Prestressing

[AntoMarinelli]

Dear STKO team,
I'm modelling a prestressed beam with a I section. I have an equal area of 1260mm2 in the bottom part and 280 mm2 in the upper part, as it is shown in the picture. The initial stress is 999MPa. I evaluated the moment induced by prestressing (not considering the selfweight) and it is equal to -349 kNm. Considering that the test is a four point bending test, the force that produce a moment equal to the prestressing one is 170 kN (M=P*a where a = 2.05). Infact, form the experimental curve at 170 kN the displacement become positive (second picture).
In the STKO curve the displacement become positive at 371 kN and I don't understand the reason.
I attached the section, the curve and the model.
(The centroid is at 395 mm and the concrete cover is 50mm ).

Thanks in advance
Regards
Antonella

[marafini.f]

Dear Antonella,
I would check your concrete characteristics. The numerical response seems to be matching the experimental one, so I think all the considerations you made in terms of geometry are correct. I would suggest you check your concrete stiffness, is it coherent with the stress and strain relationship you defined. I think it's a bit high but you should make an attempt with a lower number and check the graph again.

Let me know how it goes.
Francesca

[AntoMarinelli]

Dear Francesca,
I have already performed the analysis that you suggested considering E=36000 MPa and not the direct value proposed for C60 class (40700 MPa). The changes in the final curve are negligible. Maybe I have lo lower it again but it doesn't make physical sense because the concrete is a High Performance Fiber Reinforced Concrete so the Elastic Modulus is around the value that I set. Moreover, the software does not allow to lower much than 30 GPa.

What could be the solution?

Thank you
Regards
Antonella

[marafini.f]

What about self-weight?
There isn't any in your model

[AntoMarinelli]

I didn't model the self weight because the curve does not consider it. I have also performed a manual checking and the self weight generates a positive moment of 15.40kN which not affect so much the final moment.

The problem is that the change of displacement sign could happen at 171 kN independently from the E modulus because the system is isostatic.

Thank you for the support
Antonella

[marafini.f]

Did you try modeling your problem initially all elastic and checking the result against your hand calculations?
Also could you send me the excel file with you graph so I can make an attempt

[AntoMarinelli]

Yes, sure!
This is the excel with experimental results.

Thanks

[AntoMarinelli]

Dear Francesca,
I performed the attemp that you suggested. The first linear branch does not change. I can match the experimental curve with an half of E, but as I said before it doesn't have physical meaning. I have checked all the geometry and it seems ok.
I don't know what could be still the problem..
Thanks for the support

Antonella

[marafini.f]

Dear Antonella,

so if you calibrate your concrete parameters correctly with respect to the C60 High-Performance Fiber Reinforced concrete then that's not the mistake.

I would suggest you recheck everything restarting with a linear model, including the geometry in both directions, and see if you can recreate the linear stiffness of your experiment.

Let me know if you have any progress I am sorry we couldn't help more.
Francesca