Hello,

I simulate inst. cooling of a steel cylinder with fluent, but my results are far from matching with experimental data.

The situation is as follows:

- 2D simulation, axis symmetry

- 100 mm steel cylinder (diameter is 20 mm) starting temp. is ab 1023 K

- cooling from the front face with an impinging jet, almost 1D cooling, comparable with the Jominy end quench test, see http://info.lu.farmingdale.edu/depts...ominytest.html

- instead of water it is high pressure GAS => HTC is between 1500 and 2500 W/m2 K (that matches with data from literature)

- steel cylinder material data: -- heat conductivity temp. dependent, between 18 and 28 W /m K -- density 7900 kg/m3 -- heat capacity temp. dependent, between 450 and 630 J/ Kg K -- austenitic steel

Now, we are interested in the cooling rates at different distances from the frond face, as well as the cooling time from e.g. 800°C to 500°C (it is called "t800-500")

The simulation fits very well the measured data up to a frond face distance of ab 20 mm. BUT, the t800-500 data we simulate for higher frond face distance are more than twice as high as the experimental values!! (so the cooling is slower).

I have no idea, whats wrong. I tried a lot of different possibilities... like increasing the impinging jet turbulence, decreased the impinging jet temperature, changed the material parameters to maximum conductivity,....

Nothing helps. So there is maybe a principle problem with fluent???

Any Idea?

Ralf

What are your BC? can you be sure that energy is transported out of the domain in a satisfying way?