FHWA Resource Center
STRUCTURES TEAM
Friday, October 27, 2000
PCEF Design Parameters Subcommittee
Issue #1 - Strand Pattern Types
Background ;
The strand pattern type (e.g. Draped, straight, or straight with unbonding)
can have a significant effect on the production costs of pretensioned concrete
girders. Although it is widely recognized that draped designs are more efficient
than straight designs based purely on the number of strands required, this is
typically not the critical point of comparison from a producer's point of view.
Many issues, such as safety, cost or availability of plant tooling, setup times,
bed utilization, and plant efficiency have significant cost implications.
In addition to unbonding used to control top tensile stresses ("designed
unbonding", for lack of a better term), the use of short lengths of unbonding
(typically a combination of 6" and 36" lengths known as "crack control
unbonding") has been shown to be effective in Pennsylvania for the elimination
of hairline cracks common to pretensioned beams caused by splitting forces due
to strand expansion at the beam ends. Confining reinforcement is also used,
however it is well known that reinforcement reduces crack size rather than
contributing to the ultimate goal of elimination of the cracks.
A common misconception is that unbonding significantly reduces the shear
strength of the girder. However, while it is commonly held that the prestressing
force aids the development of shear resistance, it is also common that straight
strand patterns require approximately 10% more prestress force than deflected
strands for the same flexural capacity based on allowable tensile stresses at
midspan. Therefore, while the unbonding does reduce the prestress force at the
end of the girder, this is offset by an increase in the total force, giving a
much smaller reduction than is immediately apparent. And, since the contribution
of the vertical component of the draping force to shear capacity is small this
is also not a compelling reason. In the event that shear resistance needs to be
increased, it is more economical in most cases to add a few stirrups, or, in an
extreme case, add strands to compensate for the reduced shear capacity.
Another issue commonly cited as a perceived problem with unbonding is the
potential for contaminants containing chlorides to travel along the unbonded
length, thus penetrating deeper into the girder than would otherwise be
possible. However, given proper end treatment of the strands after removal from
the casting bed, this potential can be eliminated.
Since the major goal of this committee is to recommend areas where regional
uniformity could substantially improve quality and/or reduce costs for all
producers and owners, and since the issue of strand type selection and
modification during construction has a significant impact on costs, it is
desirable to achieve a consensus among the participating states and producers on
this issue.
From previous PCEF questionnaires, following is a listing of current
practices for the participating states ;
CT Use either draped or unbonded strands. DE Typically uses debonding and
occasionally drapes strands. MD No response. However, based on experience,
current policy seems to be to only use draping, with no alternate for straight
or unbonded strands allowed during the construction phase. NJ Use both draped
and debonded strands. NC Uses either draped or debonded strands. In cases where
either pattern could be designed, the straight strand pattern is used if the
straight pattern requires less than 6 additional strands. However, debonding is
not allowed in I-girders. PA Uses either draped or debonded strands and allows
changes during construction. VA Only uses draped strands. No debonding is
allowed. WV Uses draped strands. Did not comment on debonding, or whether it is
allowed.
Proposal ;
Given the significant usage of straight strands with
unbonding, for the states that currently do not allow unbonding as a design or
construction option, the subcommittee requests that the rationale behind this
limitation be explained so that means and methods might be proposed to alleviate
the concerns.
Given a satisfactory resolution to #1, the subcommittee recommends that the
full committee endorse draped, straight, and straight/unbonded strand patterns
as potentially equal solutions to the same problem, and also the concept that
alternative patterns may be proposed during construction.
The subcommittee welcomes comments and suggestions on this proposal. Comments
should be received by December 31, 2001 so that a draft ballot can be prepared
in time for our spring PCEF meeting.
TRAINING
