Completely insulated object?

Energy2D Forum Forums Energy2D Questions Completely insulated object?

This topic contains 5 replies, has 2 voices, and was last updated by  Charles Xie 5 months, 2 weeks ago.

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #231

    stefanoss
    Participant

    Hallo!

    My question is (I think) a very simple one: it seems that, even if I set the environment with 0 conductivity and without any convective effect, in the simulation there is some kind of interaction between an object and the environment itself.
    If one puts an object with given initial temperature in a colder environment with zero conductivity, convective option unchecked, the temperature of the objects gets lower and lower (this becomes more and more evident by increasing, for example, the specific heat of the environment which, if I understand well the simulator behavior, should not exchange energy with the object (see the attached E2D file).
    What am I doing wrong?

    Thanks for the beautiful Energy2D!

    greetings
    stefano

    Attachments:
    You must be logged in to view attached files.
    #233

    Charles Xie
    Keymaster

    Good question. I confirmed this with your model.

    Now, if the specific heat of the square is larger than or equal to that of the ambient, we can find out that the thermometer will read almost the same value or slightly smaller as initial.

    The interesting thing happens whenever the specific heat of the ambient is larger than that of the square–the thermometer starts to register a lower value. But it eventually stops dropping further and settles at a certain temperature. The larger the specific heat of the ambient is, the lower the final temperature is. Which is exactly what you reported here.

    The reason, I figured, is because of the boundary layer effect. When you have a layer that has a huge heat capacity next to the square, it will draw some energy from the square into it. The zero conductivity of the ambient stops the flow of energy through the ambient, but it doesn’t necessarily prevent the energy from getting into the contact layer.

    I attached a screenshot of a revised model and the actual model file for your reference.

    Hope this helps.

    • This reply was modified 5 months, 2 weeks ago by  Charles Xie.
    • This reply was modified 5 months, 2 weeks ago by  Charles Xie.
    Attachments:
    You must be logged in to view attached files.
    #240

    stefanoss
    Participant

    Good morning. Surely it helps, thanks for your quick reply. I raised my question since I was comparing the E2D outputs with an analytical (exact) solution and I found a systematic, even if little, discrepancy. The point is that, according to the heat equation, with zero conductivity, no matters the heat capacity value, the temperature of the square should not change. So, in view of this “leaking effect”, I am forced to alter artificially (and in quite an unpredictable way) the ambient heat capacity in order to recover the actual temperature value during the transient time. This happens also for more complex situations, of course, where the ambient-object interaction should not have any role on the behavior of the system. I notice that in proximity of the boundaries between the square and the ambient there is a very tiny, yet non-zero energy transfer even with zero point zero conductivity. This should not happen even if one follows a finite-difference numerical method as it happens in E2D, or am I wrong?

    Thanks again, cheers
    Stefano

    #241

    Charles Xie
    Keymaster

    Of course you are correct, Stefano. This problem is related to the relatively large grid cells used by default in Energy2D (100×100). If we can reduce the size, the effect can be reduced, but I suspect we can’t get rid of it completely as this is the side effect of numerical simulation based on discretization. A numerical solution is always only an approximation.

    You may notice that the leaking only happens at the interface. Although the medium has zero conductivity, in reality does it really prevent the layer from getting any energy through the contact? What would happen to the molecules of such a hypothetical (and non-existent) medium at that interface do when they are in contact with a heat source? The physicist part of me would say they would heat up no matter what. Otherwise, this would be an open violation of fundamental laws.

    So this offers an “explanation” for the numerical problem if that comforts us.

    • This reply was modified 5 months, 2 weeks ago by  Charles Xie.
    #242

    stefanoss
    Participant

    You got an important and sensible point, i.e. reality vs. model!

    I do agree on everything.

    Thanks again, if you like so I will keep you informed about my project (which makes use of a FLIR one camera and some simple, yet interesting experiments).

    Greetings!
    stefano

    #243

    Charles Xie
    Keymaster

    Please do.

    I am also working on an app based on FLIR ONE (literally as I am replying your questions). So we have overlapping interest.

Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.