a computer. It is shooting a laser beam at various prisms mounted across the construction site and on stable reference points. This instrument will measure any movement of the building (Tech) that occurs during construction. The results of the Leica survey are available in a few seconds in a lab in the Tech building thanks to a wireless connection between the roof top mounted sensor and the laboratory computer. The immediate availability of the data is very important because the contractor has a requirement that total movement of the Tech building must not exceed 1.5 inches. Conventional means of measuring this movement make use of a laser scanner and the report summarizing the data is not available for several days. To the right of the Leica is a web TV camera. This camera allows remote observation of the excavation site by accessing the camera on our web site.

Figure 5. Instrumentation on the Roof of NU’s Tech Institute

Figure 6. ITI Research Engineer Dan Hogan welding a vibrating wire strain gage on a diagonal H- brace

A total of 34 vibrating wire gages and about a dozen strain gages were installed in the excavation under typical Chicago February weather conditions (Figure 6). Tech is built on a shallow foundation, about ten feet deep. The adjacent, nearby excavation for the Ford center is being excavated down to forty feet. Tilt and vibration sensors were installed in the basement of the Tech building to monitor potential movement of this building as a result of the excavation activities (Figure 7).

Figure 7. ITI engineers Dave Kosnik and Matt Kotowski installing tilt meters on building piers in the basement of Tech.

COMMERCIALIZATION OF INSTRUMENT FOR MICRO-INCH MEASUREMENT OF CRACK WIDTH IN SUPPORT OF CONTINUOUS REMOTE MONITORING FOR BRIDGE MANAGEMENT

Principal Investigator: Prof. Charles Dowding

First-Ever Commercial Instrument for Autonomous Crack Monitoring

During the fall of 2003, researchers on this project installed and began a two-year test program of the first commercial instrument developed specifically for autonomous crack monitoring. GeoSonics of Warrendale, Pennsylvania, has produced the beta test model (Figure 8) for validation under this project as GeoSonics also pursues additional validation.

Figure 8. First-Ever Commercial Instrument for Autonomous Crack Monitoring, Developed in Partnership with GeoSonics

This parallel deployment scheme was utilized by Prof. Dowding under this project and GeoSonics to enable GeoSonics to maintain clear ownership of any independently developed hardware or software. The system of parallel codeployment has allowed significant synergism as GeoSonics and project researchers can trade experience without GeoSonics fear of issues of ownership of intellectual property.

Figure 9. Installation of Beta Test Model of First Commercial Autonomous Monitoring Device New Berlin, Wisconsin

Both systems have been installed in the test house (Figure 9), which has been loaned to the project by Vulcan Materials Company, another co-deployment partner. The test house is a significant research asset as it is close to Northwestern and it subjected to blasting vibrations. Construction of such a test structure and artificially subjecting it to vibration would be a prohibitively expensive proposition. Response of the GeoSonics system will be compared to that of the project-developed system.

A MS thesis that describes the procedure for qualification of and the performance verification of the GeoSonics system is currently being written. It is expected to be completed by the end of this fiscal year. Another MS thesis will be written to complete the comparison and document the newest demonstration of ACM technology described below. This thesis is expected to be sometime in the next fiscal year.

Instrumentation of Historic Structure for FHWA- EFLD

ITI was asked by the Eastern Federal Land District of FHWA to install an ACM system on an important historic structure to facilitate completion of a critical road reconstruction in Washington DC (Figure 10-11). This installation was necessary to document the impact of construction vibration on the structure. The Kaman sensors were installed on both the interior and exterior of the structure as shown in the photographs. The ability of several agencies to observe results immediately and simultaneously was a key component in the interest in the system. Results will be published as part of the MS theses written by ITI sponsored students.

Figure 10. Installation Plan for Exterior Crack

Figure 11. Kaman Sensor Across Interior Crack

Papers and Articles Published

"Response of Cracks to Construction Vibrations and Environmental Effects" Dowding, C.H. and Snider, M. L. Proceedings of the GeoTrans Conference, Los Angeles, CA, ASCE, July 2004

CRACK AGE DATING

Principal Investigator: Prof. Charles Dowding

Billions of dollars are being spent on claims and litigation that allege cracking damage was produced by some recent adjacent anthropomorphic activity such as traffic, construction, blasting, etc. or some recent natural phenomena, such as earthquake, hurricane, flood, etc. Some time after the disturbance the concerned party inspects the facility and notices cracks. The observer often seeks compensation for the cracks from those that caused the disturbance or those who insure against associated perils or hazards.

Once a crack becomes the issue, attention then turns to what caused the crack. The date the cracking began more often than not plays a central role in the investigation. Therefore, quantitative methods to determine crack age would have a dramatic effect on many damage claims, as old cracks could be distinguished from the new.

This new project will focus on investigation of the possibility that biological methods could be employed to determine the age of cracks. As a first step, carbonation of cracks in cement is being investigated in detail as an analog to the biological methods. The photographs below demonstrate the penetration of the carbonation process into cracks subjected to 100% CO2 in a specially constructed chamber.

Carbonation is a more analytical technique used today to age cracks, but which is applicable only to cement based materials. As shown in Figure 12, the age of cracks is determined by comparing the depth of carbonation on the adjacent face of the material to that on the crack surface (Patty, 2002). Carbonation occurs when the cement, which is very basic with a pH around 11, reacts with carbon dioxide, a slightly acidic gas to change the chemical properties of cement. By measuring the depth of carbonation on the face and on the crack surface, a ratio of the depths of carbonation indicates the ratio of the age of the face surface to the crack surface.

Figure 12. Comparison of Carbonation on Crack Surface to Adjacent Edge

Principal Investigator:  Prof. Charles Dowding

Design & Installation of Inexpensive Radio Communication for Florida Sink Hole TDR Demonstration Site

In the summer of 2003, continuous communication was established between the Northwestern University data polling computer and the previously-installed TDR-tiltmeter site on state road 66 near Sebring, Florida (Figure 13). This site has been chosen as a permanent demonstration site for both TDR instrumentation and autonomous monitoring of site data (see 2002 annual research report). The site was a pioneer in the autonomous monitoring and Internet-based display of data which led to the launch of the Computer Data Systems business (see below).

Figure 13. Location of TDR Test Site near Sebring, FL

Engineers from the Infrastructure Technology Institute upgraded the communications system at the TDR site in June of 2003. The original installation used a cellular telephone for transmitting data to our web server. The cellular signal proved to be unreliable. The nearest land line telephone connection to the site was over a quarter of a mile away. Extending the line would have been prohibitively expensive. ITI engineers developed a wireless solution using point to point spread spectrum data radios. A modem and data radio were installed on a pole off site where telephone service was available. Another data radio was installed at the main instrumentation site to complete the link. Both sites are completely powered by solar panels.

This method of communications has proven to be very reliable and FLDOT is planning to install a similar system at a TDR site in the median of a divided highway in late 2004 or early 2005.

This project is being phased out as TDR cable technology has reached near commercialization: Companies such as GeoTDR in Columbus Ohio are routinely installing TDR systems; a book that describes the technology, GeoMeasurements by Pulsing TDR Cables and Probes, has been published through the support of ITI; and special sessions on TDR instrumentation have been held at TRB. The project was funded for this year at a level only sufficient to maintain the Florida site as a permanent demonstration of TDR and radio communication. It is anticipated that future funding will be further reduced to the minimum necessary to maintain the Florida site and the web page that supports TDR cable technology at www.iti.northwestern.edu/tdr/. No graduate student support was provided to produce papers; however two presentations were made on the topic of "Infrastructure Surveillance."

Lectures/Talks Given

"Surveillance of Critical Infrastructure by Internet" Rockford College, Rockford Illinois; 15 April 2004

"Surveillance of Critical Infrastructure by Internet" Illinois Department of Transportation, District 1, Shaumburg, IL; 3 June 2004

MEMS INSTRUMENTS FOR MICRO-INCH MEASUREMENT OF CRACK WIDTH IN SUPPORT OF CONTINUOUS REMOTE MONITORING FOR BRIDGE MANAGEMENT

Principal Investigator: Prof. Charles Dowding

This new project was initiated this year to investigate the feasibility of MEMS instrumentation for ITI remote monitoring. MEMS stands for Micro Electro Mechanical Systems. Development of these systems has been supported by DARPA and venture capital for some time now and the technology is becoming robust enough for application to infrastructure surveillance.

The first application is for investigation is Autonomus Crack Monitoring inside structures because these systems will eliminate wiring through their wireless communication. The photographs in Figure 14 compare the size of the MEMS technology. Equipment has been purchased, activated, and programming tasks have been distributed to investigate feasibility.

Figure 14. Temperature and Humidity and Data Logger/Radio Instruments

MOBILE INFRASTRUCTURE CLASSROOM

Principal Investigator: Prof. Charles Dowding

An initial testing phase has been completed in this project. It involved eight engineering undergraduates, split into two teams to evaluate the feasibility of existing technology for developing a mobile PDA-GPS-Phone (PGP) technology. One team investigated a low cost, non GPS system and the other the PDA-GPS system. These students were part of the Engineering Design and Communication class at Northwestern University.

The object of this project is to ultimately develop PDA-GPS-Phone (PGP) technology as a "virtual infrastructure teacher." Such a purpose-built location-specific learning system would allow educational, civic, cultural, and infrastructure organizations to vastly improve their ability to offer multi-level self-selected information on infrastructure to high-end continuing education and K-12 students in the field. Examples include location-specific self-guided architecture tours, and bus or boat tours offering educational experiences on bridges, buildings, and other infrastructure. Eventually, PGP could allow people to ask "what if?" questions, and delve deeper into technology and history questions of interest. The technology allows full immersion in the subject itself: a city or a civil war battle field with access to a wide range of location-specific information.

Use of these two systems was tested on the Northwestern campus in a very preliminary fashion. The campus was chosen for content as the navigation challenges were similar to those that would be encountered in the downtown Chicago, where the large infrastructure would be located. One of the user interfaces is shown in Figure 15. It is planned to continue this project next year in the conjunction with the Engineering Design and Communication classes at Northwestern to both develop the system as well as to interest students in the infrastructure.

Figure 15. Self-Guided Infrastructure Tour Using PDA

NONDESTRUCTIVE EVALUATION OF CONCRETE WITH GUIDED WAVES

Principal Investigator: Professor Richard J. Finno

Purpose

The goal of this project is to develop methods to non-destructively evaluate the condition of concrete components of bridges, including columns, piers and foundation elements by extending guided wave theory to flexural wave propagation in concrete cylinders and to longitudinal and flexural waves in embedded plates (representing cast-in-place and soil mixed walls).

Many times one cannot access the top of a cylindrical element, but can only place instruments on the sides of the member, for example, at a pier for a bent of a bridge. When energy is added at the side of a column, one can generate flexural waves that have different propagation characteristics than longitudinal waves. Flexural waves in general are dispersive, i.e., their propagation velocity depends on frequency. Furthermore, the displacements associated with flexural waves also depend on frequency. To use these waves to non-destructively evaluate a member, one must have the dispersion relation in hand to know how fast the waves travel and to locate the optimal position of a transducer to measure the response. To extend non-destructive techniques to allow application of higher frequencies, and hence create the ability to sense smaller defects, this project will develop numerical solutions for flexural wave propagation in cylinders and plates and for longitudinal propagation in plates. In addition to the numerical work, experimental verification of the numerical solutions will be made on prototype cylinders and plates.

Summary of Progress

We have found numerically solutions to evaluate flexural wave propagation in a concrete cylinder or pile, assuming the cylinder is a wave guide. The theoretical evaluations include derivation and numerical solution of the governing equations and parameter analysis. The graphical representation of the dispersion curves for the flexural waves in cylinders is shown in Figure 16. The real part of the solution can be used to define the flexural wave velocity of any mode and the imaginary part of the solution can be used to define the geometric attenuation when the cylinder is embedded in the ground.

Figure 16. Non-dimensional frequency versus wave number for flexural waves

We have also extended the guided wave theory to embedded plates so that we can develop techniques applicable to in situ walls, such as structural slurry walls or soil-mixed walls, that comprise part of many excavation support systems. The question of integrity of these wall systems arose a number of times during construction of the Central Artery/Tunnel project in Boston and during construction of the secant pile wall at the Chicago-State Subway Renovation project. During our work on these projects, we found that the conventional techniques based on 1-dimensional wave propagation in a cylindrical structural element were inadequate to provide answers regarding integrity of the as-built walls.

The theoretical and numerical solutions are complete, and are the subject of the PhD dissertation of Mr. Helson Wang. He is currently writing his thesis, which should be finished by the end of summer 2004.

We also have used the results to interpret the results of flexural wave identification tests, a conventional non-destructive technique that use relative low frequency (assumed non-dispersive) flexural waves to evaluate integrity of concrete piles and shafts.

In addition to the theoretical development, we need to explore the ramifications of the results to learn how to best induce and measure flexural guided waves. The components needed to conduct the guided wave tests have been assembled and software written such that a prototype version of the system is operational. We have conducted laboratory verification tests of the system. A concrete cylinder was cast, and will be tested as both a free-standing column and an embedded pile. In tests wherein flexural waves are to be induced, two experimental set-ups were employed. First, a single shaker was mounted in the center of a cylinder with incident angles of 45° and 90° as shown in Figure 17. In these cases, multi-axial accelerometers are mounted on the top surface of the concrete to measure the response and verify the mode of the received signals.

Figure 17. Shaker and transducer arrangement for flexural wave tests

In the second arrangement, a shaker was mounted on the side of a concrete cylinder, and triaxial accelerometers were mounted on the sides of the cylinder. A photo showing this testing arrangement on the concrete cylinder cast for this work is shown in Figure 18.

Figure 18. Side-mounted shaker for flexural wave tests

The verification testing is on-going. A typical result is shown in Figure 19. The input signal is represented as the force-time trace measured at the shaker. The response of the concrete cylinder is expressed voltage measured by accelerometers placed at two locations on the side of the cylinder. The responses shown are in the radial direction for accelerometers mounted at 180 degrees from the axis of the shaker. The dispersive nature of the input energy is clearly seen by the appearance of distinct phase and group periods.

Figure 19. Typical responses measured in flexural wave test

After experimentation on the free-standing column is completed, the column will be embedded at the National Geotechnical Experimentation Site at Northwestern, and tests conducted to evaluate the theoretically-predicted attenuation characteristics of flexural waves.

Personnel

The principal investigator for this project is Professor Richard Finno. James Lynch and Helsin Wang assist him as they pursue PhD degrees.

Publications

Publication of results in journals and conference proceedings also will continue. It is anticipated that both Helsin Wang and James Lynch will complete their dissertations by the end of the project. Their theses will serve as the final report and it is anticipated that a number of papers will be published from each thesis.

AUTOMATED DEFORMATION MONITORING

Principal Investigator: Professor Richard J. Finno

Purpose:

The purpose of this work is to improve the state-of-the-art and practice of predicting and controlling ground movements associated with supported excavations and tunneling operations, and the consequent deformations of adjacent structures and utilities. To accomplish this, we are automating the data collection, data transmission and interpretation of performance-related quantities associated with construction operations. In particular, optical survey, tilt meter and strain gage measurements are autonomously obtained, transmitted to a remote location for processing, and processed automatically so that the results can be interpreted in a timely fashion.

These improvements will be tested in the field in real time during the excavation for the Ford Research Center on the campus of Northwestern University in Evanston, IL. These techniques will have application to any project where excavations may impact adjacent structures or utilities wherein deformations must be measured to verify performance or control construction operations.

Summary of Progress

As of June 2004, the excavation for the Ford Center has reached its maximum depth and construction of the building has begun. We have deployed at the site a total station to measure ground movements, tilt meters to measure structural responses of the adjacent Technological Institute and strain gages to measure forces in the temporary support members.

We have deployed a Leica TPS1100 Professional series total station and have developed software to acquire and process the measured data. The total station has been in service through the winter. A photo of the system in service atop the Technological Institute is shown in Figure 5 above. Also shown in the lower right side of the photo is the radio link that allows wireless communication between the total station and a remote computer.

The total station remotely and automatically sensed the lateral and vertical movements of 9 survey points established around the Ford Center excavation. The deformations are computed by comparing the location of a particular survey point with those of 2 fixed reference points, established on buildings near the ground surface several hundred ft from the excavation. Typical results from four points are shown in Figure 20, which plots both settlement and excavation elevation versus time. The development of the settlements as the excavation progresses is clearly seen. The accuracy of the data is approximately ± 5 mm.

Figure 20. Settlement versus time from total station data

In addition to the optical survey data, we have deployed 20 tilt meters inside the Technological Institute to help record the response of the building to the adjacent excavation. These systems have also been installed so that the tilts of the columns and walls can be recorded. A photo showing a typical installation on a column is shown on Figure 21a. Shown in Figure 21b is a typical communications center, consisting of an A/D converter and a free wave transceiver that allows automatic and remote data acquisition.

(a) Tilt meter (b) Communication center

Figure 21. Tilt meter installation

Figure 22 shows typical data recorded by the tilt meters in the Physics Annex portion of the Technological Institute. For reference, the excavation progress is also shown.

Figure 22. Typical tilt meter data

We have also installed 36 vibrating wire strain gages on the diagonal and cross-lot bracing which provides the lateral support of the temporary support structure for the excavation. These data are collected to compare with an automated, remote sensing strain gage system under development. Figure 23 shows vibrating wire strain gages installed at the quarter points of a diagonal bracing element, as well as the communications link and resistance gage for the automated system. Note that the vibrating wire strain gages have performed well to date – 31 gages remain in service after 6 months. Problems related to the transmission distance for the signals of the automated strain gages are still being evaluated.

Figure 23. Strain gage installation

Typical processed strain gage data are shown in Figure 24. The total force and extreme fiber stresses are shown during a 6 week period. The large bending stresses shown in the middle of April are a result of the excavator constructing a temporary ramp atop the brace for access to the bottom of the excavation. This is a very atypical loading for an internal bracing element.

Figure 24. Typical strain gage data from cross- lot brace

We are currently analyzing all data to develop a comprehensive understanding of the overall measured response. Additional ground movements and structural responses are expected when the internal braces will be removed in August 2004.

Personnel

The principal investigator for this project is Prof. Richard Finno who is assisted by Tanner Blackburn, a PhD candidate in the Department of Civil and Environmental Engineering. Messrs. Dan Marron and Dan Hogan from ITI helped in setting up the remote access for the total station, and installing the tilt meters and strain gages.

Publications

After only 6 months and with the excavation still underway, no articles have been prepared. However, the following abstract has been submitted:

Finno, R.J. and Blackburn, J.T., "Diagonal Supports and Corner Effects in Braced Excavations," submitted to the 16^th International Conference on Soil Mechanics and Geotechnical Engineering, Osaka, Japan, 2005.

The principal investigator will present the results of all aspects of the study to professional societies both locally and nationally. Papers will be submitted to Journals and future conferences to further disseminate the findings.

BRIDGE ASSET MANAGEMENT BASED ON LIFE-CYCLE COST CONSIDERATIONS

Principal Investigator: Prof. Raymond J. Krizek

Associate Investigators: David Novick, Prof. Ahmad Hadavi, Prof. Pablo Durango-Cohen

Bridge asset management based on life cycle cost considerations provides a resource allocation framework for cost-effective decision-making on how to build, preserve, and improve a bridge to minimize its life cycle cost while achieving a satisfactory level of service over the useful life of the bridge. The objective of this project is to use actual historical cost data from a variety of geographically distributed bridges of different structural designs to formulate a cost model for bridge life cycle cost, assess the impact of deferred maintenance on bridge total life cycle cost, and develop a supporting rationale for projecting the useful life of a bridge. Guidelines will be prepared to suggest the actions that must be implemented to achieve the target useful life. The results of this study will provide critically needed supplemental input information for currently used bridge management systems, such as PONTIS, and more recent bridge life cycle cost analysis tools, such as BLCCA; input to these systems is currently obtained primarily from expert elicitation.

Chicago Movable Bridges

Chicago has more movable bridges than any other city in the world. Evolved from earlier types of movable bridges, including swing bridges and rolling lift bridges, the double-leaf trunnion bascule type most satisfactorily met the existing conditions and became most popular in Chicago. Among the 63 original movable bridges, 32 are single-deck double-leaf trunnion bascules, 4 are single-deck single leaf trunnion bascules, and 3 double-deck double-leaf trunnion bascules. Based on popularity, bridge age, and availability of historical cost information, this study concentrates on 21 single-deck, double-leaf trunnion bascule bridges built between 1902 and 1936.

Classification of Chicago-Type Bascules

Based on bridge design, the Chicago-type trunnion bascule bridges can be classified into four generations (Hopson, 1994).

First Generation

The first generation of Chicago-type bascule bridges are those built between 1900 and 1910. They all have tall through-trusses braced over the roadway. While the operating machinery and counterweights are below the deck, each leaf is turned by the action of fixed pinions that rotate and mesh with curved racks on the tail-ends of the three trusses arching above the roadway. Originally, these bridges had simple, wood frame operator houses and were operated manually. Table 1 is the list of first generation bascules chosen for this study. Among them, the Cortland Street Bridge, built in 1902, is the first Chicago-type bascule.

¹Initial construction cost for superstructure and substructure; the actual cost before adjustment.
²Over canal

Table 1. First generation Chicago-type bascules chosen for this study.

Second Generation

The second generation of Chicago-type bascule bridges consists of those built between 1910 and 1930. In 1911 Alexander Von Babo, a city bridge engineer, patented the internal rack. By placing the rack below the bridge deck, the need for through trusses was eliminated. The newly appointed Chicago Planning Commission and its consulting architect, Edward H. Bennett, began in 1910 to work with engineers to improve the artistic quality of Chicago bridges. The efforts of this group resulted in extensive revisions to the shape of the trusses, the configuration and façade of the operator houses and pit walls; and the ornamental detailing of sidewalk railings, light fixtures and other ornamental metal elements. The new plan was heavily influenced by Parisian architecture, which, at the time, was considered to be the model of urban design. Within the same generation, the bridges located in the downtown business center or the main channel are more sophisticated, such as the Clark Street Bridge, which has two well ornamented bridge houses rather than the usual one bridge house. Table 2 is the list of second generation Chicago-type bascules chosen for this study. Among them, the Washington Street Bridge is the first trunnion bascule bridge reflecting the new aesthetics.

Table 2. Second generation Chicago-type bascules chosen for this study.

Third Generation and Fourth Generation

From an architectural point of view, the period between the great depression and World War II (1930-1945) constitutes the third generation in bridge design. After the consulting architect, Edward Bennett, was dismissed in 1930, design engineers wished to project a more contemporary image, and the state of the economy encouraged streamlining. So the bridges built during this period followed established structural and architectural models, but exhibit simpler ornamental details. There are several movable bridges built after 1945, and classified as the fourth generation of Chicago-type trunnion bascules. They are sleeker and technically more sophisticated than their predecessors. This study concentrates on older bridges which were built before 1940 and of have had a relatively long useful life. Table 3 is the list of third generation bridges which have been chosen.

Table 3. Third generation Chicago-type bascules chosen for this study.

Total Life Cycle Cost Analysis

Total life cycle cost of a bridge is the total investment on a bridge; it consists of the initial construction cost and all maintenance, repair, and rehabilitation costs throughout the life of the bridge. The design and rehabilitation of bridges based on life cycle cost considerations has the potential to save hundreds of millions of dollars in future construction costs (Novick, 1991). Therefore, it is important to determine the total life cycle costs for different types of bridges.

Data Base

This study collected all available actual cost data from the Chicago Public Library and the Chicago Transportation Department for each bridge up to the year 2002. Over this long history, the data were recorded in many different reports and it took a long time to collect and verify them. The MRR costs for the years from 1900 to 1978 are complete (except for several years). For the year after 1978, the repair costs and rehabilitation costs were collected from the capital improvement program report. There is no detailed maintenance cost report for individual bridges, but a lump sum number is reported for all city bridges (about 250). For all missing cost information, assumptions are made based on history and engineering judgment. By using the Engineer News-Record (ENR) Building Cost Index (BCI), the actual costs are converted to constant costs to account for inflation or deflation impact.

Relationship between Structure Size and Initial Construction Cost

The design identity and availability of historical costs provide a rare opportunity to analyze the relationship between structure size and initial construction cost for Chicago-type bascules. The results are expected to benefit the conceptual estimate for new bridges. In this study, bridge structure size is defined as Bridge Structure Length X Bridge Deck Width, and the initial construction cost includes the cost for the superstructure and substructure only. All costs are adjusted to year 2002. All first generation bridges listed in Table 1 are included except bridge 1-3 for which the deck type/thickness is different from the others. Of the second generation bridges listed in Table 2, 8 bridges are selected; bridges 2-6, 2-10, and 2-14 with different decks and bridges 2-2, 2-7, and 2-12 which are special, were eliminated. This study omits the third generation bridges because of their small number. Figure 25 shows that initial cost and structure size are approximately linearly related.

Figure 25. Size and initial construction cost relationship for Chicago-type bascules.

Achievable Useful Life

There is argument regarding the definition of bridge useful life. This study defines the achievable useful life as the period from where the bridge is built to its reconstruction or replacement. Most of the Chicago-type movable bridges studied have never been reconstructed, and their useful life is beyond 75 years, which is an often used bridge design life. The first Chicago-type bascule, the Cortland Street Bridge, is now 102 years old.

Cost Distribution over Time

The cost data shows that, after the movable bridges are built, they require an annual cost to maintain the operation. Three or four decades later, they usually had major repairs or rehabilitation. The annual costs fluctuate considerably, but the accumulated costs present certain trends over time. Figure 26 shows the accumulated cost distribution over time for the Cortland Street Bridge to demonstrate the general pattern.

Figure 26. General pattern of accumulated cost distribution over time.

Relationship between Total Life Cycle Cost and Bridge Age

To determine the total life cycle cost model for these movable bridges, this study considered all possible impacting factors, including the on-bridge traffic volume (ADT and ADT growth rate), bridge annual opening times, and bridge age. To normalize the analyses, the ratio of the total life cycle cost (TLCC) and the initial cost (IC) are calculated for each bridge. There are no obvious relations found between either the TLCC/IC ratio and the traffic volume or between the TLCC/IC ratio and bridge openings. However, there is a strong relationship between the bridge age and the TLCC/IC ratio for the Chicago-type bascules, as shown in Figure 27.

In general, the older the bridge, the higher the ratio of TLCC/IC, which is as expected. It is interesting to know the total cost of the bridge compared with its initial cost. For these first generation Chicago-type bascules, which are around 100 years old, the total cost is about 4.5 times the initial cost.

Figure 27. Relationship between total life cycle cost and bridge age.

MRR Impact on Total Life Cycle Cost

Although the total life cycle cost of a bridge is determined primarily by the bridge age, as illustrated by Figure 27, bridges of the same age may have a different TLCC/IC ratios, which means that the total life cycle costs are different. In practice, the bridge manager always needs to make the decision when and how to rehabilitate the deteriorated bridge; treat it now or later, extensive rehabilitation or minor repair. This study attempts to determine the impact of MRR strategy on total life cycle cost so as to benefit the bridge management based on life cycle cost considerations. Table 4 gives the scenario of major capital improvement histories for two bridges which were built at almost the same time with identical design. Bridge 1-4 was rehabilitated in 1931 and bridge 1-5 had two major repairs in 1955 and 1966. They were both scheduled to be reconstructed in 2002. By 2001, the TLCC/IC ratio and reconstruction cost for bridge 1-4 was lower than that for bridge 1-5, and bridge 1-4 was in better condition. Figure 28 shows that, although the TLCC/IC ratio for bridge 1-4 was higher after it received the rehabilitation in 1931, 22 years later the ratio was lower than that for bridge 1-5. Up to reconstruction, the total cost saving for bridge 1-4 was 4.64 million US dollars. In other words, the delayed rehabilitation cost 3.69 million US dollars more for bridge 1-5. Our analyses of other Chicago-type bridges also showed that, in general, timely and complete repair and rehabilitation will save money eventually.

¹ Real cost adjusted to year 2002
² Not including cost for reconstruction in 2002

Table 4. Major capital improvement comparison.

Figure 28. Demonstration of MRR impact on bridge total life cycle cost behavior.

Distribution of Maintenance Cost

A movable bridge usually needs routine maintenance to support its operation. Maintenance costs for movable bridges are higher than for fixed bridges. Another purpose of this study is to determine the maintenance cost and its distribution among major items. The results may help the bridge manager to budget maintenance funds. Table 5 gives the average results for Chicago-type trunnion bascules based on the historical maintenance cost records from 1929 to 1966.

^*Real cost adjusted to year 2002

Table 5. Maintenance costs and distribution among major items.

California Bridges

This research tried to collect historical cost information for highway bridges from the California Department of Transportation (Caltrans). Based on the availability of initial construction costs, 48 bridges were identified from over 300 bridges which were built before 1940 and have had no reconstruction. Furthermore, 26 bridges having clear MRR cost records were chosen for detailed analysis. These 26 bridges are cataloged into 5 groups according structural type. All historical costs are adjusted to year 2002 by using the ENR Building Cost Index. The following is a brief summary of the research completed so far.

RC Arch Bridges

Our data collection includes 7 reinforced concrete arch bridges (Table 6) which were built in the early 1930s and are the hallmark of California highway bridges.

Table 6. California RC arch bridges.

The total life cycle cost in terms of the TLCC/IC ratio over the bridge age is described in Figure 29.

Figure 29. Life cycle cost for Caltrans reinforced-concrete arch bridges.

RC Girder Bridges

According to our interview with Caltrans bridge professionals, concrete arches are rather labor intensive and concrete slabs and girders evolved into the more common bridge types, especially for shorter span bridges. Concrete girder bridges were built extensively from the 1930s into the 1950s. Our research found RC girder bridges to be dominant in our data base, as listed in Table 7.

Table 7. Caltrans reinforced-concrete girder bridges.

The total life cycle cost in terms of the TLCC/IC ratio over the bridge age is described in Figure 30.

Figure 30. Ratio of total life cycle cost to initial cost over bridge age for Caltrans RC girder bridges

RC Box Girder Bridges

The reinforced concrete box girder bridge was developed around 1940 and became widely used by Caltrans. Our study includes two of the earliest RC box girder bridges, as listed in Table 8. The Eel River Bridge is the first RC box girder bridge built in California.

Table 8 Caltrans RC box girder bridges.

The total life cycle cost in terms of the TLCC/IC ratio over the bridge age is described in Figure 31.

Figure 31. Ratio of total life cycle cost to initial cost over bridge age for Caltrans RC box girder bridges

Steel Bridges

The steel bridge is one of the oldest types of bridge in California. This study includes 3 steel bridges, as listed in Table 9.

Table 9. Caltrans steel bridges.

The total life cycle cost in terms of the TLCC/IC ratio over the bridge age is described in Figure 32.

Figure 32. Ratio of total life cycle cost to initial cost over bridge age for Caltrans steel bridges

Steel & Concrete Composite Bridges

Data for two steel and concrete composite bridges are listed in Table 10.

Table 10. Caltrans Steel and concrete composite bridges.

The total life cycle cost in terms of the TLCC/IC ratio over the bridge age is described in Figure 33.

Figure 33. Ratio of total life cycle cost to initial cost over bridge age for Caltrans steel & RC composite bridges

Basic Findings

Our study has preliminarily reached following basic findings for the above-listed old bridges from Caltrans:

• The bridges studied are 64 to 77 years old in 2003
• TLCC/IC ratios range from 1.0 to 4.6, with an average of 1.31
• Major maintenance items are
  
  • Clean and reseal joints
  • Clean bearings
  • Patch concrete spall
  • Deck overlay
  • Repaint
  • Rail repair
    
• Major repair occurs at an age of about 50 years
• Steel bridges tend to be replaced by RC bridges

New York bridges and tunnels

We recently obtained detailed maintenance and rehabilitation cost data for 4 bridges and two tunnels, as listed in Table 11, from the Port Authority of New York & New Jersey. We are still in the process of determining whether the data are complete and in sufficient detail.

Table 11. New York Bridges and Tunnels.

Cost index selection

Available cost indices include the ENR Building Cost Index, the ENR Construction Cost Index, the Consumer Cost Index, the Producer Price Index, the Highway Price Trend (by nation and individual state), and the California Bridge Construction Cost Index. Thus far, this study has used the ENR Building Cost Index (BCI) to convert historical cost data into year 2002 cost data based on the following considerations:

• The ENR Cost Indices have the longest history and are the only suitable cost indices for bridges built before 1913, such as the Cortland Street Bridge. As shown in Table 12, Product Price Index is the newest one, and obviously does not meet our needs.

Table 12. List of available cost indices.

• The Highway Price Trend includes the Structures Index, which covers bridges, but the index from 1919 thru 1960 was a mathematical conversion from the index created from the 1937 through 1941 Base Year; from 1961 to date used the actual indicator items and prices. Our study found such mathematical conversion may not be able to reflect the real cost of inflation, because these converted cost indices fluctuated less, as shown in Figure 34.

Figure 34. Cost index comparison for years 1922 to 1980.

• The ENR Building Cost Index seems to be an average value of all considered cost indices, as shown in Figure 35.

Figure 35. Cost index comparison for years 1966 to 2001.

• Both the ENR Building Cost Index (BCI) and the Construction Cost Index (CCI) are 20-city national averages and apply to general construction costs. The difference is in their labor component. The CCI uses 200 hours of common labor, multiplied by the 20-city average rate for wages and fringe benefits. The BCI uses 68.38 hours of skilled labor, multiplied by the 20-city wage-fringe average for three trades–bricklayers, carpenters, and structural ironworkers. The CCI can be used where labor costs are a high proportion of total costs. The BCI is more applicable for structures. Bridges may be regarded as structures.

The Materials Science of Concrete – A Web Site

Principal Investigators:  Prof. Jeffrey Thomas and Prof. Hamlin Jennings

Project Overview

This project is creating an electronic document and an associated web site that will provide free, up-to-date, understandable information about the materials science of cement and concrete to anyone with access to a computer and the internet. That there is a demand for such a resource is demonstrated by the popularity of a similar site on modeling of cement developed and hosted by NIST, which receives more than 10,000 independent hits per month. Motivating factors behind the creation of this web site include increasing the awareness of the importance and technical complexity of concrete as a material, improving the overall quality of concrete by promoting good mix designs and curing practices, and providing exposure, credibility, and prestige to the entity that hosts it.

Structure of the site

As presently envisioned, the home page of the site will introduce the site and give the user the following content options:

Science of cement document: This will be the heart of the site. An electronic document with several chapters devoted to the chemistry, microstructure, and behavior of cement and concrete. The content will not be "dumbed down" but technical terms will be defined carefully and writing style will be more accessible and informal style than in scientific publications. Content discussed in the next section.

Question and Answer: A list of common questions about cement and concrete with varying degrees of complexity. Each question will have short answer followed by cross-reference links to the main document for more in depth information. This is envisioned as a more user-friendly interface to draw people into the subject.

Interactive microstructure: Micrographs of cement paste and concrete at different magnifications will be labeled for the various phases. Clicking on the phases will provide information on their morphology and contributions to the properties.

All three of these sections will be designed in such a way that the site can be easily updated and expanded over time.

Content of the main document

A detailed thirteen chapter outline has been produced, and writing is under way. A brief summary of the content follows below.

Chap. 1: Introduction: Describes the purpose of the monograph and defines some basic terminology such as cement paste, curing, setting, etc.

Chap. 2: Concrete Basics: Part 1 describes the basic components of concrete (cement, aggregate, water, admixtures, etc). Part 2 gives an overview of the construction process from mix design to placement and consolidation. Written at a simpler level than remainder of document.

Chap. 3: Manufacture of Portland Cement: The focus is on the mineral composition of the clinker, and different types of Portland cement. Environmental aspects (emissions, energy use) are also discussed.

Chap. 4: History of Concrete: Brief description of the development of concrete from ancient times to the present, focusing on the technical advances in ingredients, kiln temperature, and overall understanding of the chemical processes.

Chap 5: Hydration and Microstructure of Portland Cement Paste: The hydration reactions of the individual minerals, morphology of the hydration products, the composition of the pore solution, and the overall microstructural features of cement paste (e.g. carbonation, ITZ).

Chap. 6: Hydration and microstructure of blended cement pastes: Manufacture and composition of slag, fly ash, and silica fume, the pozzolanic reaction, the specific features and properties of blended cements.

Chap. 7: The pore structure and surface area of cement paste: Classification of pores, techniques for measuring porosity and surface area, effects of curing and mix design on the pore system and surface area.

Chap. 8: The chemical and nanometer-level structure of C-S-H phases: Focus on the colloidal structure and aspects of C-S-H gel. Includes much recent research and modeling by the PIs.

Chap. 9: The kinetics of hydration of C₃S: The early nucleation and growth kinetics, activation energy, the later diffusional kinetics. Relationship between the kinetics and the properties. Effects of accelerators and retarders.

Chap. 10: Shrinkage and creep of concrete: Mechanisms of shrinkage and creep, microstructural changes on irreversible deformation, combined loading and drying.

Chap. 11: Mechanical properties of concrete: Compressive and tensile strength as a function of microstructure, elastic modulus and hardness at the micro and nano scale, the size effect in concrete. High-strength concrete.

Chap. 12: Durability of concrete: Permeability of concrete, effects of leaching, sulfate attack, alkali-silica reaction, freeze-thaw, etc. Designing durable concrete.

Chap. 13: Properties of fresh cement paste and concrete: Workability, segregation, rheology, effects of plasticizers. Self consolidating concrete.

Current Project Status

The electronic document is currently being written by the PIs. In the meantime, the framework of the web site is being put into place by an undergraduate student, Rance Barber. The initial version of the site should be finished by the end of the summer.

USING SIMULATION TO PLAN TRAFFIC INCIDENT MANAGEMENT STRATEGIES

Principal Investigators:  Prof. Joseph L. Schofer, David Schulz, &  John Wirtz

Since September 11 there has been greatly heightened concern for the security of America’s vital infrastructure systems.  Efforts have been undertaken to identify possible measures for preventing infrastructure disruptions and dealing with them if they occur. Some funds have been appropriated to begin to attack the problem and perhaps most importantly, planning for disruption prevention and management has increased dramatically, as has coordination among responsible agencies.  As indicated in Figure 36, planning to mitigate the impact of terrorism-caused disruptions occurs five levels:  incident response and management, recovery, protection, impact reduction, and preemption.

Figure 36. Conceptual Model of Infrastructure Disruption Mitigation Planning

While the current concerns regarding disruption of infrastructure systems are spurred by threats of terrorism, transportation systems are routinely disrupted by incidents, including planned special events, crashes, law enforcement actions, fire and hazardous material situations, equipment and facility failures, severe weather, and other acts of God.

This project uses a dynamic traffic assignment model to explore the congestion reduction benefits of pre-planning incident management strategies. Incidents of varying scale and duration are simulated and several different mitigation strategies are tested for a single location on the Chicago area expressway system. Visual Interactive System for Transportation Algorithms (VISTA) software is used to assign all vehicles to time-dependent shortest paths and measure the resulting vehicle density on highway links. Vehicle density data is converted to level of service (LOS), allowing disruption impacts to be quantified on individual highway links in units of lane-mile-hours at LOS "F." This allows the determination of the best response action for a given incident scenario and measures the expected spread of congestion to critical alternate routes.

The best response actions are found to vary by geographical level of analysis as well as by incident scale and duration. Three areas are examined, the directly impacted expressway, a sub-network that absorbs the diverted traffic called the indirectly impacted zone, and the northern half of the regional highway network. Closing expressway entrance ramps upstream of a simulated incident is effective at reducing traffic congestion in the indirectly impacted zone during large-scale incidents - multiple lane closures for several hours. The best response to minimize congestion on the directly impacted expressway is more dependent on the incident duration than scale, with traffic benefiting more from closures during long-duration incidents. Region-wide average vehicle delay data indicate that the impacts of any disruption may be mitigated by closing various numbers of upstream expressway entrance ramps, the number of closures increasing with scale and duration.

Table 13 shows the number of alternate routes to the expressway predicted to be impacted by congestion in the tests conducted for this research. Knowing which alternate routes will become congested during a particular incident can aid in planning rerouting and determining what to recommend to motorists via the various channels of information dissemination. In this study, parallel alternate routes are only impacted during incidents causing a full expressway closure, thus providing a definition for a "major disruption." When such large incidents are encountered, responsibility for network congestion mitigation should be transferred from first responders to a transportation agency, which has a view of and responsibility for network-wide operations. It is also notable that arterials perpendicular to the directly impacted expressway were always congested in these tests, and thus limited access to the parallel alternate routes. This suggests that pre-planning activities should include increasing capacity on these arterials, partially through improved signal timing and synchronization plans.

Table 13. Number of Alternate Routes Impacted by Incident Type.

In addition to supporting pre-planning of responses such as upstream ramp closures, simulation tests show that faster implementation of beneficial response actions might further reduce congestion. This suggests that a remotely controlled barrier gate system on freeway entrance ramps may mitigate congestion by allowing quick closures in response to major incidents. Congestion is shown versus response time for one simulated incident scenario in Figure 37.

Figure 37. Congestion vs. Response Time - Full Closure of Edens Expressway

The current research supports the hypothesis that pre-planning more creative and effective incident responses, and implementing these effective actions more quickly, will help mitigate disruption impacts. More research is needed testing various incident locations to determine the feasibility of creating generalized incident response templates, rather than location-specific responses.

This work has also shown the value of advanced simulation tools, particularly dynamic traffic assignment, for planning and evaluating incident management strategies. Eventually, models that run faster than real time would provide a basis for making more informed responses to large, complex, and unexpected roadway incidents.

NONDESTRUCTIVE DETERMINATION OF EARLY-AGE CONCRETE PROPERTIES WITH AN ULTRASONIC WAVEREFLECTION METHOD

Principal Investigator: Surendra P. Shah

Introduction

A technique for in-situ monitoring the setting and hardening of concrete was developed at the Center for Advanced Cement-Based Materials. The experimental procedure is based on high-frequency ultrasonic measurements and consists of monitoring the reflection loss of ultrasonic shear waves at the interface between a steel plate and the hardening concrete. A transverse wave pulse is transmitted into the steel and reflected at the steel-concrete interface. The loss of the amplitudes of the first and second reflections due to transmission losses at the interface between the steel and the hydrating concrete is calculated and monitored over time. The reflection process is schematized in Figure 38. The test setup is presented in Figure 39.

Figure 38. Multiple Reflection Process as Steel-Concrete Interface

a) the transducer transmits a pulse into the steel plate

b) for fresh concrete: the pulse is entirely reflected at the steel-concrete interface (T-waves do not propagate in liquids) for hardened concrete: the pulse is partially reflected and transmitted at the steel-concrete interface

c) the reflected pulse encounters the steel-transducer interface and is reflected again

Figure 39. Experimental setup

Research Objective

The previous research has shown that the wave reflection method can predict in-situ early age compressive strength of concrete. This was shown in a successful field trial in a precast plant. The current research has the objective to eliminate the need for calibration of the strength prediction procedure. This can be achieved by developing a strength prediction concept that uses and combines essential relationships determined by experiments and numerical simulations. The primary purpose of this concept is to provide to relate wave reflection measurements, key mechanical properties and fundamental microstructural characteristics of the materials tested within this study. The results of the experimental and numerical investigations conducted so far indicate that the reflection loss is governed by the properties of the cement paste or mortar phase of the tested materials. Consequently, the constitutive material model to be developed will provide the basis for determining early age properties of cement paste and mortar based on composition (e.g. w/c-ratio) and reflection loss measurements. The final goal is to develop the strength prediction model in a form that requires no further experimental calibration.

The conducted investigations were also focused on the relationship between the wave reflection measurements and viscoelastic parameters of cement paste. Viscoelastic parameters, such as viscosity, can be of importance when working with self-compacting concrete. The experiments have shown that the wave reflection method can reliably measure the viscosity of cement paste.

Strength Prediction Model

The model is based on relationships originating from two different concepts. One component includes the results obtained from the experimental studies. The investigations on the relationship between the reflection loss and microstructural parameters of the tested materials have shown that the reflection loss has a unique relationship to the gel-space ratio of the cement paste. This relationship was found to be independent from the w/c-ratio. In addition, the experiments have shown that the gel-space ratio has an unique relationship to the compressive strength of the tested cement pastes and mortars, which makes this gel-space ratio concept very useful for strength prediction purposes. The second component of the model is based on the results of the numerical simulations using HYMOSTRUC3D. It was found that the numerically determined specific contact area, as a parameter of cement paste, is uniquely related to the reflection loss measured on mortar. This relationship is also not influenced by the w/c-ratio. The specific contact area itself was found to be uniquely related to the compressive strength of the tested mortars. The outline of the proposed model is given in Figure 40.

In the following it will be shown how the mentioned relationships can be used to determine the development of the compressive strength of plain cement paste and mortar based on reflection loss measurements. First, each relationship will be analyzed and evaluated separately. For the benefit of a higher prediction accuracy the relationships will then be combined to a single model by means of multiple nonlinear regression analysis.

Figure 40. Outline of constitutive material model to predict compressive strength from reflection loss measurements

The results of the strength prediction utilizing the unique relationships between reflection loss, gel-space ratio and compressive strength are given in Figure 41. Starting with the in-situ measured reflection loss, the gel-space ratio is determined using the relationship between reflection loss and gel-space ratio, which is then translated into compressive strength. The comparison between the measured compressive strength and that computed with the gel-space ratio concept is given in Fig. 42. The quality of the correlation between the two values is described by a coefficient of determination of R² = 0.9426.

Fig. 41: Comparison of compressive strength measured and calculated from gel-space ratio concept	Fig. 42: Comparison of compressive strength measured and calculated from specific contact area concept

The specific contact area A_c describes the area that provides contact between the single cement grains and is a measure of the degree of the interparticle bonding of the cement particles considered by the HYMOSTRUC3D model. The comparison between the specific contact area calculated for cement pastes, the reflection loss, and compressive strength measured on appropriate mortars has shown that these parameters are related by unique trends independent from the w/c-ratio. The strength prediction based on the contact area concept only is given in Figure 43.

To increase the accuracy of the strength prediction it is beneficial to combine the two concepts. This combined concept incorporates the relationships between reflection loss and the parameters gel-space ratio and specific contact area in the same form as they were used before. Based on the reflection loss the gel-space ratio and the specific contact area are calculated and then used to determine the compressive strength. The improvement of the new combined concept originates from a two-parameter model that accounts for the influence of gel-space ratio and specific contact area on compressive strength at the same time. The constants of the two-parameter model are determined by multiple nonlinear regression of gel-space ratio, specific contact area and compressive strength data. The general form of the two-parameter model can be derived as shown in Figure 43(d). The strength prediction is given in Figure 44.

Figure 43. Combined concept for determination of compressive strength from reflection loss

Figure 44. Comparison of measured and calculated compressive strength for the combined concept

Conclusions (Strength Prediction Model)

The presented strength predictions were made based on relationships derived from experimental work. Based on these relationships it is possible to predict the compressive strength of cement mortars with different w/c-ratios without further calibration. It will be subject of the research in the next months to evaluate how this model can be applied to mortars with different sand contents and concrete which additionally contains coarse aggregates.

Viscoelastic Properties

The viscoelastic behavior of Portland cement paste at very early age was investigated with the rheometric and the wave reflection method. Three cement pastes with water/cement-ratios equal to 0.4, 0.5 and 0.6 cured at a constant temperature of 25°C were studied. By measuring the wave reflection coefficients and the phase angles of the reflected shear waves the viscosity of the cement paste can be calculated based on the wave reflection measurements. For verification, the viscosity calculated from the wave reflection measurements was compared with the results obtained directly from the step rheometric method at different ages.

The viscosity of the three Portland cement pastes was measured at different hydration ages. The development of the viscosity of the pastes in time is shown in Figure 45. It can be seen from the figure that for a given age the cement pastes with higher w\c-ratios exhibit significant lower values of the viscosity. The high content of solid cement particles in a paste with the low w/c ratio densifies the suspension, which reduces the flowability of the material. It should also be noted that the increase of the viscosity with time is not linear. The viscosity increases at a very slow rate at very early age. This can also be attributed to the dormant period of the hydration process.

Figure 45. Time development of the viscosity determined with the step rheometric method

The comparison of the viscosity obtained from the wave reflection and step rheometric methods on the cement pastes at the age of 15 minutes is given in Figure 46. In general, it can be stated that that the viscosity increases with a decrease in w/c ratio. The fact that the viscosity values plotted in the figure have a very small deviation from the line of equality shows that the viscosity measurements of the two methods agree very well on a qualitative and quantitative basis.

Figure 46. Comparison of viscosity values obtained with wave reflection and step rheometric method at an age of 15 minutes

To further validate the ability of the wave reflection method to provide information about the viscosity of cementitious materials, the viscosity values of the three cement pastes was measured with the step rheometric method at different ages (Figure 45). The comparison of these results with viscosity values calculated from the WR-measurements is given in Figure 47. It can be seen that similar to the previous figure the viscosity values obtained from the two different methods have a very similar trend. Based on these results it can be concluded that the wave reflection technique provides reliable information about the viscous properties of fresh cement pastes.

Figure 45. Comparison of viscosity