AnsweredAssumed Answered

Air thermal conductivity issue

Question asked by Domenic Santoianni on Sep 27, 2008
Latest reply on Oct 11, 2008 by Ian Hogg

Like • Show 0 Likes0
Comment • 6

I have a project which consists of 2 tubes (tube A and Tube b) of equal lengths. Tube B is smaller than Tube A and is placed inside Tube A concentrically. Cryogenic liquid flows through the inner tube (Tube B). The space between The outer wall of Tube B and Inner wall of Tube A is vacuum (Air at low pressure) with no flow and at room temp initially. Lids are placed at both ends and they are set to insulation material. I put a heat load (heat transfer coefficient) on the outer wall of the outer tube A.

I put boundary conditions on the inlet & outlet of the inner tube B that has cryo fluid flowing through it. And The outer wall bounder condition is set on Tube A.

The model solves nicely with no warnings BUT the vacuum air does not seem to insulate as it should. I am measuring the average surface temperature of the outer tube A as a goal. I change the air pressure between the 2 shafts but I get the same result. No matter what pressure I'm putting for the air, the outer shaft temperature is the same.

what can I do to troubleshoot this problem?

thanks in advance

No one else has this question

Outcomes

Rich Bayless Sep 29, 2008 5:04 PM

Hello,

try using 'X-Y plot' to obtain the values of the low pressure air properties, just to see if they make sense physically. Start with air pressure at ambient and then keep lowering the pressure, to see if the physical properties make sense.

You'll have to add a line through your Tube A, perhaps one on 'centerline' of the outer cylinder and another on the interface surface, and then create the x-y plot using one of the lines.

rich.

Actions

Domenic Santoianni @ Rich Bayless on Oct 4, 2008 3:25 AM

Thanks for the advice Rich. I put a vertical line extending from the outer wall of the inner tube to the inner wall of the outer tube. And I did a x-y plot , checked density, pressure of the air. Eveything is as it should be. I compared 2 simulations-- the first with air at a pressure of 1E-5 Torr and another with air at atmospheric (760 torr). Both resulted in the exact same temperature profile on the x-y plot. The low presure (vacuum) is not insulating. I'm at a loss.

I attached the file (SW 2008) for anyone kind enough and interested enough to take some time.

Attachments

SHORTY.zip122.8 KB

Actions

Ian Hogg

@ Domenic Santoianni on Oct 7, 2008 5:20 PM

Hi Domenic,

You need to be careful about the assumption of modeling a vacuum (or near vacuum) as a gas here.

Flow can calculate only a continuum. What is continuum? This can be classified by the Knudsen number (Kn) which is equal to Kn=l/L (here l is fluid molecular free length and L is typical model scale) which determines level of continuity of fluid flow . If Kn<0.01 fluid flow is continuum and such flow can be calculated with Flow. If Kn>10 such medium is called rarefied and cannot be calculated with Flow. If the Kn is inbetween, then the answer may not be correct since the fluid may not be a continuum.

Cheers,

Ian

Actions

Rich Bayless @ Domenic Santoianni on Oct 7, 2008 5:43 PM

Hello Domenic,

I ran your model, it looks well defined. The x-y plot feature allows one to select many different variables. look for the small icon at the bottom, for more parameters.

The density of air is changing due to lowering the pressure. The other properties of air are dependent on temperature. For example, from the x-y plot from your results, the conductivity of air has the same values at 1 Torr, 760 Torr, and 7600 Torr. Assuming this problem is controlled by conductivity, your results are probably reasonable.

Solution? define your own custom material properties.

Hope this helps. Rich.

Actions

Domenic Santoianni @ Rich Bayless on Oct 10, 2008 2:42 AM

Thanks Ian and Rich.

The thermal conductivity of air at 10^-5 bar is abot 3x10^-5 W/mK. so I created a new gas - cut & copied air properties and change he theral conductivity to 3x10^-5. I ran tthe model and still got no insulating affect. How close to reality would the results be if I increase the pressure to atmosphere but keep a low thermal conductivity value?

Ian's post made me realize this model isn't as simple as I though it would be. That means you can't design a dewar flask using floworks?

Actions

Ian Hogg

@ Domenic Santoianni on Oct 11, 2008 3:20 PM

Hi Domenic,

It seems strange that altering the conductivity of the air is not changing the insulation effects, even ignoring the flow continuum aspect.

As far as modeling a vaccuum flask, I'm not sure of a way in Flow to do that, since we talking about a region with no conductivity or fluid) and only radiation between surfaces (that would seem to be easier to model in COSMOSWorks). Maybe you could model it as a solid with near zero conductivity and heat capacity?

I would run this through technical support to get thme or the developers to review if this is a viable approximation or to see if there is something wrong that explains the lack of change in insulation properties.

Sorry I can't be of more help.

Ian

Actions

6 Replies

Rich Bayless Sep 29, 2008 5:04 PM
Mark Correct
Correct Answer
Hello,

try using 'X-Y plot' to obtain the values of the low pressure air properties, just to see if they make sense physically. Start with air pressure at ambient and then keep lowering the pressure, to see if the physical properties make sense.

You'll have to add a line through your Tube A, perhaps one on 'centerline' of the outer cylinder and another on the interface surface, and then create the x-y plot using one of the lines.

rich.
Like • Show 0 Likes0
Actions
- Domenic Santoianni @ Rich Bayless on Oct 4, 2008 3:25 AM
  Mark Correct
  Correct Answer
  Thanks for the advice Rich. I put a vertical line extending from the outer wall of the inner tube to the inner wall of the outer tube. And I did a x-y plot , checked density, pressure of the air. Eveything is as it should be. I compared 2 simulations-- the first with air at a pressure of 1E-5 Torr and another with air at atmospheric (760 torr). Both resulted in the exact same temperature profile on the x-y plot. The low presure (vacuum) is not insulating. I'm at a loss.
  
  I attached the file (SW 2008) for anyone kind enough and interested enough to take some time.
  Attachments
  - SHORTY.zip122.8 KB
  Like • Show 0 Likes0
  Actions
  - Ian Hogg @ Domenic Santoianni on Oct 7, 2008 5:20 PM
    Mark Correct
    Correct Answer
    Hi Domenic,
    
    You need to be careful about the assumption of modeling a vacuum (or near vacuum) as a gas here.
    
    Flow can calculate only a continuum. What is continuum? This can be classified by the Knudsen number (Kn) which is equal to Kn=l/L (here l is fluid molecular free length and L is typical model scale) which determines level of continuity of fluid flow . If Kn<0.01 fluid flow is continuum and such flow can be calculated with Flow. If Kn>10 such medium is called rarefied and cannot be calculated with Flow. If the Kn is inbetween, then the answer may not be correct since the fluid may not be a continuum.
    
    Cheers,
    
    Ian
    Like • Show 0 Likes0
    Actions
  - Rich Bayless @ Domenic Santoianni on Oct 7, 2008 5:43 PM
    Mark Correct
    Correct Answer
    Hello Domenic,
    
    I ran your model, it looks well defined. The x-y plot feature allows one to select many different variables. look for the small icon at the bottom, for more parameters.
    
    The density of air is changing due to lowering the pressure. The other properties of air are dependent on temperature. For example, from the x-y plot from your results, the conductivity of air has the same values at 1 Torr, 760 Torr, and 7600 Torr. Assuming this problem is controlled by conductivity, your results are probably reasonable.
    
    Solution? define your own custom material properties.
    
    Hope this helps. Rich.
    Like • Show 0 Likes0
    Actions
    - Domenic Santoianni @ Rich Bayless on Oct 10, 2008 2:42 AM
      Mark Correct
      Correct Answer
      Thanks Ian and Rich.
      
      The thermal conductivity of air at 10^-5 bar is abot 3x10^-5 W/mK. so I created a new gas - cut & copied air properties and change he theral conductivity to 3x10^-5. I ran tthe model and still got no insulating affect. How close to reality would the results be if I increase the pressure to atmosphere but keep a low thermal conductivity value?
      
      Ian's post made me realize this model isn't as simple as I though it would be. That means you can't design a dewar flask using floworks?
      Like • Show 0 Likes0
      Actions
      - Ian Hogg @ Domenic Santoianni on Oct 11, 2008 3:20 PM
        Mark Correct
        Correct Answer
        Hi Domenic,
        
        It seems strange that altering the conductivity of the air is not changing the insulation effects, even ignoring the flow continuum aspect.
        
        As far as modeling a vaccuum flask, I'm not sure of a way in Flow to do that, since we talking about a region with no conductivity or fluid) and only radiation between surfaces (that would seem to be easier to model in COSMOSWorks). Maybe you could model it as a solid with near zero conductivity and heat capacity?
        
        I would run this through technical support to get thme or the developers to review if this is a viable approximation or to see if there is something wrong that explains the lack of change in insulation properties.
        
        Sorry I can't be of more help.
        
        Ian
        Like • Show 0 Likes0
        Actions

Air thermal conductivity issue

Outcomes

Attachments

Related Content

Recommended Content