Engineering Using AI: Practical Tools and Techniques

This module will guide you through using ChatGPT for performing engineering calculations, verifying results, and integrating AI into your workflows for faster, reliable answers. This step-by-step guide provides precise, click-by-click instructions, ideal for experienced engineers new to AI.

Introduction

Using AI in engineering calculations can streamline workflows and save time. However, accuracy is critical. In this module, you’ll learn how to use ChatGPT to perform specific engineering calculations, request complex calculations, verify results, and ensure that your AI-driven work meets the rigorous standards of engineering practice.

Note: ChatGPT is highly flexible. You can adjust prompts, calculations, and steps to fit your specific engineering requirements.

Learning Objectives

By the end of this module, you’ll be able to:

1. Accurately perform engineering calculations using ChatGPT.

2. Prompt ChatGPT for specific calculations and modify requests as needed.

3. Verify AI-generated results for compliance with engineering standards.

4. Use Python code provided by ChatGPT for complex calculations.

5. Present organized results using industry-standard documentation.

Module Steps

Step 1: Log in to ChatGPT

1. Open your browser and go to https://chat.openai.com.

2. Log in to your OpenAI account if prompted. If you don’t have an account:

   • Click "Sign Up" and follow the on-screen instructions.

3. Ensure you’re in ChatGPT mode. This is the interface where you’ll enter questions and requests for calculations.

Step 2: Begin with a Simple Calculation Request

Example Task: Start with a basic engineering calculation, such as calculating flow rate.

1. Type your request in ChatGPT’s prompt area in clear terms:

   Calculate the flow rate given a pipe diameter of 0.5 meters and a fluid velocity of 2 meters per second.

2. Press Enter. ChatGPT will respond with a step-by-step solution, including the formula and calculation details.

3. Review ChatGPT’s Response:

   • Confirm ChatGPT provided the correct formula and applied it with your specified input values.

   • Example output:

     Flow rate (Q) = Area × Velocity

     With pipe diameter = 0.5 m and velocity = 2 m/s, Q = π(0.5/2)^2 * 2 ≈ 0.393 m³/s.

4. Ask for Further Clarification if needed:

   • Type:

     Explain why you used this formula for flow rate.

   • Press Enter for ChatGPT’s explanation.

Step 3: Requesting Complex Calculations and Using Python Code

For more detailed calculations, ask ChatGPT to generate Python code. This allows you to verify results within a coding environment.

Example Task: Calculate the pressure drop across a pipe using Python code.

1. Enter the following prompt in ChatGPT:

   Write Python code to calculate pressure drop across a pipe with length 50 meters, diameter 0.3 meters, and flow rate 3 m³/s.

2. Press Enter. ChatGPT will respond with a Python code snippet.

3. Copy the Code:

   • Highlight the code, right-click, and select Copy.

4. Run the Code in Python:

   • Open a Python IDE (e.g., Jupyter Notebook or VS Code).

   • Paste the code into a new cell or script file.

   • Run the code to see the result.

5. Verify and Modify the Code:

   • Confirm that the output matches your engineering expectations.

   • For adjustments, return to ChatGPT.
     Type:

     Modify the code to include fluid density of 1000 kg/m³.

   • Press Enter to receive the modified code.

Step 4: Ensuring Units Consistency

1. Check Units in ChatGPT’s Calculation:

   • If ChatGPT doesn’t specify units, request clarification:

     What units were used in this calculation?

2. Request Unit Conversion Assistance if needed:

   • Type:

     Convert the flow rate result to liters per second.

   • Press Enter to get the converted result with units.

3. Document Units in your final output to ensure engineering accuracy.

Tip: Specifying units clearly aligns calculations with industry standards, reducing the risk of misinterpretation.

Step 5: Cross-Verification with ChatGPT

1. Verify Calculations with Different Inputs:

   • To ensure consistency, repeat the calculation with slightly modified values.

   • Type:

     Calculate the pressure drop again with a pipe length of 55 meters.

2. Compare Results:

   • Check if ChatGPT’s results are consistent across different inputs.

   • Consistent results confirm that the calculations are reliable and accurate.

Step 6: Using ChatGPT for Explanations and Cross-Checks

For thorough understanding and verification, ChatGPT can also clarify methods and provide cross-verification options.

1. Ask for Explanations on Specific Formulas:

   • Example prompt:

     Explain why the Reynolds number is calculated using this specific formula.

     
     

2. Request Alternative Calculation Methods:

   • If you want to verify an output with a different approach, ask:

     Can you calculate the flow rate using an alternative formula?

     
     

3. Run Cross-Check Calculations in Python:

   • Generate Python code if you need additional verification of complex calculations:

     Write Python code to recalculate the flow rate with modified variables.

Additional Information for Engineers New to AI

What AI Can Offer to Engineering Calculations

1. Automation of Repetitive Calculations: AI, like ChatGPT, can quickly handle repetitive tasks like unit conversions, basic calculations, and reformatting data.

2. Enhanced Documentation: AI can generate step-by-step calculation details and even explain complex concepts in simple terms, helping you understand and document each calculation effectively.

3. Flexible Access to Information: AI allows engineers to access a range of information, from theoretical explanations to industry standards, without consulting multiple reference materials.

4. Time Savings: For tasks that require extensive calculation or documentation, AI can reduce the time spent by providing instant answers, formatting, and even code for verification.

Understanding AI Limitations in Engineering

1. Reliance on Prompts: AI’s output quality depends on how the request (or "prompt") is phrased. It’s essential to be clear, specific, and accurate with input values and units for best results.

2. Lack of Specialized Engineering Knowledge: While AI is trained on a broad knowledge base, it doesn’t specialize in specific engineering domains. AI might not always apply the correct standard or formula without proper guidance.

3. No Real-Time Validation: AI doesn’t have a real-time feedback loop. It performs calculations based on static inputs, so it doesn’t have self-checking or validation mechanisms typical in engineering software.

4. Understanding Assumptions and Context: AI might assume certain standard conditions or contexts (e.g., atmospheric pressure or temperature in °C). It’s important to specify if different conditions apply.

Best Practices When Using ChatGPT for Engineering Calculations

1. Start with Clear, Simple Prompts: Begin by asking ChatGPT for straightforward calculations to ensure clarity. For example, instead of asking for a "thermal analysis," specify each component:

   Calculate the thermal conductivity required for a heat exchanger with input values: area = 5 m², temperature gradient = 20 °C.

   
   

2. Break Down Complex Calculations: For multi-step engineering tasks, break the problem down into smaller steps and prompt ChatGPT for each one.

3. Always Double-Check Outputs: AI outputs should be validated by performing the same calculation manually or using known engineering software.

4. Request Explanations for Each Step: Especially when working with new or complex formulas, prompt ChatGPT to explain its approach.

5. Request Code for Verification: For calculations that require exact values or precision, ask ChatGPT to generate Python code, which you can run in your preferred coding environment for independent verification.

Quick Checklist for Accurate Engineering Calculations with ChatGPT

• Specify values and units clearly in each request.

• Review each calculation step to verify that the correct formula and input values were used.

• Use Python code for complex calculations and verify outputs in a Python environment.

• Ask for explanations to understand the applied methods and clarify any assumptions made.

• Perform cross-checks by recalculating with varied inputs.

Congratulations on completing this module! You have learned to use ChatGPT for essential engineering calculations, prompt effectively for specific tasks, verify outputs, and ensure accuracy across calculations. By integrating these methods into your workflow, you can streamline your processes and reduce manual calculation efforts while maintaining engineering standards.

Remember: ChatGPT is a flexible tool. Feel free to modify these steps and prompts to fit your specific engineering needs, calculation requirements, or workflow preferences. As you become more comfortable with AI-driven calculations, adapt the processes to enhance efficiency in your unique projects.

Objective:

Introduce engineers to ChatGPT, covering how to log in, navigate, and create their first effective prompt for a basic engineering calculation. By the end of this module, they’ll be able to access ChatGPT, understand its interface, and craft simple, targeted prompts for engineering tasks.

Section 1: Setting Up Your ChatGPT Account

1. Creating an OpenAI Account:

   • Step 1: Open your web browser (Chrome, Firefox, Edge).

   • Step 2: Go to https://www.openai.com.

   • Step 3: On the top-right corner, click on Sign Up.

   • Step 4: Enter your work email address or, if you prefer, use a Google or Microsoft account to sign in faster.

   • Step 5: Follow the prompts to complete the registration process, including verifying your email.

2. Logging into ChatGPT:

   • Step 1: After creating your account, go to https://chat.openai.com.

   • Step 2: Click Log In.

   • Step 3: Enter your email and password or log in with Google/Microsoft.

   • Step 4: You’re now on the ChatGPT main interface, where you’ll see a blank chat screen ready for prompts.

Section 2: Understanding the ChatGPT Interface

1. Interface Overview:

   • Left Panel: This panel shows past conversations. You can click New Chat at the top to start a fresh session.

   • Main Chat Window: This is where you type your questions or prompts and where ChatGPT replies.

   • Settings and Account Information: Click your profile in the bottom-left corner to adjust settings or log out.

2. Choosing a Model (If Available):

   • Some accounts may have access to different models (e.g., GPT-4). If prompted, select GPT-4 for the best engineering support.

Section 3: Crafting Your First Prompt for an Engineering Task

This section will guide you in creating your first prompt using a pipe flow calculation as a practical example.

Example Objective: Calculate the Pressure Drop in a Pipeline

1. Identifying What You Need:

   • To calculate a pressure drop, we need:

     • Fluid properties (density and viscosity)

     • Pipeline diameter and length

     • Flow rate or velocity

2. Writing the Prompt:

   • Step 1: In the chat window, type the following prompt exactly as shown:

     I need to calculate the pressure drop in a pipeline. The pipeline is 50 meters long with an internal diameter of 0.1 meters. The fluid has a density of 850 kg/m³ and viscosity of 0.001 Pa·s. The flow rate is 3 m³/h. Could you guide me through the steps to calculate this?

     
     

   • Step 2: Press Enter to submit your prompt.

3. Analyzing ChatGPT’s Response:

   • ChatGPT will likely start by explaining the formula for pressure drop in a pipe (Darcy-Weisbach equation) and might even perform the calculation.

   • Tip: If the response is unclear or incorrect, type “Can you clarify the calculation steps?” to get a clearer breakdown.

Section 4: Refining Prompts for Accurate Responses

As you start using ChatGPT, you may need to refine your prompts to get better answers. Here’s a quick guide on how to do this.

1. Be Specific with Units and Conditions:

   • When entering variables, specify units. For example:

     Calculate the Reynolds number for a fluid with density 850 kg/m³, viscosity 0.001 Pa·s, flowing at 3 m/s in a 0.1 m diameter pipe.

     
     

2. Request Step-by-Step Solutions:

   • For detailed explanations, add “step-by-step” in your prompt:

     Provide a step-by-step calculation for the pressure drop in the pipeline with the conditions given.

     
     

3. Clarify Engineering Standards or Sources if Relevant:

   • You can add references to known standards (e.g., “According to API 520”) to prompt ChatGPT to consider relevant guidelines:

     Using API 520 as a reference, explain the selection criteria for a pressure relief valve in liquid service.

     
     

4. Dealing with Incorrect or Incomplete Responses:

   • If the response seems off, use clarifying prompts like:

     Double-check the units and provide the formula used for the pressure drop calculation.

Section 5: Practice Exercise

Try a few variations to reinforce the process. Here are some sample prompts with instructions:

1. Example 1: Sizing a Heat Exchanger

   • Prompt: Type:

     I need to size a heat exchanger for cooling 500 kg/h of oil from 120°C to 80°C. The oil’s specific heat capacity is 2 kJ/kg°C, and the cooling water available is at 30°C. Could you guide me through the steps to calculate the required heat exchanger area?

   • Action: Review ChatGPT’s steps and check if it includes heat transfer formulas and temperature differences.

2. Example 2: Calculating Pump Power

   • Prompt: Type:

     I need to calculate the brake horsepower for a pump with a flow rate of 10 m³/h, a discharge pressure of 300 kPa, and an efficiency of 70%. Can you show me the steps?

   • Action: If the calculation isn’t detailed enough, follow up with, “Please include the formula and all substitution steps.”

3. Example 3: Estimating Reynolds Number

   • Prompt: Type:

     Calculate the Reynolds number for water flowing at 2 m/s in a 0.1 m diameter pipe. The density of water is 1000 kg/m³, and its viscosity is 0.001 Pa·s.

   • Action: Observe ChatGPT’s approach and ask, “What does this Reynolds number indicate about the flow regime?” to learn more.

Section 6: Troubleshooting and Getting Help

If ChatGPT’s responses are confusing, try the following steps:

1. Rephrase or Simplify Prompts:

   • If a response isn’t clear, simplify the question. For example, change:

     Provide the detailed steps for calculating the hydraulic power of a pump with given parameters.

     to:

     Calculate the hydraulic power for a pump with a flow rate of 10 m³/h and a discharge pressure of 200 kPa.

2. Use Follow-Up Prompts to Request Clarifications:

   • If needed, ask ChatGPT to “explain further” or “provide more detail on the last calculation.”

3. Accessing OpenAI Help Resources:

   • If you encounter issues with your OpenAI account, go to OpenAI Help for FAQs or support requests.

Module Recap and Key Takeaways

• Log in to ChatGPT and get comfortable with the interface.

• Formulate clear, specific prompts for calculations or engineering inquiries.

• Use step-by-step language in prompts for detailed responses.

• Troubleshoot by rephrasing prompts and asking follow-up questions if the output is unclear.

With this structure, engineers will have a guided, hands-on experience from logging in to refining prompts for specific engineering tasks, preparing them to use ChatGPT effectively in their work.

Learning Objective:

By the end of this module, engineers will be able to design and configure custom GPTs tailored to specific engineering tasks, such as calculations and data analysis, using best practices for clarity, accuracy, and adherence to industry standards. They’ll gain the skills to define precise instructions, embed engineering knowledge, and refine outputs to streamline technical workflows.

In this module, you’ll learn how to use ChatGPT to create and modify Excel documents effectively, using detailed, step-by-step instructions. This approach will allow you to utilize AI to assist in generating data tables, organizing information, and editing Excel content, while still having full control over your spreadsheet.

Introduction

Excel is a powerful tool for organizing engineering data, but making modifications manually can be time-consuming. ChatGPT can help streamline this process by generating and formatting data, organizing calculations, or even modifying spreadsheet structure upon your request. This guide will help you use ChatGPT to support your Excel needs efficiently.

Learning Objective

By the end of this module, you will be able to create, modify, and format Excel documents using ChatGPT by generating data and formatting instructions that you can easily apply in Excel.

Step 1: Log In to ChatGPT

1. Open your web browser and go to https://chat.openai.com.

2. Log in to your OpenAI account if prompted.

3. Once logged in, ensure you are in ChatGPT mode, ready to provide input and receive responses.

Step 2: Open Excel and Prepare to Use ChatGPT

1. Open Excel on your computer. You can do this by searching “Excel” in your start menu or applications folder.

2. Open a new workbook or an existing document you want to modify.

Step 3: Ask ChatGPT to Generate an Excel Table

1. Return to ChatGPT and type in your request to create a specific table. For example:

   • Type: “Create a table in Excel with columns for Date, Temperature (°C), Pressure (kPa), and Flow Rate (m³/hr). Populate with 5 sample rows.”

2. Press Enter. ChatGPT will generate a table format and sample data.

3. Copy the Table Output:

   • Highlight the table data ChatGPT provides.

   • Right-click and select Copy.

Step 4: Paste Data into Excel

1. Go to Excel and click into the first cell where you want the table to start (e.g., cell A1).

2. Right-click and select Paste or press Ctrl + V (Windows) or Cmd + V (Mac).

3. Adjust column widths if needed:

   • Click and drag the right edge of each column header to resize.

Step 5: Modify Existing Data in Excel Using ChatGPT

1. Identify the modification you need. For example, if you want to increase all values in the “Pressure” column by 10%.

2. Return to ChatGPT and type: “Provide a formula to increase each value in the Pressure column by 10%.”

3. Press Enter. ChatGPT will suggest a formula, such as =B2*1.1 (assuming Pressure data is in column B).

4. Copy the formula provided by ChatGPT.

5. Apply the Formula in Excel:

   • Go to the first cell in the Pressure column (e.g., cell B2).

   • Paste the formula by clicking into the cell and pressing Ctrl + V.

   • Press Enter to see the adjusted value.

   • Click and drag the fill handle (small square at the bottom-right corner of the cell) down the column to apply the formula to other rows.

Step 6: Use ChatGPT for Data Formatting Instructions

1. Ask ChatGPT for formatting help. For example, type: “How do I format the Temperature column to show only one decimal place?”

2. Press Enter. ChatGPT will provide a step-by-step answer, such as:

   • Select the cells in the Temperature column.

   • Right-click and select Format Cells.

   • Go to the Number tab, select Number, and set Decimal places to 1.

3. Follow these instructions in Excel to format your data as needed.

Step 7: Save and Organize Your Work

1. Save Your Excel File:

   • Go to File > Save As to save the document with a relevant file name.

2. Organize Files:

   • Choose a designated folder or drive to store files securely.

Quick Checklist for Excel Document Creation and Modification

• Generate data in ChatGPT and paste it directly into Excel.

• Use formulas suggested by ChatGPT to automate calculations.

• Apply formatting as directed to keep data consistent.

• Save and organize your documents for easy access.

By following these steps, you can streamline Excel tasks using ChatGPT, enabling efficient data entry and modification. Leveraging AI for Excel can reduce manual effort, allowing you to focus on higher-level engineering tasks.

In this module, we’ll walk you through using Python libraries to ensure accurate calculations when using AI, like ChatGPT, for engineering tasks. You’ll learn how to prevent errors known as “calculation hallucinations”—incorrect outputs generated by AI when doing complex math—by incorporating reliable Python code to double-check calculations.

Introduction

When using AI for engineering calculations, especially for critical values, it’s essential to validate results. Inaccurate or "hallucinated" numbers can lead to costly mistakes. By integrating Python libraries, you can ensure accuracy, performing calculations outside of AI and pasting the results back for discussion.

Learning Objective

By the end of this module, you’ll be able to use Python libraries in ChatGPT to run verified calculations, preventing calculation hallucinations and ensuring engineering precision.

Step 1: Access ChatGPT’s Code Interpreter (Python Mode)

1. Open your web browser and go to https://chat.openai.com.

2. Log in to your OpenAI account if prompted.

3. In the main ChatGPT interface, select the GPT-4 model (if available) and check if Code Interpreter or Python Mode is enabled.

   • Note: Not all versions of ChatGPT support Python directly. If unavailable, skip to the “Using a Local Python Environment” section below.

Step 2: Setting Up Python Libraries for Accurate Calculations

1. In the chat window, type “Enable Python” if the feature is available.

2. Once Python is enabled, you can type Python code directly into ChatGPT.

Step 3: Install and Import Python Libraries for Calculations (If Working Locally)

If you’re running Python locally, you’ll need to install libraries directly on your computer.

1. Open a Python IDE like Jupyter Notebook, VS Code, or Spyder (installed via Anaconda if not available on your system).

2. Install essential libraries by typing the following commands (one at a time) into the terminal:

   bash

   Copy codepip install numpy

   pip install scipy

   pip install pandas

   
   

Step 4: Performing a Sample Calculation Using Python in ChatGPT

1. In ChatGPT’s chat window, enter the following code for a sample calculation:

   python

   Copy codeimport numpy as np

   # Example: Calculate the root of a quadratic equation

   a = 1

   b = -3

   c = 2

   # Calculate the discriminant

   discriminant = b**2 - 4*a*c

   # Calculate roots

   root1 = (-b + np.sqrt(discriminant)) / (2*a)

   root2 = (-b - np.sqrt(discriminant)) / (2*a)

   root1, root2

   
   

2. Press Enter to run the code. ChatGPT will return the result, giving you the verified roots of the equation.

Step 5: Validating Engineering Calculations with Python Libraries

For more complex engineering tasks, use trusted Python libraries like scipy for advanced math. Here’s an example using scipy for a trigonometric calculation:

1. Enter the following code for trigonometric calculations in ChatGPT:

   python

   Copy codefrom scipy import constants

   import numpy as np

   # Example: Calculate the sine of 45 degrees

   angle_degrees = 45

   angle_radians = np.deg2rad(angle_degrees)

   sine_value = np.sin(angle_radians)

   sine_value

   
   

2. Press Enter and review the result. ChatGPT will display the sine of 45 degrees, calculated using reliable Python methods.

Step 6: Run Validated Code in ChatGPT (if Available)

1. Type your engineering calculation in Python code in ChatGPT, ensuring it aligns with your project's needs.

2. Click “Run” or press Enter to execute.

3. Review the output. Copy the verified result into your document or calculation sheet.

Step 7: Document Results and Verify Consistency

1. Copy the result from ChatGPT or your local Python environment.

2. Paste it into your report or worksheet, labeling the calculation to indicate it was verified with Python.

3. Always double-check units and constants to confirm alignment with your project’s specifications.

Quick Checklist to Avoid Calculation Hallucinations with Python

• Enable Python Mode in ChatGPT or use a local environment for calculations.

• Install trusted libraries like numpy and scipy.

• Run code directly in Python rather than relying on AI for complex math.

• Document results clearly, showing Python was used for verification.

By using Python libraries for calculations, you ensure accuracy and prevent AI-generated errors. This process provides you with reliable, documented results, allowing AI to support, not risk, your engineering work.