The Effect of Temperature on the Effective Prestressing Force at Release for PCBT Girders

1
The Effect of Temperature on the Effective
Prestressing Force at Release for PCBT Girders

• Timothy Wood, Research Aid
• Charles Newhouse, Assistant Professor

2
Fabrication and Instrumentation
3
Instrumentation was installed and monitored
during the fabrication.
4
Instrumentation was installed and monitored
during the fabrication.
Vibrating Wire Gage
Instrumentation Being Placed
CR23X Data Recorder
5
Instrumentation was installed and monitored
during the fabrication.
Completed Girders
Steam Curing
6
Fabrication Sequence
7
Predictions
8
Elastic Shortening and Relaxation Loss Predictions
9
Barr Temperature Loss Predictions
10
Measured Results
11
Some concrete and bed temperatures were higher
than expected.
12
Cumulative strains show interesting trends.
13
Strains at detensioning indicate more effective
prestress force than most studies report.
14
Instantaneous measured losses and required
prestress was much higher than predicted.
15
Instantaneous measured losses and required
prestress was much higher than predicted.
16
Thermal gradient predicted more losses, not the
measured gains.
17
Thermal gradient predicted more losses, not the
measured gains.
18
Exploration of Explanations
19
Variations in the Modulus of Elasticity of
Concrete

• Measured value was consistent with predicted
  value from compressive strength.
• Ec measure 5200 ksi
• Ec predicted 57,000vfc 57,000 v(7800psi)
  5030 ksi
• If Ec 3460 ksi
• The measure strain at the centroid of the section
  would match predictions
• But curvature would be grossly overestimated,
  even above the measured value
• Initially predicted curvature 10.6 µe/in
• Measured curvature 12.5 µe/in
• Predicted curvature with lower Ec 16.0 µe/in
• Though this may contribute it is unlikely that it
  controls.

20
Variations in the Coefficient of Thermal
Expansion of Concrete

• Strain gage temperature corrections due to
  differences between concrete and gage coefficient
  of thermal expansion are required to determine
  absolute strain.
• Cylinder measured values of the coefficient of
  thermal expansion varied greatly depending on the
  amount of water in the specimen.
• Recommended value for concrete 10.4 µe/C (5.8
  µe/F)
• Minimum value for granite aggregate concrete
  6.84 µe/C (3.8 µe/F)
• Value for water 70 µe/C (38.9 µe/F)
• measured value for concrete 9.18 µe/C (5.1
  µe/F)
• A time dependent value might be more precise.
• Does not affect short-term, relative observations.

21
Temperature correction has a significant
influence on the final vibrating wire gage
reading.
22
Restraints Prior to Release may explain the
observations the best.

• The concrete set in an expanded state as
  illustrated by tensile strains due to
• Early increase in temperature
• Higher coefficient of thermal expansion
• Concrete in a plastic state

23
Restraints Prior to Release may explain the
observations the best.

• The concrete experienced restraint forces from
  prestressing steel and formwork due to
• Shrinkage
• Cooling
• This shrinkage is indicated by compressive strain
  measurements.
• In larger beams, this shrinkage produced
  pre-release cracks.
• Measured strain of 111 µe produced by only
  18.5F.
• Cracking would have occurred at 131 µe.

24
Restraints Prior to Release may explain the
observations the best.

• At release, the prestressing steel must overcome
  the tensile forces and close the cracks in larger
  beams, and then undergo elastic shortening
  losses.
• total strain movement from closing of
  pre-release cracks and tension movement from
  elastic shortening loss.
• Results in higher than predicted strain and
  camber.
• Unclear as to whether significant gains or losses
  are actually produced.

25
Conclusions
26
Further Research Directions

• Gain better understanding of early, effective
  modulus of elasticity for concrete.
• Develop testing and prediction procedures for the
  development of the coefficient for thermal
  coefficient of expansion.
• Further explore causes and consequences for
  pre-release cracks.

27
Conclusions

• Current publications on temperature effects may
  predict greater losses, but this is not what is
  often seen in the field.
• More understanding is needed in the area of
  temperature and shrinkage effects on prestressed
  concrete beam fabrication.
• Questions?