November 6th, 2019, 12:12 PM  #1 
Newbie Joined: Nov 2019 From: UK Posts: 1 Thanks: 0  Thermal Conductivity
Hello All Currently trying to get through the question below, but any formula I find is too complex for the brief detail given, with regard thermal conductivity of gases. It is proposed to use a katharometer to measure the amount (about 5%) of oxygen in nitrogen, in the presence of a small amount (0.5%) hydrogen. How constant would the proportion of hydrogen have to be in order to limit errors in measurement of % oxygen to + 0.1%? The thermal conductivities are: nitrogen 0.993 oxygen 1.052 hydrogen 6.993 
November 6th, 2019, 06:15 PM  #2 
Senior Member Joined: Jun 2019 From: USA Posts: 386 Thanks: 211 
Question 1. Are you using a specific model for the thermal conductivity of a mixture of gases? If we were looking for specific heats, energy, or similar properties of a mixture of ideal gases, we would use a massweighted average. In this type of model, the thermal conductivity would look something like this: $\displaystyle k_{mixture} = [1f_{O2}f_{H2}] k_{N2} + f_{O2} k_{O2} + f_{H2} k_{H2}$, where $f_i$ (or $mf_i$, or $c_i$) is the mass fraction of gas $i$. However, I also know of literature that claims that kinetic theory does not predict the thermal conductivity of gas mixtures well, and slightly more complex empirical formulas have been proposed. Either way, you are going to have some equation $k_{mixture} = k(f_{N_2},f_{O_2},f_{H_2})$. Since no information was given about the accuracy of the katharometer or the Wheatstone bridge, I assume the question is about keeping the conductivity of the mixture sufficiently constant. If $f_{H2} = 0.005 \pm u_{fH2}$, then barring any other sources of uncertainty, $\displaystyle k_{mixture} = k(0.945,0.05,0.005) \pm \frac{\partial k_{mixture}}{\partial f_{H2}} u_{fH2}$. Whatever equation you use for the thermal conductivity of the mixture, take the partial derivative with respect to the hydrogen mass fraction and evaluate at the nominal conditions. Then, $\displaystyle u_k = \frac{\partial k_{mixture}}{\partial f_{H2}} u_{fH2}$. But then, you are going to measure (indirectly) the thermal conductivity, and back out the mass fraction of the oxygen, and it is this number you want to keep within ±0.1 %. $\displaystyle u_{fO2} = \frac{\partial f_{O2}}{\partial k_{mixture}} u_k$. So, evaluate both partial derivatives at the nominal conditions, and solve, $\displaystyle \left \pm 0.001 \right \geq \left \frac{\partial f_{O2}}{\partial k_{mixture}} \frac{\partial k_{mixture}}{\partial f_{H2}} u_{fH2} \right$ for the maximum allowable uncertainty $u_{fH2}$. If your question was specifically about how to calculate the thermal conductivity of the mixture, then I can guide you to some literature, but I am not an expert and could not guarantee the reliability of one model over another. 

Tags 
conductivity, thermal 
Thread Tools  
Display Modes  

Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Thermal Expansion  Grntxavier  Physics  1  October 16th, 2017 03:48 AM 
Another thermal expansion question  hyperbola  Physics  4  September 16th, 2015 03:02 AM 
Thermal expansion question  hyperbola  Physics  0  September 12th, 2015 09:47 AM 
Thermal Equilibrium  rnck  Physics  3  June 15th, 2011 07:55 AM 
Define thermal conductivity . Also on what factor 'K' depend  rsoy  Physics  0  March 17th, 2011 06:22 AM 