Welch Teaching & Coaching · AP Physics C: Mechanics

Motion Detector & Gravity Inquiry

Two complementary virtual labs: reproduce motion graphs with a virtual walker, then estimate gravitational acceleration from noisy detector data using regression and uncertainty analysis.

Lab A · Match the Motion Detector

In the physical lab, you walk in front of a motion detector to make a graph. Here you control a virtual student's velocity, not position directly. Decide when to move toward the detector, away from it, stop, or turn around. Then compare your measured motion to the target.

Conventions: position is distance to the detector; positive velocity means moving away. A horizontal position–time line means stationary; a horizontal velocity–time line means constant velocity.

📡🚶
0 m1 m2 m3 m4 m5 m6 m
Elapsed time0.00 s
Detector distance2.00 m
Actual velocity0.00 m/s
Target match score (after run)—

Live motion graphs

Dashed gold = target; blue = recorded motion. The velocity and acceleration traces are estimated from noisy position measurements with local fitting, similar to real detector software.

Position · x(t)

Velocity · v(t)

Acceleration · a(t)

Prediction and graph interpretation

Complete a run to estimate slope and area.

Lab B · Find Gravity Using Motion Graphs

Model a ball tossed vertically upward in front of a motion detector. In this simulation upward is positive, so the true acceleration is approximately −9.81 m/s². The detector samples height at fixed time intervals; the graphing software estimates velocity from measured positions.

Investigation: predict the shape of the graphs, collect data, choose a fitting interval, use the slope of v(t) to estimate g, and independently fit y(t) to a quadratic. Compare results and discuss uncertainties.

Detector data and fitted models

Fit range is adjustable

Measured position y(t)

Estimated velocity v(t)

Collect a trial to obtain measurements and calculate regression results.
Important: regression standard errors quantify random scatter under the fitted model. They do not include calibration bias or model misspecification from drag. Velocity points are derived from overlapping position measurements, so their errors are correlated; ordinary least-squares standard errors are only approximate.
Time (s)Measured height (m)Estimated v (m/s)
No data yet

Repeated trials and uncertainty

The table stores fitted gravity magnitudes from separate data collections. Keep your acquisition settings and fitting interval fixed for repeatability comparisons.

Trialg from v slope (m/s²)g from quadratic (m/s²)SE of velocity fit (m/s²)
No fitted trials yet
Run and fit at least two independent trials to calculate the mean, standard deviation, and uncertainty of the mean.

AP Physics C · Derivatives, fitting and propagation

Assuming constant acceleration, fit y(t) = A + Bt + Ct². The fitted acceleration is a = 2C. From the velocity data, fit v(t) = mt + b and estimate a = m.

For a quadratic least-squares fit, propagate the uncertainty in its curvature coefficient: u(a) = 2u(C). For the linear fit: u(a) = u(m). The two estimates are correlated because they originate from the same position readings; don't combine them as independent measurements.

Extension · Are the fit residuals random?

Change the drag model while keeping random noise small. In real experiments, systematic curvature in the v(t) residuals suggests that constant acceleration is an incomplete model. Compare early and late fitting windows to identify model dependence. A calibration scale error can also shift both estimates without increasing random scatter.

Lab Notebook

Record evidence from both activities and export your responses. All calculations happen in your browser; the page does not submit student work to a server.

Motion detector evidence

No walk runs saved yet.

Gravity evidence

No gravity results saved yet.

Cross-lab synthesis

Next Step: Model Motion with Python

After finishing the lab, use your measured acceleration and time in a constant-acceleration model. Compare predicted velocity and displacement with your observations and explain the assumptions.

Predict first, build and run your program in the browser sandbox, then use the results to explain your model. Download your lab record when you finish.