[Dee Kay]

Hi All,

I do have a steady-state problem with 4 different surfaces from different bodies in an assembly exposed to radiation from coal combustion gas.

I've Number of Histories = 50 000 000 on each of the surfaces exposed to radiation and I do have %SD Sur > 30 % on three surfaces as attached.

Does that mean I have to keep increasing the Number of Histories? If yes, I guess that will be an order of magnitude. As the set-up is currently, computationally it is taking a while to go through iterations. Is there a way of making this more efficient with a huge Number of Histories?

Thank you for your help with this.

[Opaque]

Originally Posted by Dee Kay

Hi All,

I do have a steady-state problem with 4 different surfaces from different bodies in an assembly exposed to radiation from coal combustion gas.

I've Number of Histories = 50 000 000 on each of the surfaces exposed to radiation and I do have %SD Sur > 30 % on three surfaces as attached.

Does that mean I have to keep increasing the Number of Histories? If yes, I guess that will be an order of magnitude. As the set-up is currently, computationally it is taking a while to go through iterations. Is there a way of making this more efficient with a huge Number of Histories?

Thank you for your help with this.

Interesting setup. You have 4 disconnected volumes where radiation is activated. Is that intentional, or a mistake setting radiation boundary conditions at some domain interfaces?

Also, you do not have 50 000 000 histories coming out of the exposed surfaces. That is not how the model works. You have 50 000 000 histories available for tracking for every radiation group you have indirectly defined. How many histories are used per surface exposed is a function of the relative temperature between the participating surface/volumes. That is, for very cold surfaces, you get less number of histories. For example, for a surface at 0 [K] you get 0 histories out that surface.

You can change how often the radiation model is solved. Say you solve radiation every 5 iterations. It seems you have a steady-state model, so you should be fine.

[Dee Kay]

Originally Posted by Opaque

Interesting setup. You have 4 disconnected volumes where radiation is activated. Is that intentional, or a mistake setting radiation boundary conditions at some domain interfaces?

Also, you do not have 50 000 000 histories coming out of the exposed surfaces. That is not how the model works. You have 50 000 000 histories available for tracking for every radiation group you have indirectly defined. How many histories are used per surface exposed is a function of the relative temperature between the participating surface/volumes. That is, for very cold surfaces, you get less number of histories. For example, for a surface at 0 [K] you get 0 histories out that surface.

You can change how often the radiation model is solved. Say you solve radiation every 5 iterations. It seems you have a steady-state model, so you should be fine.

Thank you for the response, Opaque.

I think I've to give a bit of context. These 4 bodies are part of an assembly sharing topology with connected interfaces and edges (Set-up). These bodies are made of different materials i.e copper, Ni-Cr alloy, steel and refractory material.

There is water flow through two of the solid domains as shown in the attachments (Set-up_Mesh).

The core of the assembly is exposed to flowing combustion gas and there is temperature development which increases the temperature as shown (Set-up_Radiation).

I've modelled those 4 surfaces with both convective and radiative heat transfer as shown in the attachment (is this correct for the 1500 C exposed domain?). The temperature given is the known gas temperature from previous work but am not modelling gas and my intention is to have the effect of that gas temperature and radiation to the surface (Set-up_Domain and Set-up_Surface).