Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Huge Energy Savings Possible w/ Curved Tunnels...

This site may earn commission on affiliate links.
There is a physics problem called the "Brachistochrone problem" which seeks to identify the "path of quickest descent" for a frictionless body on a ramp with the only force acting being gravity.

For bodies with 0 initial velocity, the problem was solved centuries ago by Isaac Newton.

The problem is, due to friction, without an initial velocity, there is no possibility of descending a ramp and rising back to the initial height, where the start and end points are horizontal to each other.

To start and end on the same height, one needs an initial velocity, but that is precisely where things get complicated.

It turns out the solution for a given velocity is "irrational" meaning it would take the fastest computer an infinite amount of time to solve precisely, but it can be solved "for all practical purposes" in the blink of an eye.

The quickest path between 2 horizontal points with an initial velocity is a portion of a curve known as a "cycloid." The cycloid can be thought of as the path that a point on the edge of a wheel takes as the wheel rolls along the ground. In order to identify the portion of the cycloid that is the quickest path between 2 points with an initial velocity, one must remove the "cusps" of the cycloid such that a free-falling body would reach the initial velocity after falling that vertical height, and the new, truncated cusps must traverse the desired horizontal distance.

M6le1.gif


To find the curve, one must identify 2 variables: the cycloid generating circle radius, and the initial angle.

It turns out the initial angle is irrational but it can be worked out to arbitrary precision with numerical computation methods as there is no algebraic solution.

Where A = -0.05102 m, the depth that accelerates objects to 1m/s in 9.8m/s^2 gravity), and B is the horizontal distance (B = 1 meter):

A=-(B/2(pi-t+sin(t)))(1-cos(t))

-0.05102=-(B/2(pi-t+sin(t)))(1-cos(t))

therefore:

t=0.8117451055281

Therefore the initial angle of the cycloid is 0.8117451055281 radians and the final angle is 2pi-0.8117451055281 radians.

Next, the cycloid generating radius is found via:

r=B/(2(pi-t+sin(t)))

Therefore:

r=0.1636480760880076136181 meters

Parametric Graph: parametric graphing

Next we find the travel time of the whole cycloid without the cusps removed with 0 initial velocity:

time_whole_cycloid=2*pi*sqrt(r/g)

0.81193690616139705994 seconds = 2*pi*sqrt(0.1636480760880076136181/9.8)

Next we find the travel time that it takes to accelerate from the cusps with 0 initial velocity to the desired initial velocity:

dtime=sqrt(a/g)*dtheta

dtime = sqrt(0.1636480760880076136181/9.8)* 0.8117451055281

0.1048967645154483811775 seconds = sqrt(0.1636480760880076136181/9.8)* 0.8117451055281

Next we subtract the time it takes to accelerate to and from the initial velocity from the time it takes to traverse the un-truncated cycloid:

0.602143377130500297585=0.81193690616139705994-0.1048967645154483811775-0.1048967645154483811775

Giving an answer of:

0.602 seconds to traverse the quickest path between 2 horizontal points separated by 1 meter with an initial velocity of 1 meter per second.

There is a more lengthy discussion on the topic here:
Quickest Route

Downloadable Excel Spreadsheet
quickest_route - Shared with pCloud

ramp-1-gif.133372


As a consequence of the fact that a distant horizontal point can be reached in less time than the straight path with the initial velocity, huge energy savings or performance increases are possible.

For example, using the spreadsheet linked above, I find the following results:

quickest_2-jpg.136423


In summary, a trip between 2 horizontal points separated by 5km with an initial velocity of 9 meters per second, takes 10.1x longer going in a straight line than on the optimal curved path.
 
Perfect vacuum?

With 0 initial velocity, you’d need a perfect vacuum to make it all the way to the end of the tunnel.

But with an initial velocity, as long as the friction in the tunnel doesn’t remove all of the initial kinetic energy, the vehicle can make it all the way through and back out again.

The optimal path calculated above does assume perfect vacuum as does the original “Brachistochrone” problem, so factoring friction, the optimal path will be different from the path assuming perfect vacuum.

ramp-1-gif.133372


^As seen in the above image, the friction in the “tunnel” is not sufficient to remove all of the initial kinetic energy the steel ball has on the flat section, so the ball that takes the curved path makes it through the entire “tunnel” and back out again with energy to spare.
 
Last edited:
It would be yet more efficient to obtain the “initial kinetic energy” via the oberth effect at the bottom of the tunnel, rather than pushing against the ground at the surface.

Instead of using electricity to push against the ground at the top, the vehicle starts with 0 velocity, descends the ramp & uses the same amount of electricity it would’ve used to get the initial velocity at the top, but instead of pushing against the ground it uses the energy to eject a mass “off the back.” When the ejection mass (kg) is chosen correctly the ejection force will be sufficient to stop the mass on the tracks. What makes it more efficient than pushing against the ground is the Oberth effect. The vehicle obtains all the mechanical energy of the “push” as well as the all the kinetic energy the ejected mass had before the push, via conservation of energy & momentum.

If the mass is a tank of water and tank mass is low to negligible, the water can be left in an underground chamber to evaporate back to the surface from geothermal heating, potentially run through a combined cycle turbine to generate power, and the tank can be lifted with negligible energy.

If friction is negligible, when the vehicle exits the tunnel, it will have the mechanical energy of the “push” plus the kinetic energy the tank + water lost at the bottom, and will arrive sooner than if the same amount of electricity had been used to push against the ground at the start of the tunnel.
 
Last edited:
  • Informative
Reactions: pilotSteve
Suppose we have a flat 100km track w/ negligible friction environment and we are using $27.76 worth of $2.14/gal gasoline at 25% efficiency to accelerate a 30000kg shipping container from 0 to 658.69mph, and factoring a 1g acceleration time, the total trip time is 5.91 minutes, and we will emit 117.69kg of CO2 into the atmosphere.

If we have a 2450m tunnel depth engineering limit based on the Gotthard Base Tunnel, and we want to use water as fuel for an underground Oberth effect maneuver, and cover the same horizontal distance in less time, for lower fuel cost, with less C02 emissions, I make the following findings:

-The travel time is reduced to 4.48 minutes
-The fuel cost for 25000kg of water is $9.89 at a price of $1.50 per 1000 gallons
-The average horizontal speed is 830.93mph
-The CO2 emissions are reduced to 0kg

Parametric Graph of Entry and Exit Tunnels:
parametric graphing
(Entry Tunnel in Orange, Exit Tunnel in Purple)

The vehicle begins its journey with a passenger vehicle mass of 30000 kg, water mass of 25000 kg in a tank plus tether with a mass of 3000 kg (28000kg total ejection mass).

The vehicle descends the entry ramp to a flat section at a depth of 2450 meters with 0 initial velocity via a half-cycloid shaped ramp with a cycloid generating circle radius of half the depth or 1225 meters.

At the bottom of the ramp, the vehicle exerts a mechanical impulse of 1300623295.36J on the tank which has 672736187.25J of kinetic energy. The entire mechanical impulse + tank kinetic energy transfers to the passenger section via the oberth effect, giving the vehicle a coasting velocity of 423.8m/s. This 1g acceleration takes 20.86 seconds and traverses 6708m horizontal distance. After the impulse the water tank is stopped on the tracks. After coasting an additional flat 80411.37 meters, the vehicle reaches the exit tunnel.

Based on the velocity at depth and the height to the surface, the final velocity at the surface is 362.70m/s. The exit tunnel parameters (cycloid generating radius and final theta) are chosen based on an initial velocity parameter of 362.70m/s and the max tunnel depth parameters in the spreadsheet linked above.

The exit tunnel parameters are 4578.77m cycloid generating radius and final theta 4.22885330717959 rad.

At the 100km finish line, both vehicles can use regen braking with 70% kinetic to kinetic efficiency to recover some of the kinetic energy. The gasoline powered vehicle has a deficit of 108.38kWh which gives us the $27.76 fuel cost for the next vehicle assuming 25% energy conversion efficiency and $2.14/gal.

The Oberth vehicle has more energy than the mechanical impulse on account of the lowered gravitational potential energy of the water tank left at the bottom. Factoring 70% kinetic to kinetic efficiency to recover some of the kinetic energy, and the energy required to lift the empty tank, there is still a recovered kinetic energy surplus of 2.416 kWh, therefore it is not necessary to purchase additional energy, only additional water ($9.89) for the next vehicle. The 2.416 kWh energy surplus does not include the additional energy that can be obtained by using geothermal heating on the 25000kg water at the bottom of the tunnel and running the steam through a turbine for additional energy production.

In summary, the water powered oberth vehicle has lower fuel costs and C02 emissions than a gasoline powered flat trajectory vehicle in a negligible friction environment.

oberth1.jpg

oberth2.jpg

oberth4.jpg
 
Perfect vacuum?

...I’ve been having some interesting conversations over at physics forums adding friction into the equation.

Suppose you want a “traversable” pathway optimized for travel time between 2 points horizontally separated by 5000m in 9.8m/s^2 gravity in a vacuum, and your magnetically levitated vehicle has a 250:1 lift to drag ratio, it turns out the minimum initial velocity is about 43mph.

I can’t give an exact travel time but in the no friction case the optimal route with 43mph initial velocity takes about 1/5th the travel time compared to the flat route, or 52 seconds optimal vs 256 seconds on the flat route with the same initial velocity.


Parametric Graph: parametric graphing

Quickest Route Between Horizontal Points using the Initial Velocity & Friction

Is it possible to solve for “t?”

Brachistochrone Problem w/ Initial Velocity


79a78967-5851-498d-981d-76d5a487013a-jpeg.137848


55b9101a-dc21-4c1e-88b6-bb9511f0623d-jpeg.273443

72a12b15-9625-43fe-bbe5-29fc0d2e0cf5-jpeg.273444
 
Last edited:
True but similar travel time on the flat route takes 25x the initial kinetic energy.

For longer routes the practical depth limit, say 2450m, precludes fully optimizing the travel time.

Even so, 2450m depth gives ~490mph of acceleration almost for free, then most of the rest of the tunnel could be flat. Like this 100km route:

Parametric Graph: parametric graphing

oberth1-jpg.603885


underground Oberth effect
 
Last edited: