Characterization and modeling of a thermal conductivity sensor for liquids

Accurate thermal conductivity measurements are difficult to achieve and require intricate and costly devices. Microfluidic solutions could aid in cutting down on costs while improving speed and efficiency due to the microscale test volume. In this paper, we aim to characterize a micromachined thermal conductivity sensor. We start by modeling the sensor to validate the design. Simultaneously, the model provides insight into the effects that relevant sensor parameters have on the thermal conductivity readings. Next, we perform measurements that validate the findings obtained through simulation. Lastly, measurements of liquids with distinct thermal conductivity are carried out to find a characterization coefficient. This coefficient facilitates the conversion of the sensor output into the measured thermal conductivity. We conclude that the sensor shows promising results, with a full-scale deviation of as low as 2%. However, further research is required to pinpoint the cause for inconsistent readings that prevent commercial viability.