Free convection from a cylinder.

[Elevated]

I am running a free Convection heat transfer experiment in which I am trying to determine the heat transfer coefficient of a long horizontal cylinder. The experiment requires that I find this coefficient by means of measuring the surface temperature and using that along with the ambient temperature and dimensions of the cylinder to determine the heat transfer coefficient. I am currently using a 120W Cartridge heater with 5/8" diameter and 6" long and varying the wattage I put into the heater.
When I calculate the heat transfer coefficient from Newtons law of cooling Q=A*h*(Ts-To) I get one value and when I use the correlations for heat transfer of a long horizontal cylinder I get a significantly lower number. I have tried this at different wattages and have come up with the same results of between 40%-55% error between the two heat transfer coefficients. Can anyone help me with what I might be doing wrong or why this error is occurring?

 

[IRstuff]

How hot are you running the surface temperature? Note that Newton's law never considered radiated emissions. Usually, when you include radiation, the effective heat transfer coefficient is between 150% and 200% of the calculated convective coefficient.

Note, also, a single surface temperature measurement is probably inaccurate, since the surface temperature is expected to vary depending on where on the cylinder you measure it.

TTFN

FAQ731-376

 

[sailoday28]

How are you measuring surface temperature? If the surface measuring device(s) are fixed, can you rotate the cylinder?

 

[dvd]

You're within an order of magnitude. You aren't taking radiation into account, nor are you mentioning conduction with whatever is holding the cylinder. What properties are you using for your air in the calculations? Is it dry air? What is the relative humidity in your testing environment? Are you sure that you've reached steady-state?

 

[vpl]

Is this a school project? If so, your advisor or other students would be a good place to start.

Patricia Lougheed

******

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[Dave41A]

Free convection correlations are for "quiescent" conditions, which are likely hard to achieve in your (apparently) simple experimental setup. The real coefficient will typically be higher, due to mixed free/forced effects. Like the others mentioned, add radiation to the mix as well.

Dave

 

[Elevated]

After reviewing the replies I ran another test. I have set up the test again putting 10 Watts into the heater. I am covering the heater in a tunnel like lexan square so there is no forced convection from air movement. I let the cylinder heat up for 30 min to ensure that it has reached steady state and took 10 surface temperature measurements and averaged them. These are the values I am using and the results.
Tsurface=376K
Tambient=297K
Tfilm=336K
B=.0030 (1/K)
v=19.51*10^-6 (m^2/s)
alpha=27.83*10^-6(m^2/s)
k=28.96*10^-3 (W/m*K)
emissivity=.17 ( stainless steel)
Rayleigh number =17075.7
Nu=5.49
h=9.99
Qconv=6.00793 W
Qrad=.9 W
Q=Qconv+Qrad=6.9 W
10-6.9/10=.31*100=31% error

So with the inclusion of radiation I am still 31% off from theory to measured values. I can assume that this 31% is not loss in conduction as the plate holding the cylinder only went up 4 degrees. If the loss is due to radiation the emissivity of the stainless steel would have to be around .76. Does anyone have any other reason these values do not match?

 

[dvd]

You have a lexan heat shield that does not allow all of the radiation from the cylinder to emit freely. You could study up on heat shields in a radiation book like:

http://books.google.com/books?id=O389yQ0-fecC&pg=PA1&dq=Thermal+radiation#v=onepage&q&f=false

When you average your surface temperature, what is the range? does it lead you to think that the surface temperature is uniform? Would you expect some variations in radiation (T^4) with your surface temperature variations? How did you determine the .17 emissivity for stainless? I can find a lot of higher values with a 30 second internet search.

Also, how are you taking surface meassurements? With an infrared thermometer? Are you measuring through the lexan? What is the emissivity of lexan? What is the emissivity of stainless steel? What is the emissivity setting for the instrument? blah, blah, blah

 

[IRstuff]

Energy MUST be conserved. If you put 10W in, then you should get 10W out. That means that either your htc is too low, or your temperatures are off, or your emissivities are in error.

However, this is also part of the "correlation" process. You try to account for all the non-idealities, and what's left over is "correlated" to the actual results.

TTFN

FAQ731-376

 

[ione]

For polished stainless steel the emissivity you have used (0.17) could be ok, but if the surface is oxidized the value is by far greater than that.