14 replies to this topic

[hamidun]

Hi all,

 

I just want to ask if anybody know, how to find heat transfer coefficient of water in a pipe/tank at resting condition or flowing.

 

Thanks for any help!!

[PingPong]

Note that the overall heat transfer coefficient U not only depends on the partial heat transfer coefficient from water to wall, but also on heat resistance of wall, and partial heat transfer coefficient of wall to the medium on other side. If present, the heat resistance of any insulation also has a huge impact on value of U.

 

You have to define your problem in much more detail than just one sentence.

[hamidun]

PingPong,

 

I'm calculating heat transfer in a 10" OD PVC-pipe (k = 0.19 W/m.K, t = 15 mm, l = 3 m) with 500 m^3/hr water (40 deg C) flowing through the pipe (35 deg C ambient temperature). I'm thinking to calculate it by either forced convection or steady state loss (thermal circuit concept). For steady state loss, I need the heat transfer coefficient from the fluid to the inner wall and from outer wall to ambient. I found a website that provide heat transfer coefficient for water, but it doesn't seem to be clear, because it only gives the range value and doesn't specify for the temperature, condition, etc. What I am trying to do is to find the heat transfer coefficient (h) of water based on condition, properties given. 

 

Anyone could help with this? 

 

Thanks!!

[PingPong]

Every book on Heat Transfer gives formulas to calculate h for flowing fluids inside pipes.

h for flowing water inside a pipe will be very high, in the order of 1000 or more W/m².K.

 

What matters however is not h but the overall heat transfer coefficient U, which in your case will be mainly determined by the PVC pipewall and the transfer of heat from the outer wall to the ambient air by natural convection (if no wind) and radiation. That U will be very small, I suspect in the order of 5 - 10 W/m².K (depending on wind speed). If there is a lot of sun radiation it gets more complicated and the pipe may even gain heat during parts of the day instead of lose heat.

 

As the length of your pipe is only 3 meters, I wonder why you would want to go through the trouble of trying to calculate a tiny heat loss (or gain), also in view of the fact that you seem to have no knowledge on the subject.

 

Anyway: if you like, you can learn about the subject by reading a book on Heat Transfer, like the one you can freely download from the MIT website: http://web.mit.edu/l...load-ahtt.shtml

Edited by PingPong, 11 July 2014 - 04:06 AM.

[hamidun]

PingPong,

 

What about the fluid is resting? How to find h, when the fluid is resting. Because based on my calculation that I've done using internal forced convection, the thermal resistance from PVC is quite dominant, so that it holds the heat transfer to outside pipe.

 

What is actually the difference between overall heat transfer coefficient, U and heat transfer coefficient, h? Isn't is the same? U is just the result of 1/sum of h, right? At the end of the day, we have to find h first, in order to get U, am I right? What I don't know is how to get h of water when water is resting in a pipe/tank based on type the conditions applied in the system (such as, flow properties, fluid properties, temperature gradient). Is there such way to calculate h as I wanted?

 

That's right, I don't have knowledge of this, that's why I find for help here. 

 

Thanks!!

[PingPong]

U is not the same as h.

 

If, for simplicity, we assume that the pipe inner surface A_i is equal to its outer surface A_o then the formula for U simplifies to:

 

U = 1 / ( 1/h_i + x/k + 1/h_o )

 

in which

h_i = heat transfer coefficient (filmcoefficient) for medium inside tube to tubewall, W/m².K

h_o = heat transfer coefficient (filmcoefficient) for medium outside tube to tubewall, W/m².K

x = thickness of tube wall = 0.015 m

k = thermal conductivity of tubewall = 0.19 W/m.K for PVC

 

h_i for streaming water is > 1000 W/m².K, for resting water it is > 100 W/m².K

h_o for resting ambient air is in the order of 5 to 10 W/m².K including radiation heat loss

 

This gives a U = ~ 5 W/m².K irrespective whether the water is streaming or resting. If there is wind U may be about 10 W/m².K

 

It should be obvious from the U formula that even if the water streams very fast and the wind is a storm the value of U can never be higher than 13 W/m².K due to the heat resistance of the PVC pipe wall (0.19 W/m.K / 0.015 m = 13 W/m².K).

 

The above values for hi and ho come out of my (experienced) thumb. You can try to calculate them more accurately using the formulas in the MIT book that I mentioned before, but the result will be roughly the same:

- the water streaming velocity has no impact on U because its h is always very many times higher than 13 W/m².K

- the expected value of U is in the order of 5 - 10 W/m².K depending on the wind speed.

[hamidun]

PingPong,

 

Thanks for your response, it's very helpful.

 

So, for hi  and ho, we just assume roughly? This is actually what I want to know, if there is such way to calculate these coefficients based on conditions applied?

 

Ok, I think, I understand a bit about this. If let say, we have case 1 which is the same with what you have made, but with high air velocity, say 10 m/s. Will that help?

 

Or case 2, let say we have good material pipe, but with low air velocity, say 2 m/s. Will this be much better?

 

Again, thank you!!

[PingPong]

So, for hi  and ho, we just assume roughly? This is actually what I want to know, if there is such way to calculate these coefficients based on conditions applied?

We do not assume something, only I assume something because I know from experience what h_i and h_o will roughly be, and that in this case with a PVC pipe wall the value of h_i and h_o hardly matter.

 

Air velocity will have impact on h_o and therefor some impact on U.

You can calculate impact of air velocity with the formulas for natural convection and forced convection from the MIT book. Give it a try.

 

Water velocity has an impact on h_i but that has absolutely no impact on U in this case with PVC pipe because 1/h_i is very very small compared to x/k and 1/h_o

Also in case of a steel pipe water velocity will have hardly any impact on U, as U will then be determined mainly by the fairly low h_o of the air.

 

Heat transfer is not an easy subject. It is often underestimated by young engineers. If you want to learn you have to do effort and study.

[hamidun]

PingPong,

 

This is getting complicated. I tried to find natural convection and external forced convection of the pipe, but then both of them need Ts of the pipe which is outside wall temperature of the pipe. How actually to find this? 

 

I tried to use steady heat loss, to find the surface temperature, but then again the same problem appeared. To get the total resistance, we need to have Ri = 1/hi Ai (where hi is fluid heat transfer coefficient), I got stuck finding this hi. I have been referring to yunus cengel textbook, but in that book, there is no formula to calculate that hi, instead it only has some range of typical values of heat transfer coefficient for gases and liquids. Is it possible to use that?

 

Or my method of finding this surface temperature is not right? If it's right, the only problem is to find the hi.

 

You are right, heat transfer is never easy. 

[PingPong]

Ts_o (and Ts_i) can only be found by trial-and-error:

Assume a value for Ts_o of say 37 ^oC (and Ts_i of say 39.5 ^oC), calculate R_i = 1/h_i and Rw = x/k and R_o = 1/h_o

Temperature profile from T_i via Ts_i and Ts_o to T_o follows profile of R's as you can find in any heat transfer book.

That gives you new value for Ts_o (and Ts_i). And based on those you calculate again R_o and R_i

et cetera.

 

I have been referring to yunus cengel textbook, but in that book, there is no formula to calculate that hi

I am sure there are formulas for h_i in that book. Look again.

They are also in the MIT book that I mentioned before.

Edited by PingPong, 17 July 2014 - 08:44 AM.

[hamidun]

Why it can only be found by trial and error? I think if we can get all resistance in the pipe and know Ti and Tambient, we can get Ts_i and Ts_{o ,} by using steady heat loss/thermal circuit concept, is that right?

 

 

I am sure there are formulas for h_i in that book. Look again.

They are also in the MIT book that I mentioned before.

 

Can you just let me know, what chapter should I read to know how to calculate hi? Because I have been trying to find it, but I couldn't find it, in most examples of chapter 3 in Yunus Cengel textbook, it always has value of hi, therefore, they don't calculate hi.

[PingPong]

Why it can only be found by trial and error?

For natural convection of water and ambient air Ts_i and Ts_o can only be found by trial and error, because h_i and h_o are a function of Ts_i and Ts_o

 

In case of forced convection of water and air no trial-and-error is required.

 

 

Cengel chapter 9 is about natural convection to calculate h_i and h_o

Cengel chapter 8 is about forced convection.inside tubes to calculate h_i

Cengel chapter 7 is about forced convection.outside tubes to calculate h_o

[hamidun]

 

 

For natural convection of water and ambient air Ts_i and Ts_o can only be found by trial and error, because h_i and h_o are a function of Ts_i and Ts_o

 

But I think if we use steady heat loss, we can get each temperature on pipe's wall, is this right? Ts_i is inner wall temperature and Ts_o outer wall temperature, right?

 

 

 

Cengel chapter 8 is about forced convection.inside tubes to calculate h_i

 

I am looking formula to find h_i when water inside pipe is resting. This chapter, I think for flowing water inside the pipe.

[PingPong]

But I think if we use steady heat loss, we can get each temperature on pipe's wall, is this right?

No, not for natural convection in resting fluids.

Ts_i is inner wall temperature and Ts_o outer wall temperature, right?

Yes.

I am looking formula to find h_i when water inside pipe is resting. This chapter, I think for flowing water inside the pipe.

Cengel chapter 9 is about natural convection to calculate h_i and h_o

[hamidun]

 

Quote

But I think if we use steady heat loss, we can get each temperature on pipe's wall, is this right?

No, not for natural convection in resting fluids.

I don't get what you mean here.

 

 

 

 

Cengel chapter 9 is about natural convection to calculate h_i and h_o

I don't know why you said there is where I couldn't find how to calculate h_i.