Concept Visualization in Thermodynamics

Taxonomy Dimensions

• Primary Purpose: Conceptual Understanding, Visualization
• Integration Depth: Guided Integration
• Student Agency: Scaffolded Autonomy
• Assessment Alignment: Meta-Learning, Comparative Analysis
• Technical Implementation: Prompt Engineering, Error Management
• Ethics & Professional Development: Critical AI Literacy

Course Context

Junior-level thermodynamics course where students struggle with connecting abstract thermodynamic concepts (entropy, availability, exergy) to real-world applications and visualizing microscopic phenomena.

Implementation Description

Activity Overview

Students use AI tools to generate multiple visual and conceptual representations of thermodynamic principles, then analyze and refine these representations to develop deeper understanding.

Step-by-Step Implementation

1. Preparation Phase:

   • Instructor creates a library of effective prompt templates focused on thermodynamic concepts
   • Students receive training on effective prompting for scientific concepts
   • Example concepts: entropy generation, availability, second law analysis

2. Guided Exploration (Week 1):

   • Students use provided prompt templates to generate multiple explanations of a complex concept
   • Each student requests 3-4 different explanations/analogies of the same concept
   • Students analyze differences between explanations and identify strengths/weaknesses

3. Visual Representation (Week 2):

   • Students use image generation tools to create visual representations of microscopic phenomena
   • Compare AI-generated visualizations with textbook representations
   • Identify misconceptions or inaccuracies in the generated visuals

4. Conceptual Integration (Week 3):

   • Students generate explanations connecting microscopic and macroscopic phenomena
   • Apply principles to specific engineering applications (power plants, refrigeration, etc.)
   • Develop their own analogies with AI feedback

5. Reflection and Synthesis:

   • Students document their learning journey through the different representations
   • Create a concept map integrating multiple perspectives
   • Evaluate which representations most enhanced their understanding

Example Prompts

Initial Concept Explanation Prompt

Explain the concept of entropy generation in a thermodynamic system in 5 different ways:
1. Using a statistical mechanics perspective
2. Using an energy quality/degradation perspective
3. Using a probability and states perspective
4. Using a real-world engineering application
5. Using an analogy accessible to a non-engineer

For each explanation, include key equations, identify the core principles, and explain how this perspective connects to the second law of thermodynamics.

Visual Representation Prompt

Create a detailed visual representation showing entropy generation during heat transfer across a finite temperature difference. 

The visualization should:
1. Show molecular-level phenomena
2. Illustrate energy states before and after the process
3. Represent the irreversibility visually
4. Include a system boundary
5. Indicate how this relates to the second law of thermodynamics

Create this as a side-by-side comparison of reversible vs. irreversible heat transfer.

Conceptual Connection Prompt

Analyze the relationship between entropy generation and lost work potential (exergy destruction) in:
1. A heat exchanger with a 50°C temperature difference
2. A throttling valve in a refrigeration system
3. Fluid flow with friction in a pipe

For each case, explain:
- The microscopic mechanism of entropy generation
- How to quantify the exergy destruction
- Engineering approaches to minimize the irreversibility
- Real-world consequences of this exergy destruction

Assessment Strategies

1. Process Documentation (30%):

   • Students maintain a structured journal documenting their prompt iterations
   • Record conceptual evolution and insights gained from different representations
   • Include reflection on which AI-generated explanations were most helpful and why

2. Concept Mapping Assessment (30%):

   • Students create a comprehensive concept map integrating multiple perspectives
   • Map must show connections between microscopic and macroscopic phenomena
   • Include practical applications and relate to core equations

3. Peer Teaching Activity (20%):

   • Students use their developed understanding to teach peers
   • Create original explanations and analogies based on AI-assisted learning
   • Evaluated on clarity, accuracy, and creativity

4. Traditional Problem-Solving (20%):

   • Closed-book assessment without AI access
   • Problems require application of thermodynamic principles
   • Questions emphasize conceptual understanding over calculation

Implementation Considerations

Required Resources

• Access to ChatGPT, Claude, or similar LLM
• Image generation tool (DALL-E, Midjourney) for visualizations
• Template library for effective thermodynamics prompts
• Learning management system for documentation

Common Challenges

• Students may accept AI explanations without critical evaluation
• Scientific accuracy of generated visualizations varies
• Different AI tools have varied capabilities for technical content
• Some students may rely too heavily on AI without developing personal understanding

Integration Tips

• Start with highly structured prompts then gradually reduce scaffolding
• Incorporate verification against textbook sources as a required step
• Focus assessment on the analysis and refinement process, not just final outputs
• Pair with hands-on lab activities that demonstrate the same concepts physically

Faculty Experience Required

• Basic familiarity with AI prompt engineering
• Ability to evaluate scientific accuracy of AI-generated content
• Understanding of common misconceptions in thermodynamics

---

This example was developed as part of the "Strategies for Integrating Generative AI in Engineering Education" workshop materials.