An initial T_wall is required from me to obtain the optimal spacing for a finned cylinder, through empirical correlations.
I have a choice to expose said cylinder under 1m/s or 0m/s wind current. I was wondering whether there is an analytical method or otherwise to obtain Twall, without knowing [h] heat transfer coefficient.
https://www.sciencedirect.com/search?qs=wall%20temperature%20of%20an%20inclined%20cylinder%20under%20free%20convection&show=25&sortBy=relevance
Why you are looking for an analytical method, if that can be easily predicted by a CFD simulation nowadays?
First approximation can be vertical cylinder. Wall temperature on inclined cylinder will be something different. But you have to calculate heat transfer coefficient and by iteration get the true wall temperature.
Or you can use CFD.
I do plan to use CFD. A correlation-analytical solution is for validation purposes alone, as I do not have the resources for an experiment.
But validation can be done on a first simulation with a non-inclined cylinder, where for sure lots of literature (experiments, correlations) exist. Once validated the CFD approach can be applied to other configurations.
Helloi Christos Andrea Christou
You can do with ANSYS FLUENT. You must select the wall and introduce in report then you can analysis.
Thomas Frank in terms of validation, will a Mesh Independence study count as validation? The Logic being that the Mesh Indep. result will be the most accurate value capable of obtaining by means of CFD regardless of whether I have a Published Paper to compare it with or not(which I don't).
Additionally, merely similar models to mine in Literature can be used as further validation of the range of sensible values(order of magnitude) I should obtain, correct?
No, unfortunately not. A mesh independency study only eliminates - if carefully carried out - the numerical errors from the CFD solution, like there are the round-off error, the iteration error and the discretization errors in space and time.
But what finally a mesh independency study cannot eliminate or help to quantify is the remaining model error. After doing it, nobody could tell whether for example the CFD result with a k-epsilon turbulence model or a SST turbulence model or a LES is the more accurate one. This can only be found out and be quantified, if the different CFD results are compared to real-world, i.e. experimental data.
And I'm pretty sure, that a thorough literature search will reveal some experimental data about convective heat transfer for a heated cylindrical body. It is quite a generic test case which most likely has already been studied in the past. I have the paper not to my disposal, but I did such a validation study already by myself with a Master student and there were some Nusselt number distribution over the cylinder surface for comparison.
Thomas Frank How closely should the published model be to mine? I am modelling a thermosyphon at 20degrees from the horizontal. I know I can change the direction of gravity, and material thermal conductivity, but what if I can't find the exact outer diameter(22mm) or condenser length(200mm)?
- Badges
- Science topic
- Similar topics
- Philosophy
- Moral Philosophy
- Free Will