Dataset for Validation of Two-Box Spacer Modeling Across Spacer Widths

This dataset contains the underlying simulation models, raw data exports, and comparative analysis worksheets for the research paper "Validation of Two-Box Spacer Modeling Across Spacer Widths for Fenestration Thermal Analysis," submitted to Energy and Buildings.

The study evaluates the accuracy and limitations of two simplified two-box spacer modeling approaches across a range of spacer widths (8-20 mm) and three spacer families, using detailed THERM simulations as the baseline.

The data was gathered using an automated simulation framework that interfaced with LBNL THERM 7.8.71.0 and WINDOW 7.8.71.0.
25 discrete widths were modelled for three generic spacer families using detailed geometries.
Corresponding two-box models were generated, one using width-specific calibration and one using nominal-width calibration (calibrated at 8 mm only).
A total of 900 full-frame simulations were performed to evaluate the total window thermal U-factor and Condensation Index.

The data demonstrates that width-calibrated two-box models are stable across all widths and spacer families. However, for the intermediate conductivity spacer family (metal, double seal), the data shows a linear increase in error as the width deviates from the calibration point. Specifically using an 8 mm calibration for a 20 mm metal double seal spacer resulted in errors exceeding industry-accepted tolerances for U-factor (0.01 W/m²·K).

Files are organized into folders mapping to the figures in the associated manuscript. The dataset includes:
.xlsx: Calculation spreadsheets for U-factor aggregation and error plotting.
.thm: THERM simulation files (serialized binary)
.o THERM simulation output file (node data, used for CI Tool)
.thmx THERM simulation output file (XML format .thm file, used for CI Tool)
.txt Text log file
.csv Result CSV file (result data exports)
.mdb Access database files (used by WINDOW)

Researchers can use this data to:
Benchmark the accuracy of other FEA-based thermal modeling tools.
Validate alternative simplified spacer modeling approaches.
Test how different calibration tolerances affect full-window results.
Development of standards and acceptance criteria for two-box spacer models in certification workflows.