A copper coil is immersed in a water bath. The water bath provides heat to heat the nitrogen gas when it passes through the copper coil. Nitrogen gas at room temperature is passed through the copper coil at a set flow rate. The heated nitrogen gas is then supplied to a drying vessel. I want to find the temperature of the nitrogen gas when it emerged from the outlet of the copper coil.

I used the heat conduction formula to calculate it but it is not the right method. $q = kA (T_1-T_2)/d$ where $q$ is the rate of heat transferred, $A$ is the area of the surface, $k$ is the thermal conductivity, $d$ is the thickness and $(T_1-T_2)$ is the temperature difference. But since the temperature in the copper coil is not constant because the temperature of the nitrogen gas will increase while it flows through the copper coil as it gains more heat, this approach is wrong. Is there a better approach to find the temperature of the nitrogen gas when it emerged from the copper coil?

  • $\begingroup$ Is the water bath mechanically agitated or just allowed to naturally convect? Does anything regulate the average temperature of the water? $\endgroup$ Commented Apr 13, 2019 at 10:52
  • 1
    $\begingroup$ @DaveInCaz the water bath is automatically regulated by a heat exchanger that is set to a certain set-point. $\endgroup$ Commented Apr 13, 2019 at 12:26

1 Answer 1


Nitrogen heating

Firstly, some assumptions:

  1. water temperature $T_{\infty}$ is uniform and constant,
  2. nitrogen gas flow is highly turbulent, plug flow. Because nitrogen viscosity is very low, that is a reasonable assumption,
  3. nitrogen gas is heated by convection only. Longitunal conduction is negligle. This is a reasonable assumption because nitrogen's thermal conductivity is low,
  4. thermal expansion of nitrogen gas is considered neglibible.

Now look at the heat balance of an element $\mathbf{d}x$, mass $\mathbf{d}m$: it receives a heat flux $\frac{\mathbf{d}q}{\mathbf{d}t}$, in accordance with Newton's cooling/heating Law:

$$\frac{\mathbf{d}q}{\mathbf{d}t}=2 \pi r_o R\mathbf{d}x[T_{\infty}-T(x)]\tag{1}$$

Where $R$ is the thermal resistance, I'll come back to it extensively below. $r_o$ is the outside radius of the pipe.

Now we also know that:


Dividing both sides by $\mathbf{d}t$ and because $\frac{\mathbf{d}m}{\mathbf{d}t}=\dot{m}$, we get with $(1)$: $$c_p\dot{m}\mathbf{d}T(x)=2 \pi r_o R\mathbf{d}x{R}[T_{\infty}-T(x)]\tag{2}$$

where $c_p$ is the constant pressure heat capacity of the nitrogen gas and $\dot{m}$ its mass flow through the pipe.

Now for ease of reading, set:

$$\alpha=\frac{2\pi r_o R}{c_p\dot{m}}$$

And with $(2)$, rearrange slightly:

$$\frac{\mathbf{d}T(x)}{T_{\infty}-T(x)}=\alpha \mathbf{d}x\tag{3}$$

Integrating $(3)$ between $(0,L)$ and $(T_0,T(L))$ we get:

$$-\ln\Big[\frac{T_{\infty}-T(L)}{T_{\infty}-T_0}\Big]=\alpha L$$

Slight reworking then yields:

$$\boxed{T(L)=T_{\infty}-(T_{\infty}-T_0)e^{-\alpha L}}\tag{4}$$

Note that for $L \to +\infty$, $T(L) \to T_{\infty}$.

Thermal Resistance:

A full treatise of the concept of themal resistance is outside the scope of this answer. May I suggest you read up on it here.

In our pipe problem we have three components to the thermal resistance:

  1. outer convection zone: water to pipe ($R_1$),
  2. conduction zone: heat conducting through circular pipe wall ($R_2$),
  3. inner convection zone: pipe to nitrogen gas ($R_3$).

The overall thermal resistance $R$ is obtained from:



  1. $R_1=h_o$, the outer (convection) heat transfer coefficient,
  2. Conduction zone:


where the $r$ are outer and inner radii of the pipe and $k_{Cu}$ the thermal conductivity of copper.

  1. $R_3=h_i$, the inner (convection) heat transfer coefficient.
  • $\begingroup$ thanks for your help. From the thermal resistance equation which you provided, it implies that the system is in parallel? Is it not in series? I have looked up ways to determine the heat transfer coefficient but I need the values of temperature for that, which I have to calculate first. Is there a better way to determine R1 and R3? $\endgroup$ Commented Apr 13, 2019 at 19:15
  • $\begingroup$ Hi. The resistances are 'in series' but they are calculated as above nonetheless. See also this bit of theory (on a composite wall): fc.civil.tamu.edu/resources/en/engr214/chapter7/node13.html. If your $R_1$ and $R_3$ are temperature dependent, you need to apply an iteration process. $\endgroup$
    – Gert
    Commented Apr 13, 2019 at 23:06
  • $\begingroup$ One simplification is to assume the inner wall temperature is constant and equal to the water bath temperature. The calculation is a little easier than. $\endgroup$
    – Gert
    Commented Apr 13, 2019 at 23:08
  • $\begingroup$ I wrote this equation, equating the amount of heat needed to raise the temperature of the gas passing through length $dx$ of tube per second by $d\theta$ to the heat coming in through a length $dx$ of tube per second: $$\rho v \pi r^2 c\ d\theta=\kappa\ 2\pi r\ dx \frac{(\theta_B-\theta)}d.$$ Here $\theta_B$ is the bath temperature. The equation is essentially the same as Gert's and integrates up to give the same result. $\endgroup$ Commented Jul 23, 2022 at 23:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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