Skip to main content
Sandbox Physics

Roller-Coaster Energy Designer

Can a roller-coaster car with enough energy to reach the top of a loop still leave the track? Yes, even without friction. Reaching a loop’s top is an energy condition; staying on its inside rail also needs enough speed. At the top, gravity can exceed the inward acceleration the curve requires. Contact would then need an upward pull, which a one-sided rail cannot supply, so the car leaves earlier.

Shape a circular or elliptical loop, release a sliding particle and watch its energy change. Try to keep it on the rail, limit the peak support load and retain enough finish speed.

Interactive modelRoller-Coaster Energy Designer
Playback clock time00
Particle speed00
Normal support force00
Valid-prefix end speed00
Selected support over weight00
Peak support over weight00
Kinetic energy00
Potential energy00
Rail friction heat00
Energy-conservation residual00
Valid-prefix end time00
Refinement time difference00
Gravity from raw speed and height00
Contact, load and finish challenge00
Contact state00
Acquired stations00

Enough speed to stay on the loop

BackgroundAt the top of a loop, gravity and the rail must bend the particle’s motion along the curve. Reaching that height is only part of the problem.

Why it mattersA higher release helps maintain contact at the top of a circular loop but increases the support load at the bottom. A successful design must satisfy both conditions and retain enough finish speed.

One-sentence intuition
The rail can push the particle along the normal. If following the curve would require a pull, contact ends.

Core mathematical model

Track direction, normal and support force

t=drdsn=(−ty,tx)N/m=v2κ+g tx\begin{aligned}\mathbf t&=\frac{d\mathbf r}{ds}\\\mathbf n&=(-t_y,t_x)\\N/m&=v^2\kappa+g\,t_x\end{aligned}

The signed curvature and the left normal come from the designed centreline itself. A positive support force pushes along the left normal.

Speed and heat along the track

d(v2)ds=−2gdyds−2μNmdQds=μN\begin{aligned}\frac{d(v^2)}{ds}&=-2g\frac{dy}{ds}-2\mu\frac{N}{m}\\\frac{dQ}{ds}&=\mu N\end{aligned}

Kinetic friction acts only while the support force is positive. Squared speed and heat are integrated separately.

Energy balance

E0=12mv02+mgH12mv2+mgy+Q=E0\begin{gathered}E_0=\tfrac12mv_0^2+mgH\\\tfrac12mv^2+mgy+Q=E_0\end{gathered}

The normal force does no work while contact holds. The heat, integrated separately, makes up the difference, so both sides match.

Staying on a circular loop

Ntop/m=vtop2/R−gvbottom2≥5gR\begin{aligned}N_{top}/m&=v_{top}^2/R-g\\v_{bottom}^2&\ge5gR\end{aligned}

On a frictionless circular loop, staying in contact at the top takes more energy than just reaching the top.

Estimating gravity from recorded data

vi2=α−2g yiv_i^2=\alpha-2g\,y_i

Fit a straight line, with free intercept and slope, to the recorded heights and speeds. Friction makes this estimate invalid, so the Lab turns it off when friction is present.

Common difficulties

An idealized ball

Typical misconceptionThe ball on screen rolls, has suspension and behaves like real coaster hardware.

Better mental modelThe model is a point particle that slides. The thickness of the rail and the marker is only for drawing.

Run the experiment

  1. 01

    Tell energy and contact apart

    Compare Complete circular loop, Ideal contact threshold and Energy without contact.

    What to observe: The bright part of the rail ends where support first disappears; the untraversed part is dim.
  2. 02

    Design the loop

    Change both loop radii, the friction, the load limit and the required finish speed.

    What to observe: Having enough energy alone does not pass the three-part challenge.
  3. 03

    Record the valid part of the run

    Open the default raw record, then change the geometry and use Acquire sensor record before exporting the CSV.

    What to observe: No sensor readings are made up after the particle leaves the rail or stalls.
  4. 04

    Check the numerical accuracy

    Double the integration intervals, then compare the refinement time difference and the separately integrated heat.

    What to observe: The time difference shows how much the result changes when the steps are refined. It is not a strict error bound.
Something broken, confusing, or missing on this page?Tell us
What kind of feedback?

Sent with this page’s address, its Lab ID, the site language and your device type. Please leave out personal details. Feedback is kept for up to 12 months.Privacy policy