
measure position against time for a cart on a ramp with phone video, show by linearizing that distance grows with time squared, and compare the measured acceleration with g sin θ using an uncertainty estimate.
Instruction
Why a ramp? A car rolling toward a crosswalk on a downhill street speeds up even with the engine off. A ramp slows this motion down enough to measure it with a phone. Today you test a prediction: if acceleration is constant, distance from rest grows with the square of time.
Where the prediction comes from. An object starting from rest with constant acceleration a has velocity v = at. Its velocity-time graph is a straight line through the origin, and the area under it after time t is a triangle: ½(t)(at) = ½at2. So x = ½at2, and x ∝ t2. Doubling the time should give four times the distance. Tripling it should give nine times. Galileo noticed the same pattern: in equal successive time intervals an object starting from rest covers distances in the ratio 1 : 3 : 5 : 7.
Measuring time with video. A phone records a fixed number of frames per second (fps). If the cart is released at frame 0 and crosses a tape mark at frame n, then t = n/fps. At 30 fps each frame is 1/30 s = 0.033 s. Slow-motion modes (often 120 or 240 fps) shrink that step. Check the real frame rate of your file, because some phones vary it. Your timing uncertainty is about ±1 frame, which is much better than a hand stopwatch, where the reaction time of the person pressing the button is typically 0.1-0.3 s.
Worked example (illustrative data). A toy car on a 2.00 m board raised 0.15 m at one end, filmed at 30 fps:
| x (m) | frame | t (s) | t2 (s2) |
|---|---|---|---|
| 0.20 | 23 | 0.767 | 0.588 |
| 0.40 | 33 | 1.100 | 1.210 |
| 0.80 | 46 | 1.533 | 2.351 |
| 1.20 | 56 | 1.867 | 3.484 |
| 1.60 | 65 | 2.167 | 4.694 |
A graph of x against t curves upward, which is hard to test by eye. So we linearize: graph x against t2. If x = ½at2, this graph is a straight line through the origin with slope ½a.
A least-squares line through the origin gives slope 0.341 m/s2, so a = 2 × 0.341 = 0.68 m/s2. The points lie close to the line, which supports constant acceleration.
The model to compare with. On a frictionless incline at angle θ, only the component of gravity along the slope speeds the object up, so a = g sin θ. (You will derive this from forces in Unit 2.) Here sin θ = 0.15/2.00 = 0.075 (θ ≈ 4.3°), so the model predicts 9.8 × 0.075 = 0.735 m/s2. The measurement is 7.1% lower. That is a physical result, not only an error. Friction in the wheels and axles and air resistance both reduce the acceleration, and they always act in that direction.
Misconception: a ball is not a cart. A solid ball rolling without slipping has to spin up as well as speed up, so its acceleration is only (5/7)g sin θ, about 0.525 m/s2 on this ramp. Use a wheeled cart or toy car if you want to test g sin θ, or state the ball model if you use a ball.
Uncertainty. One frame is 0.033 s. At the first mark (0.767 s) that is about 4% of the time, but at the last mark only about 1.5%. That is why the longest distances carry the most reliable information. Report your acceleration with an uncertainty range, and repeat each run at least three times.
📖 Read for this lesson
The reading opens over the lesson; close it to come back to exactly where you were. The checks below draw on it.
Lab: Phone-video ramp lab: does distance grow with time squared?
Question. Does a cart released from rest on a ramp travel a distance proportional to the square of the elapsed time, and how does its measured acceleration compare with the model a = g sin θ?
Materials
- A smooth board, shelf or two taped meter sticks about 1.5-2.0 m long as a track
- Books or blocks to raise one end 10-20 cm
- A low-friction toy car or dynamics cart
- Masking tape, a marker and a meter stick or tape measure
- A phone or tablet camera (slow-motion optional) and any free video player that steps frame by frame
- A folded towel as a soft stop; a stopwatch as a backup timer
- Put a folded towel or box at the bottom of the ramp so the cart cannot roll off the table; keep the floor and walkway clear of the ramp and cords.
- Stack books on a stable table, not on chairs, and do not stand on furniture to film; keep fingers out of the cart’s path at the bottom.
- Keep phones on a stable surface or tripod away from the table edge.
Procedure
- Set up the track and raise one end. Measure the track length L along the slope and the height h of the raised end above the low end, and record both.
- Put tape marks at 0.20, 0.40, 0.80, 1.20 and 1.60 m from the release line (use shorter spacings if your track is shorter). Label each mark clearly so it is readable on video.
- Place the phone so the whole track is in view, side-on and as far back as practical, to reduce parallax. Start recording, then release the cart from rest at the release line without pushing.
- Repeat for at least three runs. Step through each video frame by frame. Record the frame number when the cart is released and when its front edge crosses each mark, and find the frame rate of the file (in the file’s details, or by filming a stopwatch for 10 s).
- Compute t = (frame − release frame)/fps for each mark and average the three runs. Compute t2.
- Graph x against t and then x against t2 (paper or a free spreadsheet). Fit a straight line through the origin to x against t2 and find a = 2 × slope.
- Change the height (for example double it), predict the new acceleration from g sin θ, then measure it with one more set of runs.
Data
| Mark x (m) | Frame, run 1 | Frame, run 2 | Frame, run 3 | Mean t (s) | t2 (s2) |
|---|---|---|---|---|---|
| 0.20 | |||||
| 0.40 | |||||
| 0.80 | |||||
| 1.20 | |||||
| 1.60 | |||||
| Illustrative sample, one run at 30 fps, L = 2.00 m, h = 0.15 m: x = 0.20, 0.40, 0.80, 1.20, 1.60 m at frames 23, 33, 46, 56, 65 (t = 0.767, 1.100, 1.533, 1.867, 2.167 s). Fitted slope 0.341 m/s2, so a = 0.68 m/s2; model g sin θ = 0.735 m/s2. | |||||
Analysis
- Claim-evidence-reasoning: is the cart’s acceleration constant? Use the shape of your x-t graph and how straight your x-t2 graph is as evidence, and x = ½at2 as your reasoning.
- Compute the percent difference between your measured a and g sin θ. Is your measurement above or below the model, and which physical effects explain the direction of the difference?
- Estimate the uncertainty in each time from ±1 frame. Which mark gives the least reliable time, and how would you change the setup to reduce the uncertainty?
- Name two further sources of error (for example parallax, the cart’s starting position, or a warped board) and say whether each would make a too large or too small.
- Did doubling h double the acceleration? Explain using sin θ and your data.
If you cannot do this lab in person: Choose one. All three earn full credit and end in the same two graphs, the same value of a with an uncertainty range, and the same comparison with g sin θ. Option A — measure your own ramp. The procedure above, with a board, a toy car and a phone. If all you have is a sloped driveway or hallway and a rolling can or ball, the same method works, but compare with the rolling model (5/7)g sin θ rather than g sin θ, and say so in your analysis. Option B — the two ramp runs recorded in this lesson. No equipment and no internet needed. Both runs use the same 2.00 m board filmed at 30 fps, with tape marks at 0.20, 0.40, 0.80, 1.20 and 1.60 m. The frame numbers are illustrative teaching data, not real measurements. Run set 1, h = 0.15 m: frames 23, 33, 46, 56, 65. Run set 2, h = 0.30 m: frames 16, 23, 32, 40, 46. For each set, convert every frame number to t = frame/30, square it, graph x against t2, fit a line through the origin and take a = 2 × slope. You should get about 0.68 m/s2 for set 1 and about 1.36 m/s2 for set 2, against models g sin θ of 0.735 and 1.47 m/s2: both about 7% low, and the acceleration doubles when the height doubles. Answer every analysis question from these two sets, using ±1 frame as your timing uncertainty. Option C — PhET Projectile Motion. The PhET simulation Projectile Motion, at phet.colorado.edu/sims/html/projectile-motion/latest/projectile-motion_all.html, lets you turn air resistance off and set the initial speed to 0 m/s, so the object simply falls. Set the cannon height to 2, 4, 6, 8 and 10 m in turn, fire, and read the time from the sim’s time readout (or use its timer tool). This is the same experiment on the steepest possible ramp, θ = 90°, so the model is a = g sin 90° = g. Graph height against t2, fit a line through the origin, and check that 2 × slope comes out near 9.8 m/s2. Because this sim has no friction, expect your percent difference to be near zero; say what that tells you about where the 7% in the ramp data comes from, and use the smallest time step the sim will show you as your timing uncertainty instead of one video frame. Cite the sim and the date you used it.
Formative check
Work through these before moving on. They are not graded — they tell you whether the standard below has landed yet.
A phone films at 30 fps. The cart is released at frame 0 and crosses a mark at frame 45, so t = s. If the video were shot at 240 fps, one frame would last about s.
Your graph of x against t2 is a straight line through the origin with slope 0.40 m/s2. What does this tell you?
A solid ball rolling down a ramp should have an acceleration of exactly g sin θ, so any smaller value is measurement error.
Write your lab analysis: state your claim about whether x ∝ t2, give your measured a with an uncertainty range, compare it with g sin θ as a percent difference, and explain two sources of error and which direction each pushes your result. Include both graphs.
How confident are you that you can measure an acceleration from video data by linearizing, and compare it with a model using an uncertainty estimate?
Practice
Work these on paper or in your notebook, then open Check your answer. Aim for all of Fluency and Application; try at least one Challenge.
Printable version: this unit’s practice workbook (PDF)
Fluency
Build speed and accuracy with the core skill.
- An object starts at 8 m/s and accelerates at 2 m/s2 for 7 s. How far does it travel?
Check your answer
Answer: 105.0 md = v0t + ½at2. - An object starts at 5 m/s and accelerates at 1 m/s2 for 11 s. How far does it travel?
Check your answer
Answer: 115.5 md = v0t + ½at2. - An object starts at 12 m/s and accelerates at 3 m/s2 for 6 s. How far does it travel?
Check your answer
Answer: 126.0 md = v0t + ½at2. - A phone films at 30 frames per second. A cart released at frame 0 crosses a mark 0.30 m down the ramp at frame 21. Find the time and, assuming constant acceleration from rest, the acceleration.
Check your answer
Answer: t = 0.700 s; a = 1.224 m/s2t = 21/30 = 0.700 s. From x = ½at2, a = 2x/t2 = 2(0.30)/0.7002 = 1.224 m/s2. - A graph of x against t2 for a cart released from rest has slope 0.26 m/s2. What is the acceleration?
Check your answer
Answer: 0.52 m/s2x = ½at2, so the slope is a/2 and a = 2 × 0.26. - A ramp is 1.60 m long and raised 0.20 m at one end. What acceleration does the frictionless model g sin θ predict?
Check your answer
Answer: 1.225 m/s2sin θ = 0.20/1.60 = 0.125; a = 9.8 × 0.125 = 1.225 m/s2.
Application
Use the skill in context. Show your reasoning.
- For the ramp above, the measured acceleration was 0.52 m/s2. Find the percent difference from the model and say whether friction explains the direction of the difference.
Check your answer
Answer: −57.6%; yes, friction makes the measured value lower(0.52 − 1.225)/1.225 × 100 = −57.6%. Friction and air resistance always reduce the acceleration, so a lower measurement is expected. - A mark is crossed after about 0.90 s, timed by video at 30 fps with ±1 frame uncertainty. What is the percent uncertainty in the time, and roughly what is it in an acceleration computed from 2x/t2?
Check your answer
Answer: about 3.7% in t; about 7.4% in aOne frame = 0.033 s; 0.033/0.90 = 3.7%. Because a depends on 1/t2, its percent uncertainty is about twice: 7.4%. - From rest with constant acceleration, a cart covers 0.10 m in the first 0.50 s. How far does it travel from rest in 1.00 s and in 1.50 s?
Check your answer
Answer: 0.40 m and 0.90 mx ∝ t2: doubling t gives 4 × 0.10 = 0.40 m; tripling gives 9 × 0.10 = 0.90 m.
Challenge
Stretch problems. Expect to think before you write.
- A student measures a = 0.66 m/s2 on a 2.00 m ramp. What height h of the raised end would make the frictionless model give exactly this value, and what does it mean if the actual height is 0.15 m?
Check your answer
Answer: h = 0.135 m; the real ramp is higher, so friction removed about 10% of the model accelerationg h/L = 0.66 gives h = 0.66 × 2.00/9.8 = 0.135 m. At h = 0.15 m the model gives 0.735 m/s2, so the measurement is 10% lower.
Review
Keep earlier skills sharp.
- Find the slope of the line through (4, 6) and (9, 5).
Check your answer
Answer: m = −1/5m = (5 − 6) ÷ (9 − 4) = −1/5 - Find the slope of the line through (1, −4) and (8, −2).
Check your answer
Answer: m = 2/7m = (−2 − (−4)) ÷ (8 − 1) = 2/7 - Solve x = ½at2 for t, then find t when x = 1.2 m and a = 0.60 m/s2.
Check your answer
Answer: t = √(2x/a) = 2.00 sMultiply by 2, divide by a, take the square root: √(2.4/0.60) = √4 = 2.00 s.
Ramp Lab Frame Record: Three Runs at Each Height, and a Ball for Comparison
SLS original frame-count records for the lesson’s ramp (2.00 m board, 30 fps), three runs at each of two heights plus a rolling-ball set; constructed teaching data, not measurements · SLS original (CC BY 4.0); constructed teaching data, labeled illustrative
Why this reading. The lab’s Option B gives one set of frame numbers per height and asks you to answer every analysis question from it, including the uncertainty from repeated trials. One run has no spread to estimate. Here are three runs at each height, with the frame-to-frame disagreement a real phone video shows, plus a set for a rolling ball so the misconception in the lesson can be tested rather than asserted. The numbers are constructed so that the run means reproduce the lesson’s figures exactly; they are illustrative, not measured.
Setup
Board length along the slope L = 2.00 m. Tape marks at 0.20, 0.40, 0.80, 1.20 and 1.60 m from the release line. Video at 30.00 fps (frame interval 0.0333 s); the release frame is frame 0 in every run. Timing uncertainty ±1 frame = ±0.033 s. Mark positions were taped to ±0.005 m.
Set 1: raised end h = 0.15 m (sin θ = 0.075, θ = 4.30°)
| Mark x (m) | Run 1 frame | Run 2 frame | Run 3 frame | Mean frame | Mean t (s) | t2 (s2) |
|---|---|---|---|---|---|---|
| 0.20 | 22 | 23 | 24 | 23.0 | 0.767 | 0.588 |
| 0.40 | 33 | 32 | 34 | 33.0 | 1.100 | 1.210 |
| 0.80 | 46 | 47 | 45 | 46.0 | 1.533 | 2.351 |
| 1.20 | 56 | 57 | 55 | 56.0 | 1.867 | 3.484 |
| 1.60 | 65 | 64 | 66 | 65.0 | 2.167 | 4.694 |
Set 2: raised end h = 0.30 m (sin θ = 0.150, θ = 8.63°)
| Mark x (m) | Run 1 frame | Run 2 frame | Run 3 frame | Mean frame | Mean t (s) | t2 (s2) |
|---|---|---|---|---|---|---|
| 0.20 | 16 | 15 | 17 | 16.0 | 0.533 | 0.284 |
| 0.40 | 22 | 23 | 24 | 23.0 | 0.767 | 0.588 |
| 0.80 | 32 | 33 | 31 | 32.0 | 1.067 | 1.138 |
| 1.20 | 41 | 40 | 39 | 40.0 | 1.333 | 1.778 |
| 1.60 | 46 | 45 | 47 | 46.0 | 1.533 | 2.351 |
Set 3: a solid ball instead of the cart, h = 0.15 m
| Mark x (m) | Run 1 frame | Run 2 frame | Run 3 frame | Mean frame | Mean t (s) | t2 (s2) |
|---|---|---|---|---|---|---|
| 0.20 | 26 | 25 | 27 | 26.0 | 0.867 | 0.751 |
| 0.40 | 37 | 38 | 36 | 37.0 | 1.233 | 1.521 |
| 0.80 | 52 | 51 | 53 | 52.0 | 1.733 | 3.004 |
| 1.20 | 64 | 65 | 63 | 64.0 | 2.133 | 4.551 |
| 1.60 | 74 | 73 | 75 | 74.0 | 2.467 | 6.084 |
Fitting a line through the origin
The model x = ½at2 has no intercept, so fit a line through the origin. For points (ti2, xi) the least-squares slope is k = Σ(xiti2) / Σ(ti4), and a = 2k. A spreadsheet’s trendline with “set intercept = 0” does the same sum.
- Set 1: k = 0.3414 m/s2, so a = 0.68 m/s2. Model g sin θ = 9.8 × 0.075 = 0.735 m/s2. Percent difference (0.735 − 0.683)/0.735 = 7.1%, measured below the model.
- Set 2: k = 0.6818, a = 1.36 m/s2. Model 9.8 × 0.150 = 1.47 m/s2; difference 7.2%, again below. Doubling h doubled sin θ and the acceleration went from 0.68 to 1.36, a ratio of 2.00.
- Set 3 (ball): k = 0.2636, a = 0.53 m/s2, which is 28% below g sin θ but within 0.4% of the rolling model (5/7)g sin θ = 0.525 m/s2. A ball that falls 29% short of g sin θ is not a bad measurement; it is a different object.
Reading the spread
At the 0.20 m mark the three runs of Set 1 disagree by 2 frames (0.067 s) on a mean time of 0.767 s, about 9%. At the 1.60 m mark they disagree by the same 2 frames on 2.167 s, about 3%. Because t is squared, the relative uncertainty in t2 is about twice that: 18% at the first mark, 6% at the last. This is why the fit should be trusted more at the far marks and why a longer board, or a slow-motion mode with more frames per second, is the first improvement to make. Parallax pushes the near marks and far marks in opposite directions unless the camera is square to the middle of the board; a cart released with a slight push arrives early at every mark and makes a come out too large; a warped board changes sin θ along the run.
Your turn
- Graph x against t and x against t2 for Set 1 using the mean times. State, with the straightness of the second graph as evidence, whether the acceleration was constant.
- Fit Set 1 using only run 1, then only run 3. Report both accelerations and use their difference as one estimate of your uncertainty range.
- Compute the percent difference for Set 2 and say, with a physical reason, why both sets come out below the model rather than scattered above and below it.
- Did doubling h double the acceleration? Answer with the two fitted values and sin θ for each height.
- Using Set 3, decide whether the ball’s shortfall is measurement error or physics, and say what evidence in the numbers settles it.
Sources: NIST, standard acceleration of gravity g_n = 9.80665 m/s^2 (exact) (physics.nist.gov)