shrinkage of concrete

[visitor]

I am designing a continuous composite girder bridge.

Please tell me how can I analyse the composite girder for shrinkage
of concrete and temperature difference ? If you brief me shortly, it will be helpful to me.
I saw one report , converting the shrinkage into equivalent temperature difference in deck slab and designing for tempereature difference. I want to know how to analyse this.

 

[ishvaaag]

The old instruction for bridge roads in Spain instructs to use engineering -not FREE, but actual- unitary shrinkage between

0.000150
and
0.000400

Normally, a composite deck relying on studs iss less able to pass its retraction to the steel than an only RC structure, then the lower values should be the more likely to represent the actual behaviour.

Say you choose 0.000200 engineering shrinkage be the one applicable to your case.

Since the coefficient of thermal expansion of concrete is

0.000011 m/(m·ºC)

you will be using a drop of temperature in the deck of...

200/11=18.18 ºC

to represent the shrinkage action.

I quote literally from the Structural Engineering Handbook 17-12, Arches and Rigid Frames...

"Shrinkage effects may normally be calculated as equivalent to a 15º temperature drop."

If Farenheit would be too low so I assume it refers to centigrade scale.

In any case you see that even the assumption above is conservative respect the SEH recommendation.

On how to analyze the easy way is to use one program able to specify temperatures for members, such RISA 3D and many others. You simply specify for the deck such lower temperature than in the steel. You specify separately deck and steel parts as members joined by rigid or whatever links in this particular procedure.

Otherwise you will need to use notional compressive loads along the axis of the deck at both ends that cause the unitary shortening expected. Then from the results in the analysis you simply substract such axial compressive load of the results for the axial forces in the deck; that should be approximate enough.

If the structure is not wholly symmetrical what in the paragraph above should be better than to impose horizontal displacements at the ends, since you will be entering as input a guess.