University of Puerto Rico-Mayagüez Campus Civil Infrastructure Research Center

Strategic Plan
Computer Facilities



New Projects

Past Projects

What's New
Progress Reports

Contact Us


Last revision:
November, 2013

Today is

Summary of completed projects


Physical Transport of Explosives Chemicals in Soils
Sponsor: DOD
Amount: $235,001
Researcher: Dr. Ingrid Padilla (PI), Dr. Ivonne Santiago

Establishment of a Center for Assessment of Potential Odor Problems on McCook Reservoir, Chicago, IL
Sponsor: EPA
Amount $70,713
Researcher: Dr. Ingrid Padilla

Evaluation of Advance Oxidation Processes for Degradation of MTBE
Sponsor: EPA
Amount: $30,000
Researcher: Dr. Ingrid Padilla

Dynamic Detection of DNAPL in Subsurface Environments using Cross-Well Radar
Sponsor: NSF-ERC CenSSIS
Amount: $165,000
Researcher: Dr. Ingrid Padilla

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Riesgo Sísmico
Sponsor: PR Insurance Company
Amount: $156,750
Researchers: Dr. José A. Martínez (PI), Luis Suárez, Ricardo Ramos & Miguel Pando

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Parámetros que Afectan la Capacidad a Terremotos
Sponsor: PR Insurance Company
Amount: $333,669
Researchers: Dr. Ricardo López (PI), Dr. José A. Martínez Cruzado & Dr. Alí Saffar

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Daños de Vientos Huracanados
Sponsor: PR Insurance Company
Amount: $154,162
Researchers: Dr. Luis Godoy (PI), Dr. Arsenio Cáceres & Dr. Raúl Zapata

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Pérdida Máxima Probable
Sponsor: PR Insurance Company
Amount: $269,669
Researchers: Dr. Luis Godoy, Dr. Alí Saffar & Dr. José F. Lluch

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Vulnerabilidad de Residencias
Sponsor: PR Insurance Company
Amount: $212,767
Researchers: Dr. Daniel Wendichansky (PI) & Dr. Felipe Acosta

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Daños Causados por Inundaciones
Sponsor: PR Insurance Company
Amount: $132,333
Researchers: Prof. Ismael Pagán Trinidad (PI) & Dr. Raúl Zapata

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Administración
Sponsor: PR Insurance Company
Amount: $143,257
Researcher: Dr. Ricardo López

Determinación de la Vulnerabilidad de Estructuras Sometidas a Desastres Naturales: Diseño Elástico
Sponsor: PR Insurance Company
Amount: $52,523
Researcher: Dr. José O. Guevara (PI), Dr. Jose F. Lluch

Construction Drawings for Cantilevered and Counterfort Retaining Walls to be Used in Emergencies
Sponsor: PRHWA
Amount: $76,678
Researcher: Dr. Daniel Wendichansky

Continuing Host Sites for the Year 2004 Summer Transportation Institute
Sponsor: FHWA & South Caroline University
Amount: $41,999
Researcher: Dr. Didier Valdés

Manual para Estimar Propiedades
Sponsor: PRHWA
Amount: $19,125
Researcher: Dr. Miguel Pando

Base de Datos de Pruebas de Carga de Fundaciones Profundas Utilizadas en Obras Viales en PR
Sponsor: PRHWA
Amount: $34,925
Researcher: Dr. Miguel Pando

Multi-Hazard Application of Regional Damage Synthesis
Sponsor: NSF
Amount: $79,436
Researcher: Dr. Ricardo López

Construction Specifications and Cost Estimating for PRSA Projects, PRASA
Sponsor: AAA
Amount: $47,733
Researcher: Dr. José F. Lluch

Geotechnical database and Ground Motion Evaluation for the City of Mayagüez, Puerto Rico
Sponsor: USGS
Amount: $41,000
Researcher: Dr. Miguel Pando, Co-PI's: Luis Suarez, Eugenio Asencio, and Jose A. Martinez

Testing of lightweight concrete
Intramural Practice
Sponsor: International Construction Solutions Inc.
Amount: $10,375
Researcher: Dr. Ricardo López & Dr. Felipe Acosta

Multi-Hazard Application of Regional Damage Synthesis
Sponsor: NSF
Amount: $79,436
Researcher: Dr. Ali Saffar, Dr. Arsenio Cáceres & Dr. Ricardo López

Probabilistic Design Factors for Fiber-Reinforced Polymers for Infrastructure Applications
Sponsor: PR-EPSCoR
Amount: $165,000
Researcher: Dr. Felipe J. Acosta

Capacity-demand analysis for seismic buckling of steel tanks
Sponsor: PR-EPSCoR
Amount: $ 36,000
Researcher: Dr. Juan C. Virella


Thrust I: Transportation

Thrust II: Structural and Geomechanical Systems

Developing Thrust Area: Water Resources and Environmental Engineering

Human Resources Development Project

Thrust I: Transportation

Traffic Signal Control Based on Fuzzy Logic. Dr. Gerson Beauchamp-Báez (PI). 2 years, 1996-1997. 2 master’s students and 1 undergraduate.

This project will develop a methodology for designing fuzzy logic controllers for traffic signals operating at isolated intersections, arterial roads, and closed networks. The main objective is to establish traffic signal control strategies based on fuzzy logic for a closed network of controlled intersections. The goal is to reduce the costs associated with the signal operation such as the number of stops, delay, and queue length.

The fuzzy logic controller for an isolated intersection is based on a Fuzzy Phase Sequencer (FPS) developed in this research. Incorporating the fuzzy phase sequencer to the Bisset and Kelsey controller (Proceedings of the 1992 European Simulation Multiconference, June 1992) is the main result of this part of the research. Simulations were run for pretimed and the different fuzzy control strategies. The performance index used for comparison was the average stopping delay per maneuver. A computer program using Visual Basic was developed to simulate the performance of the controller.

Simulation results show that the joint fuzzy phase sequencer and fuzzy termination controller (FPS-FPT) performs better than the pretimed and Bisset and Kelsey controller except at a flow rate of 720 veh/hr. Results suggest that a maximum phase time limit should be incorporated in the FPS-FPT controller. Fine tuning is being planned for this controller. This work resulted in a master’s degree for an electrical engineering student.

The arterial fuzzy logic controller is under development. A search of the available literature on traffic control, fuzzy logic fundamentals, and fuzzy logic based traffic control has been completed. Most of the literature found has been reviewed but some material is still under revision. The development of the controller is being carried out using MATLAB and its Fuzzy Logic Toolbox. The arterial system has been studied and the fuzzy rule base for this system is being developed. There remains to be developed a simulation environment of an arterial road system. Through this simulation the controller will be tested, fine-tuned and validated. Once this is accomplished, the results will be published and presented appropriately.

Development on the closed network controller will begin in August 1996 and involve one graduate student.

Back to Current Research Projects

A Detailed Examination of Quantitative and Qualitative Aspects of Urban Rail Transit Systems. Dr. Felipe Luyanda (PI) and Dr. Jaime Gutiérrez-Sánchez (Co-PI, Social Sciences). 3 years. 3 master’s student, and 2 undergraduates.

The purpose of this project is to perform a comprehensive detailed examination of essential quantitative and qualitative aspects related to the operation of urban rail transit systems and their impacts in the urban areas. The research being carried out covers five principal categories, namely (1) design and construction, (2) operation and service characteristics, (3) costs and benefit analysis, (4) demand management, and (5) land use strategies and social, economic and environmental impacts.

The following tasks have been accomplished:

1. Initial contact has been established with metropolitan transit agencies in the U.S. and Colombia, the Federal Transit Administration (FTA), Bureau of Transportation

Statistics (BTS), and American Public Transit Association (APTA), and the Latin American Association of Metros and Subways.

2. A literature review was performed using various sources.

3. Reports, data, environmental impact statements, and proceedings were obtained.

4. A form was designed to summarize qualitative information.

5. Quantitative information was summarized in tabular form.

6. A form was designed to summarize newspaper information.

7. Trips were made to Medellín, Colombia, and Miami, Florida, to study their respective transit systems.

Considerable data has been obtained using the following sources:

Silver Platter CD ROM database with +400,000 records, including OECD and TRB

Local library and interlibrary loans

Publications and data directly from agencies and operators

EIS of more recent projects

Access through the Internet to U.S. Libraries and agencies

Theses from University of Michigan Microfilm Service

Latin American Association of Metros and Subways

Tasks Remaining

The tasks remaining have been divided into those that will be accomplished in the near-term (within the next summer and first semester of the next academic year) and short-term (during the rest of the project’s period of duration).


Design of a questionnaire for operators to obtain additional information about their systems.

Make arrangement for a trip during the summer months to study additional systems, obtain first hand information, perform structured interviews and obtain additional data.

Prepare a paper on quantitative characteristics, emphasizing comparisons between systems when variable are standardized by population, service area, directional route miles, number of vehicles, number of station, and others.

Prepare a local article summarizing the information that has appeared in the El Nuevo Día and the San Juan Star newspapers.


Final selection of systems for case studies.

Prepare a report summarizing the case studies results and with an annotated bibliography on the qualitative aspects studied.

Prepare a paper on the main human and organizational issues associated with the operation of urban rail transit systems.

Prepare a paper summarizing the most important aspects of the case studies.

Develop a proposal to obtain funding for a Conference/Workshop on urban rail transit systems.

Develop a graduate course on transit systems.

Back to Current Research Projects

Methodology for Development of Performance and Survivor Curves for Flexible Pavements in Puerto Rico. Dr. Benjamín Colucci (PI, Thrust Area Leader) and Dr. Nazario Ramírez (Co-PI). 2 years, 1996-1997. 2 master’s students.

The usual method of developing performance and survivor curves for flexible pavements cannot be applied in Puerto Rico because of insufficient historical data on pavement deterioration and traffic loads. This project will apply a novel methodology requiring minimal road-use data compared to the standard approach; the methodology involves two complementary approaches:

1. Performance curves will be developed on the basis of PSI and ESAL measurements which will be obtained at a single point in time over different highway segments belonging to the same homogeneous cluster.

2. The Weibull probability density function will be used to express the probability that a specific roadway survives when a given traffic load is applied.

The scope of this project is summarized in six major tasks, listed here along with the results obtained during the reporting period:

1. Decomposition of the pavement network into homogeneous regions by means of multivariate techniques (i.e. hierarchical cluster analysis).

Several data sources were searched in the process of establishing the climatological variables for 76 towns in Puerto Rico. Five discriminate variables were used to generate the homogeneous climatic zones namely: population density, elevation, annual rainfall, average temperature, and type of soil. As a result of the clustering analysis, it was concluded that the appropriate number of clusters is four. The Statistical Analysis System (SAS) and the centroid method were used to delineate four homogeneous climatic zones, namely: high, dry, wet, and humid zones. Each cluster will be divided into three sub-clusters containing roadway segments in good, fair, and poor condition. It is expected that at least each sub-cluster will contain two roadway segments.

2. Use of Delphi technique to estimate initial and failure serviceability index, failure mode, shape of performance curve, and traffic variability within each region.

The Delphi technique encompasses several procedures to facilitate collective decisions via a series of questionnaires administered to all members of a panel and accompanied by summaries of the panel's earlier response. The panel was composed of Puerto Rican professionals with expert knowledge of the interrelationships of traffic loads, roughness, climatic effects, material characterization and pavement evaluation and management concepts. The major conclusions reached through the Delphi technique are summarized below:

A. The panel estimated traffic to failure for highways in the different regions to range between 7.0 and 15.4 million ESAL.

B. The dry region was estimated to have a lower variability (12%) as compared with the wet region (18%).

C. The serviceability parameters P0 and Pf were estimated as 4.0 and 1.5, respectively.

D. Fatigue cracking combined with induced moisture damage associated with poor drainage practice is the principal failure mode on the observed pavements.

E. Initial quality control varies significantly from region to region.

3. Estimation of the present serviceability index (PSI) by means of roughness measurements.

Data is being collected to determine the roughness of representative highway segments within each homogeneous region. Roughness is measured at high speed using the Automatic Road Analyzer (ARAN) owned by the Puerto Rico Highway and Transportation Authority (PRHTA). A relationship between the roughness measurements obtained from the ARAN and the international roughness index (IRI) obtained from the Dipstick, considered to be more accurate, has been developed by related research projects at UPR Mayagüez. Based on the modified IRI, the PSI will be estimated at several levels of deterioration in each homogeneous region.

4. Estimation of the cumulative ESAL18.

Analysis and evaluation of historical data from 18 truck weigh stations was carried out using Microsoft Access and a custom-made data collection and analysis program for truck weight data (Pesaje) used by the PRHTA. The records of average daily traffic (ADT) were used to estimate the rate of traffic increase. The cumulative traffic load up to the analysis time, N18, can thus be estimated for each representative segment in the homogeneous regions.

5. Determination of generic performance functions for each climatic region.

Performance curves are being developed on the basis of PSI and cumulative ESAL18 measurements obtained at a single point in time over different highway segments which belong to the same cluster. Each cluster will be organized into good, fair, and poor segment conditions, depending on the traffic load and PSI values. The mathematical formulation of the performance function was initially developed at the AASHO Road Test. The required data for parameter estimation are the current PSI (P) and the cumulative ESAL18 (W). The SAS computer package was used to conduct parameter estimation. However, the nonlinear estimation subroutine, SYSNLIN, fails to converge for this particular data set. Thus, the strategy used to conduct parameter estimation consisted of linearizing the model to obtain an initial point, and then using the Levenberg-Marquardt algorithm to conduct nonlinear parameter estimation. Performance curves for the humid region have been developed. A paper on this and the previous research tasks has been submitted to the Transportation Research Board.

6. Determination of survivor functions for each climatic region.

A survivor function is a mathematical model that estimates the probability that an individual mile of pavement will survive fully functional for a considered traffic load level. Let W be the cumulative traffic load to failure. For simplicity or because in most cases W is known with certain precision it has been considered as a deterministic variable. However, in our case where W is an unknown variable, a more realistic representation could be that W behaves as a random variable. Thus, it would be meaningful to determine a stochastic relationship that estimates the probability that an individual mile actually survives when the performance function reaches a given critical value. The Weibull distribution is being used for this research since it is flexible and fits all possible failure rates. Furthermore, the Weibull distribution is one of the probabilistic models which has been applied to various failure problems, such as electron tube failure, the fatigue life of deep-groove ball bearing, floppy disk failure. A paper on this aspect of the research has been submitted to the Transportation Research Board.

Back to Current Research Projects

Asphalt Chemical Characterization Studies Using High Performance Liquid Chromatography. Dr. José Colucci (PI) and Dr. Benjamín Colucci (Co-PI). 3 years. 2 master’s students.

The main objective of this project is to study the relationship between chemical composition of asphalt physical fractions and laboratory performance of crumb rubber modified asphalt. This overall objective can be further divided into the following specific objectives:

1. Development of analytical techniques using HPLC to characterize asphalt into several chemical fractions.

2. Development of performance-chemical properties correlations. Specifically, laboratory performance tests will be evaluated such as viscosity, resiliency, penetration, etc.

In this study HPLC is used to perform chemical composition analysis of asphalt physical fractions. Experimental variables such as reaction temperature, time, mixing rate, and physical and chemical compositions are being studied in order to predict laboratory properties such as viscosity, resiliency, etc. The rubber modified asphalt laboratory properties will be correlated to several operational and analytical properties.

The project is right on schedule. During the first six months, the focus was on purchasing and installing all the necessary equipment. The new HPLC unit was acquired at a cost of $25,350 from Hewlett-Packard, Inc. Two separation columns and two guard columns were purchased from Phenomenex Co. at a total cost of $2,273. During the next three months training in the use of new HPLC was carried out.

Graduate Research Assistant Rafael A. Díaz is now developing the separation method for the characterization of asphalt. The method development is done with an existing HPLC belonging to another project. The columns used for the study are Phenomenex Phenogel® (10-4 Å, 5 micron pore size and 10-3 Å, 10 micron pore size), both with dimensions of 300 mm X 7.8 mm.

We will also continue to improve the Polystyrene separation with the new HPLC equipment to adjust the resolution and then analyze asphalt samples with and without CRM. A recently acquired X-Ray Fluorescence apparatus will also be used to characterize the asphalt samples with regard to nitrogen and sulfur content. The full laboratory work got underway in August 1996, as was originally proposed.

Back to Current Research Projects

Technology Assessment for Bridge Deck Condition Surveys. Dr. Houssam Toutanji (PI). 3 years. 2 master’s students, 2 undergraduates.

The purpose of this research project is to investigate methods for detecting the major distresses affecting Puerto Rico's bridge decks: shattered edges, reactive aggregate, corrosion, and debonding. During the first phase of this project, information from the Federal Department of Transportation (USDOT) was reviewed to identify the inspection technologies that appear to be most promising for solving the specific needs in Puerto Rico. The ultrasonic wave velocity technique was selected for investigation. The importance of this method is that it can detect areas of internal cracking, internal delamination, and strength parameters (i.e., elastic modulus approximations). Receiver and computerized data acquisition components are used in the overall testing system. In general, this method has been shown to be effective, but it also time consuming and labor intensive.

The main problem associated with the use of the ultrasonic wave velocity method is the difficulty of analyzing and interpreting signals from concrete bridge decks. Based on wave velocity and travel time, some properties can be determined such as modulus of elasticity and Poisson's ratio. However, the technique cannot detect the kind of flaws or crack presented in the concrete slab. The technique is one of most used techniques for concrete evaluation because it is very easy to use and the device is relatively small so that it can be carried from on site to another with minimal effort.

Direct and indirect ultrasonic measurements were taken for the six concrete slab specimens. The waves received from the direct and indirect measurements differed; however, detection of anomalies in the concrete specimens was possible using both techniques. The variation of the ultrasonic signals from plain concrete to cardboard, sponge, and honeycombed embedded concrete was significant. The waves received from the slabs with these artificial cracks show irregular damping with low voltage; however, for the plain concrete the waves showed regular damping and high voltage.

Using the indirect method, the cracks can be determined with some accuracy by plotting time versus wave travel distance. When the slope of this curve deviates from a straight line it indicates the presence of flaws and cracks.

These are preliminary results. More results and data collected over summer are being analyzed. Currently, the collected data are normalized; thus a better comparison between good and anomaly concrete will be made. A new ultrasonic device was acquired for this project. The new device enables us directly to measure the Poisson’s ratio, density, elastic modulus, and to control the number of pulses. In addition, the device provides correction factors for humidity, which is very important due to the high humidity of the climate in Puerto Rico. Most importantly the device will give more precision in collecting data and will replace the one we have been using which is over 10 years old.

Back to Current Research Projects

Thrust II: Structural and Geomechanical Systems

Design Earthquakes for Puerto Rico. Dr. José A. Martínez-Cruzado (PI), Dr. Klaus Bataille (Co-PI, Geology Dept. and PR Seismic Network), James Joyce (Co-PI, Geology), Milton R. Martínez-Delgado, UPRM Collaborator. 3 years. 3 master’s, 1 Ph.D. graduate students, 1 undergraduate.

This project is developing a set of design earthquakes for Puerto Rico’s most populated areas based on the known seismological and geotechnical conditions affecting the Island and the surrounding region. Since strong motion data in Puerto Rico are scarce, several models will be developed and analyzed to generate reliable alternatives to Puerto Rico’s code for earthquake design, which will lead to safer and more cost-effective methods of earthquake resistant design and retrofitting.

I. Earthquake Recurrence Model and Development of PGA Contours for Puerto Rico

Data on 3,628 tremors occurring over 6 years were obtained by the Puerto Rico Seismic Network of the Geology Department (1990-1995). Data of tremors classified as aftershocks and swarms were removed. With the remaining data, nine seismic zones were identified and delineated. For each seismic zone the earthquake recurrence parameters (a and b values) were obtained. The recurrence parameters indicate that the southwestern part of Puerto Rico is the most active zone. However, because of the high b-value, major earthquakes should not be expected in this zone. On the other hand, although the Muertos Trough shows the lowest seismicity, major earthquakes could be expected in that particular zone as well as in the Puerto Rico Trench.

Service, Damage, and Ultimate design levels having a 10% chance of exceedance in 10, 50 and 100 years, respectively, have been selected. Also, maximum credible earthquake was assumed as the one with 10% chance of exceedance in 250 years (return period of 2,373 years). Earthquake magnitudes were obtained from the recurrence plots for each design level. These preliminary results show maximum credible earthquakes smaller than what has always been expected.

The next step will be to find attenuation laws from the eastern USA or eastern Canada which better match the geological characteristics of Puerto Rico. The minimum epicentral distance from several towns to each seismic zone needs to be calculated. With the selected attenuation laws and the minimum epicentral distance to selected towns, maps of expected peak ground accelerations would be developed for each one of the selected earthquake levels.

II. Extrapolation from Imported Earthquake Acceleration Records

More than 160 acceleration records have been obtained worldwide, mainly from California. This data was organized according to magnitude, mechanism, and focal depth. The acceleration records obtained for each earthquake were organized according to the type of soil and the epicentral distance.

From the small earthquakes that have been recorded since 1990 to 1995 by the Puerto Rico Seismic Network, nine seismic zones were delineated. Maximum credible earthquakes were established for each seismic zone (10% probability of being exceeded in 250 years).

Three main cities were selected (San Juan, Ponce, and Mayagüez) for the derivation of artificial accelerogram records. The maximum and minimum distance to each earthquake zone is now being calculated.

The next step will be to select, from among the acceleration records available, those earthquakes whose magnitude, mechanism, and focal depth provide a close match to the characteristics of earthquakes possible in each seismic zone in Puerto Rico, and whose epicentral distances fall within the ranges established between the seismic zones and the three cities for which design earthquakes are being developed. Once the complete set of accelerogram records has been selected, the response spectra will be calculated for each accelerogram and the envelope will be drawn. Finally, from the envelope of each set of response spectra a synthetic acceleration record will be derived.

III. Extrapolation from Available Native Acceleration Records

On February 29, 1996 (0h 29m 26s), a small earthquake struck at Barrio Limón of Mayagüez at a focal depth of 28 km. The tremor registered a Coda Magnitude of 3.9 and was recorded by accelerometers located in Mayagüez (PGA = 1.4%g) and San Germán (PGA = 2.5%g). This is the first earthquake ever recorded in a digital accelerometer in Puerto Rico. The records have been processed so that the corrected acceleration, velocity, and displacement histories are already available.

Using these recent seismic records and any others that may be obtained in the near future, moderate and major earthquakes will be modeled by conveniently scaling up the amplitude and/or the time scale of the prototype accelerogram.

The scaling of the amplitude will be based on the desired earthquake magnitude and expected peak ground acceleration, while the scaling of the time will depend upon the expected predominant period of vibration of the site under consideration.

IV. Correlation of Artificial Acceleration Records

Once synthetic acceleration records have been obtained at the selected sites by both approaches described above, a comparison study will be carried out. Differences among artificial records will be identified and explained. This comparison might help us to improve the models.

Back to Current Research Projects

Patterns of Inelastic Deformation in Reinforced Concrete Frames Subjected to Strong Earthquakes. Dr. Ricardo López (PI, Thrust Area Leader) and Dr. M. Saiid Saiidi (Co-PI, University of Nevada, Reno). 2 years, 1996-1997. 2 master’s and 2 Ph.D. students.

This project will use available technology (like nonlinear analysis programs) to develop a procedure useful to practicing engineers. Such a procedure based on elastic analysis is not available. Its development would result in substantial savings of time and materials without compromising safety.

The following tasks have been accomplished or are in progress:

1. Modify the existing computer program LARZW, to account for dead load effects.

This program was developed by the PI’s, Saiidi and López, respectively. The program has been modified and gravity load has been included in the analysis.

2. Select a database of earthquake records that represents historical earthquakes in South, Central, and North America.

Five records have been selected. The ground motions cover a wide range of dominant frequencies, time duration, and maximum accelerations. Table 1 summarizes the groud motion identification information. Plots of ground acceleration vs. time are given in Figure 1, and plots of response spectra for displacement and acceleration are shown in Figure 2. The spectra were calculated assuming a damping ratio of 5% of critical.

3. Select a series of typical R/C frames of varying number of stories (5 to 15), dimensions, configuration (regular or irregular), and strength.

The following parameters have been selected for the study: 5, 10, and 15 stories; 2, 3, and 4 bays; 10%, 20% and 30% design base shear strength as percentage of weight; details of reinforcement as specified in 1987 (most recent) and 1968 Puerto Rico building code.

4. Design the frames.

Three 5-story frames have been designed according to modern code (UBC 1994). One 15 story frame has also been designed. Design of the frames has been a task that takes more time than originally expected.

5. Analyze the frames using the modified program LARZW for the earthquake records selected in task 2.

The 5-story frame was analyzed statically and dynamically. The static analysis was performed with inverted triangular load distribution to failure.

6. Study the patterns and the magnitude of nonlinearity in the joints and elements of all the frames for all the earthquakes and identify the areas of nonlinearity as a function of the frame and earthquake parameters.

The locations of plastic hinges in the 5-story frame as a result of the static and dynamic analyses were determined. It could be observed that the earthquakes induce extensive hinging in the frames, and that the hinges appear in almost every floor of the frames. The 15-story frame did not yield when subjected to the San Salvador ground motion. Hinging was limited to the lower two thirds of the building, which indicates that for taller buildings the procedure proposed for development in this study could be useful.

Work should continue with design and analysis of the remaining frames, as outlined in task 3. After most analyses are completed, correlation of the results with simpler elastic analysis will be carried out, with the objective of developing guidelines to help the designer identify “non-seismic” joints in conjunction with an elastic static analysis.

Back to Current Research Projects

Seismic Response of R/C Frames with Friction Composite Moment Connections. Dr. Leandro Rodríguez-Agrait (PI) and Dr. Ali Saffar (Co-PI). 3 years. 1 Ph.D. graduate student and 1 undergraduate.

The ductile connections in reinforced concrete frames designed for earthquakes are the most important component in guaranteeing the safe, nonlinear behavior of the structure. These connections are difficult to build and, thus, expensive. The improvement in the earthquake behavior of frames with friction connections has been well documented in the literature. The objective of this project is to study the feasibility of using composite friction type connections in reinforced concrete structures. To do this, complementary analytical and experimental approaches are proposed.

In the analytical part, FORTRAN programs were developed for the design of beams and columns for typical reinforced concrete framed buildings. The design of three buildings of 4, 7, and 20 floors have been completed. In addition, a program was developed that performs the complete design of the friction steel connection to replace the concrete one.

The program that performs the non-linear dynamic analysis of frames with friction connections has been completed. The program was submitted to a convergence process, in terms of the increment time of the numerical integration. The program was also tested in some tasks with well-known programs like Drain-2D. The results obtained were very satisfactory.

With this program the general behavior of the structure is being studied. Important parameters like the lateral displacement of the structure, the work imposed by the earthquake, and the base shear are being observed.

In order to manage the huge quantity of results generated by the main program two post-processing tools were developed in VISUAL BASIC. One shows the response of the structure in time, the hysteritic loop, etc.; the other one shows the plastic hinge sequence, giving the possibility of observing if a collapse mechanism is formed.

In the experimental part, two steel friction connections have been fabricated; they were tested this summer. The results are now being analyzed.

Back to Current Research Projects

Seismic Retrofitting Performance and Cost-Effectiveness of Steel Bracing Systems with Supplemental Damping Devices. Dr. Ali Saffar (PI) and Dr. Leandro Rodríguez-Agrait (Co-PI). 3 years. 4 master’s and 1 Ph.D. student.

Steel bracing has shown major advantages over other retrofitting schemes. Although both concentric and eccentric K bracing systems have been studied in the past, their cost effectiveness has not been compared. Nor have comparisons been made with other bracing systems with supplemental damping devices, some of which are already used in practice. A detailed response analysis of these systems is essential for making qualified decisions on the best retrofitting scheme for a building. Of course, one cannot establish the cost-performance criteria for seismic retrofitting before one ascertains the appropriate level of upgrading.

A key component of any economic model measuring the appropriate level of upgrading for an existing building are the associated capital stock figures. During the first year of this study, most of the variables needed for a capital stock model were quantified. A cost survey on the businesses and manufacturing in Mayagüez area were designed to supplement the existing data. The survey is now being carried out with the assistance of the municipality, which has shown a great deal of interest in this project. We are also in the process of selecting an advisory committee to help estimate the building downtime for different occupancy types, which is critical to our model.

Some prototypes of existing buildings in Puerto Rico have already been selected with a more extensive set to follow. The computer programs needed to carry out the numerical analysis of the design frames are already in place, although the modifications required are not completed. This phase of the project will now run concurrently with the on-going development of the economic model previously discussed. The expected completion date is no later than August 1997. The last year of this project is then devoted entirely to exploring the possibilities for adding an economic importance factor to the seismic design regulations in Puerto Rico.

Back to Current Research Projects

Behavior of Damaged Structural Components. Dr. Luis A. Godoy. 2 years, 1995-1996. 2 master’s and 1 Ph.D. student.

Theobjectiveofthisresearchistostudythestressredistribution, buckling, and post-buckling behaviorthatoccurinstructuralcomponentsasaconsequenceofdamage, andtoprovidetechniquesofanalysisadequatetomodeltheproblem.

During the period covered by this report, we have developed a formulation for sensitivity analysis of structures based on perturbation techniques. The formulation takes into account the fundamental path, the critical buckling state, and the post critical equilibrium states. Two forms have been used: one based on the equilibrium condition that employs stiffness matrices and can be readily implemented in finite element analysis; and a second approach based on the total potential energy of the system. The second approach is powerful in the study of buckling and postbuckling states, because the general theory of elastic stability is based on potential energy considerations.

The solution of the problem employs perturbation techniques leading to sensitivity results of the desired behavior. Additionally, damage of real tanks has been assessed in St. Thomas after hurricane Marilyn, which struck the island in September 1995. The PI traveled to St. Thomas and inspected several tanks showing severe damage.

Back to Current Research Projects

Developing Thrust Area: Water Resources and Environmental Engineering

Development of a Decision Support System for Urban Watershed Management in Puerto Rico. Dr. Jorge Rivera-Santos (PI) and Dr. Raúl Zapata-López (Co-PI). 3 years. 3 master’s students.

The scope of the this research includes the adaptation and/or development of a conceptual rainfall-runoff hydrologic computer model with an unsteady uniform hydraulic component, and its application to a highly developed urban watershed in Puerto Rico. The specific objectives of the current research are to:

Develop a rainfall-runoff model to be used in a decision support system. This will provide civil infrastructure administrators with the tools needed to analyze a construction permit application for development of areas which are not necessarily within the delineated flood prone areas but could still put human life and property in greater jeopardy from increased flooding.

Provide a research tool that can serve as the foundation for continuous research in the field of computer hydrologic modeling in tropical zones.

This research effort has been divided into several phases. The current project addresses only the first phase. A future second phase will include the development of specific calibration and verification procedures and training for the prospective users of the proposed system

In Phase 1, then, the work has been divided into the following tasks:

Tasks completed:1. Selection of a representative urban watershed. After an analysis of various potential watersheds, the Río Piedras basin in the Metropolitan area of San Juan has been selected. It contains a streamflow gaging station at the PR-1 bridge and a rainfall station at the Botanical Gardens of the Central Administration of the University of Puerto Rico. It is subjected to rapid urban development with an estimated 80% already developed. The climatological and hydrologic instrumentation is operated and maintained by the National Weather Service and the U.S. Geological Survey, respectively.

2. Compilation of the required data for the H&H simulations. These data include, but are not limited to: geomorphologic data of the watershed, rainfall records, streamflow data, and channel cross-sections. The cross-sections, describing the floodable portion of the basin, have been obtained from data used by the U.S. Army Corps of Engineers for the Río Puerto Nuevo channelization project. The watershed topography has been acquired from the U.S. Geological Survey in ARC/INFO format. Watershed boundaries, soil types, and land covers have been determined by graduate students and processed for GIS use. Other specific data required by different H&H models is compiled and prepared for use as the evaluation of the models advances.

3. Screening of H&H computer models for possible adaptation and/or modification. A large number of H&H models are available. The first steep was to identify all possible models and prepare a data base with the characteristic and features of each model. From previous work conducted by the P.I. and from the efforts of graduate students, 212 hydrologic and 50 hydraulic models have been identified. All hydrologic models have been classified as continuous, event oriented, lumped, or distributed. The most frequent application, when available, has been also identified as rural, urban, forest, mountain, or agriculture. Hydrologic components such as surface runoff, infiltration, precipitation, subsurface flow, chemical movement, depth storage, interception, and flood routing have been considered and the models which feature them identified. References and contact persons, when available, have also been included.

A preliminary screening was conducted on the models data base and the number of models for evaluation was reduced to a manageable number. Those passing the screening were subjected to more conscious and thorough examination. The finalists for this round were the SSARR from the Corps of Engineers, which is a conventional type hydrologic model, the PRMS from the U.S. Geological Survey, which is a water balance model, and the TOPMODEL from the University of Landcaster, UK. As a mater of comparison, the HEC-1 and the HEC-2, from the Corps of Engineers, are being used to simulate the hydrology and hydraulics of the Río Piedras watershed. If no model is found to simulate adequately the basin, then modifications, or even a new model, will be made to meet the particular requirements of the project.

Tasks to be done:

1. Adaptation/modification and/or development of the computer model. The computer model selected in the previous step will be adapted for inclusion in the computer scheme envisioned. To implement the system some modifications are unavoidable, but only those that are strictly necessary will be performed. GIS will be utilized to manage the data base for the model.

2. Implementation and testing of the developed system. The hydrologic and hydraulic models selected will be assembled together to create a working integrated computer model. The data base with the required information to carry out the simulations will be prepared using GIS techniques. The system will then be tested in the Rio Piedras watershed at different stages of its development and the results assessed with real events of the past and present conditions.

3. Technology Transfer. At least two technical papers will be prepared, in addition to the final formal report of the investigation. Short in-school presentations are possible throughout the duration of the research. Partial results have already been presented in the Sigma Xi Student Poster Day held early in April of 1996. A technical paper is being prepared on the development and use of the H&H model data base.

Back to Current Research Projects

Detection of Water Losses in Piping Systems. Dr. Walter F. Silva. 3 years. 2 master’s and 1 undergraduate student.

This project focuses on the development, implementation, and evaluation of a method to estimate the water loss in a pipeline system by combining measured data with numerical computations. Numerical comparisons and experimental data will be used to evaluate the sensitivity of the numerical model to inaccuracies in the estimation of the parameters. The study pretends to determine the potential use of the model in field applications. The objectives of the project are:

1. Develop a methodology for the estimation of water losses in water distribution systems.

2. Determine the sensitivity of the numerical predictions to inaccuracies in the estimation of the energy losses, the magnitudes of the water demand, and limited availability of data.

3. Establish criteria for the application of the model to field conditions.

After the approval to proceed with the project (July 1995), the P.I. and his graduate student, Marcela Durán, began to collect information on leak detection algorithms available in the petroleum, nuclear power, and water industry. A report summarizing this information was prepared (in Spanish) during October 1995. The main conclusion was that most of the water utilities do not invest in leak detection programs because water audits require extensive and expensive surveys. To the P.I.’s knowledge, no attempt has been made by the water industry to use numerical techniques for the estimation of leakage in water systems.

The project involves analytical and experimental phases. The analytical part refers to the derivation of the unsteady state governing equations for the new model and a methodology to solve them by using typical field data as input. It was found that variations in the numerical procedures should be implemented depending upon the assumptions made during the derivation. Several attempts have been made to find a simple and reliable methodology for the solution of the equations. Some of the trials presented numerical difficulties and were not attractive for implementation in accordance with the purposes of the project.

Recently, a simplified form using a uniformly distributed demand was derived, and a solution procedure using the method of characteristics is being tested. The distributed pipe-flux model assigns a uniform distribution of water consumption along a pipe reach. Many residential areas follow this pattern closely. This numerical algorithm only requires the time history of pressure at three equally spaced points along a distribution line plus the initial conditions. In the present stage of development, it is been tested along one pipeline only.

Parallel with the unsteady state algorithm, a sensitivity analysis of the distributed demand model compared to a discrete demand under steady state conditions is being carried out. Discrete demand is the individual water consumption at each household along a distribution line. Different pipe roughness conditions, pipe lengths, and demand patterns can be tested in this model. Preliminary results using this algorithm indicate that friction losses, spatial distribution, and amount of the demand flow are parameters affecting the accuracy of the results.

Two main tasks are to be done during summer and fall 1996. The first is to finish the development of the numerical algorithm for unsteady state flow and to determine its limitations. Further improvement of the numerical model is expected. A detailed simulation of a time-dependent discrete demand will be possible with the implementation of another computer program yet to be developed. It must be remembered that, even though there are a few commercial unsteady pipe flow codes available, this project presents a very specific application for which no commercial software is really appropriate. The P.I will supervise Marcela Durán during the completion of this job. She is expected to finish her master’s thesis in the fall of 1996. Her thesis will report the results of the numerical evaluation of the computer model and make recommendations for the implementation of the model in complex systems.

The second task to be done is the experimental part of the project. The numerical algorithm will be verified using an experimental setup with instrumentation available in the market. The purpose is to determine the applicability of the new model to field conditions. The graduate student Israel Pabón joined the P.I. team in January. He has a strong background in computer applications and also has taught fluid courses at the UPRM General Engineering Department. His job is to work on the design of the experimental setup and in the selection of instrumentation and data acquisition equipment (DAS). The project’s limited budget and the increase in price of materials and equipment demanded a careful and well planned design. Also, the water supply system for this experiment is being rehabilitated at the Fluid Mechanics Laboratory. That work is scheduled to be finished in May of 1996.

Israel is graduating from Mechanical Engineering next summer and a new graduate student will join the project next semester. He will be working for his Master’s Degree until the end of the project. The P.I. will provide a partial assistant ship to this student through the Fluid Mechanics Laboratory of the General Engineering Department. The construction of the experimental apparatus is planned to start in summer and the pre-testing period by the end of the fall semester. Training the graduate student in programming DAS appears to be a critical task for keeping the project on schedule. The P.I. began conversations with the Electrical Engineering Department to ask for help in this matter. This experience will be integrated into the Fluid Mechanics Laboratory course to provide the students with some background in this important topic.

Final experimental runs, model verification, and preliminary results are expected by summer of 1997. In the fall of 1997 we will be analyzing the results and obtaining final conclusions. By that time, the new graduate student should be finishing up his or her thesis. A final report on the project and the draft of a paper for submission to a peer-reviewed journal will be completed during the first months of 1998.

Back to Current Research Projects

Human Resources Development Project

UPR/MIT Education and Human Resources Development Program in Conjunction with the Tren Urbano Project. Dr. Antonio A. González-Quevedo, Ph.D. Program Coordinator. 20-24 UPR students, undergraduate and graduate, per year.

This program, which will be identified as the Tren Urbano Internship Program (TUIP), is a multidisciplinary education and human resources development effort initiated by the University of Puerto Rico (UPR) with the cooperation of the Puerto Rico Department of Transportation and Public Works (PRDTPW) through its Tren Urbano Office (TUO), and involving the Center for Transportation Studies at the Massachusetts Institute of Technology (MIT). The program which started in the summer of 1994, takes advantage of unique opportunities for training and technology transfer presented by the project now underway to design and build a rail transit system known as Tren Urbano (“Urban Railway”) in the San Juan Metropolitan Area. This innovative program is administered by the CIRC and currently involves no NSF or Puerto Rico EPSCoR funding. The participation of the University of Puerto Rico’s students and faculty in the TUIP is funded entirely by the UPR Office of the President; additional funding is being sought for the program from the Puerto Rico Economic Development Administration, through the PRDTPW. MIT’s participation is funded by the General Management and Architectural and Engineering Consultant (GMAEC), a joint venture contracted by the PRDTPW to help the Tren Urbano Office design and manage the project.

Each year students from engineering, architecture, and urban planning from each university are selected to participate in the program for one to two years. The Tren Urbano students participate in two ten-day symposia, one at MIT and one at the University of Puerto Rico, where experts in development of mass transit rail systems, and specifically San Juan’s urban railway, discuss different aspects of the development of these types of major infrastructure projects. The students develop research projects related to Tren Urbano, which they carry out individually or in collaboration with counterparts from the other university, under the supervision of faculty from UPR or MIT. The TUIP emphasizes the interdisciplinary nature of major infrastructure projects, so that students learn the importance of looking at all the aspects of a project, such as engineering, transportation, systems, architecture, social factors, urban planning, safety, and public participation.

The program has been in operation since the summer of 1994. In the summer of 1995, UPR chose its second group of ten students to participate in the program, five from civil engineering and five from architecture and urban planning. Three symposia have been held during the reporting period: two in Boston, hosted by MIT, and on in San Juan, hosted by UPR. The organization of these activities is a tremendous undertaking, which includes all the logistics of hosting a group of students and professors for 10 days of lectures, site visits, cultural activities, traveling in the metropolitan areas of Boston and San Juan and, in Puerto Rico, to the western part of the Island.

Even though it was not part of the original program, the UPR group visited Caracas in November of 1994, and Miami with a second group of students during spring break of 1996. Both visits were short but very successful. Our students have developed a very critical mentality and, when exposed to operating systems, they can readily understand what functions well and what does not. The visit to Caracas was a student’s initiative, to which the CIRC gave its moral support. The expenses for this trip were absorbed by the participants who decided to go. The trip to Miami, on the other hand, was sponsored by the program. The second group of students displayed their own special initiative in the form of a first-ever forum on transportation and the urban environment at UPR Río Piedras. Some of our students organized and participated in this activity.

The CIRC will receive a special grant from the Puerto Rico Economic Development Administration, through the Puerto Rico Department of Transportation and Public Works, which it will use to expand the Tren Urbano Internship Program to include more students and more activities.

Back to Current Research Projects


Return to beginning of page

 © 1991-2005 Civil Infrastructure Research Center, UPRM.
All rights reserved. - v4.1