EPQ Ideas for Engineering

What makes a strong engineering question

• One thing changes, one thing is measured. If three things change at once, you cannot say which one mattered.
• You can repeat it. Five repeats of a cheap test teach you more than one repeat of an expensive one, because the spread of your results is evidence too.
• A simple model makes a prediction. The project is the comparison between what the model says and what you measured.
• It fits the time you have. Your school's EPQ coordinator sets the deadlines and supervises the project, so agree the question with them early.

Six illustrative ideas

These are examples to show the shape of a good question. They are illustrations, not projects we have run or results anyone has reported.

• Mechanical: How does the shape of a bottle rocket's nose change how high it flies? Measure launch height for three nose shapes, five launches each, timed from video. Why it works: one variable, cheap to repeat, and the scatter between launches is part of the answer. Launch only outdoors in a clear area, with an adult or at school, using a launcher made for water rockets and with eye protection on.
• Aerospace: How much does a paper wing's camber change its lift at a fixed air speed? Measure the force on a wing held in a fan's airflow, read from a kitchen scale. Why it works: it forces you to control the air speed, which is the hard part of every wind tunnel test.
• Civil: Which truss shape carries the most load per gram of material? Measure the load at failure of lollipop stick trusses of equal mass. Why it works: you compare designs fairly only if you fix the mass first, a habit engineers use every day.
• Chemical: How does stirring speed change how fast sugar dissolves? Measure time to dissolve at four stirring speeds, with temperature held steady. Why it works: it is mass transfer, the heart of chemical engineering, in a kitchen.
• Electrical: How does a solar panel's output fall as it tilts away from the light? Measure voltage and current across a fixed resistor at angles from 0 to 80 degrees. Why it works: a cosine model makes a clear prediction that your data can confirm or contradict.
• Materials: How does the width of a strip of card change the load at which it buckles? Measure the buckling load for five widths, compared with what a simple model predicts. Why it works: the model will be wrong in an interesting way, and explaining why is the project.

Turning an idea into a project

1. Write the question in one sentence, with the one thing you will change.
2. Predict the answer from a simple model before you measure anything.
3. Run a small pilot test, then fix whatever went wrong before the real runs.
4. Record every reading, including the ones that look wrong, and say what you think caused them.
5. Compare the results with your prediction, and spend most of your write up on why they differ.

Where we can help

If you want someone to talk the question through with you as it develops, our project mentorship pairs you with a mentor for a project of your own. Your school still supervises and assesses any EPQ.

Keep going

• ESAT practice: what the test covers and how to prepare
• Oxbridge interview questions, all subjects