Convection from a braided hose

[patdh1028]

Hello all,

I'm attempting to work through a heat transfer problem and I was hoping to get some advice and clarity about how to model it.

I have a corrugated, double-braided SS tube carrying hot (500-800degC) gas at about 30 bar over a length of a few meters. Tube motion is minimal, so it is just sitting in standard conditions. The purpose of the calculation is to estimate temperature drop in the gas and figure out how much insulation we need to minimize it.

So I have a few heat transfer conditions. From the gas to the tube is forced convection, then conduction through the tube, then free convection to surrounding air from the outer tube surface.

The way I was thinking of approaching this was taking a small length of hose, assuming that its surface temp was equal to gas temp, and determining what the rate of heat transfer to surroundings would be, removing that energy from the gas, and then iterating that method for the full length (i.e. the temperature of the gas at the outlet of my first small length of pipe would be the new assumed surface temp of the next small length of pipe).

The thing I'm not sure about is where free vs. forced convection comes in. Does the fact that convection is free on the outside of the tube mean that the total heat transfer is limited to that rate, or does the turbulent, high temp. flow inside the tube mean that the rate of heat transfer to the surroundings is higher?

 

[racookpe1978]

Your real problem will not be calculating the thickness of insulation needed, but in finding a suitable insulation "blanket wrap" or style - and the protective cover for that insulation to shield it from the outside environment of snow, rain, ice, water, humidity and chemicals and UV light - that will also "flex" with the flex hose as it moves and is disconnected.

 

[patdh1028]

The hose won't be outside; my apologies for not being clear. It will be inside and should not be particularly close to any corrosive materials or bad weather. By standard conditions I meant 1 atm, room temperature.

 

[patdh1028]

Rather than iterate, which would require punching all my hand calcs into excel (the only math software I have) and working out how that would go (time consuming, error-prone), I did the analysis for a 1-cm section of the tube accounting for radiation, which gave me a T-in (800C) and a T-out(797.3C).

Then I fit an exponential decay to those two points with the variable being the length in cm. Can anyone comment on how close that would be to actual?

 

[StoneCold]

patdh1028
you can simulate this as a pipe in 3E Plus. You can download that for free.
It says for personell protection you need 5.5" of high temperature blanket to get the surface temp down to 140F.


Regards
StoneCold

 

[chicopee]

I have no experience with heat transfer thru braided hoses but in my area of Springfield, MA there is a company named Titeflex that specializes in the braided hose technology and these products are used in aerospace products, transporation and automobile industry so you could try to get answers from these people.

 

[patdh1028]

Thanks for pointing me to 3E Plus, it looks very handy. I'm not really concerned with getting the outer temp down to personell protection levels; the hose is going to be in a sealed robotic cell. If some personell get in there they'll have more to worry about than the hot hose.

 

[racookpe1978]

I'm surprised you're concerned about insulating against a <2 degree temp drop ... Will the heat buildup in the cell be the problem?

 

[patdh1028]

It's two degrees per cm. I assumed that the hose surface temp was equal to gas temperature (since its an SS hose, I don't think that is too outrageous. I then calculated free convection and radiation from that 1-cm section to get the heat energy flow out, then subtracted that thermal energy from the gas thermal energy exiting the 1-cm section and solved for temperature there.

The 3E software does what I was trying to do anyway, which is figure out insulation thicknesses, so my own HT equations don't have to be fantastically accurate.