feat: Major update with validators, skills, dashboard, and docs reorganization

- Add validation framework (config, model, results, study validators) - Add Claude Code skills (create-study, run-optimization, generate-report, troubleshoot, analyze-model) - Add Atomizer Dashboard (React frontend + FastAPI backend) - Reorganize docs into structured directories (00-09) - Add neural surrogate modules and training infrastructure - Add multi-objective optimization support 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-25 19:23:58 -05:00
parent 74a92803b7
commit e3bdb08a22
155 changed files with 52729 additions and 37 deletions
--- a/studies/drone_gimbal_arm_optimization/NX_FILE_MODIFICATIONS_REQUIRED.md
+++ b/studies/drone_gimbal_arm_optimization/NX_FILE_MODIFICATIONS_REQUIRED.md
@@ -0,0 +1,146 @@
+# NX File Modifications Required for Drone Gimbal Arm Study
+
+## Overview
+
+The study uses the same beam model as `simple_beam_optimization` but requires modifications to:
+1. Add modal analysis (frequency extraction)
+2. Update loading conditions for the 850g camera payload
+3. Ensure material properties match Al 7075-T6
+
+## Critical Modifications
+
+### 1. Simulation File (Beam_sim1.sim)
+
+**REQUIRED: Add Modal Analysis Solution**
+
+You need to add a **second solution** for modal analysis:
+
+1. **Open** `Beam_sim1.sim` in NX Simcenter
+2. **Create New Solution**:
+   - Solution Type: `SOL 103 - Normal Modes`
+   - Name: `modal_analysis`
+   - Number of modes: `10` (we only need the first, but calculate more for safety)
+   - Frequency range: `0-500 Hz`
+
+3. **Use Same Mesh** as the static solution
+   - Link to existing FEM file: `Beam_fem1.fem`
+
+4. **Boundary Conditions**: Use same constraints as static analysis
+   - Fixed constraint at base (same as static)
+   - No loads needed for modal (it finds natural frequencies)
+
+### 2. Static Analysis Modifications
+
+**Update Load Magnitude**:
+
+The existing static analysis load needs to represent the **850g camera payload**:
+
+1. **Open Solution 1** (static analysis)
+2. **Modify Force Magnitude**:
+   - Old value: (whatever is currently there)
+   - **New value**: `8.34 N` (850g × 9.81 m/s²)
+   - Direction: Downward (negative Y or Z depending on your coordinate system)
+   - Location: Tip of beam (where camera attaches)
+
+Note: 120 MPa stress limit provides safety factor of 2.3 on 6061-T6 yield strength (276 MPa)
+
+### 3. Material Properties
+
+**Verify Material is Al 6061-T6**:
+
+1. **Open Part File**: `Beam.prt`
+2. **Check Material Assignment**:
+   - Material: `Aluminum 6061-T6`
+   - Yield Strength: ~276 MPa
+   - Young's Modulus: ~68.9 GPa
+   - Density: ~2700 kg/m³
+   - Poisson's Ratio: ~0.33
+
+3. **If not Al 6061-T6**, update material assignment to match drone application requirements
+
+### 4. Results Configuration
+
+**Ensure these results are requested**:
+
+**For Static Solution (Solution 1)**:
+- Displacement (VECTOR, all components)
+- von Mises Stress
+- Mass properties
+
+**For Modal Solution (Solution 2)**:
+- Natural frequencies
+- Mode shapes (optional, for visualization)
+
+## What You DON'T Need to Change
+
+The parametric design variables are already set up correctly in the beam model:
+- `beam_half_core_thickness` (20-30mm)
+- `beam_face_thickness` (1-3mm)
+- `holes_diameter` (180-280mm)
+- `hole_count` (8-14)
+
+These parameters will be automatically updated by the optimization loop.
+
+## Verification Steps
+
+Before running optimization, verify:
+
+1. **Two Solutions Exist**:
+   ```
+   Solution 1: Static Analysis (SOL 101) - displacement and stress
+   Solution 2: Modal Analysis (SOL 103) - natural frequencies
+   ```
+
+2. **Load is Correct**:
+   - Static load = 8.34 N downward at tip
+
+3. **Material is Al 7075-T6**
+
+4. **Both solutions solve successfully** with baseline parameters:
+   ```
+   beam_half_core_thickness = 25mm
+   beam_face_thickness = 2mm
+   holes_diameter = 230mm
+   hole_count = 11
+   ```
+
+## Quick Test
+
+Run a manual solve with baseline parameters to verify:
+
+Expected Results (approximate):
+- **Mass**: ~140-150g
+- **Max Displacement**: ~1-2 mm
+- **Max Stress**: ~80-100 MPa
+- **First Frequency**: ~120-140 Hz
+
+If these are wildly different, check your setup.
+
+## Extraction Configuration
+
+The optimization engine will extract:
+- **Mass**: From Solution 1 mass properties
+- **Displacement**: Maximum displacement magnitude from Solution 1
+- **Stress**: Maximum von Mises stress from Solution 1
+- **Frequency**: First natural frequency (mode 1) from Solution 2
+
+All extraction is automated - you just need to ensure the solutions are configured correctly.
+
+## Optional Enhancements
+
+If you want more realistic results:
+
+1. **Add Gravity Load**:
+   - Apply -9.81 m/s² gravity in addition to tip load
+   - Represents arm's own weight during flight
+
+2. **Add Damping** to modal analysis:
+   - Structural damping ratio: ~0.02 (2%)
+   - More realistic frequency response
+
+3. **Refine Mesh** at stress concentrations:
+   - Around holes
+   - At base constraint
+   - Better stress accuracy
+
+But these are NOT required for the optimization to run successfully.
--- a/studies/drone_gimbal_arm_optimization/README.md
+++ b/studies/drone_gimbal_arm_optimization/README.md
@@ -0,0 +1,193 @@
+# Drone Camera Gimbal Support Arm Optimization
+
+## Engineering Scenario
+
+**Application**: Professional aerial cinematography drone
+**Component**: Camera gimbal support arm
+**Goal**: Lightweight design for extended flight time while maintaining camera stability
+
+## Problem Statement
+
+The current production arm weighs **145g** and meets all requirements. Marketing wants to advertise "30% longer flight time" by reducing weight. Your task: optimize the arm geometry to minimize weight while ensuring it doesn't compromise camera stability or structural integrity.
+
+### Real-World Constraints
+
+- **Camera Payload**: 850g (camera + gimbal mechanism)
+- **Maximum Deflection**: 1.5mm (required for image stabilization systems)
+- **Material**: Aluminum 6061-T6 (aerospace grade)
+- **Safety Factor**: 2.3 on yield strength (276 MPa)
+- **Vibration Avoidance**: Natural frequency must be > 150 Hz to avoid coupling with rotor frequencies (80-120 Hz)
+
+## Multi-Objective Optimization
+
+This study explores the **trade-off between two competing objectives**:
+
+### Objectives
+1. **MINIMIZE Mass** - Every gram saved increases flight time
+   - Target: < 120g (17% weight savings)
+   - Current: 145g baseline
+
+2. **MAXIMIZE Fundamental Frequency** - Higher frequency = better vibration isolation
+   - Target: > 150 Hz (safety margin above 80-120 Hz rotor range)
+   - Trade-off: Lighter designs typically have lower frequencies
+
+### Constraints
+1. **Max Displacement** < 1.5mm under 850g load
+2. **Max von Mises Stress** < 120 MPa (Al 6061-T6 yield = 276 MPa, SF = 2.3)
+3. **Natural Frequency** > 150 Hz (hard requirement)
+
+## Design Variables (Parametric Beam Model)
+
+- **beam_half_core_thickness**: 20-30 mm (affects stiffness and weight)
+- **beam_face_thickness**: 1-3 mm (face sheets for bending resistance)
+- **holes_diameter**: 180-280 mm (lightening holes for weight reduction)
+- **hole_count**: 8-14 (number of lightening holes)
+
+## Expected Outcomes
+
+- **Pareto Front**: Shows designs on the optimal trade-off curve between mass and frequency
+- **Weight Savings**: 10-20% reduction from 145g baseline
+- **Constraint Analysis**: Clear visualization of which constraints are active/limiting
+- **Design Insights**: Understand how design variables affect both objectives
+
+## Study Configuration
+
+- **Protocol**: Protocol 11 (Multi-Objective Optimization)
+- **Sampler**: NSGA-II (genetic algorithm for Pareto fronts)
+- **Trials**: 30 (sufficient for Pareto front exploration)
+- **Duration**: ~1-2 hours (2-4 minutes per trial)
+
+## Files in This Study
+
+```
+drone_gimbal_arm_optimization/
+├── 1_setup/
+│   ├── model/
+│   │   ├── Beam.prt              # Parametric beam geometry
+│   │   ├── Beam_sim1.sim         # Simulation setup (needs modal analysis!)
+│   │   └── Beam_fem1.fem         # Finite element mesh
+│   ├── optimization_config.json   # Multi-objective config
+│   └── workflow_config.json       # Extractor definitions
+├── 2_results/                     # Created during optimization
+│   ├── study.db                   # Optuna database
+│   ├── optimization_history_incremental.json
+│   └── ...
+├── run_optimization.py            # Main execution script
+├── NX_FILE_MODIFICATIONS_REQUIRED.md  # IMPORTANT: Read this first!
+└── README.md                      # This file
+```
+
+## Before You Run
+
+**CRITICAL**: You MUST modify the NX simulation files before running.
+
+Read [NX_FILE_MODIFICATIONS_REQUIRED.md](NX_FILE_MODIFICATIONS_REQUIRED.md) for detailed instructions.
+
+**Summary of Required Changes**:
+1. Add **Modal Analysis Solution** (SOL 103) to extract natural frequencies
+2. Update **static load** to 8.34 N (850g camera payload)
+3. Verify **material** is Al 7075-T6
+
+## Running the Optimization
+
+```bash
+# Quick test (5 trials, ~10-20 minutes)
+cd studies/drone_gimbal_arm_optimization
+python run_optimization.py --trials 5
+
+# Full study (30 trials, ~1-2 hours)
+python run_optimization.py --trials 30
+
+# Resume existing study
+python run_optimization.py --resume
+```
+
+## Monitoring in Dashboard
+
+While optimization runs, monitor in real-time:
+
+1. **Start Dashboard Backend** (separate terminal):
+   ```bash
+   cd atomizer-dashboard/backend
+   python -m uvicorn api.main:app --reload --port 8000
+   ```
+
+2. **Start Dashboard Frontend** (another terminal):
+   ```bash
+   cd atomizer-dashboard/frontend
+   npm run dev
+   ```
+
+3. **Open Browser**: http://localhost:3003
+
+4. **Select Study**: drone_gimbal_arm_optimization
+
+### Dashboard Features You'll See
+
+- **Real-time trial updates** via WebSocket
+- **Pareto front visualization** (mass vs frequency scatter plot)
+- **Constraint violation tracking** (which trials failed which constraints)
+- **Progress monitoring** (30 trials total)
+- **New best notifications** when Pareto front expands
+
+## Interpreting Results
+
+### Pareto Front Analysis
+
+The Pareto front will show:
+- **Lower-left designs**: Lighter but lower frequency (more prone to vibration)
+- **Upper-right designs**: Heavier but higher frequency (better vibration isolation)
+- **Middle region**: Balanced trade-offs
+
+### Selecting Final Design
+
+Choose based on flight profile:
+- **Stable hovering flights**: Select lighter design (mass priority)
+- **Dynamic maneuvers**: Select higher frequency design (vibration priority)
+- **Balanced missions**: Mid-Pareto design
+
+### Constraint Active Check
+
+Look for designs where:
+- Displacement constraint is just satisfied (1.4-1.5mm) = efficient use of deflection budget
+- Frequency constraint is marginally above 150 Hz = not over-designed
+- Stress well below limit = safety margin confirmed
+
+## Why This is Realistic
+
+This scenario reflects real engineering trade-offs in aerospace:
+
+1. **Weight vs Performance**: Classic aerospace dilemma
+2. **Multi-Physics Constraints**: Static strength + dynamic vibration
+3. **Safety Margins**: Realistic stress limits with safety factors
+4. **Operational Requirements**: Specific to drone camera applications
+5. **Pareto Decision-Making**: No single "best" design, requires engineering judgment
+
+## Comparison with Bracket Study
+
+Unlike the bracket study (single objective), this study shows:
+- **Multiple optimal solutions** (Pareto set, not single optimum)
+- **Trade-off visualization** (can't optimize both objectives simultaneously)
+- **Richer decision support** (choose based on priorities)
+- **More complex analysis** (static + modal)
+
+## Next Steps After Optimization
+
+1. **Review Pareto front** in dashboard
+2. **Select 2-3 candidate designs** from different regions of Pareto front
+3. **Detailed FEA verification** of selected candidates
+4. **Fatigue analysis** for repeated flight cycles
+5. **Prototype testing** to validate predictions
+6. **Down-select** based on test results
+
+## Technical Notes
+
+- Uses **NSGA-II** multi-objective optimizer (Optuna)
+- Handles **3 constraints** with penalty methods
+- Extracts **4 quantities** from 2 different solutions (static + modal)
+- Fully automated - no manual intervention during run
+- Results compatible with all dashboard visualization features
+
+## Questions?
+
+This study demonstrates the full power of multi-objective optimization for real engineering problems. The Pareto front provides engineering insights that single-objective optimization cannot offer.
--- a/studies/drone_gimbal_arm_optimization/reset_study.py
+++ b/studies/drone_gimbal_arm_optimization/reset_study.py
@@ -0,0 +1,16 @@
+"""Reset drone gimbal arm optimization study by deleting database."""
+import optuna
+from pathlib import Path
+
+study_dir = Path(__file__).parent
+storage = f"sqlite:///{study_dir / '2_results' / 'study.db'}"
+study_name = "drone_gimbal_arm_optimization"
+
+try:
+    # Delete the study
+    optuna.delete_study(study_name=study_name, storage=storage)
+    print(f"[OK] Deleted study: {study_name}")
+except KeyError:
+    print(f"[WARNING] Study '{study_name}' not found (database may not exist)")
+except Exception as e:
+    print(f"[ERROR] Error: {e}")