Discussion Forum

Hi

How do you define your cooling rate ?

is it the temperature derivative for a transient analysis ? this should be Tt = dT/dt in COMSOL notation

be sure your time derivatives is turned, that your physics do solve for T and that you are in a transient analysis

in v4.1 HT you have also the (ht.gradTphi, ht.gradTr, ht.gradTz) and ht.gradTmag

--
Good luck
Ivar

Hi Ivar,

Thank you for the quich reply!

Yes, I mean the cooling rate (CR) is the first derivative of temperaure, dT/dt [K/s]. I use Comol 3.4 release, transient heat transfer module. Are ht.gradTphi, ht.gradTr and ht.gradTz expressions available in v3.4?

If not so, could you please advise what sort of formula (expression, function) should I use to determine the cooling rate at 700 C (CR700) in the whole domain?

thank you again,
best

Imre

Hi

no, all prefixed_dot_ are v4 notations, in v3 it was suffix _ht

then I suppose that Tt should be defined as dT/dt (your version is old, I have forgotten in the mean time ;)

But in 3.5 I remember that the time derivatives were only calculated on demand, there is perhaps also a check box to select in the solver menu

--
Good luck
Ivar

if the casting is under equilibrium condition the cooling rate will be low and you must use implicit or explicit schem and not tansient integration

Thank you very much for your responses.

I think my question was not clear, therefore I would like to ask some more help:

I am able to calculate the dT/dt (cooling rate, CR) in every time instance for the whole spatial domain by using the global expression ie: CR = diff(T,t).

However, I like to determine the CR given at 700 C (CR700). In other words, I would need the value of dT/dt when the temperature is T = 700 C for every points of the model .

So, what expression should be introduced to calculate CR700?

Thanks for your help in advance!

if you want to keep the temperature constant at 700C, i thimk it a state change: liquid to solid for exemple, so you must use heat latent, so dT/dt = 0 for few second.

Hi

I'm afraid that in 3.5 you need to go through matlab to have full control on the results.

In v4 you can easier use "filters" to limit the amount of data to i.e. a range of the bolean expression (T>LowerLimitOfT)*(T<HigherLimitOfT) and then plot Tt as variable. Then the question is to export this data, and if you need the localisation of these conditions. There again you would neeed Matalb , I believe

--
Good luck
Ivar