CSM – Geometric kinks during span-by-span bridge construction with Vorschubrüstung

SOFiSTiK
, , ,
1

Hello everyone,

I am modelling an existing multi-span prestressed concrete bridge (Talbrücke), originally constructed span by span using a movable scaffolding system (Vorschubrüstung).

In CSM, I activate the stiffness of each new construction segment first (ICS1) and its self-weight later (ICSD), using CTRL CANT 0.

My problem is that noticeable geometric kinks appear at the construction joints when the next segment is activated.

According to the original construction documentation, the fresh concrete load is initially transferred to the cantilever of the preceding segment. After the new concrete hardens, the Vorschubrüstung is lowered, the temporary load is removed, and the new segment’s self-weight becomes active.

I also tested this sequence in a simplified three-span model. The temporary load is correctly removed, but its influence on the deformation is not completely reversed after the next segment’s stiffness has been activated.

Interestingly, when I use tangential erection (CTRL CANT 2), the kinks disappear, but the bridge progressively deflects upward as new segments are activated.

My questions are:

  1. What is the recommended CSM approach for modelling this construction method?

  2. Is CAMB MODE LINE the appropriate solution for the geometric kinks, or should the segment activation/erection method be modified?

  3. Are there any SOFiSTiK examples specifically dealing with bridges constructed using a Vorschubrüstung?

Any advice or references to relevant examples would be greatly appreciated.

Thank you!

2 Answers

2
2

Dont worry

Kinks is what you should get! When you do span by span erection there should be displcemnt jumps at jointa. Otherwise solution is wrong

Thank you for your reply! Just to clarify: do you mean that these kinks in the uncorrected CSM displacement profile are expected even when the temporary suspension reaction from the Vorschubrüstung (including the fresh-concrete load) is applied to the preceding cantilever and later removed? Also, would you recommend CTRL CANT 1 for this type of span-by-span in-situ construction, followed by CAMB MODE LINE for the intended final geometry? I am mainly trying to distinguish the expected erection geometry from an incorrect construction-stage modelling assumption. Thank you!

 –
3

I would use cant 0 unless geometric nonlinerity is needed to be accounted