4 replies to this topic

[process.eng]

Regarding design calculations for the heat transfer process in a heat exchanger,

Do we need to take average values for each physical property such as heat capacity, density, thermal conductivity and viscosity ?


For instance,if the heating utility/hot stream, high pressure steam is entering as saturated vapour and leaving as a saturated liquid (condensate) there would be a highly significant difference in its inlet and outlet density, heat capacity, thermal conductivity and viscosity.


As per my understanding we need to take an average of these values for each of these properties.


Please correct me if I'm mistaken; a little confusion has risen in my mind as a friend was telling me that we only take average values for certain physical properties

Edited by studentsphere, 19 April 2010 - 05:41 AM.

[Andres Valencia Michaud]

Regarding design calculations for the heat transfer process in a heat exchanger,

Do we need to take average values for each physical property such as heat capacity, density, thermal conductivity and viscosity ?


For instance,if the heating utility/hot stream, high pressure steam is entering as saturated vapour and leaving as a saturated liquid (condensate) there would be a highly significant difference in its inlet and outlet density, heat capacity, thermal conductivity and viscosity.


As per my understanding we need to take an average of these values for each of these properties.


Please correct me if I'm mistaken; a little confusion has risen in my mind as a friend was telling me that we only take average values for certain physical properties

Well, Donald Kern in his book "Process Heat Transfer" described the solution for your problem in chapters 13 and 14.

[kkala]

Regarding design calculations for the heat transfer process in a heat exchanger, Do we need to take average values for each physical property such as heat capacity, density, thermal conductivity and viscosity ?
For instance,if the heating utility/hot stream, high pressure steam is entering as saturated vapour and leaving as a saturated liquid (condensate) there would be a highly significant difference in its inlet and outlet density, heat capacity, thermal conductivity and viscosity.
As per my understanding we need to take an average of these values for each of these properties.

Looking at McCabe-Smith-Harriot, we see in the Chapter of "Heat transfer to fluids without phase change, Average value of hi in turbulent flow : unless hi varies over the length more than about 2:1, an average value of hi is calculated and used...computed by evaluating the fluid properties Cp, K, and μ at the average fluid temperature, defined as the arithmetic mean between the inlet and outlet temperatures.
Note: This is mainly for hand calculation. Computer software usually does not do it, but divides the whole exchanger into many successive parts, assigning values of each physical property (average?) to each part to obtain more precise results.
So average values can be used, on the condition that fluids do not change phase. If they do, situation / calculation get more complex (critical temperature drop, dropwise or film condensation, physical properties of both vapor and condensate used) and temperature of two phase fluid more "constant".
specifically for steam: At the beginning (in a small part of exchanger) superheated steam approaches saturation point. Then there is saturated steam and condensate till their separation in the steam trap. Since condensing steam has very little resistance to heat transfer, we usually assume a value for steam side (e.g. 11 kW/m2oC) without calculating corresponding hi.This "error" is usually well covered by the uncertainty of hi of the other fluid.

Edited by kkala, 19 April 2010 - 04:14 PM.

[Zauberberg]

How do you imagine to use "averaged" properties for heat transfer applications where the phase change is involved? If you really look at it from a common sense point of view, you'll see that it's impossible and illogical.

As Kostas said, this approach applies only for those applications not involving phase change, whether it is partial or complete.

Also to emphasize, "averaging" fluid properties should be considered only in those applications where there are no drastic changes of fluid's transport and thermo properties with temperature (viscosity in particular). If you have something like e.g. cooling vaccum residue or slurry, I would be extremely reluctant to use average values between the inlet and outlet temperature.

[fatimah]

Hi there

This is my findings for 2 cases of HTRI design with one phase fluid only. Yes, for specific heat, density, thermal conductivity, the graph shows straight line, while for viscosity, a curve generated but nearly approach a straight line.