This is the main repository for compressibleVoF (compressibleInterFoam) https://cdn.discordapp.com/attachments/726377781377630318/1372206758159712340/image.png?ex=6825eea1&is=68249d21&hm=3bd7240720c4693b6964a6e5b56136bd5b94ed695bc1b7d5f079f70d1d39a237&

and basically you must go in this "fvModels" folder, copy the cavitation repository (since it's phase change on pressure), change the parameters to link it to temperature, and link the good libraries. Finally you recompile the whole https://cdn.discordapp.com/attachments/726377781377630318/1372207030734946376/image.png?ex=6825eee2&is=68249d62&hm=486b18eecc00a05a97041ce8c7cdf09cddb0f3304249e869e078ffd9574020e7&

if I can do it, I will share it here 👍 although I don't know how usefull it will be for you guys 🫡

Hi everyone So, I have a stirred tank with a coil for heat exchange. A solid enters and melts into liquid, and I want to determine the temperature in the bulk. However, there's the effect of latent heat, which removes energy from the system and lowers the temperature I tried creating a source term to extract heat from the medium, but it didn’t work. Then, chatgpt suggested using an equation for effective specific heat: Cp_eff = Cp + (latent heat / deltaT) I’d like to know if this approach is actually valid. I’m not sure how, but it worked 😅

Can you define effective specific heat?

The division by deltaT must be compensated by a remultiplication in a close by equation

Because q= mCpdeltaT1_liq +mH_latent+mCpdeltaT2_sol Where deltaT1+deltaT2 = deltaT total and H_latent would occur at some intermediate point but basically removes a constant amount of heat.

From what you've written I'm guessing that deltaT2 is zero and then everything is getting divided in chatGPT expression and then remultiplied to get total heat... Things would be slightly more sophisticated if there was also a deltaT2 non zero. Because then you'd have two Cp's

By following this kind of multiplication and division logic across all the relevant equations, it should become evident what is happening and how it ends up being the correct answer.

According to what chatgpt explained, this deltaT would represent the small temperature range over which the phase change occurs — for example, if the latent heat is released between 140  and 150 °C, then deltaT would be 10 °C

Is this an alloy?

It seems quite consistent with the theory, after all, Cp = dQ/dT, so Cp_eff = Cp (without latent heat) + Cp due to phase change (dQ/dT). But I haven’t found any case studies, papers, or textbooks that use this method to account for latent heat removal, so I’m still a bit uncertain about its theoretical foundation

is a polymer, polypropylene

I didn't want to trust an AI 😅

And it has a range of melting? I'd typically expect that the temperature stays constant until melting is complete (that's how melting, freezing,boiling work in general)

Or are you saying that the 10 degree drop is on the other side ? Because in that case it would make sense because H = Q = Cp(of material from which heat was removed)deltaT, and thus Cpeff of that material would be it's Cp + H/deltaT.

Which side's Cp and which side's H is being talked about?

That's a very good idea. it's been going around suggesting people to change turbulence model constants to get right answer , so you've got to be careful.

There is a temperature range in which the material exceeds the melting point and actually melts, deltaT would be this "range" where the melting point exists, so it would be the latent heat/melting temperature

yes, the idea is very good and in the simulation it is working better than a source term, but I haven't found anything similar

This is a little confusing in terms of whose temperature is changing, and why the temperature is exceeding if it hasn't melted (which is contrary to expectation since any excess heat supplied at melting point will go into melting so there should be no rise in temperature before 100% melting has occurred )

You don't get ice at 10 degree C. The ice gets to 0 degree and then turns to water at 0 degree. The water then goes to 10 degree. The ice doesn't go to 10 degrees and then melt, do you get what I mean?

it makes sense

but what chatgpt wrote: "The ΔT represents the smoothed temperature range around the melting/solidification point, where the latent heat is distributed. It is an artificial thermal bandwidth used to avoid numerical discontinuities in the simulation. ΔT is a numerical, not physical, parameter that defines the smoothing of the solid-liquid transition. It does not depend on the actual temperature of the fluid, but rather on the thermal range in which the latent heat is "diluted". Its choice affects"

IDK how to evaluate that on face value. This is basically vibe-coding your simulation. But I didn't imagine it'd be this hard to get latent heat into picture because it is a known constant so if you know material throughput you can estimate the heat requirement rate to keep converting that throughput from solid to iquid

I mean this isn't something I encountered in my heat trasfer class and LLMs whole deal is producing sensible sounding string of words. So, may be if it works it works, but I would highly encourage verification against theory and some standard texts, basic analytical examples type situation where this kind of smoothing might have it's value directly apparent (since that was your original point).

One way would be to identify any open ended choices involved, e.g. choice of this band of dilution, and how it affects the solution. That'll give you and appreciation of numerical vs physical aspect of it.

I’m going to try to validate this better, I found the approach interesting, and it’s not something I remember seeing in standard heat transfer classes, but there are definitely a few points that don’t quite make sense to me. I’ll look into how this temperature range actually influences the process

Quick question : the equivalent of an adiabatic wall boundary condition is "zeroGradient" boundary condition in /T right ?

right