Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Simulation analysis of temperature field
Posted Mar 26, 2022, 9:01 a.m. EDT Electromagnetic Heating, Heat Transfer Version 5.6 1 Reply
Please login with a confirmed email address before reporting spam
hi，i am a COMSOL greener The carbon nanofibers I'm working with are electrically heated, similar to incandescent lamps, where the temperature changes are transient warming-thermal Dynamic equilibrium-rapid cooling after power failure.
I USE COMSOL software to simulate a series of steady-state temperature with the applied voltage, resistance data, to fit a T (u, R) three-dimensional surface graph, but according to the equation of (u ^ 2)/R = A * t + b * t ^ 4 + C, the goodness of fit is very low, only about 0.4, the fitting effect is not good
(U^2)/R=AT+BT^4+C
On the left side of the equals sign is the joule heat production power expression
On the right side of the equals sign are the expressions for convective and radiative heat dissipation
At steady-state temperatures, there is a heat production-cooling equilibrium, so I wanted to use this formula to fit, but it didn't fit very well
Now there are two cases, one is my COMSOL model built wrong, resulting in simulation of the temperature data are all wrong, and can not be fitted
Another is the above-mentioned physical model expression error, the lack of some items, that is, the physical model is not accurate, resulting in bad fitting results
In addition, there's another pitfall here, which is this R, which is not a stable value, this R is just the resistance at room temperature, and as the temperature goes up, this R is going down, carbon-based materials have a negative temperature property, with higher temperature and lower resistance. In COMSOL, the simulation is in the form of conductivity, which is a function of temperature, which I've got
Please login with a confirmed email address before reporting spam
I now suspect that the R-T nonlinearity affected my fitting because the carbon-based material has a negative temperature property, so I measured the R-T relationship experimentally, and the nonlinear conductivity-temperature relationship is obtained through the s-curve fitting in the classical nonlinear fitting. The nonlinear relationship is very strong. In this case, it actually affects my original model, u ^ 2/r = A * t + b * t ^ 4 + C
The R in this formula is actually a function of T, and if you take that into account, then the formula will change for some reason