Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Insulation of steel beams - heat transfer

  1. Oct 6, 2014 #1
    I'm working on a problem involving steel structural members on the exterior of a building. I need to figure out how much insulation (for summer conditions) and possible heat tracing (for winter conditions) would be required to maintain minimum and maximum temperatures. I know I need to use the equations for conductive and convective heat transfer, but I'm not sure how to go about it.

    Any pointers would be much appreciated.

    Thank you,
  2. jcsd
  3. Oct 7, 2014 #2
    To further explain this I have figured out the heat transfer to the beams (for summer design conditions) and from the beam (for winter design conditions) based on the allowed temperature ranges (52°F-92°F) with and without insulation (I have not factored in the additional heat from the heat trace for winter design yet). This gives me btu/h, but what does this tell me? How do I know if I need more or better insulation? Design conditions are 99°F in the summer and 9°F in the winter. I am using the equation for conductive heat transfer (Qdot=kAdT/x) though I know the outer layer of insulation would experience radiative heat transfer. I am using conduction for simplicity for now.

    Thanks again,
  4. Oct 7, 2014 #3

    You can only have one layer of insulation and it must be used for both seasons. Figure the maximum amount of insulation needed to keep the beams within design conditions during the ambient temperatures for each season. Whichever is greater will be the amount for used both seasons.

    It seems to me that steel beams would gain or lose heat through radiation or convection. I would think that the only things that the beams would be in contact with, to lose or gain through conduction, are the points where they attach to other structure. So far you haven't mentioned the R value of the insulation.

    Think about winter loss by radiation. And summer gain by convection. We won't even go into sunlight I presume.
    Last edited: Oct 7, 2014
  5. Oct 7, 2014 #4
    I was considering conduction through the layers of insulation. There will likely be a polystyrene insulation with an aluminum type duct wrap on the outside. You are right about the radiation and convection. I feel like I am making this way more complicated than it should be already. Does the sun's radiation apply to only surfaces facing the sun, or the entire exposed surface?
  6. Oct 7, 2014 #5
    Sunlight will primarily affect those objects that it strikes, otherwise it only heats ambient air and you already have that temperature ( I may have made things more difficult than need be by mentioning it).

    Basically I don't think that it would be possible or practical to determine the conduction through foam insulation. It is possible but it is difficult because the rate of conduction is constantly changing with the thickness of the insulation. Save that for when you design your first nuclear reactor. .)

    Here is an aid to finding conduction through various materials.


    From the information that you have given I really think that just determining your heat loss or gain and then matching it up with insulation of the proper R value is all you need.
  7. Oct 9, 2014 #6
    You're saying I should focus on convection (with the hot or cold air around the beams) and radiation (from the sun). I looked into calculating the sun radiation and it gets very complicated when you consider the different angles throughout the day. Is there a simplified approach?

    Thank you for all the help thus far.
  8. Oct 9, 2014 #7
    In the convection equation (or any of the heat transfer equations) q/A=h(Tf-Ts) I don't know what the heat transfer will be or what the surface temperature of the beam will be because they are both based on time. Is there another useful equation that I am missing?
  9. Oct 9, 2014 #8
    The sun will only be a factor if the beams are directly exposed to sunlight. If this is a theoretical project the sun probably will not be a factor. If the sun is involved you may have already solved the problem with a reflective "aluminum type duct wrap."
    Last edited: Oct 9, 2014
  10. Oct 9, 2014 #9
    Let me step back a second, Will the beams be insulated from the main structure or in contact with the structure.?

    If there is contact with the main structure there will be conduction involved where the steel beams touch the other structure. This will be heat conducted to the structure in the summer which is why the insulation is needed. In the winter there will be conduction from the structure to the beams. If the insulation is not sufficient in the winter to keep the heat loss less than the gain from the structure then a heat trace is used on the beams to keep the temperature within specs. ( I think I just made this more complicated .)

    You can simplify the time factor in several ways. The temperature of the buildings internal structure is a constant.

    Could you just figure the amount of insulation needed to prevent a 7 degree temperature rise in the summer. [A bit more would be better] For the winter determine enough insulation to prevent a loss to keep within specs, plus a trace if needed.

    Without knowing the starting temperature of the beams at any given time I think that determining the R value of the insulation is the easiest way to go.
    Last edited: Oct 9, 2014
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook