Skip to main content

Importing, Modifying, and Configuring Blender Models

This tutorial will demonstrate how to import, configure, and parametrically modify Blender models directly in the Luminary platform.

Updated over a week ago

Reading Time: 10 minutes

Target Audience: intermediate to advanced users with a basic understanding of Blender


In this Tutorial

Welcome to this tutorial on Luminary’s Blender integration! This tutorial covers the following key concepts:

  • Preparing a model for parametric modification in Blender

  • Importing a blender model into Luminary

  • Modifying the model in Luminary

  • Swapping part/configurations

By the end, you’ll understand the fundamentals required to modify a geometry for design exploration, optimization, and dataset generation for physics AI model training.


Background

Blender is a popular 3D modeling software that is widely used by industrial designers. It’s an extremely powerful tool that enables users to freely or parametrically morph geometric shapes. Pairing such a tool with a massively scalable simulation platform like Luminary unlocks many possibilities for realtime design feedback and optimization.

You can read about how we leveraged this integration to build the SHIFT physics AI foundation model here.

The parameterized Blender model was created from the original AeroSUV STLs (https://www.ecara.org/drivaer-1) following this procedure:

  • Initial import

    • Create a new Blender project

    • Import the AeroSUV STLs

    • Organize the different meshes into collections for easier management

  • Create a cage around the model:

    • Create a Cube mesh

    • Iteratively subdivide and model the mesh to create a tightly fit cage around the model. Note that the model should be entirely enclosed inside the cage

    • Create 4 Cylinders around the brakes and suspensions

    • Create shape keys on the Cage that correspond to desired parameterized deformations

  • Deform the model

    • For each mesh of the model (except the wheels; see below), create a MeshDeform modifier and bind it to the Cage.

    • For each of the 4 suspensions, create a Vertex Group of the vertices that are closest to the associated wheel

    • For each wheel, create a Child Of constraint that targets the suspension mesh and the associated Vertex Group, with Location X and Y enabled. This will allow the wheels to slide along the xy-plane, together with the suspension but without being deformed

  • Hide any meshes you do not want to import into Luminary, e.g., the Cage and any other other unwanted configurations.

Note that this model was decimated in Blender to reduce the mesh size. Features might not be exactly respected. Read more about applying shape keys in Blender here.


Create a Project

Let’s start by creating a project:

  • Click the Create Project button.

  • Navigate to the Tutorials tab.

  • Select the project Blender: AeroSUV Variation


Explore the Geometry

This model is a variation of the AeroSUV [link], an open-access, SUV reference model developed by FKFS (Forschungsinstitut für Kraftfahrwesen und Fahrzeugmotoren Stuttgart) as an extension of the DrivAer concept.

In the geometry tree you can explore the volumes and surfaces in the model. On the right side you can view the parameters that have been imported from the model. These parameters are based on commonly used in automotive design.


Blender Import Overview

Luminary supports a number of file formats that can be exported from Blender. Of these formats, Blender, Universal Scene Description, Filmbox, and Graphics Library Transmission all support shape keys, which are used to modify the geometry after importing into Luminary.

Note that Luminary imports meshes that are visible in the viewport, i.e., both the meshes and their parent collections are visible.

Modifying parameters in the geometry tab

To modify the imported parameters:

  1. Click on the Modification List under the “Import AeroSUV..” dropdown

  2. In the Import panel, input values or change the slider associated with each parameter you wish to modify

  3. (optional) change the overall scale of the model using the Scaling input field

  4. Click Apply at the bottom of the Import panel

Setting up and modifying a Configuration

Configurations allow you to easily swap between different, previously set-up geometry variations. When you create a Configuration, you set up each configuration as a set of 0, 1, or more volumes and give it a unique name.

You then choose an active configuration. All volumes associated with the active configuration are kept, while volumes not associated with the active configuration are removed from the model. You can later choose a different active configuration, and the model will be updated to display the volumes associated with that configuration instead.

To modify a Configuration:

  1. Open the Configurations panel:

    • If it’s a new configuration, click on the Configuration button in the toolbar banner

    • If it’s an existing configuration, click on it in the Modification List

  2. In the Configurations panel, you can:

    • Add new empty configurations or remove existing ones

    • Rename configurations

    • Add or remove volumes to/from a configuration

    • Choose the active configuration

  3. Click Apply at the bottom of the “Configurations panel

Did this answer your question?