🧵 ParSim: A Multithreaded Particle Collision Simulator

A high-performance 2D particle simulation using multithreading in Python. Simulates particle motion and collisions in a bounded space with real-time grid-based spatial partitioning for efficient collision detection.

---

🚀 Features

• Simulates 10,000 particles (customizable)
• Multithreaded for better CPU utilization (default: 8 threads)
• Realistic collision handling with elastic response
• Spatial grid for optimized neighbor checks
• Per-step logs of:
  • Duration
  • Collisions resolved
  • CPU utilization
  • Grid rebuild time
  • Active threads
• Built-in performance metrics: FPS, total simulation time, CPU average

---

🧠 How It Works

1. Initialization:

   • Each particle is randomly assigned position, velocity, radius, and mass.
   • The 2D space is divided into grid cells based on particle size.

2. Each Simulation Step Includes:

   • Particle movement update
   • Boundary collision management
   • Grid repopulation
   • Particle-particle collision resolution across cells (with lock-based safety)
   • Performance logging (timing, CPU, etc.)

3. Multithreading:

   • Uses ThreadPoolExecutor for parallel updates and collision checks.
   • Dynamically distributes work across threads.

---

📦 Files

File	Description
physics_simulator.py	Main simulation script
output.txt (optional)	Console logs of each simulation step

---

▶️ How to Run

✅ Requirements

• Python 3.7+
• psutil: for measuring CPU usage

Install required package:

pip install psutil

🔁 Run the simulation

python physics_simulator.py

You’ll see logs printed step-by-step in the console, e.g.:

Step 1/50 | Duration: 0.048s | Collisions: 2400 | CPU util: 99.7% | Active threads: 9 | Grid Rebuild: 0.005s
...

At the end, summary metrics like total time, FPS, and total collisions will be shown.

---

⚙️ Configuration

You can tweak simulation parameters at the top of physics_simulator.py:

no_of_particles = 10000        # total particles
no_of_sim_steps = 50           # how many steps to simulate
no_of_threads = 8              # parallel threads to use
width = 1000                   # width of the space
height = 800                   # height of the space
particle_max_initial_velocity = 100

---

📈 Output Metrics

• Total simulation time
• Average, max, min step duration
• Estimated FPS
• Total collisions handled
• Average CPU usage per step

---

🧪 Benchmarking Tips

• Try changing no_of_particles or no_of_threads to see how it scales.
• You can redirect output to a file:

python physics_simulator.py > output.txt

---

📌 Notes

• Particle locking ensures safe concurrent updates.
• CPU utilization may show >100% due to multithreading on multi-core CPUs (normal in Python’s psutil).
• Grid-based collision detection significantly boosts performance compared to brute-force.

---

🧊 Future Improvements (Suggestions)

• Add visualization (e.g., using pygame or matplotlib)
• Profile memory usage
• Try multiprocessing for CPU-bound steps
• Experiment with alternative data structures (QuadTrees, etc.)

---

🧑‍💻 Author

T Jaiwanth

---

Happy Simulating! 🚀