慕容青草的博客

哲学园地

首页文章列表博文目录

个人资料

慕容青草 (热门博主)

给我悄悄话

博客访问：

Verification of the Unexplainability of Dzhanibekov Effect by th

(2025-11-15 19:32:21) 下一个

Rongqing Dai

Abstract

Based on all the videos of the Dzhanibekov effect from space stations of different countries available to the public, it is a reasonable assertion that the Dzhanibekov effect cannot be explained by the existing known physics, while the mainstream academia has been trying to avoid admitting this in their erroneous analyses of the Dzhanibekov effect over the past three decades. Nevertheless, due to the shocking nature of the said assertion and its logical openness, it would be desirable to get it verified through advanced computational analysis and lab experiments. In this paper, a conceptual design for an AI-aided experimental verification of the said assertion is proposed, and the outcome of its realization is discussed.

Keywords: Dzhanibekov Effect, AI, Experiment, Verification, Known Physics

1. Background

Concerning the famous Dzhanibekov effect, since early 2025 I have pointed out that the academic explanations of the Dzhanibekov effect over the past 30 years are completely wrong; accordingly, based on all the videos of the Dzhanibekov effect that have been sent back from space stations of different countries and available to the public, I have made such a reasonable (preliminary though) assertion that the Dzhanibekov effect cannot be explained by the existing known physics. I do not need any additional verification for my statement that “the academic explanations of the Dzhanibekov effect over the past 30 years are completely wrong” since I have clearly identified where they went wrong [[1]]. However, regarding the unexplainability of the Dzhanibekov effect by the existing known physics, although I am fully confident in the logic of my analysis, it would be beneficial to get it verified through some AI-aided experiments in order to convince the public, due to its negative nature of denying the existence of certain open possibility. This type of negative conclusions about a certain possibility usually requires special verification because of its logical openness. For instance, if someone points to a petal on his floor and says, “I see a petal over there,” he is making a positively definite statement and its verification is a straightforward simple job; but if someone says, “there isn’t any petal in this entire building”, then he needs to provide some special verification to prove that his conclusion is correct due to its negative openness.

This paper proposes a conceptual design for verifying the assertion that “the Dzanibekov effect cannot be explained by the existing known physics” through AI-aided experiments. The verification process mainly consists of two parts: training the AI and using the AI to verify the assertion.

2. The conceptual design for the verification

2.1. AI training

To verify the assertion in question, we need to first train the AI through the following three main steps:

Step One

First, educate the AI with relevant knowledge such as Newton's theory of gravitation, rigid body mechanics, some aerodynamics knowledge, as well as the mechanics of the motion of the space station and all objects within it. However, the Euler equations of rigid body dynamics should be excluded due to its defect [[2]].

Step Two

Let the AI that has learned the above knowledge solve examples of the movement of rigid bodies in the atmospheric field under the influence of the Earth’s gravity, such as the flight of bullets, the fall of objects from a height, and even the movement of quasi-rigid bodies with little deformation like volleyballs and balloons, as well as movement of a space station and the objects within it.

The main training objective of this step is: based on the known parameters of air temperature, pressure, humidity and wind speed, as well as the distribution of the mass and moment of inertia of the moving object, calculate under the given initial conditions: 1) the coordinates of the trajectory of the object’s center of mass; 2) the variations of Euler angles of the object; 3) the velocity, angular velocity, acceleration, angular acceleration of the object at different time points.

Step Three

Unlike the previous step, in this step, the AI is not given initial conditions. Instead, it needs to observe the motion of actual objects in the atmospheric field under the influence of Earth’s gravity through video recording and provide the following information based on this “visual” observation:

1) The equation of coordinates of the motion trajectory of the object’s center of mass:

(x,y,z) = (X(t),Y(t),Z(t))

2) The equation of the variation of Euler angles of the object:

(α,β,γ) = (A(t),B(t),G(t))

After having the above two sets of equations, it can calculate:

3) The velocity, angular velocity, acceleration, angular acceleration of the object at different time points;

Then, based on the known distributions of mass and moment of inertia of the moving object and the actual measured parameters of air temperature, pressure and humidity at that time, etc, calculate:

4) The forces and torques acting on the object at different time points.

5) Verify the conservational status of momentum and angular momentum according to the above calculations.

Once the AI has been able to pass through these three steps very proficiently and reliably, we can move to the next step to use it to verify the said assertion about the Dzhanibekov effect.

2.2. Verification of the assertion

Under the premise that the AI is already capable of handling the previous step three, the verification process should be relatively simple: have the AI watch the videos of the Dzhanibekov effect recorded on the space station, and inform the AI of the position of the space station and its speed at the time of video recording, as well as the relevant physical parameters inside the space station and the mass and moment of inertia of the rotating object.

3. Discussion

3.1. Regarding the training steps

For ease of discussion, I have made the calculation of an object’s motion in the atmospheric field under the influence of Earth’s gravity with the given physical conditions a separate step from the calculation by the AI based on its own visual observation in combination with the given conditions. However, in actual training, it is best to merge these two steps for the following reasons:

With the traditional engineering mechanics, calculating the free motion of a rigid body in an atmospheric field under the influence of gravity is not an easy task, especially when the Euler equation of rigid body mechanics is not allowed to be used due to its defects. Engineers need to use tools of computational mechanics such as finite difference and finite element, and many technical challenges might arise when solving for a completely free rigid body, especially under the premise that the gravitational field itself could change. Even if the AI is equipped with superb computing power, it is probably not a wise choice to do the required calculation solely by traditional methods.

On the other hand, allowing the AI to visually observe the movement of objects can significantly reduce the difficulty faced by the traditional computing. In traditional calculations of engineering mechanics, the calculation of an object’s position, deformation, velocity and acceleration, as well as the calculation of the forces acting on the object, are determined by a highly coupled nonlinear system of equations. Once AI can visually observe the three-dimensional coordinates and shape changes of an object from all angles by itself, it can directly calculate the position, deformation, speed and acceleration of the object’s movement. The remaining tasks are only the calculations of forces (including internal and external forces as well as temperature and pressure, etc). To verify the said assertion on the Dzhanibekov effect, as long as enough cameras are installed in the experimental section of the space station, the AI can conduct a comprehensive visual observation of the movement of the object and thus record all the required parameters.

3.2. Two different possible results

There could be only two types of outcome of the verification:

1) AI discovers that when the Dzhanibekov effect occurs, angular momentum is completely conserved.

When this happens, since my previous theoretical analysis indicates no possibility of such outcome, if it is verified that the AI’s calculation is error-free, then we need to analyze what kind of forces could cause the periodical flipping around the axis of the least moment of inertia of the rotating object, based on the forces calculated by the AI at the selected time points, from which we should be able to know why the mainstream academics over the past 30 years, as well as myself in the past few months, have missed the forces identified by AI that would cause the Dzhanibekov effect. In case such a situation really occurs, it should be helpful for us to better understand aerodynamics.

2) AI discovers that when the Dzhanibekov effect occurs the conservation of angular momentum is violated, which validates my aforementioned assertion that the Dzhanibekov effect cannot be explained by the existing known physics.

4. Final Remarks

From the videos of the Dzhanibekov effect of space stations recorded by various nations, we have not been able to identify any type of forces that can cause the periodical flipping of the object around its axis of least moment of inertia (and thus the changes in the angular momentum). However, if all the potential dynamic factors involved are not tested to the greatest extent possible, there will always be a concern deep in people’s hearts: since the combined effect of air and gravity could normally be very complicated, maybe something has been overlooked during pure theoretical analysis for the cause behind the Dzhanibekov effect if it is conducted solely based on human visual observations. With the computing power of AI, once the conceptual design proposed in this paper is successfully implemented, it will be able to determine with certainty whether the assertion that “the existing known physics cannot explain the Dzhanibekov effect” is correct or not.

If the proposed AI-aided experiment verifies the assertion that “the existing known physics cannot explain the Dzhanibekov effect”, it will undoubtedly be a shocking result, because it means that we have confirmed that Newtonian mechanics has a loophole, and the next step will be clear: to figure out how to modify the Newtonian mechanics based on the Dzhanibekov effect.

However, even if the proposed AI-aided experiment denies the assertion that “the existing known physics cannot explain the Dzhanibekov effect”, all the efforts made to train AI as proposed above would still be meaningful. This is because the AI trained according to the design given in this paper is clearly not only capable of verifying the assertion in question, but also has great practical potential of doing many things.