This document summarizes a study on assessing the remaining prestress force in pretensioned concrete bridges using a surface strain relief method. The method involves measuring the elastic rebound of concrete after coring or notching its surface using strain gages or laser speckle imaging. Testing of full-scale beams found coring to a depth of 1 inch provided strains within 8% of theoretical values on average. Notching to a depth of 1 inch with a spacing of 3.5 inches and length of 3 inches provided results within 20% of theoretical on average. Finite element models matched experimental results and can predict optimal testing methods.
Assessing the Damage Potential in Pretensioned Bridges Caused by Increased Truck Loads Due to Freight Movements (Phase I)
1. Assessing the Damage Potential in
Pretensioned Bridges Caused by Increased
Truck Loads Due to Freight Movements
Robert J. Peterman, Ph.D., P.E.
Martin K. Eby Distinguished Professor in Engineering
Kansas State University
2. Disclaimer
The contents of this report reflect the views of the
authors, who are responsible for the facts and the
accuracy of the information presented herein. This
document is disseminated under the sponsorship of the
U.S. Department of Transportation’s University
Transportation Centers Program, in the interest of
information exchange. The U.S. Government assumes no
liability for the contents or use thereof.
3. Other Contributors
Steven F. Hammerschmidt, CE Dept.
Dr. Weixin Zhao, MNE Dept.
Dr. B. Terry Beck, MNE Dept.
Dr. John Wu, Ph.D., IMSE Dept.
5. Introduction
•Many bridges are approaching their design life expectancy and/or
exposed to larger demands (10-15% are currently deficient).
•In order to accurately assess the condition of a prestressed
concrete bridge (highway or railroad), the remaining prestress force
level must be known.
•Time dependent losses decrease the prestress force in a member.
•The project’s goal was to develop an efficient, and inexpensive
way to determine the existing stress in a prestressed concrete
bridge member, thus the condition of these bridges can be
accurately assessed.
6. Surface Strain Relief
Major Steps:
2) Set up initial strain measurement device
•Electrical resistance strain (ERS) gages
•Laser speckle imaging (LSI) device
3) Core or notch to relieve strain
4) Measure elastic rebound of the concrete
5) Relate rebound of the concrete to the average prestress
force
7. Surface Strain Relief
Electrical Resistance Strain (ERS) Gages
•Gage length of 2”
•Epoxy used to mount gage to surface
•Gages protected with polyurethane
coating and microcrystalline wax
•Four pin terminal block was connected
to the lead wires attached to the strain
gage with silicone
8. Surface Strain Relief
Laser Speckle Imaging (LSI) Device
•Device developed at Kansas State University
•Images the speckle pattern produced by a laser reflection off the
surface which serves as the “fingerprint” of the location
•Subsequent images are related to the reference images and the
amount of displacement is calculated
LSI Device with a 2” Gage Length Speckle Pattern
9. Coring/Notching Procedure
•Used a 3” outside diameter dry coring diamond bit
•Used a 4.5” diameter dry diamond cutting wheel
•Core and notch temperature was monitored using a
non-contact thermometer
10. Procedure
•Locations were marked on the beam and gages attached
•All gages were initially set to zero microstrain or the LSI device was
used to take initial readings
•Coring guide was clamped into position on the surface of the beam or
layout lines were drawn on the beam with a distance of 3.5” between
notches
•Core locations were cored to an initial depth of ¾” and then 1”
•Notch locations were cut to an initial depth of 1” and then 1¼”
•There was a 10 minute delay between any increase in depth to allow the
entire location to reach equilibrium with the surrounding area
13. Calculating the Average Prestress
Force
•The relief strain is a positive or tensile strain so a sign change is
needed
•Relief stress related to the relief strain through Hooke’s Law
σ =ε·E
•The modulus of elasticity was determined in accordance with
ASTM C469 and by the load deflection response of the beam
15. Test Specimens
Rectangle Beams Beam 1
•Cast in 2010
•Strands initially stressed to
202.5 ksi Beam 2
•Average 28-day compressive
strength: 7,440 psi
16. Test Specimens
T-Beams
•Cast in March of 2002
•Lightly reinforced in
longitudinal direction
•Strands initially stressed to
202.5 ksi
•Average 28-day compressive
strength: 7,040 psi
17. Finite Element Models
•Models created:
•Varying depth of cores: 0.75”, 1”, and 1.25”
•Varying notch depths, spacing, and lengths: Length
•Depths of 1”, 1.125”, 1.25” Depth
•Spacing of 2.5”, 3”, and 3.5”
•Lengths of 2”, 3”, and 4” Pin Support
•Beams restraint as a pinned, roller
Roller Support
21. Finite Element Models
Variable Notch Depth
Simpson's Rule % Relieved
Notch Depth (in)
Calculated Stress (psi) Stress
1 -352.85 76%
1.125 3.97 100%
1.25 317.93 121%
*Spacing between notches 3.5” and length of notch 3”
22. Finite Element Models
Variable Notch Spacing
Simpson's Rule % Relieved
Notch Spacing (in)
Calculated Stress (psi) Stress
2.5 283.18 119%
3 81.45 105%
3.5 -352.85 76%
*Depth of notch is 1” and length of notch is 3”
23. Finite Element Models
Notches on T-Beams
Core perpendicular Core parallel to Notch
to web bottom of beam
*T-beam properties were the same as the rectangle beam models
30. Conclusions
•A 3” core bit used with a 2” strain gage resulted in an almost
complete rebound of the surface strain when coring to a depth of
1”, with an average error of less than 8%
•A notch depth of 1”, spacing of 3.5” and length of 3” provides
more varied results, with an average error of around 20%
•The Laser Speckle Imaging device provided a quick and accurate
way to measure the strain.
•Multiple locations can be tested to reduce the overall error, and
taking the average of 4 cores is recommended.
31. Conclusions
•Strain drift due to temperature change can be mostly eliminated by
allowing 10 minutes after coring and 5 minutes after notching.
•Finite element models successfully predicted the amount of
relieved strain similar to the experimental results, and could be
used to determine the optimal method for other geometries and
strand configurations.
•Reinforcement around the core area significantly affects the
measured relief strain, and steps should be taken to prevent coring
in the immediate vicinity of stirrups.
The process of determining the average prestress force in the member by the surface strain relief method involves four major steps: 1) Setting up the initial strain measurement device whether it would be the use of Linear-resistance strain gages or the use of a Laser Speckle Imaging Device, then coring around the strain gage or cutting notches on each end of the strain gage or the area measured using the LSI device. Once the area is notched or cored the relaxation of the concrete is measured and then the measured strain relaxation can be used to calculate the average prestress force in the member.
2” Linear Strain gages were used due to the large size of aggregate present in many concrete mixes. The gages were attached to the beam using a low creep epoxy to minimize any error in the measurement. In order to core around the gage, the terminal wires needed to be disconnected so a four pin terminal block was attached to the gage after the gage had be protected with two protective layers; polyurethane coating and microcrystalline wax.
Dr. Peterman, you know more about the Laser Speckle than I do so feel free to expand on what I have put on the slide
A 3” outside diameter dry diamond coring bit show in the lower right picture was used for the coring method and a 4.5” diameter dry diamond cutting wheel attached to a hand held grinder fitted with a cutting and dust extraction guide was used for the notching method. To monitor the effect of temperature increases and correct for this a non-contact thermometer was used to measure the temperature throughout the process
The gages were attached to the beam in a configuration as show in the figure with the gages aligned to the maximum strain in the beam and at the same height as the centriod of the prestressing steel
Once the gage has been zeroed, the lead wires are disconnected at the terminal block and a coring guide is attached to the beam, as shown in the upper right photo, next using the coring bit and drill the gage is cored around as shown in the left figure. The lower right figure shows the gage once the process is complete
The notching process is similar to the coring procedure with a few differences, the LSI device was used with the notching procedure to achieve a higher accuracy. In the upper left corner, the LSI device is mounted to the beam and initial reading taken. Then the LSI device is removed and the notches are cut as in the figure on the right side. After cutting the notches, the LSI device is reattached to the beam using the mounts attached to the beam and a measurement is take.
The relief of strain is a positive or tensile strain due to the gage initially being set to zero strain, and then when the compression is relieved in the core or between the notches, a tensile or expanding strain is measured. To relate the relieved strain or residual stress, a sign change is needed to represent the compressive force in the beam
Beam elements with an initial stress equal to the average prestress stress were used for the prestressing steel and structural hex elements were used for the remaining beam. Individual properties were given to the steel and the concrete used in the models. Each model consisted of approximately 15,000 elements, with a average mesh size of 1 inch for the majority of the model and a mesh size of .25 inch around the core or notch. This was done to reduce the computational time of each model. Red = Maximum Tensile Stress Orange = Approximate Zero Black = Maximum Compressive Stress
To determine the measured relief strain from the finite element models, the strain values were plotted across the length of the core or notch then using Simpson’s Rule the measured strain was calculated across the area of where the gage would be positioned as shown in the figure.
Three models were created with the only variable changing was the depth of core (3/4”, 1”, and 1 ¼” ) One inch core provided the closest to 100% relief of stress
Next using the finite element models, various notch configurations were modeled
Finite Element models were made using the optimal method determined from the rectangle beam models to compare the results, also the direction of the core was investigated
The direction of core did not create a significant variation in the relieved stress, this was modeled due to the tapered web of the T-beams to see the error created if the core was not drilled parallel to the base of the beam Overall the relieved stresses were comparable to the rectangle beam models
Beam 1a and 1b were casted at the same time, so theoretically each beam should have the same properties. Beam 1a was tested 3 months after casting while beam 1b was tested around 10 months after casting to compare the effect of time on the method.
Beam 2a and 2b were casted at the same time, so theoretically each beam should have the same properties. Beam 2a was tested 3 months after casting while beam 2b was tested around 10 months after casting to compare the effect of time on the method. Core 2 on beam 2b was damaged while increasing the depth from ¾” to 1”
The error with the notching procedure was generally larger (around 10% larger) than the coring procedure. The LSI device had smaller errors that the strain gages, due to the LSI device having a larger gage length so it captures more of the double hump as shown in the finite element models Beam 1a was not tested with the notching procedure due to lack of room on the beam and the amount of cores taking from the beam
The T-beams had errors similar to beam 1b and 2b which were tested approximately 10 months after casting Reinforcement running through the core significantly effects the results of the core as shown in Cores 3 and 4 on T-beam 2
Larger overall errors than with the coring method, similar to with the rectangle beams data Notch location 4 and 5 at a depth of 1.25” and error was seen with the LSI device making a larger error and were not used to compute the average prestress force in the beam as shown on the next slide
This would be the last slide, click the text to change it to your information.