SpritZ - Tool for the management of GT-SUITE models

This tool integrates an innovative methodology for the development and analysis phases of any GT-SUITE model using the git versioning tool.

After 10 years of experience using GT-SUITE as an automotive powertrain and vehicle engineer, I believe that GT-SUITE still lacks an integrated versioning system. In my opinion, such a versioning system is essential to ensure consistency among results and models and to establish a clear 'reference' model for developers and teams. For these reasons, I developed this tool.

SpritZ - Tool for GT-SUITE Model Management

The project aims to create a versioning tool for GT-SUITE models using MATLAB. GT-SUITE is a commercially available 1D-CFD software widely used in automotive research and industry for system model simulations.

The core idea behind developing this tool is that GT-SUITE is typically used in two main phases:

1. Model development: during this phase, models are created, updated, and validated against experimental tests. Multiple modifications and variants are often explored. However, not all development paths, or "branches," yield fruitful results. Only one version becomes the reference model.
2. Analysis: the model developed in the first phase serves as a virtual test rig for sensitivity analysis, component changes, and various operating conditions.

To efficiently utilize GT-SUITE, it's crucial to manage model versions in a structured manner. That's why I decided to harness the versioning capabilities of the well-known Git software (https://git-scm.com/).

This tool offers the following capabilities:

• ensures that a single, unique model serves as the reference for a given project;
• manages versioning of files, keeping a comprehensive model history;
• organizes GT-SUITE result files (.gdx, .glx) and renames them with codes from the model history;
• launches models on the distributed solver;
• facilitates sensitivity analysis with an accompanying Excel file for updating model parameters;
• integrates a timer for scheduling simulations to avoid overloading the distributed server during working hours;
• can be used to retrieve results from the distributed solver.

I hope you find this tool helpful, and I'm open to any feedback!

Alessandro

P.S. This tool was developed at the beginning of 2019 during my PhD thesis and was invaluable for keeping track of versions and models. If you're interested in my PhD work, you can find my full thesis here: https://iris.polito.it/handle/11583/2842511