Frictionless experiment with a torsion balance apparatus

Using metrology similar to Henry Cavendish’s 1706 experiment to measure the force of gravity between masses


If the idea/principle/method/law disagrees with the experiment it’s wrong, that all there is to it (Richard Feynman describing what science is).


Experiment we must


Fig 10


We need an airtight structure (box), and we must turn on the propellers for x seconds at different positions.

Position 1, near the box’s opposite side.

Position 2, far from the box’s opposite side.


If the resulting force (which we can deduce by box’s change in position/velocity) is the same regardless of the propellers position, I am wrong and will never bring this idea up again.


If the resulting force decreases as the distance increases, well that means we have a new method of space propulsion.


The method has been tested horizontally on ball bearings, dry ice plucks and floating on water, however in order to eliminate the influence of friction I shall suggest the following “test rig”


Simple Dynamic Tests Rig I.


Fig 11


We put a motor and propeller (representing mass M2) balanced by a counter weight hanging from a string (Fig 11), this permits the assembly to easily rotate around the vertical axis and will represent the movement of the inner free floating mass (M2).



Fig 12


The balancing motor assembly is inside a perimeter cylinder structure (fig 12), balancing motor assembly and perimeter cylinder structure may rotate independently (until the balancing motor assembly bumps into the separator borders).


Fig 13


Cutaway (Fig 13) illustrates the balancing motor can turn on its vertical axis independently of the surrounding perimeter / support / separators for an approximately 40º angel before colliding with one of the separators.


We can replace the balancing propeller with a wheel in direct contact with the perimeter cylinder as a control experiment (see Note 2)




Turning on the motor/propeller near (almost touching) “rear” separator

(As in position 1 fig 10)


If we position the motor/propeller assembly as close as possible to the “rear” separator (fig 14a), turning on the motor blows air directly against the “rear” separator giving the perimeter cylinder a clockwise movement (green arrows), the balancing motor/propeller assembly will accelerate in the counter clockwise direction (blue arrows) (fig 14b) until it collides with the “forward” separator bringing the perimeter cylinder’s clockwise spin to a standstill (fig 14c).


In the described configuration it is difficult to obtain an increment in rotation of the perimeter cylinder, the spacecraft will not accelerate.



                                       Fig 14a                                                                        Fig 14b

                               Fig 14c

Video of the experiment.


Breakdown of results with motor/propeller near (almost touching) the “rear” separator/target surface.


Fig 15


Turning on the motor/propeller as far from the “rear” separator as is possible


If we position the balancing motor assembly’s initial position far form the “rear” separator (fig 16a), with a approximate 35º angle (θ 3), when we turn on the propeller the balancing motor assembly accelerates in the counter clockwise direction (fig 16b) as in the previous experiment, but the perimeter cylinder’s clockwise acceleration is significantly less than the previous experiment, this is because the force exerted against an object by a (non laminar) gas flow is inversely proportional to the distance separation gas source from the target surface (separator) (see).


The instant of collision between the balancing motor assembly and the “forward” separator (fig 16c) finds the perimeter cylinder has made only a small clockwise movement, and the momentum transferred by the balancing motor assembly to the perimeter cylinder is sufficient to accelerate it in the counter clockwise direction beyond its original position and continue turning in the counter clockwise direction (fig 16d).


By repeatedly inverting the direction of the propeller (Note 3) it is possible to continually “bump” the “forward” separator constantly increasing the perimeter cylinders counterclockwise velocity. (See cycle for generation thrust on a spacecraft)



                                         Fig 16a                                                                                         Fig 16b






                                           Fig 16c                                                                                    Fig 16d



Video of the experiment II, another view


Breakdown of results with motor/propeller as far as possible from the “rear” separator/target surface.


Fig 17


What is observed?


When we turn on the motor, the clockwise angular displacement of the perimeter cylinder is depended on the distance that separate the propeller from the “rear” separator, (Videos 1, 2 and 3), this is because the force exerted against an object by a (non laminar) gas flow is inversely proportional to the distance separating gas source from the target surface.


The balanced motor assembly continues to gain angular velocity until it collides with one of the separators (Fig 17c), it transfers momentum to the perimeter cylinder that gains velocity in the counter clock wise direction, the balanced motor assembly bounces in the clockwise direction, the cycle can be repeated indefinitely by changing the direction of the propeller (see Note 4) at the appropriate moment (see videos 2 and 3)


Some are disturbed because the experiment appears to contradict the law of conservation of linear momentum, others that have given the matter some thought believe it does not.


Simple Dynamic Tests Rig II


Again I mention that ideally the proposal should be tested in a micro gravity environment, as the international space station is not available we constructed the test assembly setup illustrated in fig 18.


A square aluminum frame (2m) is attached by 4 cables to a long cord hung from the ceiling; this setup permits the frame to rotate on its central axis with minimal friction.


On one end of the aluminum frame we have a transparent airtight box (Fig 18) that will serve as our closed system, at the opposite side a ballast (counter weight) for balance. The assembly has freedom to rotate around the rotation axis.






Fig 18


Fig 19


To illustrate the effect a mass interacting inside a closed system (the transparent box), we will first use a toy R/C car (Fig 19), when the toy car (internal mass) is turned on it gains acceleration by directly interacting with the transparent box (fig 22).


Figs 20 and 21 show the toy car’s initial position at rest, at this moment all the elements on the test bed are at rest.



Fig 20


Fig 21


If we turn on the car’s wheels in a counter clockwise direction, the car will move in the –X direction and the transparent box moves in the +X direction making the frame rotate in the counter clockwise or +X direction, fig 22 shows a side view of the described actions, fig 23 shows a top view of the actions portrayed.



Fig 22


Fig 23


The frame will continue to rotate in the +X direction until the car collides with the transparent box’s inner –X surface (fig 24) exerting a force (F1), that cancels the frame’s rotation.


Fig 24


If we reverse the toy car’s direction the objects will return to their original position. It is not possible to give the frame a constant angular acceleration by any combination of movements we command the toy car to perform, at most we can obtain a cyclical clockwise/anticlockwise oscillation, but the frame will not gain rotational velocity.


This so far confirms that a closed system (spacecraft) cannot accelerate by interaction/movements of masses inside the spacecraft if no mass is expelled from the system.


Let’s do something different



Fig 25


If we replace the toy car with a propeller car (Fig 25), the wheels of the propeller car are “free moving” and as frictionless as possible. If the propellers are turned on the “car” is set in motion by the collision of air molecules and propeller.




Fig 26



Fig 27


Figs 26 and 27 illustrate the test bed’s elements in their initial position at the start of the experiment, the propeller car is positioned against the transparent box’s inner +X side.


Fig 28


We will use the R/C control to create a 3 part cycle (1-blow air in the +X direction, 2-blow air in the –X direction, 3-collition of car and box’s inner +X wall)


Part 1: We turn on the propellers so they blow air in the +X direction against the transparent box’s inner +X side, this results in two actions: (Fig 28)

1-The propellers collision with surrounding air molecules propel the propeller car in the –X direction (fig 28 ∆2)

2-Collisions of air molecules against the transparent box’s inner +X wall give the box a slight acceleration in the +X direction (fig 28 ∆1) and the frame rotates counter clock wise (fig 29).



Fig 29


Part 2: As soon as the propeller car moves a few centimeters in the –X direction, the propellers direction is inverted so that air is blown in the –X direction, this makes the car change direction and accelerate in the +X direction (figs 30 ∆ 3 and fig 31).

We observe that although the frame’s counter clock wise rotation slows slightly it does not stop (fig 31).



Fig 30




Fig 31






Fig 32


Fig 33


Part 3: the propeller car collides with the transparent box’s inner +X (figs 32 and 33) giving a hard “bump” that increments the frame’s counter clock wise rotation velocity.


Video of the experiment




After repeating the experiment numerous times we can:


Confirm that it is not possible to increase the frame’s rotation by acceleration of a contained mass by any means that requires direct contact with the box (figs 20 to 24) (various other mechanisms were tried but not shown in the document)


It is possible to increase the frame’s rotation if the contained mass is made to accelerate with no direct contact with the cylinder by means of air currents (figs 25 to 33).


This experiment is presented to observe a simple method of propulsion for spacecraft and can be easily replicated.


No insights on apparent conflicts with the conservation of linear momentum law are offered in this particular document.




Note 1 What about Newton’s third law (To every action there is always an equal and opposite reaction)?


Fig 34


The motor spins the propeller (A), when the propeller collides (B) with a air molecule (D) 2 equal and opposite forces interact, F1 pushes the propeller assembly in the +X direction, and F2 hurls the molecule (D) in a –X direction.


Fig 35


With every collision (we are taking billions and billions of collisions) two equal forces hurl the colliding molecules in different directions, the magnitude of the forces does not change but the vector direction are randomized with every collision. See kinetic theory of gases.



 Note 2 Control Experiment



                          a                                                b













If we attach the wheeled assembly to the motor (b) in such a manner that the wheel is in contact with the perimeter circle so that when the motor is turned on the wheel will “roll” on the cylinder (Fig 36)










Fig 37a                                                                                                                                     Fig 37b



Fig 37b illustrates what happens when we turn on the motor so that the wheel turns in a clockwise direction, the motor / wheel assembly pushing against the perimeter cylinder will move in a counter clockwise direction (θ 1)  while the perimeter cylinder will turn in a clockwise direction (θ 2).


The ratio between the counter clockwise angular displacement of the balancing motor assembly (θ 1)  and the clockwise angular displacement of the perimeter cylinder (θ 2) will depend on the ratio between the masses of the objects, if they have equal mass (this can be controlled by adding mass where necessary) the angular displacements will be equal. (Fig 37b)


If we reverse the direction of the motor the objects will return to their original position. It is not possible to give the perimeter cylinder a constant angular acceleration if the balanced motor assembly is restrained to a fixed angle by separating panels.


At most we can obtain a cyclical clockwise anticlockwise oscillation



Note 3


To control the propeller motor (on-off-left spin-right spin) it is best to use a RC controller so that no external cables protrude from the assembly.


Suggested method:


Obtain an inexpensive toy RC car or other toy

Remove the RC circuit (generally includes the battery compartment) (fig 38)

Fig 38


Connect the control wires of one of the toys motors (most have two or more) to the propeller motor, in this example a normal LEGO motor (fig 12)

This setup will control the propellers direction remotely.


Fig 39