Guest
Comments on the Cable Analysis Solution Algorithm and
Assumptions in GTSTRUDL v25
1. The problem:
Occasionally, GTSTRUDL cable analysis will predict compression stresses in the cables.
2. Explanation:
a. The initial implementation of cable analysis in GTSTRUDL was based on the understanding
that design code specifications strongly suggest that all cables maintain a sufficient state
of tension under all loading conditions. Such a requirement is intended to prevent
undesirable aerodynamic behavior of the cables such as galloping, slack-taught pounding,
and aeolian vibration (singing).
b. Since codes generally require cables to maintain a state of tension, it is necessary that the
engineer specify initial prestressing conditions which would ensure that a state of tension
will exist in all cables under all loading conditions.
c. The initial implementation of the nonlinear cable analysis solution algorithm in GTSTRUDL
did not include a "tension-only" feature. This was done in order to improve the convergence
efficiency of the algorithm. However, if the initial cable prestress conditions are not
sufficient, it is possible that the solution procedure will predict compression in some cables.
If this condition occurs, it usually occurs during the nonlinear analysis continuation for the
service loads which follows the cable prestress nonlinear analysis. Compression stresses
may indicate the possibility of a galloping or pounding problem.
3. Corrective Action:
a. If design codes require that all cable stresses are to be in tension under all loading
conditions, it is not the responsibility of the analysis program to change initial prestressing
conditions in order to achieve the desired state of final stress under all such loading
conditions.
b. Therefore, if compression stresses are predicted in the cables during the nonlinear analysis
for one or more loading conditions, the initial prestressing conditions must be changed by
the engineer so that the final state of computed cable stress results in tension only in all
cables. This may require multiple trials of initial conditions before success is achieved.
4. Future Versions of GTSTRUDL:
We are planning to add an option in the version of GTSTRUDL that will be released after the next
new version in which the user may specify that cables are strictly tension-only. However, even with
this feature, it will still be the responsibility of the engineer to change the initial prestressing
conditions so that the final state of computed cable stress results in tension only in all cables (i.e.,
so that no cable is in an unstressed slack state).
