0
$\begingroup$

I have a question, which seems simple, but for the life of me I cannot figure out. So, let's say I have a very thin cylinder (consider an interface), between two fluid with different temperatures. The fluid inside the pipe is flowing with a mass rate of $ \dot{m}=\rho U S $, where $ \rho $ is the density of the fluid, $ U $ is the velocity and $ S $ the section of the cylinder.

Naively I could say that the rate of heat transfer, $Q$, between the two fluid is $ Q=hA(T_o-T_i) $, where $h$ is the heat transfer coefficient, $A$ the "contact area between the two fluid" (or the lateral surface of the cylinder), and $ T_o, T_i $ are respectively, the temperature outside and inside the cylinder.

Since I am interested in the heat added per unit mass, naively (again) I would define it as $$q=\frac{Q}{\dot{m}}=\frac{hA(T_o-T_i)}{\rho U S }=\frac{2L}{r}\frac{h(T_o-T_i)}{\rho U}$$

where $L$ is the cylinder length and $r$ its radius.
From this equation one could conclude that, if we increase the cylinder radius, the heat per unit mass will decrease, even if $Q$ is proportional to $r$.
If $r$ increases, the volume of the cylinder increases, so even if we add more heat, it will be "distributed" on a larger volume.
But this reasoning breaks down if we increase $L$. In this case, we have both $q$ and $Q$ are directly proportional to the length. Even if, we should also see the effect of the volume increment if we increment $L$.

Have I been too naive in the use of the equations? Or is there something that I haven't caught ?

$\endgroup$

1 Answer 1

0
$\begingroup$

Heat transfer coefficient is used for heat transfer between fluid and the wall surface, not conduction through the wall. For a steady state, you would need 2 heat transfer coefficients $h_i$, $h_e$ and also wall conductivity $\lambda$. $$\dot{Q} = \left(T_{fi}-T_{wi}\right)\cdot A_i\cdot h_i = \left(T_{wi}-T_{we}\right)\cdot \frac{2\pi \lambda L}{\ln\left(\frac{r_e}{r_i}\right)} = \left(T_{we}-T_{fe}\right)\cdot A_e\cdot h_e$$ Increasing length is a little bit different than increasing radius, because fluid temperature will probably change along the length (in some situations this can be negligible). For inlet and outlet fluid temperatures $T_{fi0}$ and $T_{fi1}$ and thermal capacity $cp$: $$\dot{Q} = \left(T_{fi0}-T_{fi1}\right)\cdot \dot{m}\cdot cp$$

I am not sure what the question is, but differences in sensitivities to $r$ and $L$ are easily explained by $r$ influencing heat transfer area $\propto r$ but also cross-section area $\propto r^2$, so the net effect for $q\propto \frac{1}{r}$. On the other hand, $L$ influences just the heat transfer area, so $q\propto L$.


Edit: More accurate dependence on length

If you are interested in dependence on pipe length, it is better to use a more precise model, where a very short section of pipe with length $dx$ transfers small amount of heat $d\dot{Q}$ into the environment with constant temperature $T_e$ ($U$ is overall heat transfer coefficient per unit length of the pipe), which also changes temperature of the fluid in a pipe by $dT$:

$$d\dot{Q} = \left(T(x)-T_e\right)\cdot U = -dT\cdot \dot{m}\cdot cp$$

This is a simple differential equation, which you can directly integrate:

$$\int\limits_{T_{i0}}^{T(x)}\frac{1}{T(x)-T_e} dx = -\frac{U}{\dot{m}\cdot cp}\cdot \int\limits_0^x dx$$

This leads to a internal fluid temperature profile along the pipe $T(x)$:

$$T(x) = T_e+\left(T_{i0}-T_e\right)\cdot \exp\left(-\frac{U}{\dot{m}\cdot cp}\cdot x\right)$$

Here is an example: enter image description here

$\endgroup$
1
  • $\begingroup$ Thanks for the comment! My question is related to physical intuition. We have a temperature difference between the pipe and the fluid. Consider a very long pipe, infinite if you like. From the formula seems that the heat added by unit mass would be infinite. But since $q$ is a quantity "by unit mass" I would have expected not to have this strong dependence on $L$. That is why I was asking if the equation that I derived is correct. $\endgroup$
    – GMV871
    Sep 5, 2022 at 7:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.