Abstract
Differential Scanning Calorimetry (DSC) is often used to characterize thermophysical properties associated with phase transformation of metals and alloys. In such devices, however, the practical design of the instrument does not allow for direct temperature measurements of the sample material. As a result, the contact conductances and radiative interactions among system components yield thermal lags between the collected temperature data and sample temperature. Therefore, a direct association of the recorded thermocouple readings to the sample site may produce erroneous results. In order to account for these heat transfer mechanisms, a new parameter estimation method has been developed utilizing a lumped heat transfer model for the key DSC components. Preliminary results using a benchmark numerical problem have been obtained which show accurate recovery of the system parameters and demonstrate the robustness of the new method. A calibration sequence is proposed to reduce the set of parameters being resolved by the numerical method. This sequencing increases accuracy in the parameter predictions and maintains stability as the induced thermocouple error is increased. Based on these encouraging numerical results, the method is presently being applied to real experimental situations.
Original language | English |
---|---|
Pages | 527-536 |
Number of pages | 10 |
State | Published - 2004 |
Event | 2004 TMS Annual Meeting - Charlotte, NC, United States Duration: Mar 14 2004 → Mar 18 2004 |
Conference
Conference | 2004 TMS Annual Meeting |
---|---|
Country/Territory | United States |
City | Charlotte, NC |
Period | 03/14/04 → 03/18/04 |
Keywords
- Function Decomposition Method
- Heat-Flux DSC
- Inverse Problems
- Parameter Estimation