# Simulation CFD

#### Forced Convection, Equation of state, Radiation

255 Views, 9 Replies
12/20/2017 3:10 PM

Hello,

I'm  currently running a thermal simulation and after 20 iteration my model started to diverge.

I turn off auto-forced convection and i enable radiation. to avoid black body interaction i change the air emissivity to ''0.3''

The main purpose of turning off ''Auto forced convection'' was basicly to have more realistic simulation where air proprieties changes with temperature.

For air density i use equation of state for the variation method.

By changing these two solving option i'm observing a completely new pattern of temperature distribution and even some ''high buble temperature'' near the opening. Few questions related to these topics :

1. I still have difficulty to include radiation in my analysis to have tangible result, which practice i need to be taking care of ?

2. if i want a dependance of air proprieties regard to surrounding temperature which variation method are the more appropriate :

equation of state , 4th order polynomial interpolation ?

Merci
Fred

Report Inappropriate Content
Message 1 of 9 ( Views: 255 )

#### Re: Forced Convection, Equation of state, Radiation

12/20/2017 3:32 PM

I see one main issue here

You cannot have internal boundary conditions. So either apply a volumetric load to those tubes or use a surface one and suppress them from the mesh.

Get steady state working before trying to get transient running, it is usually a far shorter path to success ;)

What do you expect to happen at the inlets where you have a P=0? Does a lot of flow really pass there? The problem you might have is if you have air recirculating over the P=0, the solver is really going to struggle. Might just a temp be OK?

EDIT: Where else is heat lost? You are adding a lot but it can't really escape. Can none leave anywhere other than the P=0's?

• I still have difficulty to include radiation in my analysis to have tangible result, which practice i need to be taking care of.
• The above will help a lot, right now the tubes cannot radiate properly
• If i want a dependance of air proprieties regard to surrounding temperature which variation method are the more appropriate
• Could you explain this a little more? Which parts and which surroundings?

Thanks,

Jon

Report Inappropriate Content
Message 2 of 9 ( Views: 249 )

#### Re: Forced Convection, Equation of state, Radiation

12/20/2017 5:00 PM

Hello John.

I see one main issue here

1.   Understood,  i will use the same boundary condition and then supress the mesh from the rod, thx for the advice :)

Should I add more wall layer around them, like five instead of three ?

2. because i was running a thermal simulation i was assuming that i was not allowed to run steady state calculation.

3. The wall defined by P=0 are connected to the exterior environment , air are free to go or free to enter, and yes i have recirculating air across the surface. I will say it's important to consider this aspect because cold air could flow in and it lower the performance of the oven.  The velocity are low 3 m/s maximum over the oven gate,  see picture ''Air_oven'' and ''Air_1'' to have a better idea.

If i want a dependance of air proprieties regard to surrounding temperature which variation method are the more appropriate

• Could you explain this a little more? Which parts and which surroundings?
• The density of air change a lot over the whole range of operating temperature. It may change the way that heat is propagading or the convection pattern (just a guess)?
•  The next step of this project is to add some bloc at 25degC and pass it throught the oven and calculate the convection factor with a maximum precision, i know temperature can have an effect on this parameter because physical propreties change for different Temperature.

Thank for sharing this knowledge :)

Fred

Report Inappropriate Content
Message 3 of 9 ( Views: 243 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 7:49 AM

Only suppress the rods if you are OK to discount radiation. Otherwise you need to mesh the rods and assign a volumetric heat load to each part.

We need the solids meshed to be able to radiate to and from them (which makes sense I think).

Thermal is fine as steady state :)

I would be really careful about those openings. Ideally we want to ensure that flow only passes one way over a P=0...

What sort of environment does this sit in? Might you be able to model a larger air domain at all? Hypothetical right now although I have an idea...

Use air and change to variable (from fixed) to allow the density to change with temperature. I would think it makes very little difference with two fans in place, other than to slow down the solve but it might be worth an experiment.

If you plan to run motion in the end you will have to run flow and thermal transiently at some point, just FYI. It can work well though.

Are the over walls adiabatic or can heat also escape there?

Thanks,

Jon

Report Inappropriate Content
Message 4 of 9 ( Views: 226 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 3:13 PM

Hello John,

• Only suppress the rods if you are OK to discount radiation. Otherwise you need to mesh the rods and assign a volumetric heat load to each part. We need the solids meshed to be able to radiate to and from them (which makes sense I think)
• Yeah it make sense
• I would be really careful about those openings. Ideally we want to ensure that flow only passes one way over a P=0...

What sort of environment does this sit in? Might you be able to model a larger air domain at all? Hypothetical right now although I have an idea...

•

• Usually the environment is the define by the atmosphere characteristic at the sea level (P=101.325KPa T=25degC) but it may change a litlle if it's used at high altitude.
• I already consider to add 2 air volume next to those opening. Should I go for a huge cube of air to modelise the ''outside surrounding''? The specification could be confusing because this volume need to reprensent atmosphere.  I guess T and P need to be constant no matter the amount of hot air who's pourring in. Really not sure how to handle this in term of boundary condition.
• Are the over walls adiabatic or can heat also escape there?
• No, they're not ! Natural convection and radiation occurs on all surface of the oven.  All wall have isulation material, so heat is first run throught the insulation by conduction and then escape by natural convection and radiation. Those loss are negligible but for sure they will be consider eventually.

Merci Beaucoup
Fred

Report Inappropriate Content
Message 5 of 9 ( Views: 223 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 3:40 PM

How about an approach like this?

Ground Based Natural Convection

I know it will still have flow in and out of a surface, just the top one and we know from experience that it can work OK :)

Plus then you can easily include the conduction through and radiation from the walls just by enabling radiation!

I think we are getting there, you?

Report Inappropriate Content
Message 6 of 9 ( Views: 220 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 4:06 PM

We're getting there :)
I'll try this right away.

last question:

by using the ground based model should a define a thickness on the oven wall ? otherwise it will be air-air conexion

and  does P=0 should be supressed on the openning of the oven ?

Fred

Report Inappropriate Content
Message 7 of 9 ( Views: 218 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 4:35 PM

Yup, model it close to reality. give it a thickness and the right material.

Remove all internal surface boundary conditions, so the flow just naturally moves as it will (plus those fans of course).

Does that help?

Report Inappropriate Content
Message 8 of 9 ( Views: 212 )

#### Re: Forced Convection, Equation of state, Radiation

12/21/2017 8:23 PM

I'll try

but include a large amount of air is really slowing  my calcul.  I will simply put two distinct air volume  at the opening.

Thank you

Report Inappropriate Content
Message 9 of 9 ( Views: 210 )