# st-tech/ppf-contact-solver

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/st-tech-ppf-contact-solver).**

_How this analysis was created: the description and tags below were written by an AI model that read this project's README and public documentation pages; stars, license and language come straight from the GitHub API. The model does not read the source code._

4,595 stars · 346 forks · Python · Apache-2.0

## Links

- GitHub: https://github.com/st-tech/ppf-contact-solver
- awesome-repositories: https://awesome-repositories.com/repository/st-tech-ppf-contact-solver.md

## Topics

`cloth` `collision` `contact` `physics` `simulation`

## Description

ppf-contact-solver is a physics engine for contact-rich simulation of deformable and rigid materials. It handles cloth and thin shells, tetrahedral solids, rods, rigid bodies, and sand in a single solver, using finite element methods and a GPU-accelerated contact solver. Every successfully solved step is free of intersecting geometry, and strict stretch limits keep fabric from behaving like rubber. The solver runs on NVIDIA CUDA, AMD ROCm, Apple Metal, or a SIMD-optimized CPU backend, and the matching build is chosen at run time.

Beyond the core solver, the project provides a Blender add-on and a chainable Python API for building scenes. Scenes are assembled from material groups, pinned vertices, invisible colliders, and keyframed motion. Scripts, JupyterLab notebooks, headless runs, or an AI tool server driven by a language model can all set up scenes, tune parameters, and generate batches of variants. Simulations can run on a remote GPU server while the interface stays local, and results are fetched back and can be baked into ordinary keyframes.

Materials include fabric presets calibrated against real textile measurements. Drape, cantilever bending, and tensile stretch tests are checked against published reference data, and the project can generate calibration reports. The elasticity models use closed-form, eigen-filtered Hessians for shells and volumetric solids, so no iterative eigen solver is needed. Diagnostics cover solver convergence, line search, contact load, and stage timing.

The solver can be packaged as a self-contained, relocatable distribution that bundles binaries, a Python interpreter, and example notebooks. It can also be built as a container image, and the documentation covers deployment on cloud GPU providers including Google Cloud, RunPod, Scaleway, and Vast.ai.

## Tags

### Game Development

- [Contact-Rich Multi-Material Solvers](https://awesome-repositories.com/f/game-development/rigid-body-physics-engines/multi-body-dynamics-simulators/hybrid-rigid-deformable-simulations/contact-rich-multi-material-solvers.md) — Resolves collisions and contact between cloth, solids, rods, rigid bodies and sand in a single GPU-accelerated finite element contact solver. ([source](https://github.com/st-tech/ppf-contact-solver#readme))
- [Barrier-Based Penetration-Free Contact](https://awesome-repositories.com/f/game-development/barrier-based-penetration-free-contact.md) — Rejects any step causing intersections using continuous collision detection and a cubic barrier potential, so every accepted step stays penetration-free.
- [Specialized Physical Simulations](https://awesome-repositories.com/f/game-development/physics-simulation-engines/specialized-physical-simulations.md) — Resolves collisions and contact between cloth, solids, rods, and other materials, demonstrated through ready-to-run example notebooks. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/frontend))
- [Hybrid Rigid-Deformable Simulations](https://awesome-repositories.com/f/game-development/rigid-body-physics-engines/multi-body-dynamics-simulators/hybrid-rigid-deformable-simulations.md) — Resolves collisions between shells, solids, rods, rigid bodies, and sand in one solver, with no intersecting geometry after each solved step. ([source](https://st-tech.github.io/ppf-contact-solver/jupyterlab_api/material_parameters.html))
- [FEM-Based Deformable Body Simulators](https://awesome-repositories.com/f/game-development/soft-body-simulations/fem-based-deformable-body-simulators.md) — Models deformable materials with the finite element method, using symbolic force Jacobians and fabric presets calibrated against real measurements. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/README.md))
- [Simulation Scene Composition](https://awesome-repositories.com/f/game-development/simulation-scene-composition.md) — Assembles registered mesh assets into a scene and builds it into a form the solver can run. ([source](https://st-tech.github.io/ppf-contact-solver/jupyterlab_api/module_reference.html))

### Scientific & Mathematical Computing

- [GPU-Accelerated Physics](https://awesome-repositories.com/f/scientific-mathematical-computing/gpu-accelerated-physics.md) — Runs a GPU-accelerated contact and elasticity solver on CUDA, ROCm, Metal, or a SIMD CPU backend, resolving collisions without interpenetration.
- [Chainable Scene Scripting APIs](https://awesome-repositories.com/f/scientific-mathematical-computing/chainable-scene-scripting-apis.md) — Builds groups, pins, curves, colliders, and time-varying scene parameters through chainable Python calls to automate full simulation setups. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/blender_addon))
- [Closed-Form Elastic Hessians](https://awesome-repositories.com/f/scientific-mathematical-computing/closed-form-elastic-hessians.md) — Derives eigen-filtered force Jacobians of isotropic energies from singular-value or invariant derivatives, keeping the Hessian positive semi-definite without iterative eigen solvers. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/articles/eigensys.md))
- [Multi-Element Deformable Models](https://awesome-repositories.com/f/scientific-mathematical-computing/multi-element-deformable-models.md) — Solves shells, solids, rods, rigid bodies and sand in one solver, with element types assigned to material groups carrying per-group parameters.
- [Cloth Strain Limiting](https://awesome-repositories.com/f/scientific-mathematical-computing/cloth-strain-limiting.md) — Enforces strict upper bounds on how far any single triangle can extend, so cloth and shells do not behave like rubber. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/README.md))
- [Contact Parameter Tuning](https://awesome-repositories.com/f/scientific-mathematical-computing/contact-parameter-tuning.md) — Sets per-group physical values such as density, stiffness, friction, shrink, inflate pressure, and contact gap across up to 32 independent groups. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/blender_addon))
- [Eigen-Filtered Hessians](https://awesome-repositories.com/f/scientific-mathematical-computing/eigen-filtered-hessians.md) — Derives elastic force Jacobians analytically from singular values or invariants, keeping Hessians positive semi-definite without any iterative eigen solver.
- [Fabric Drape Validation](https://awesome-repositories.com/f/scientific-mathematical-computing/fabric-drape-validation.md) — Compares simulated drape coefficients with target ranges per fabric, reporting fold counts and bend droop angles to confirm realistic behavior. ([source](https://st-tech.github.io/ppf-contact-solver/fabric-report/index.html))
- [Fabric Material Calibration](https://awesome-repositories.com/f/scientific-mathematical-computing/fabric-material-calibration.md) — Checks simulated fabric presets with drape, cantilever bending, and tensile stretch tests against published reference data and generates calibration reports.
- [Hardware Compute Backend Selection](https://awesome-repositories.com/f/scientific-mathematical-computing/hardware-compute-backend-selection.md) — Selects a CUDA, Metal, ROCm, or SIMD CPU backend for the target machine and reports a clear error when no toolkit is found. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/Cargo.toml))
- [Material Groups](https://awesome-repositories.com/f/scientific-mathematical-computing/material-groups.md) — Creates named material groups of type solid, shell, rod, static, rigid body, or sand, assigns objects, and tunes per-group parameters and colors. ([source](https://st-tech.github.io/ppf-contact-solver/blender_addon/integrations/python_api.html))
- [Scriptable Simulation Frameworks](https://awesome-repositories.com/f/scientific-mathematical-computing/scriptable-simulation-frameworks.md) — A framework driven through a chainable Python API and JupyterLab notebooks to automate scene setup, parameter control, headless runs and batch variants.
- [Shell Elasticity Models](https://awesome-repositories.com/f/scientific-mathematical-computing/shell-elasticity-models.md) — Extends closed-form eigen-analysis to triangular meshes with non-square deformation gradients, giving stable elastic Hessians for cloth and thin shells. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/articles/eigensys.md))
- [Workflow Automation Scripts](https://awesome-repositories.com/f/scientific-mathematical-computing/simulation-input-scripting/workflow-automation-scripts.md) — Automates scene building, parameter tuning, and batch generation of simulation variants from Python scripts, notebooks, headless runs, or language-model prompts. ([source](https://st-tech.github.io/ppf-contact-solver/blender_addon/workflow/index.html))
- [Simulation Material Groups](https://awesome-repositories.com/f/scientific-mathematical-computing/simulation-material-groups.md) — Assigns objects to solid, shell, rod, rigid body, sand or static groups and tunes per-group density, stiffness, friction and strain limit. ([source](https://st-tech.github.io/ppf-contact-solver/blender_addon/workflow/index.html))
- [Volumetric Solid Elasticity Models](https://awesome-repositories.com/f/scientific-mathematical-computing/volumetric-solid-elasticity-models.md) — Evaluates ARAP, Symmetric Dirichlet, MIPS, Ogden, and Yeoh energies on 3D deformation gradients using singular values and a closed-form decomposition. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/articles/eigensys.md))

### Part of an Awesome List

- [Cloth Modeling and Simulation](https://awesome-repositories.com/f/awesome-lists/ai/cloth-modeling-and-simulation.md) — Models how silk, cotton, wool, denim, and leather drape, fold, and droop using bending stiffness, Young's modulus, and Poisson ratio. ([source](https://st-tech.github.io/ppf-contact-solver/fabric-report/index.html))
- [Blender Scripting](https://awesome-repositories.com/f/awesome-lists/devtools/blender-scripting.md) — Controls the whole simulation pipeline from Python scripts in Blender's scripting editor for procedural scene setup and batch variant generation. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/README.md))
- [JupyterLab Extensions](https://awesome-repositories.com/f/awesome-lists/devtools/jupyterlab-extensions.md) — Provides a browser-based JupyterLab interface with a chainable Python API for building meshes and scenes and viewing results interactively. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/README.md))
- [Invisible Collider Boundaries](https://awesome-repositories.com/f/awesome-lists/devtools/physics-and-collision/invisible-collider-boundaries.md) — Defines invisible walls and spheres with adjustable friction, contact gap, and keyframed position or radius to confine simulated objects. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/blender_addon))

### Artificial Intelligence & ML

- [Hardware-Accelerated Simulations](https://awesome-repositories.com/f/artificial-intelligence-ml/hardware-accelerated-simulations.md) — Runs contact and elasticity solvers in parallel on NVIDIA, AMD, or Apple silicon GPUs, with a SIMD-optimized CPU fallback. ([source](https://github.com/st-tech/ppf-contact-solver#readme))
- [Hardware Backend Selection](https://awesome-repositories.com/f/artificial-intelligence-ml/hardware-agnostic-accelerators/compute-backend-abstractions/hardware-backend-selection.md) — Bundles CUDA, ROCm, and CPU solver builds and picks the only GPU build present or the first reporting a usable device at run start. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/build-linux-native))

### Data & Databases

- [CPU-GPU Backend Switching Abstractions](https://awesome-repositories.com/f/data-databases/storage-scaling/gpu-acceleration-backends/unified-gpu-backend-abstractions/cpu-gpu-backend-switching-abstractions.md) — Compiles one solver codebase for CUDA, ROCm, Metal and SIMD CPU targets and selects the matching build at run time.

### Development Tools & Productivity

- [Natural Language Automation](https://awesome-repositories.com/f/development-tools-productivity/cross-platform-automation-frameworks/natural-language-automation.md) — Exposes add-on tools through a server so a language model can build scenes, tweak parameters and run simulations from text prompts. ([source](https://github.com/st-tech/ppf-contact-solver#readme))
- [Interactive Notebook Environments](https://awesome-repositories.com/f/development-tools-productivity/interactive-notebook-environments.md) — Provides a browser-based JupyterLab environment with a chainable Python frontend for building scenes, viewing results interactively, and exporting them. ([source](https://github.com/st-tech/ppf-contact-solver#readme))
- [Remote Simulation Connectivity](https://awesome-repositories.com/f/development-tools-productivity/remote-simulation-connectivity.md) — Runs simulations on a remote GPU server while the interface stays local, then fetches results back, even from a laptop. ([source](https://github.com/st-tech/ppf-contact-solver#readme))

### DevOps & Infrastructure

- [Cloud Container Deployments](https://awesome-repositories.com/f/devops-infrastructure/cloud-container-deployments.md) — Packages the solver in a compact container image that launches on major cloud GPU services and shares notebook links with collaborators. ([source](https://github.com/st-tech/ppf-contact-solver#readme))
- [Remote Simulation Hosting](https://awesome-repositories.com/f/devops-infrastructure/cloud-deployment/gpu-training-deployments/remote-simulation-hosting.md) — Runs simulations on remote or rented GPU servers via containers, SSH, and cloud guides, then fetches results back into the local interface.
- [Self-Contained Distribution Builds](https://awesome-repositories.com/f/devops-infrastructure/self-contained-distribution-builds.md) — Compiles the solver on Linux and packages backends, a relocatable Python, frontend packages, notebooks, and a video encoder into one directory needing no toolkit or root. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/build-linux-native))
- [Remote Server Connections](https://awesome-repositories.com/f/devops-infrastructure/self-hosted-collaboration-servers/remote-server-connections.md) — Links the Blender add-on to a remote GPU server, such as a rented cloud instance or Windows machine, so simulations run off the local computer. ([source](https://st-tech.github.io/ppf-contact-solver/blender_addon/tutorial.html))

### Graphics & Multimedia

- [Blender Add-ons](https://awesome-repositories.com/f/graphics-multimedia/blender-add-ons.md) — Ships a Blender add-on that builds simulation scenes, runs them on a remote GPU server, and bakes results into keyframes. ([source](https://github.com/st-tech/ppf-contact-solver/blob/main/install-blender-addon.ps1))
- [Blender Physics Simulation Workflows](https://awesome-repositories.com/f/graphics-multimedia/blender-physics-simulation-workflows.md) — Sets up pinned vertices, colliders, and keyframed motion in Blender, runs simulations, and bakes the results into standard animation keyframes.
- [Simulation Cache Baking](https://awesome-repositories.com/f/graphics-multimedia/keyframe-animations/simulation-cache-baking.md) — Converts fetched point-cache simulation results into ordinary keyframes so scenes play back without the add-on installed. ([source](https://st-tech.github.io/ppf-contact-solver/blender_addon/workflow/index.html))
- [Material Behavior Modeling](https://awesome-repositories.com/f/graphics-multimedia/particle-physics-simulations/gpu-accelerated-physics-simulations/interactive-physics-engines/material-behavior-modeling.md) — Assigns each simulated body a material type such as shell, solid, rod, static, rigid body, or sand with its own model. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/blender_addon))

### Software Engineering & Architecture

- [Chainable Scene-Building APIs](https://awesome-repositories.com/f/software-engineering-architecture/chainable-scene-building-apis.md) — Provides a chainable Python API that assembles scenes, groups, pins, colliders and parameters and drives the solver from notebooks, Blender or headless scripts.
- [Client-Server Remote Execution](https://awesome-repositories.com/f/software-engineering-architecture/client-server-remote-execution.md) — Uploads geometry and parameters from a local frontend to a remote GPU solver host, runs the simulation there, and fetches frames back.

### Testing & Quality Assurance

- [Programmatic Simulation Control](https://awesome-repositories.com/f/testing-quality-assurance/environment-simulations/real-time-simulation-modes/programmatic-simulation-control.md) — Offers a scripting interface and session management to configure parameters, run simulations, and inspect results programmatically. ([source](https://github.com/st-tech/ppf-contact-solver/tree/main/frontend))
