Theoretical Asymmetric Inertial Propulsion System – Seeking Technical Feedback

[RotorRiddle]

Hello NASASpaceFlight community,

I’m an independent researcher based in the U.S., and I’ve recently completed a manuscript proposing a theoretical inertial propulsion system that generates net thrust without expelling mass. The concept utilizes phase-locked, counter-rotating rotor sets to create time-asymmetric centripetal force interactions with center of mass displacement, resulting in a net impulse per cycle.

The system is designed to comply with conservation laws through rotational energy depletion and constructive-phase isolation. While experimental validation is pending, I recognize the speculative nature of the concept and believe the mathematical modeling and configuration merit technical discussion.

I’m posting here to invite feedback, critique, and discussion from the community.

Full manuscript available here: https://doi.org/10.31219/osf.io/c5mz4_v1 or https://doi.org/10.36227/techrxiv.174838066.62088627/v2

Manuscript is also attached to this post.

Key topics covered:
- Time-asymmetric inertial force generation
- Conservation of momentum and energy
- Hann-window force modeling
- Cardioid force trajectory
- Phase-selective clutch mechanism for impulse isolation

I welcome all constructive feedback and look forward to engaging with this brilliant community. If anyone is interested in collaborating or exploring experimental validation, I’d be glad to connect.

Best regards,
James A. Wells III

[InterestedEngineer]

Under classical mechanics, the momentum of a closed system remains constant unless acted upon by an external force.
The propulsion system described here challenges this principle by generating net directional impulse through precisely
timed internal inertial interactions. Unlike conventional propulsion methods, which rely on the continuous expulsion of
mass, this approach leverages asymmetry in the timing and distribution of internal forces

probably belongs in the New Physics for Space Tech thread

https://forum.nasaspaceflight.com/index.php?board=73.0

[InterestedEngineer]

As the frame's velocity increases, more power is required to maintain a constant acceleration, because the instantaneous
power delivered by the propulsion force is proportional to the product of force and velocity (F · v). To

You don't show that anywhere.  In fact earlier in the paper you show deltaV is merely proportional to the impulse divided by the mass, which makes this a net energy generator as you'd get the same deltaV for the same number of joules, when in any conventional physics system (e.g. acceleration of a particle along an electric potential, a rocket nozzle, etc) the power added goes up by the square of the velocity, and thus has limits.  It's why the rocket equation is a log function.

I'm not sure this is a perpetual motion machine unless you can run this in around a shaft (too many rotating frames from my brain to contemplate), but it is an infinite energy machine. 

Let's say I have a 5kg mass acceleration at 2m/sec² at a constant 100J as you show in your paper (and completely independent of the motion relative the reference frame).

Over 86400 seconds, my 5kg mass has achieved a velocity of of 172,800m/sec and have a kinetic energy of 75GJ, all while having expended 86400 * 100 = 8.6MJ.  Nice bonus there, where'd the energy come from?  Leaving aside relativity, you should be able to get this thing to the speed of light.

It's not enough to state that energy input to the rotating mechanisms has to increase with reference frame velocity - you have to show it both mathematically and mechanistically.  You have done neither.

How does your system "know" what its velocity against the reference frame is?  In a rocket, it "knows' because it threw mass out the back and there's a limit to how much mass you have left.  In a field, the gradient is either decreasing with the square and going to zero quickly (outbound), or increasing with the square (inbound), the latter being bounded by colliding with the source of the field.

Your system has no bounds.  At least, you didn't show any other than a bald assertion without any math involved.

[RotorRiddle]

Thanks for the thoughtful replies—really appreciate the engagement.

@InterestedEngineer: Your breakdown of impulse and force mechanics is spot-on and extremely helpful for framing the discussion. You're absolutely right that in conventional propulsion systems, mass ejection drives acceleration, and the impulse-time relationship governs force output.

In my proposed system, however, there’s no expelled mass. Instead, the net impulse arises from internal mass displacement and time-asymmetric force interactions. The cardioid trajectory and Hann-window modeling are designed to isolate constructive force phases while minimizing counterproductive ones. In fact, the constructive phase generates approximately 60% more force than the recoil phase. The phase-selective clutch mechanism—though optional—can cancel the destructive phase entirely, further enhancing net impulse without violating conservation laws.

I recognize this challenges conventional interpretations, especially regarding momentum conservation in closed systems. Although I’ve worked out the net force resultants over a full cycle, I plan to develop MATLAB simulations to visualize force vectors, center-of-mass shifts, and energy depletion over time. The system’s energy input is rotational, and I’m exploring how internal energy loss translates into sustained impulse generation.

Would love to hear your thoughts on how internal energy depletion might factor into momentum conservation, especially in systems where mass remains constant but force asymmetry is introduced through timing and geometry.

Thanks again for the thoughtful critique—this is exactly the kind of discussion I was hoping for.

Best, 
James

[Robotbeat]

Welcome!

Because this isnt congruent with conventional interpretations, I think it belongs in this forum section:
https://forum.nasaspaceflight.com/index.php?board=73.0

[InterestedEngineer]

I recognize this challenges conventional interpretations, especially regarding momentum conservation in closed systems. Although I’ve worked out the net force resultants over a full cycle, I plan to develop MATLAB simulations to visualize force vectors, center-of-mass shifts, and energy depletion over time. The system’s energy input is rotational, and I’m exploring how internal energy loss translates into sustained impulse generation.


James

I bet your energy input completely cancels any linear momentum or energy gain you thought you were extracting from the syste

[catdlr]

Welcome!

Because this isnt congruent with conventional interpretations, I think it belongs in this forum section:
https://forum.nasaspaceflight.com/index.php?board=73.0

Yes, it does.  Locking it temporarily so I can move it over.

Move completed.  Thank you, gentlemen, carry on.  (This and the previous post from me will be deleted in a day.)

Tony

[Paul451]

I recognize this challenges conventional interpretations, especially regarding momentum conservation in closed systems.

Quite the contrary, it fits very neatly into a broad category of ideas that are all based on the principle that if you add enough moving parts to a system, you will eventually lose track of one or more energy vectors and suddenly believe you've discovered magic.

[RotorRiddle]

Thanks for the reply. I understand the skepticism—this field has seen its share of overcomplicated systems that obscure rather than reveal. That said, I’ve taken care to rigorously track all energy and momentum vectors in my model, and I’m not claiming any violation of conservation laws. The goal is to explore whether certain internal dynamics can produce net external motion without violating known physics.

If you believe a specific vector or interaction is being misrepresented or overlooked, I’d welcome your technical feedback. Dismissing it as “magic” is clever but doesn’t move the conversation forward.

[laszlo]

So just to save everyone a lot of work, I've extracted the high-level, non-mathematical description from the document and attached it. You don't need math to see what's wrong.

Looking at figures 1,2 & 3, all the rotors do is cyclically move the device's center of mass around. Since both rotors repeat their motions, eventually the center of mass will return to the starting point and do it again. The net result is that the device wobbles. There is no thrust. The phased forces that generate thrust are a fictitious result of mixing up different frames of reference. It's the classic Dean Drive.

In fact, you can get exactly the same results by keeping the lower rotor stationary and driving the upper rotor at the difference in rotational rate of the original setup. This restates the problem in the frame of reference of the lower rotor and the fictitious force disappears, leaving an easier-to-analyze wobbly device.

Finally, if you still don't buy this, consider satellites in orbit that have rotating components. They change their attitudes, not their altitudes as a result of the rotations.

[InterestedEngineer]

Thanks for the reply. I understand the skepticism—this field has seen its share of overcomplicated systems that obscure rather than reveal. That said, I’ve taken care to rigorously track all energy and momentum vectors in my model, and I’m not claiming any violation of conservation laws. The goal is to explore whether certain internal dynamics can produce net external motion without violating known physics.

If you believe a specific vector or interaction is being misrepresented or overlooked, I’d welcome your technical feedback. Dismissing it as “magic” is clever but doesn’t move the conversation forward.

I did give technical feedback above.  First you show that you violate conservation of energy laws by showing a constant acceleration for a given impulse and energy input, and then claim you don't, without showing any work on how you don't.

Until the velocity of the motion of the entire system relative to "start" of operation is taken into account in your energy expenditure and acceleration rates, you violate the law of conservation of energy.

Do as I did - run it out 86400 seconds and make sure energy balanced. It sure didn't when I followed the equations you gave.

BTW the same idea has been presented on this forum about 3 times so far (I lost track, they keep deleting the threads instead of locking them, so we learn nothing. but I digress).

If you fail to answer the question of conservation of energy from root principles, you will be judged as a perpetual motion machine and the thread will be locked or deleted, if history is any judge on this forum. Hopefully locked, so that there's a record of how to deal with these issues that's short on name calling and long on basic physics.

[RotorRiddle]

Thanks again for the thoughtful response and diagram—it’s genuinely helpful to see the concept visualized from another angle.

To clarify a key point from my manuscript: the system’s behavior over a full rotation of the slow rotor is central to the analysis. The asymmetric control phases—where one rotor is slow relative to the other—are designed to produce a non-zero net resultant over each complete cycle. This isn’t a transient imbalance; it’s a cumulative effect that arises from the timing and angular velocity differential between the rotors.

In conventional systems, symmetric rotation yields no net displacement—internal forces cancel out over a full cycle. But in this design, the slow rotor’s angular velocity differential introduces a persistent asymmetry in the resultant force vectors. These vectors are calculated from the angular separation and timing of the inertial masses throughout the cycle. There is only one instance where the resultant force vectors balance out: when the slow rotor is at 180 degrees and the faster rotor is at 360 degrees.

This asymmetry is key—it prevents the system from returning to a perfectly balanced state after each rotation, allowing a directional bias to accumulate over time. The effect isn’t instantaneous; it builds gradually through controlled imbalance, and I’m working on quantifying it more rigorously.

I’m not claiming reactionless thrust in the traditional sense, and I’m certainly not invoking fictitious forces. The goal is to explore whether internal asymmetries, when carefully timed and executed, can result in a net inertial effect that’s observable and repeatable—without violating conservation laws.

I’m currently refining the mathematical model to better capture these resultants and would welcome any feedback or suggestions on how best to approach that—especially from those with experience in dynamic systems or experimental setups.

Thanks again for engaging with the concept. I appreciate the dialogue and look forward to learning from the community.

[redneck]

Finally, if you still don't buy this, consider satellites in orbit that have rotating components. They change their attitudes, not their altitudes as a result of the rotations.

I went down that rabbit hole for a time. If a body(tether) in orbit is rotating such that the outer tips are moving at half orbital velocity, The additive side moving at 1.5 orbital velocity has 2.25 times the orbital energy at that altitude while the negative side at 0.5 orbital velocity for that altitude has 0.25 of the energy required to orbit there. 2.25+0.25= 2.5. The 2.5 energy divided by 2 gives the whole system 1.25 times the energy required to maintain that orbit.

One of the disproof's of the concept is that satellites with rotating components do not translate to higher orbital planes. Other than being infeasible, it really seemed to be an answer to reboost in LEO.In case anyone is skimming without comprehension, It doesn't work.

[RotorRiddle]

To address Interested Engineer’s concern about energy conservation:

You're absolutely right to flag the apparent discrepancy between energy input and kinetic energy output. If the system truly delivered constant acceleration with fixed energy input per unit time, it would accumulate kinetic energy indefinitely—clearly violating conservation of energy. That’s not what I’m claiming.

The issue likely stems from an oversimplified impulse-to-deltaV relationship presented early in the manuscript. That formulation doesn’t yet account for how the energy required to sustain internal motion must scale with the system’s inertial velocity. In conventional systems, the energy required to produce a given deltaV increases quadratically with velocity, and I agree this scaling must be reflected in any physically valid model.

In this system, energy is input through the rotational actuation of internal masses. As the system’s translational velocity increases, the inertial resistance to internal actuation also increases. That means the work required to maintain the same internal motion must grow accordingly. This scaling—likely quadratic—is essential to preserving energy conservation, and I’m actively working to model it from first principles.

As for how the system “knows” its velocity relative to the inertial frame—it doesn’t, and it doesn’t need to. The energy cost of actuation increases naturally due to the changing inertial dynamics, not due to any external awareness. That’s the part I need to model more carefully: how internal actuation translates into external motion, and how that energy scales with velocity.

I’m not proposing infinite energy or perpetual motion. I’m exploring whether internal asymmetries—when carefully timed and executed—can produce net motion, but only if the energy accounting holds up. If it doesn’t, then the concept fails, and I’ll accept that. I’m here to test it rigorously, not defend it dogmatically.

Thanks again for the challenge. I’ll revise the energy model and share updates as I go.

[RotorRiddle]

Thanks for sharing that example, Redneck. I appreciate the reference to rotating tethers in orbit—it’s a fascinating system, and I agree it’s often cited as a cautionary tale for internal propulsion concepts.

That said, I believe the scenario you described—where the tether tips have different orbital velocities and energies—illustrates a redistribution of energy within the system, not a net gain. The average energy of the rotating components may exceed the energy required to maintain the orbit, but unless there's an external interaction (like mass ejection or tether retraction), the system’s center of mass doesn’t climb to a higher orbit. That’s consistent with conservation laws.

In my case, the concept doesn’t rely on orbital mechanics or rotating frames in space. It’s focused on whether internal asymmetries in inertial mass motion, when carefully timed and actuated, can produce a net translational bias in a closed system. The goal is not to extract energy from nowhere, but to explore whether momentum can be redistributed in a way that results in observable motion—without violating conservation of energy or momentum.

I’m working on a more rigorous energy model to ensure the input scales appropriately with system velocity. If it turns out the concept fails under strict conservation analysis, I’ll accept that. But I think it’s worth testing with an open mind and solid math.

Thanks again for engaging with the idea.

[RotorRiddle]

For context, the system I’m exploring shares conceptual similarities with the Thornson device, which was famously demonstrated by propelling a canoe across a pool using internal mass motion. However, my design does not rely on the flailing arm mechanism used in that demonstration. Instead, it uses controlled rotational actuation of internal masses to generate asymmetric force vectors over time.

The goal is the same: to investigate whether internal motion—when carefully timed and executed—can result in net translation of a closed system. But I’m approaching it with a more rigorous energy model and without relying on impulsive or chaotic mass movements.

[CoolScience]

To address Interested Engineer’s concern about energy conservation:

You're absolutely right to flag the apparent discrepancy between energy input and kinetic energy output. If the system truly delivered constant acceleration with fixed energy input per unit time, it would accumulate kinetic energy indefinitely—clearly violating conservation of energy. That’s not what I’m claiming.

The issue likely stems from an oversimplified impulse-to-deltaV relationship presented early in the manuscript. That formulation doesn’t yet account for how the energy required to sustain internal motion must scale with the system’s inertial velocity. In conventional systems, the energy required to produce a given deltaV increases quadratically with velocity, and I agree this scaling must be reflected in any physically valid model.

In this system, energy is input through the rotational actuation of internal masses. As the system’s translational velocity increases, the inertial resistance to internal actuation also increases. That means the work required to maintain the same internal motion must grow accordingly. This scaling—likely quadratic—is essential to preserving energy conservation, and I’m actively working to model it from first principles.

The bolded phrase is where you go completely wrong. There is no such thing as a system's inertial velocity. Velocities are always relative to something else. There is no such thing as a universal inertial reference frame. Your device cannot scale with something that doesn't exist.

You cannot solve this from first principles if you start from something nonexistent.

I’m not proposing infinite energy or perpetual motion. I’m exploring whether internal asymmetries—when carefully timed and executed—can produce net motion, but only if the energy accounting holds up. If it doesn’t, then the concept fails, and I’ll accept that. I’m here to test it rigorously, not defend it dogmatically.

You need to stop saying that you are not proposing perpetual motion. Your claims of something generating net motion without pushing off something else or expelling mass have long since been proven to be precisely equal to perpetual motion. There are countless explanations of this available online.

The bolded statement in particular is something that has been carefully reviewed by countless scientists, and you can find the answer with a simple google search : No. It does not matter what the internals of your device are and there are no changes you can make to avoid this. It has been proven in the general case.

Your mistakes have already been pointed out, yet you continue trying to claim your device does something it does not and cannot.


You claim you are working on a "more rigorous energy model," but we already have a perfectly rigorous energy model called Newtonian mechanics. (Neither special relativity nor general relativity are relevant here.)

So instead of making up some new model, you can just answer whether you have any disagreements with Newtonian mechanics. If you do, then that is the problem, if you don't then the answer is already known:

If you claim the center of mass of your device ever has a non-zero velocity, (A Box moving from point A to some non-overlapping point B means that it did) then your claims violate both momentum and energy conservation, and you have made a mistake.

[Bob Shaw]

Please could the mods simply stop allowing reactionless drive posts on NSF! They are a total waste of space, time and resources!

[DanClemmensen]

Please could the mods simply stop allowing reactionless drive posts on NSF! They are a total waste of space, time and resources!

I'm perfectly happy to allow any discussion of these drives after they have been published in either Science or Nature or if a Nobel prize has been awarded.

[RotorRiddle]

Momentum Conservation and Frame Displacement
The displacement of the frame is not a violation of momentum conservation—it is a manifestation of it. The system is closed, and all internal forces are equal and opposite at any instant. However, the temporal distribution of impulse is asymmetric due to shaped force profiles (e.g., Hann window modulation). This results in a net momentum transfer to the frame during the constructive phase, which is not immediately canceled by the destructive phase.

Over a full cycle, the total momentum of the system remains conserved, but the center of mass of the frame is displaced. This is analogous to certain internal actuation systems—such as asymmetric mass drivers or vibrational locomotion—where net displacement occurs without violating conservation laws, provided the internal force asymmetry is time-dependent and cyclic.

Energy Conservation
This is a subtler issue, and I appreciate your attention to it. The energy required to generate the asymmetric impulse profile—particularly during the constructive phase—is not free. In the manuscript, this energy is drawn from the rotational kinetic energy of the counter-rotating masses. As the system imparts net impulse to the frame, it transfers energy from rotational to translational modes.

To maintain constant rotational energy, a motor continuously drives the rotation axes. Thus, the system is externally powered, and energy conservation is upheld:

Input energy: rotational kinetic energy (via motor)

Output energy: translational kinetic energy of the frame + internal losses (e.g., friction, heat)

The system does not create energy; it transforms it through internal dynamics. The key is that this transformation is directionally biased due to the shaped impulse profile, enabling net displacement without external mass ejection.