Data from: Quantification and mitigation of uncertainties in thermal conductivity measurements using a modified ASTM D5470 thermal resistance tester
Data files
Aug 13, 2025 version files 247.16 MB
-
ned3-008_ASTMD5470.zip
247.15 MB
-
README.md
3.82 KB
Nov 11, 2025 version files 253.23 MB
-
ned3-008_ASTMD5470.zip
253.22 MB
-
README.md
5.06 KB
Abstract
In this paper, we have investigated the effect of thermal interface materials (TIMs) on the accuracy and uncertainty of thermal conductivity measurements. A modified ASTM D5470 thermal resistance tester (TRT) has been developed to measure the out-of-plane thermal conductivity of pyrolytic graphite (PG) and titanium grade 2 (TiG2) with and without TIMs. Compared to the ASTM D5470 standard, this modified design uses three thermocouples per side to support regressive analysis of the temperature profile and quantifies the uncertainty of the measurements with and without TIMs. Nine PG samples and four TiG2 samples of varying thickness have been tested to obtain thermal resistance as a function of sample thickness. The steady-state temperature profiles were used for heat flux and thermal resistance calculation. The results reveal that TIMs significantly reduce measurement uncertainty for both samples, i.e., 38.8% for TiG2 and 27.8% for PG. The effect of TIMs on the measurement accuracy diverges, with a far more pronounced effect on TiG2 than PG. This can be owed to the higher out-of-plane thermal conductivity of TiG2 (~ 16.2 W/m-K) than PG (~ 2 W/m-K). Contact resistance is expected to play a more critical role in tests of materials with higher thermal conductivity, and TIMs can effectively mitigate contact resistance. For lower conductivity materials, the weight of contact resistance is lower, and the effect of TIMs is thus less pronounced.
Dataset DOI: 10.5061/dryad.mcvdnckck
Description of the data and file structure
This dataset includes temperature measurements over time for 6 thermocouples for steady state analysis of thermal resistance and thermal conductivity of Pyrolytic Graphite and Titanium Grade 2. The temperature data is measured using OMEGA TJ36-CPSS-032G-3 T-type thermocouples with a 0.032” probe diameter connected to a NI-9210 Thermocouple DAQ with Spring Terminals, attached to a NI-cDAQ-9174 DAQ Chassis which is connected to a computer running Windows 10. Time data is sampled at 1000 Hz (every 1 ms), starting at 0 s and continuing until manually stopped. The data acquisition setup is used in the facility described in the manuscript.
Files and variables
File: DatasetsAndCode.zip
Description: The zip file "ned3-008_ASTMD5470.zip" includes 3 folders, i.e., “DataForPaper,” “PythonCode,” and "CAD."
DataForPaper:
The “DataForPaper” folder further contains two folders, i.e., PG for the Pyrolytic Graphite data, and TiG2 for the Titanium Grade 2 data.
Each folder is then divided into sub-folders for “WithTIM” for datasets with TIMs applied, and “NoTIM” for datasets without any TIMs applied.
Each “WithTIM” and “NoTIM” folder is then sub-divided into “Run1”, “Run2”, and “Run3” which contain the data for each material, i.e. 9 files for PG corresponding to the 9 PG samples, and 4 files for TiG2 corresponding to the 4 TiG2 samples.
All the Data files are .lvm files with 23 lines of headers before the actual data starts. The data is divided into 7 columns delimited using “tabs” as follows: Time, Temperature at TC1, Temperature at TC2, Temperature at TC3, Temperature at TC4, Temperature at TC5, Temperature at TC6. Time is measured in seconds (s) and the temperature is measured in Celsius (degree C).
PythonCode:
The “PythonCode” folder contains multiple .ipynb files that are used for the steady state analysis, uncertainty analysis, thermal conductivity, and thermal resistance calculations.
“SSCalc.ipynb” checks if the data collected has reached steady-state and compares the ΔT (change in temperature) with respect to time over a period of 120s or 2 minutes. It also checks the ΔT with respect to time over a period of 30s, 60s, and 180s to ensure that minor changes in T or sensor noise does not affect long term steady-state temperatures.
“AnalysisCodeTRT.ipynb” requires a defined material, TIM state (With or No TIM), and the repeated “Run Number” for the tests in Cell 2 of the code. It outputs the measured thermal conductivity of the materials and the uncertainty in the thermal resistance measurements.
“ScedasticityCheck.ipynb” uses the Uncertainty in each Thermal Resistance measurement from “AnalysisCodeTRT.ipynb” and the sample thicknesses to numerically check for homoscedasticity or heteroscedasticity in the data and visualize it using plots.
“YorkRegression.ipynb” uses “YorkRegression.py”, a modified version of “fit_bivariate.py” by Mikko Pitkanen, to compare regular Least Squares Regression methods to York’s Regression mentioned in York et al. 2004 and provides thermal conductivity and uncertainty values based on York’s Regression.
CAD:
The “CAD” folder contains CAD model of the modified ASTM D5470 thermal resistance tester. It includes two folders, "SolidWorks_SLDPRT_SLDASM" and "eDrawings_EPRT_EASM." The folder "SolidWorks_SLDPRT_SLASM" contains 1 Assembly file named "TRT Assembly.SLDASM" which is the complete assembly for the thermal resistance tester used in the paper. The "CAD" folder also contains 3 folders which are named after the respective sub-assemblies of the setup mentioned in the paper. Each folder contains parts of the sub-assembly as mentioned in the paper. The "eDrawings_EPRT_EASM" includes the same CAD models as the "SolidWorks_SLDPRT_SLDASM" folder, but instead of .SLPART and .SLDASM files, it uses .EPRT and .ESAM files, respectively.
PythonCode:
The “PythonCode” folder contains multiple .ipynb files that are used for the steady state analysis, uncertainty analysis, thermal conductivity, and thermal resistance calculations.
Code/software
The LabVIEW measurement files (.lvm files) can be viewed using, text editor, LabVIEW, Excel, Python, and MATLAB.
The Python notebook files (.ipynb) and code (.py) can be opened using Google Colab and Jupyter Notebook, and VS Code.
The SolidWorks part and assembly files (.SLDPRT and .SLDASM, respectively) can be opened and edited using SolidWorks 2024 or newer.
The eDrawings part and assembly files (.EPRT and .EASM, respectively) can be opened using eDrawings Viewer, which is free software.
Changes after Aug 13, 2025: Added CAD models to the dataset. Specifically, a folder "CAD" is added to the main directory. README is updated to include a description of the added files.