Disclaimer: Students who read this are forewarned that it is informal speculation, not hard science. It deals with things such as gravity shields which may not even be possible in our world.

Warning! Do not try to build a gravity shield at home. Consider the unintended consequences if it should actually work, though that outcome is highly unlikely. H. G. Wells played with the idea (fictionally only) in his 1901 book The First Men in the Moon. It supposes that a gravity shielding material, called "cavorite" might be used in a flying machine, and even to lift a flying mchine to the moon. You can read about one unintended consequence in Chapter 2.

Elementary observations.

• If the wheel were to experience a rotation through a small angle, it would be the same in every respect as it was before. So we have no reason to expect it to initiate such rotation of its own accord.
• Stevin's principle shows that it won't work.
• The laws of thermodynamics show that it won't work.

Then there's these somewhat more interesting observations:

• If the wheel of the diagram is imagined to be rotated clockwise from a stationary position, through a small angle, to a new stationary position, a small sector of mass near the top moves from a region where its mass is effectively zero (on the left top) to a region where it again has "normal" mass (at the right top). But at the same time an equal sector of mass has been moved from the bottom right into the shielded space, and loses that same amount of mass. So, whatever static position the wheel assumes, it has the same mass distribution, and there's no change of its center of mass caused by such displacement.
• This same argument applies whichever direction the wheel is rotated, leading us to conclude that the arguments for "one-way" operation of the uniformly symmetric wheel version of this machine were flawed.

The second observation recalls and refutes the original rationale/explanation of why this machine should work, which predicted only one-way operation of the machine.

But what about the other form of this engine: the eccentric weight? One who thinks this ought to work might argue as follows:

If this one starts at the top, falling in the unshielded region, it gains speed and attains a certain speed at the bottom. On entering a completely shielded region it experiences no gravitational force, so its speed is unchanged (constant) all the way up. It gains more speed falling on the right, then going back into the unshielded region. Thus it begins each cycle with an additional amount of speed (and kinetic energy) more than when it started the previous cycle.

This modified machine casts serious doubt that our previous "solution" can be applied to this version.

We also observe that this rationale predicts that the wheel would gain kinetic energy when revolving in one direction, but lose kinetic energy when moving in the other direction.

All of this leaves nagging questions. What, exactly, does a gravity shield do, and is a gravity shield even a possibility?

More detailed observations.

We know that we can make electric field shields, and even partial magnetic shields. Is there any fundamental proof that gravity shields are impossible? There are experimenters right now (2002 CE) who claim to have built an apparatus which reduces the gravitational field slightly in a small region. But their experimental claims are suspect. Theoreticians produce very persuasive arguments why such a shield is impossible. We'll let the theoreticians and experimentalists argue about this one. In our usual spirit of generosity to inventors, let's grant that somehow they could obtain a gravity shield. Could they use it to make a perpetual motion device? Suppose that part of the region above the shield is really gravity-free or at least has significantly reduced gravitational field.

Just what would a gravitational shield do?

Since no one has yet made a gravitational shield, and we don't know whether anyone can, we should rephrase the question. Just what do we want a gravitational shield to do?

We want it to reduce (or increase) the effective gravitational force in a localized region of space, in particular, in the region encompassing just part of the path of a mass moving in a closed path. This is a modest goal. We don't need a perfect or complete shield. One way we might imagine to do this is with a device that produces another gravitational field strictly localized in that finite localized region of space.

Field shields must also obey the fundamental laws of physics. "Magical" shields are not allowed.

If the fundamental principle of vector addition of forces is still valid, then this field of the shield, together with the earth's gravitational field, would allow us to have part of the path of the moving mass pass through a greater or lesser strength field than the other part.

Our problem reduces to answering the question: "What energy changes occur as the body passes through the boundary between the fields?" This could be a messy problem. I don't have a simple answer.

I have attempted to avoid calculus or higher mathematics in these documents. That's difficult, and error-prone. Perhaps in this case it is impossible. One reader wanted to dig deeper, and here's a mathematical treatment he wrote. Analysis of the gravity shield engine, by Dr. Peter Moomaw. His document helped me to correct and remove some indefensible comments I had perviously had in this section of this web page.

For another take on this interesting machine, see Kevin Kilty's analysis.

Most recent edits May 2008.

—Donald E. Simanek

Return to The Museum of Unworkable Devices.