Thermal Analysis Trivia

7th : What is the Measurement Errors in TMA Analysis ?

Accurate TMA analysis requires careful attention to both sample dimensions and instrument drift.

"Sample length" and "drift" are cited as sources of error when calculating the expansion rate or average coefficient of thermal expansion, CTE from TMA measurement results. Since the expansion rate and the CTE are calculated by dividing by the “sample length”, the sample length must be measured accurately using a micrometer or a similar device. Another factor, "drift", is the amount of change that depends on the instrument, and to investigate the amount of drift, perform blank measurements. Blank measurement can be performed using a differential expansion TMA by placing reference samples — SiO₂ (or Al₂O₃ for high-temperature applications) — on both the sample and reference sides. In differential expansion TMA, the measurement result is calculated as:

　　sample change − reference sample change + drift

If the sample expansion is sufficiently larger compared to the amount of drift, , the resulting error is small. However, when the sample expansion is small and drift becomes a concern, the drift can be canceled out by subtracting the blank measurement result from the TMA sample measurement result. This operation can be performed within the software by selecting [Analysis] - [Correction/Calibration] - [Subtracting Blank]. (The process of adding the “change in the reference sample” to the TMA measurement results to obtain only the “change in the sample” is called reference sample correction.) Single-rod expansion methods such as the tensile load method, which measures only the sample cannot perform blank measurements; thus, the ability to conduct blank measurements can be considered a distinct feature of the differential expansion method.