Copyright 2001 by David Eugene Cowlishaw, all rights reserved.



* OLD NEWS * . NEXT UPDATE .* INDEX *. * Site List *

Driven Cone Gear
Improvement, by

12:32 PM, Tuesday, 26 December 2000

Hello again, many were wondering if "they" got me, it's been so long since I've updated my web site (nearly a YEAR! :)! No, they didn't get me, but after WELL over three years at this, I'm no longer as anxious to get us launched yesterday at my own and total expense! I have maintained (though delayed and spotty) my email efforts, and passed along the gains to those hard-core email group adherants. The IIC Group (International Inertial Consortium) are folks that are willing to put up with my singular ability to talk around any subject with fluff and bother, before getting to the point! The IIC email Group also includes most of the pioneers of this project.

Many things have happened over that year, resulting in a good number of improvements, both in theory, and in machine development, which resulted from the better understandings. Pictured above is the latest configuration of parts, that I believe will be THE winner! A very elegant and easy to implement MATH relationship of all moving parts (for engineering use) fell into my lap, but only after many virtual models were generated, and similarities in each, shouted out for me to pay attention! I will tell you what that is, and what it can do for us, later in this update (bear with me, HEY, it's BEEN a while since I talked with you last, I gotta lot to say!)

My past updates and previous documents detail the earliest understandings, and are spiced with examples of the personalities that made it possible. You are urged to explore! (see): DavidC's site list!. Also see the previous web site, in a compressed folder of the beginnings of this project at:

In there I detail examples of the work of many other unchained intellects, that have helped advance this work with me. These include over 20 physical verifications of my concepts and findings by others through several "incarnations" of machines that worked, and then we set about understanding WHY! :). They've convinced me that I was on the right track, and kept my spirits up in this battle (yes, I see it as such)!

Here in this update, I will give you my latest attempt at conveying the principles of operation as currently understood, assuming you have a rudimentary (high school) knowlege of motion physics. I will not attempt the Greek translation, you can find a recent one at: Constantine's math analysis. This is a viable one, according to several responses from many in my international research group that can "sling the Greek"!

SO, on with the show! :)

Principles of Operation

The most promising parts configuration of the concept is called the ALFAmax GIT (MAXimized Angular to Linear Force Accelerator, a Gyroscopic Inertial Thruster project series development).

The ALFAmax consists of one or more ring driven planetary geared ring masses of variable pitch diameter, that travel about a stationary eccentric sun gearing, which in combination, provides for an inversely matched cyclic "cross-over" of momentum exchange between the linear and angular accelerations of those orbitals, with the linear and angular mass characteristics of the frame. This effects an externally demonstrated thrust on that frame. (in other words, wind this baby up with a lawn mower engine, a coal-fired (or even, shudder... nuclear!) steam engine, or WHATever, and it GOES! :)

If you get nothing else from reading this page, understand this: An eccentric mass in a variable velocity orbit (an off-center lump that speeds up and slows down repeatedly in a circular cycle), is a POINT MASS that will interact with BOTH the LINEAR MASS aspect of the frame (it's linear position/velocity in space), AND the ANGULAR MASS (moment of inertia) of the frame, which describes it's angular position/velocity in relationship to space.

A mass's linear (point) mass aspect, accelerating in an angular (circular) path, crosses the supposed "barrier" between linear and angular momentum! AS such, it can become a bridge for us to carry angular (spin) accelerations out into the linear world, and "force rectify" them to produce external (linear) thrust!

Here's the set up:

Create a machine that has a "variable velocity orbit" of an eccentric weight, but NO orbital spin accelerations as yet. It speeds up on one side, and slows down on the other, a big chunk of mass generating an action on the frame, that has a high velocity, angularly constrained mass at one end of the cycle, and a much slowed down mass swinging around at the other in a frame, HOPEFULLY to generate a forward surging "centrifugal" thrust!

 Ideal 2 body motions of
 variable velocity orbit,
 Inertial thrust attempt. At the right is a graphic depicting this attempt at inertial propulsion. Note that the projectile is traveling in the top section at twice the angular (and instant linear) velocity of it's slow pokey velocity around the bottom loop. The purple sections are say, servo coils that accelerate and decelerate the point mass orbital. This motion example is relative to the tubing frame only, and not how it really works.

The assumption is that the faster traveling mass will produce a greater centrifugal force at the front of the orbit, and a lesser centrifugal force at the rear where it's poking along. While that's true, many ignore the fact that the bilateral side acceleration zones, also both apply negative impulses to the system (where it's slowed down, left in this graphic, and sped up again, shown on the right). All together, the action/reaction sum of this simple attempt goes no where!

Nope, variable velocity alone won't work, simply because you have to observe empirical fact (they ain't flying)! The momentum exchange between the orbital mass, and the frame mass, is NOW expressing a back and forth (even a bit oval shaped) point mass oscillation (in real space), along with an angular oscillation, a so called "washing machine motion" or "the twist" (a 1950's dance for you whipper-snappers, somewhere between Bach and the Beatles in music history! :)

When the orbital slows against the side of the frame, the frame moves backwards a bit (slows it's forward lunge actually), and starts turning in the direction of the orbit, to absorb a share of momentum from that action of a point mass deceleration, as a spin reaction by the frame.

At the rear of the orbit, where the eccentric mass is expected to be at it's slowest velocity in the frame, the frame (along with the moving center of system mass following the eccentric about), is now turning (relative to the universe), along with it! This obviously complicates the "minimum centrifugal force" assumption that would result, if we only examined the point mass relative to the frame!

Using only the point mass aspect of the orbital in this setup, the actual motions are very similar to those shown in the next animation below and to the right.

On the acceleration side of the orbit, the frame begins turning opposite of the orbit direction, and thus the "real space" maximum forward velocity of the eccentric mass is actually slower, than if we were only examining it relative to the frame.

 Actual motions of
 variable velocity orbit,
 Inertial thrust attempt. In short, the "angular momentum" of the point mass interacts only with the "angular mass" aspect of the frame, and the left-over instant linear velocity change components interact with the linear (point mass) component of the frame.

Our (frame rotation weakened) forward centrifugal lunge, is negated by three negative linear impulses. These include both negative tangential impulses on either side, adding to the strengthened negative centrifugal turn around at the tail, to completely cancel any net external linear velocity change. It dances a hula, and does the twist at the same time, but goes NO WHERE!

NOW, consider that much of the instant linear velocity that the orbital's POINT MASS imparts to frame spin, is actually absorbed in a mass aspect of the frame that does NOT interact directly with the external frame of reference, as far as net external LINEAR acceleration goes! The real space velocity of the orbital is changed by reacting with a mass EQUAL to the sum of BOTH the linear AND angular transfers to the frame! In other words, the frame absorbs impulses on it's sides, and that portion of negative momentum of the point mass transfer to the sides of the frame (that results in a reversing SPIN acceleration of the frame), represents the action/reaction source, from which we will extract our thrust!

NOW ADD THE ANGULAR MASS OF THE ORBITALS into that dance to nowhere, and we can do an amazing thing, we can conga (actually "The Swim") right on outa here! :)

By increasing the spin rate of the orbital as it slows down on one side of the orbit, the orbital's two mass aspects generate a pair of competing (head to head) torques on the frame. This is an angular to angular momentum exchange, between mass expression aspects of the same object, through an external angular intermediary, which is the moment of inertia of the frame (it's angular mass), which is all proper and legal of course! (~8,' )

On the other side we decrease the spin rate of the orbital, which dumps orbit direction torque into the point mass of the orbital (through the frame), while it speeds up again in the orbit. On the orbit velocity acceleration side, we again put the two torque aspects "head to head".

If we exactly match the orbital spin torques on the frame in each side cycle, to completely cancel out the tangential frame torques of it's own point mass velocity changes on that frame (inverse on either side), we can completely cancel out the "washing machine motion" in each cycle! Our ride just got a bit more comfortable!


By adding in an "outside mass" aspect of the orbital's own moment of inertia to interact with the original three material players in our dance (I call it angular mass to point up it's self-evident existance!), we can actually transfer forward orbital point mass momentum (in real space) into the frame as an ANGULAR result! The resulting frame spin will NOT interact directly with the linear aspect as a negative linear impulse. We THEN turn around and absorb THAT transfer (as real space motion expression) with another mass, as a countering angular exchange in a spin moment change of the orbital! We in effect, suck up some of the instant linear velocity change as an even MORE internal angular result (and externally non-expressive as a linear result), as an orbital spin change!

The spin change of the orbital absorbs and releases some of it's own negative (instant linear) momentum changes, into and out of that orbital's own angular mass (moment of inertia), simply by "bouncing it off of a backboard" of the frame's angular mass! Hey, it's called conservation of angular momentumum uhhm ... err... sort of! :)

That spin change does not interact directly with the external linear forces in play on the frame, but DOES represent a place to "hide" some of the negative point mass velocity changes towards the back where we don't want them, and "show" them again towards the front, to effect a net linear thrust (potential) result! (Hey it's only potential until the craft moves, THEN we have to worry about "work" and fuel consumption, more on that later!)

Now as the maximum frame velocity of the orbital passes the nose, that frame is nailed (angularly) in place, instead of having the frame turning anti-orbitward, to defeat it's highest (real space) velocity aspirations! The fully calculated forward centrifugal lunge (if only considering the orbit frame), imparts it's gift of motion to the frame in the forward to backward turn-around as forward acting "centrifugal force"!

Instead of the decelerating orbital first stopping the anti-orbitward spin of the frame as in the simple example, then sending the frame spinning orbit-ward, (to tag along with the slowed orbital), NOW, by changing the SPIN rate of the orbital, it will SWALLOW up it's own negative (angular "definition") point mass velocity changes, as a spin velocity change ON BOTH (negatively acting) SIDES!

At the tail, the frame is again held angularly still, and the former (larger) negative centrifugal addition to the negative will (that the washing machine dance added to keeping us still), is more in line with calculating the smaller negative centrifugal force, relative to the frame only! Less to hold us back, more to propel us into the future, HEY, I call that THRUST! :)

A respectable amount of the two negative impulses on either side get transfered away from negative external linear effect with the spin change mass aspects of the same orbital, and drop out of the back side of the originally balanced linear "tug of war"! We are counter-twisting the frame with the orbital spin changes, to work in opposition to the twists that the side accelerations of the point mass of the orbital places on each side of the frame.

The short explanation for finding the amount of thrust this thing will generate, is to break out the traditionally calculated "angular momentum" exchanges that the eccentric places on the frame, THEN add the two sides together. We must take care NOT to directly cancel the side torques by sign, as is traditionally done, since the tangential impulses are now no longer acting in direct opposition to one another through time. Then subtract that (since they are now hidden in equally acting orbital spin changes), from the rear end of the tug of war in each orbit cycle!

It still dances a hula, but without the twist (with a single orbital), yet now takes a net step forward MORE (beyond momentary velocity) with each swing of the hips! Each orbit cycle adds a net forward impulse to the "rest velocity" of the craft, and only time and mass can tell us how fast this thing will accelerate! Consider that it takes one year at one gravity of acceleration (nuclear sub tech at most!) to reach light speed, we can visit a lot more than the local planetary neighborhood! :)

That's the wordy explanation as to what we are doing, and takes into consideration most of the objections of "establishment science" based on momentum conservation principles. I hope it got you a little closer to understanding why this thing works, perhaps so you can clearly show others how our future can be a LOT brighter, if we look for the solutions, instead of trying to ignore them!

NOW the only objection remaining is the pervasive concept that linear and angular momentum can not be exchanged directly with one another in a "closed system". To the best of my knowlege this misconception is not a "law" attributed to any science icon personality, it just sort of crept into our limited (dogmatic) understandings of our universe! Hey, it took a LONG time for the world to become round, I guess 4 years at this without global acceptance is not such big deal after all! ;)

Now for a more in-depth exam of the latest machine variant, and the math that has folded in around it!

Sr = (Cr+1)/2 : Tr = Sr/Cr : Rr = (Sr^2) / (Tr^2)

Driven Cone Gear
Improvement, by
DavidC At the right is a shrunken version of the topside graphic (if your browser is working properly! :) showing the latest version of improvements. This model uses spur gears at the tips of the orbitals (where the cone gear reaches it's smallest diameter), to engage the ring drive with, which also drives them about the stationary center eccentric sun gears.

Also included on each side of the bilateral orbital is a cylindrical bearing face, that rides on a matching cylindrical surface on the ring gears. The cylindrical bearing faces keep the orbital teeth from grinding the ring drive gears into dust, and transfers the centrifugal force from the point mass of the orbital, through ring retainer roller bearings, to the air frame or other vehicle mass.

By eliminating the "yoyo axle" split in the orbital (previous ALFA incarnation), fewer parts are needed, it strengthens the ultimate loads it can endure, and simplifies balancing, among other attributes. It also happens to generate a simply elegant math relationship among the mass and dimension aspects of the orbitals in relationship to the frame!

In every tip driven ALFAmax that I have examined, from a Cone ratio from two to eleven (ratio of the largest pitch cone diameter divided by the smallest), for every orbit diameter tested, and for every angular moment and maximum diameter of the orbitals in relationship, ALL obey the same relationships described herein:

Cone ratio = Cr = Largest cone diameter divided by smallest. Ie, a 5" max cone diam with a 1" minimum = a Cr of 5.

Spin ratio = Sr = (Cr + 1) / 2. This is the ratio of the spin rate of the orbital at the tail (highest spin velocity), divided by it's spin velocity at the nose, which is ALSO equal to the Cone Ratio plus one, divided by two, FOR ALL TESTED VARIANTS! For our 5 cone, that would be (5+1)/2 = 3, or for all tip driven orbitals of 5 to one, the orbital will be spinning 3 times faster at the tail than it is at the nose!

Tangential ratio = Tr = Cr / Sr. The tangential ratio represents the instant orbit velocity of the orbital at the nose where it's greatest, divided by it's (smaller) orbit velocity at the tail. It is also equal to the Cone ratio divided by the Spin ratio! Wow, what a find for me (and now you! :)

For our 5 cone example: Tr = Cr/Sr = 5/3 = 1.667 Tangential ratio. For all orbit diameters and average orbit velocities, the orbital will be traveling 1 and 2/3rds faster at the nose than it does at the tail (or 3/5ths as fast at the tail than it travels at the nose)!

Momentum and force calculations are now MUCH simpler to derive using Delta V (change in velocity over time or distance factors) to find accelerations, which will integrate into our dimension and mass factors!

Hey, not my job, I'm an IDEA guy! :)


Using the factors above, we can NOW integrate mass and distance into a palpable and useful thrust expectation, by observing traditional physical parameters, and juggling them for our desired result, which (last time I checked my email), is Inertial Propulsion!

Radius Leveraging

Some intuitive characteristics of the ALFA become apparent. One is that for a larger orbit relative to the inertial moment of the frame, a greater share of the angular expressions of the point mass velocity changes will operate on that frame. This can be utilized for positive and negative thrust difference, and it's negative linear aspects become smaller (actually get transfered to a non-linear accelerated mass expression), resulting in greater thrust potential. The hollow center shown will likely be a good place to put the motor! :)

Smaller orbitals (for any given orbit diameter), can provide for greater side accelerations (more room to hit the gas and brakes! :), and/or an increased number of orbitals. This gives us more thrust potential for any considered mass as parts. Unfortunately, as the orbital gets smaller in diameter than the orbit, it has to change it's spin rate considerably more, in order for the small radius of the orbital's angular mass, to compete with the "leverage" advantage that the point mass has in swinging about it's halting arc of travel!

We are working with torque (inch pounds, meter kilograms, whatever! :). When pitting the twist that the point mass aspect has on it's "lever arm" actions on the frame (in opposition to the smaller diameter of the orbital's moment of inertia), we have to increase the "pressure" on the shorter moment arm of the orbital's angular mass. This is to balance the lower pressure, but longer arm of the point mass acting on the frame's moment (at the orbit radius).

For all machines having orbits greater than the diameter of the orbitals, we will need greater spin changes than tangential changes, in order to make an efficient machine. Incidently, for every variation of the tip driven ALFAmax, the Tangential ratio is "infinitely approaching 2", and we will never have a Tr greater than or equal to 2 for this machine definition (shades of Zeno's paradox! :).

An analogy would be a 300 lb. weight on a one foot arm, balancing a 100 lb. weight on a three foot arm, balanced, equal in effect about the pivot point, but clearly unequal in force at the points of force application! REALLY small orbitals would have to change their spin velocities REALLY hard (compared to it's orbit rate changes), in order to become efficient. Increasingly smaller orbitals (per orbit size) increases the needed difference of largest teeth to smallest teeth on the cone gear, and raises material strength concerns. High efficency will become synonymous with high stress factors for the parts.

We need one more ratio to calculate our machine, and I call that the Radius ratio. For any cone gear definition, an ideal relationship can be found between the orbit radius (point mass relevant calcs), and the length of the moment arm of it's matching angular mass. Remember that I=M (moment of inertia = mass) for a thin ring mass or hollow cylinder, but we will be dealing with a "ring mass equivalent radius" in other shaped orbitals. In example, a solid disk of 3 inches radius, would have a Ring mass equivalent (Rme) of a 1.5 inch radius thin ring, as I=.5m for a disk.

If we use our 5 cone example, and assume a ring mass of 5 inch diameter (actually, the cone teeth extend to nearly 6 inches, so a fairly thick ring can achieve an angular moment of M*2.5" or better), we can then find the ideal orbit diameter.

Using our lever (actually balance beam) analogy, and having found the Spin ratio and Tangential ratio, we can filter out the canceling entities and get right to a number. Force is acceleration is pressure, and analogous to "weight" on the unequal balance beam. Assuming the mass quantity of the linear and angular componants of the orbital are equal (at the "ring mass equivalent" radius), we then can drop out mass, and concentrate on pressure to achieve balance.

The smaller moment arm of the orbital's angular mass aspect, has to have a "heavier" effect (greater acceleration force), to balance the point mass on a larger (orbit radius) moment arm, but since the mass is equal, the only thing we can change, is how fast we accelerate them!

The Spin ratio represents end spin velocities of the orbital. The 5 cone has the orbital spinning 3 times faster at the tail than at the nose, while it's orbit velocity is 1 & 2/3rds faster at the nose than at the tail. The actual acceleration forces can be found by "Delta V" (change in velocity) method. The acceleration ratios are a fixed relationship of velocity differences that holds true for all constant ring drive rotation rates, and are bound together by mass and circumstance! :)

Mass being equal, and having known velocity differences, we now only need to find the orbit radius (with a smaller velocity change difference), that will match the assumed or engineered length of the orbital's angular moment. We need to find the exact radius relationship to give us matching (and opposed) torques on the system frame.

F=MA (force = mass * acceleration) and also = mV^2 (mass times velocity squared). Since the mass quantity on both sides of a torque contest is equal, we can cross cancel that out, leaving us with acceleration (also v^2), and radius on "each side" to get us to equal!

The shorthand of all my verbal dancing is that I square the Spin ratio and Tangential ratio numbers (velocity differences squared), divide the larger by the smaller (Sr^2)/(Tr^2), to achieve an ideal length of the orbit radius for any orbital's moment of inertia (Rme).

Rr = (Sr^2) / (Tr^2) = Radius ratio
= Spin ratio squared, divided by the Tangential ratio squared.

Now all of the elemental math elements are in place for engineering effective tip driven ALFAmax thrusters! (Ok, I coulda used the word "archetypal", but thought that a bit stuffy! :)

For our 5 cone example, the Radius ratio is (3^2) / (1.667^2) = 9 / 2.778 = 3.24. Multiply the 2.5" Rme radius of the orbital by 3.24, and we get an orbit radius of 8.1 inches, or an orbit diameter of 16.2 inches. This relationship SHOULD exhibit NO "washing machine motion" and will "force rectify" the orbital's spin accelerations to give us an efficient linear thrust result!

Obviously we can have different angular moments for the orbitals, so a smaller Rme would need a smaller orbit, and for a larger Rme of the orbital, we would need a proportionately larger orbit, to keep the torques balanced (where our linear to angular exchanges take place! :).

While "overbalancing" the spin torque can result in a larger net thrust impulse per orbit, I believe power transfer will suffer. It will do a "washing machine" dance, but in the opposite direction(s) of it's simpler version's performance. For those of you with electronics experience, I use an "impedance matching" analogy for the most efficient power transfer, in arriving at my decision to exactly match spin torque to orbital torque in design element justifications.

I will follow up on power considerations after the next section.

A Force Balanced Model is already modeled!

While I'm still on the subject of the tip driven ALFAmax variant, here is one already mapped out in virtual reality for you to do the math on (to find actual thrust and stress forces for). Hey, I've other things to do! :) I'm currently working on a smaller 7 cone version I might be able to afford to model in reality. I may be making this larger one available in a mold set (eventually), if there is sufficient interest BEFORE I get one self-lifting ('cause when that happens, I think I may get busy elsewhere! :). I'll repeat the graphic for your visual reference convenience (no need to retrieve it from the net, it should already be in your hard drive temp files! :).

Driven Cone Gear
Improvement, by

The pictured model is a 5" to 1" cone gear (one I already have generated in casting material), in a 16.667" orbit, and would require an Rme for the cone of (16.667"/3.24)/2 = 2.572" ring mass equivalent radius (5.144 thin ring mass diameter equivalent). It has 99 teeth in the eccentric sun gears, and 213 inside teeth in the ring drive, so the ring turns (99+213)= 312 teeth for one complete orbit, 527.323944 degrees, or 1.465 ring revs/orbit. I chose to add a tooth to the "normal" drive ring tooth count (at 17.667" pitch diameter, 12 teeth per circular inch pitch), to make the orbit numerically divisible by 2, 3, or 4 orbitals (for equal time spacing).

The 99 eccentric sun gear teeth divided by the 12 of the orbital, gives us 8.25 contact orbital spin, plus one for the orbit, which results in 9.25 actual (real space) revolutions of the orbital per orbit.

The flat geared tail section is equal to 47 teeth (a 90 degree wedge section of a 188 tooth flat spur gear). This is shown as the lower left blue circle sections of the eccentric sun gears in the rendering.

For a 90 degree travel of the orbital through the flat spur gear nose section (7 teeth), it has 300 degrees of (actual) spin. The same ring advancement that gives us the front "centrifugal impulse", has the orbital traveling 54 degrees in the tail section, while spinning 900 degrees. The ring turns 101.831 degrees (.282864 turn) during the complete 90* nose transition.

The two left over acceleration zones (parabolically curved to distribute acceleration rates evenly) occupy the remaining 180 degrees of orbit, split into two bilateral inverse acceleration zones.

If you like math puzzles, this one doesn't require fancy math skills, and can give you an appreciation for the power this machine might achieve once one is built! If you're proud of your work, post it on the net, and tell me about it, so I can link it in! :)

Power transfer considerations for Inertial Propulsion

Speaking of power transfer, I have seen the "work ethic" hold true for this device. Move a mass, energy is "consumed" (actually converted to another form), and through standard physical definitions (divorced from the non-supported but supposed barrier between angular and linear momentum), energy only changes forms.

Picture the above "Tube bullet" animation as an example (ok, if your imagnation is that poor after being exposed to earlier input, scroll up! :), and also refer to the earlier animation showing no frame rotation changes. If the airframe moves during the forward centrifugal turn around (momentum transfer at a linear level), and that energy expression that is not totally recaptured in the recycling of the thrust reaction masses at the tail, energy then is "consumed" as far as we are concerned. Forward acceleration of the craft will cause a decay in the average orbit and ring drive rates, so we must put that much gas into the tank, PLUS our machine friction losses, to restore the decaying (transfered momentum depletion) ring drive rotation rate!

Interestingly, for a free body in linear space, work may be relavant to reclaimed work, as in "inertial storage" reuse technology of motion exchanges with "potential" storage of energy. One thought is in dropping from orbit height in a controled fall, we MIGHT have to leave room in the batteries to store the change in potential energy to motion (drop in height), or burn it off! So far this is only speculative, but something to watch for as this project progresses into better testing (SOON I hope!)!

One thing I'm fairly comfortable with, is that if we achieve an average orbital turning rate, that exactly cancels out the total gravitational pull on the craft (a hover), the fuel cost will be equal to the machinery friction losses needed to sustain a "one gravity thrust" as against a wall for example! If work is not performed in the classical sense (wall doesn't move, work isn't done), I forsee some rather cheap fuel costs for just hanging around! :)

That same one G turned sideways can boost a lifting body into a stellar performer, and maximum forward velocity in air will be dependant on airframe friction losses, along with machine friction needed to keep a steady one G thrust pushing on a craft that is no longer accelerating. In other words, some very interesting energy problems and solutions still need to be found as far as efficiency is concerned, and all I'm doing is pointing out a few other "impossible" (but now probable) places to look! :)

A few more thoughts from the year involving this concept:

In machine design, a greater number of orbitals will smooth out the "bumps" that result in the lop-sided motion changes, but will also cancel the thrust potential of one another on either side, if two or more are allowed to be under opposing acceleration at the same time. The tangential accelerations of each, will directly cross-cancel, and the orbital spin changes will also transfer directly, leaving the frame out of their mutual battles as a neutral angular intermediary.

Of the machines I've sculpted in virtual reality, most have a two orbital system that appears to be the most efficient. The delay factor of the slower orbit rate at the tail can have both orbitals in the flat gear tail zone at the same time (at constant angular and linear velocity) , one just arriving, as the other is just leaving. A single orbital makes the tail to nose turn around (hitting the gas and the brakes in the go-round) all by itself, so the other orbital never gets a chance to do direct momentum battle with it.

Another concern of mine is in the recurring theme of locking two ALFA's together with opposed orbit directions and timing, to hopefully effect a cancelation of the side to side movements of the combination. This might work if the opposed orbitals don't exactly mirror one another (first one, then the other pass the nose for a double forward punch), but I'm fairly certain that the exactly mirrored machine will not work.

I'm also concerned about a "hard mount" coupling of the air frame's angular moment into that of the thruster, which would really bollix up the momentum transfers under consideration! I see at least a "sliding mount" that would allow side to side oscillation of the thruster seperate of it's air frame, and feel better about a "spin coupling" break of the slide bar with the air frame. For instance, if the thruster is in a "sloppy" circularly changable mount (we can point it anywhere around the 360 degrees), then we can insure linear to angular transfer with the orbital, instead of one or the other twisting the air frame. If not, a slight mismatch in the orbital's linear and angular transfers might not be allowed to fully properly transfer, and instead, add unwanted angular vibrations in the cabin. An angular isolation of the thruster can prevent a momentum path and expression we don't want.

Oh, from a recent letter (thanks Rich Schweitzer!) I need to address "Time versus force" in the centrifugal examination of these machines. A plausible source of "thrust negation" is embodied in the thought that the slower moving orbital at the tail, while having a lower force, applies negative thrust for a longer period of time. This actually happens, some of the negative thrust CAN be blamed on time difference front to back, but for a radially constrained mass, we have to observe current facts.

If we use the above simple graphic animation example, we have a nose velocity of twice that at the tail. This means that the "centrifugal force" at the tail lasts for TWICE the time of that at the nose. HOWever, centrifugal force is equal to mass times velocity SQUARED, divided by the radius! That double speed orbital is placing FOUR TIMES the force on the "main axle" than that applied at the tail turn around! If we have 1 "bump per second" at the tail on the main axle, and it lasts for two seconds, that gives us a negative 2 bumps per second total. Compare that with the nose, at twice the velocity (half the time), a one second nose time travel gives us 4 bumps per second! The nose wins by two bumps per second (in examining ONLY the "centrifugal force" in an angularly stablized variable velocity orbit)!

For centrifugal force considerations, time is merely additive, while force is squared for differing velocity examinations.

Well, I need to get this on the net (so you can read it)! I may add more to this update a bit later, once reportable progress is made. I am still trying to get a cone gear and variable sun gear made in material, but work is stalled awaiting a "miracle" of money for the accuracy and material strengths I need. There has been talk of investment, but I'm not holding my breath! :)

I have made an offer to corresponding folks interested in seeing money put into this project, and will make it an open offer. Since published machines and improvements can NOT be patented (by me or anyone else), all I can legitimately offer is to grant title to the first professionally generated prototype, in exchange for it's construction expense.

I will provide the design in digital format (most CAM and Rapid Prototype formats can be generated quickly). I keep design rights for the exact model (but am willing to license or transfer copyrights if the price is right), and I get to test it after taking molds of the parts.

I can NOT guarantee performance on an as yet untried model, prior to testing. Hmmm... no patent rights, can't say it will actually work as expected... like I said, I'm not holding my breath! :)

Until then, DO take the time to read all the past material on my site, and if you don't mind waiting a week or more for an answer, write me at, and let me know what you think! 11:24 PM, Thursday, 11 January, 2001

New Alphabetic listing of this site will be updated regularly

Use the Old News link to download previous updates, a LOT of good info is to be found there and nowhere else!

Forums and Chat at:

* OLD NEWS * . NEXT UPDATE .^ TOP of Page ^ . * INDEX * . * SiteList *