I have a simulation of a transition piece/connector piece between a concrete shaft and a steel T-flange. The T-flange rests on a leveling plate and is connected to the concrete through bolts on both sides. The concrete has post-tensioning cables.

I wished to find a better way to mimic the real process of first tensioning the cable, then adding grout on top of the cable and then tensioning the bolts.

My idea was to apply a very small Young's modulus to the grout-part while tensioning the cable. Then I changed the Young's modulus of the grout to a suitable value in it's own step using field parameters before tensioning the bolts in the next step.

I wanted to check that the top of the grout was level so that there was still contact with the leveling ring after tensioning the bolts and I realised I got a very strange deformation contour. I have attached a snip showing the U3-deformation for two analyses. They are exactly the same except for the change in material stiffness. To the left is the contour for a simulation with the same material stiffness applied to the grout throughout the analysis and to the right is the same contour for the simulation where I've changed the stiffness (increased it) after applying tension to the cable.

Why does this happen?

The stress contours and contact pressure contours seem fine and are comparable between the two analyses. But my bolt forces are significantly higher in the analysis with changed stiffness of the grout. The contact pressure contours also reveal that more of the applied force (applied to a tower sitting atop the T-flange) goes into the bolts rather than relieving contact between flange and levelling ring.

I wish to figure out why I get different bolt loads for the same applied level of loading. The biggest deviation between the analyses (apart from the bolt load of course) are the deformation contours I have attached.

More Ingrid Holte's questions See All
Similar questions and discussions