I’m currently working on the analysis of the propagation of the temperature inside a composite steel-concrete bridge cross-section (when the air temperature is variable).
For this, I put myself into Ansys for the 2D thermal modeling of a symmetric cross section (two steel I-beams under a concrete slab). I used elements PLANE55 and designed only one halt of the section. The materials properties that I’m using are:

-Density: 7800 kg/m3 for steel and 2300 kg/m3 for concrete
-Specific heat: 465 J/(kg K) for steel and 900 J/(kg K) for concrete
-Thermal conductivity (isotropic): 54 W/(m K) for steel and 0.55 W/(m K) for concrete

Now, for the application of a convection load on the lines, I have to declare a film coefficient for the contact material-air. And that’s where I’m blocked. Does it depend on the sort of the material? What about the orientation of the line (vertical/horizontal)?
I was advised to use a value between 5 and 25 W/(m2 K) for free convection with air.

Am I on the good way so far?

And it would be very interesting if I could insert also a radiation load (equally variable – daylight), but I still don’t know how.
If someone could help me, it would be great!

Thank you in advance!

Film coefficient is mostly dependent on the air itself, but there are some modifications for turbulence, etc., that are dependent mostly on the length of the structure.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Thank you for your answer IRstuff.
Finally I applied only one film coefficient for both materials!

I’ll create a new thread at the Ansys section with a link for this one to see if I can get any help with the rest of the modelling!

Thank you!

The Nusselt number and so the heat transfer coefficient depends both on air and surface material temperature. For still air (natural convection) the range you’ve reported is reasonable. If you the take the higher values of that range you can safely assume not only natural convection but also radiation is taken into account.

Thank you for your answer ione !

Actually I don’t know the surface material temperature… I have a table with the air temperature every 10 minutes in 24h. The goal of my analysis is to get the variation of the temperature inside the bridge during a day.

With these results I’ll be able (I hope so!) to see the interference of a (possible) thermal gradient between the concrete and the steel at the bridge deformation.

The problem is that I need to suppose almost everything about the materials properties (we don’t even have the resistance of the concrete!).

I’ll suppose that at 4AM everything is at the same temperature (air and structure), and after that I’ll change the air temperature during the next 24h. I think that if there’s a mistake at the initial temperature of the elements, it will be regulated in less than 24h…