Box girder - effects of horizontal tendon eccentricity

[robkor]

I'm designing a 4-span (30m+2*38m+30m) post-tensioned bridge – box girder, which shall be constructed span by span. The thickness of webs is 40cm (midspan) and 60cm (above supports +/-7.0m).

I want to have a HORIZONTAL tendon profile as simple as possible (only few horizontal deviations; possibly straight horizontal profiles). As a result the horizontaly straight tendons, which are in the centreplane of the thiner web(40cm), they become eccentric on the thicker web(60cm), with an eccentricity of 10cm.

Now the problem. I don't know how to calculate these efects of eccentricities on the web. I am afraid that the web will deflect and crack in the inner part after tensioning. I don't have a software with 3D solid finite elements but only shell elements. I tried to model one span with these but the problem is in how to model the transition from thiner to a thicker web, because of the centreplane eccentricities. I tried to model the transition with an angle of 45° but the results were catastrophic as I couldn't calculate the reinforcement because of huge inner forces in the transition zone.

Please, could someone suggest me how to consider these effects.
Thank You very much for your response

Robert

PS: I'd like to attach a file for explanation, but I don't know how.

 

[wiktor]

1. The eccentricity of 10cm is still within "the core" of the section, so although there is some horizontal moment involved, it will not create any tension.
2. The "huge inner forces" in the transition zone point out to some errors in the model - you can calculate your forces assuming 40 cm section and 60 cm section, and smooth transition in between.
3. The issue of local bending of the web (still in full compression - see above) is, in my opinion, superficial, as the girder will have no eccentricity, and to rigorous calculation of the stresses in the points of discontinuity typically lead to completely misleading results.

 

[Dinosaur]

I worked on a project where we used 45m 5 span continuous concrete PT box span-by-span with 200 mm thick webs and that project was horizontally curved. It would seem that your problem is not too severe.

Good Luck

 

[robkor]

Thank you very much for your response!

When explaining my problem I took 1 tendon, but in reality there are 4 tendons on each web with a force of cca. 14000kN (28000kN both) with total eccentricity of 7cm. Also the bridge is curved with R=666m.

I understand, that the whole web is under compression if you take a tendon and then calculate the stresses from Np/Aweb +- Np*e/Wweb but in reality the compression force that "belongs" to a single web is only a part of the prestressed force let's say Fp,web=Aweb/Awhole*Fp, but I suppose that the moment that the tendon produces because of it's eccentricity on the web is still Np*e (particularly if the bridge is curved). I suppose I'm not right here, but if I am, then I get tension in the inner part, otherways I get very big compression in the outer side (30-40MPa) which exceeds permisible pressure for concrete C35/45(EC2). Increasing concrete strenght class above C40/50, for structures of such spans, would not be normal with our regulations/expectations.

Again, thank you for your explanations.
Robert