As the washers and crossbars are tightened, they move in the direction shown by the red arrow.    Their current location is fortuitous because it looks like the final position of the crossbars will max out  right over the top of the stanchions.    This is ideal because the unsupported portions of the end caps will not be subject to the massive pressure coming from the crossbars.    All my earlier springs  had limited  linear tension cranked into them,  so the spiral in the spring was much greater, and that allowed the crossbars to  rotate around far enough to end up midway between the stanchions.   As mentioned in an earlier posting, this caused some minor bending of the end caps (now all straightened out) .    In response to this minor disaster, the vernier plates were  upgraded with some spring grade 4140.   However,  with the new straighter springs, it appears  I could have saved myself the trouble.  (What is a vernier plate or an end cap?  see posting from 12/25/2008,  “The case for the vernier plate”).

With the crossbars looking like they will have their rotation max out so close to the stanchions, it does kind of raise the question:  is this just good luck?   Or are we looking at some fundamental elegance inherent in  the original Roman design?   I am betting that any torsion spring worth it’s salt (nylon, sinew, or horsehair), and installed with a maximum level of  linear tension, will naturally have the crossbars max out pretty close to the stanchions.   Besides,  if they rotated much beyond the stanchions, the springs would begin to lose that  desirable condition of,  “straightness”.  No matter how you look at it,  it does seem like a positive development.   I sense that the original designers of the Orsova machine, were very much aware of these dynamics.

One Response to “Is it just luck? or is it Memorex?”


  1. Captain Harpoon. says:

    Am I correctly understanding from your above description of tightening that you turning both top and bottom crossbars in the same direction?

    In later posts I have been searching for indications that you are indeed tightening both the top and bottom crossbars in a single direction. I remember that in one of your videos that the top and bottom crossbars when tightened are indeed perfectly alligned.

    Because the limbs load in a clockwise direction, it would serve little purpose to load or tighten the rope bundle in the same direction. As you go to rotate the limb or put load on it, it would feed slack into the equation, negating the tightening effect. Therefore is makes sense to me (initially anyway) That the top crossbar must be rotated counter clockwise, and the bottom, since it is a mirror image (backward)must be rotated the OTHER way…

    While mentally engeering twist into the rope bundle without the limb inserted is easy enough as it is very plain that to introduce twist the easiest way is to rotate each crossbar in the opposite direction to ensure an even twist.
    In this example, there is only twisting point which is in the middle, evenly spaced between the crossbars.. (A lazy an incorrect way would be to rotate only either the top or bottom.)

    The mental engineering and modeling is then complicated by the addition of the limb or arm, which by virtue creates in inverted image,and one which (in theory)acts the opposite of the other.

    (I seem to believe that somewhere in your machine is one component which is out of whack, robbing your machine of its true potential or maximum effeciency. I am searching for that component(s).

    While it is indeed possible to “tighten” up the springs by turning both top and bottom crossbars in the same direction,
    only one half of the spring will provide resistance or twisting power (the top half).

    The result would be an unequal loading with all of the twisting power coming from one half of the spring, which over time might result in twisting the machine to one side and causing structural failure in the Kamirion, stanchion an support post, folding the machine inward at the top. A non tensioned or loose string I would think would only compound that tendency due to the effect of the stanchion or stop acting like a fulcrum instead of a stop.

    When the limb is inserted into the bundle, it also introduces a second point point of twist. One per each side of the limb, top and bottom.

    Now here’s the kicker (remember that an unbalanced system will always seek to restore balance…). In an ideal situation, the twist point on each side of the limb should be in the middle or exactly between the two points of contact (limb and crossbar, or collar perhaps)where it is perfectly in balance.

    To move that twisting point to one side or the other (closer to either bar or to limb) creates an imbalance. If the twisting point is VERY close to the crossbars (like your current setup)it really screws things up so that “huddling” of the ropes occur which we have learned is a loss in power and performance.

    However, if the twisting point is moved toward the limb, then performance is improved resulting in increased power and more importantly limb speed.

    This fact can be readily and very easily checked for yourself (recommended) by a simple experiment I conducted today with a brick, a piece of string, and a couple of hooks.

    Most red clay brick have a hole or several holes in them. All that is required is to put the string through the hole in the brick, and then tie both ends of the string to a sincle contact point, so that the string is triangle shaped.

    Give the brick a quarter or half twist and watch the speed which it returns to zero or rest.

    Now, take a second hook and set it the width of the brick apart from the first hook. Attach each end of the string to a hook and repeat the experiment. The width of the hooks and the width of the brick are the same, so the twist point is now in the middle and the machine in balance.

    The final experiment is to set the distance apart of the hooks to twice the width of the brick.

    Notice how rotational speed increased by a multiple when you went from one hook to the two? And rotational speed again increased when the distance of the hooks were wider than the width of the brick?

    HOw does this relate to the performance of your machine? Thats easy, substitute the widthor thickness of the crossbar to the hooks, and the width or brick to that of the thickness of the limb.

    The fix should be to both increase the width or thickness of the crossbar (rounded rope contact surface is good)while similtaneously reduce the thickness of the arm or limb.

    There is still two potentially performance robbing items of the current limb shape. It is my thinking that the limb should be round as opposed to flat – in order to increase distance between the two contact points of crossbar and limb.

    The current limb seems to be approximately five inches thick(high?) so that it takes away from the length of both top and bottom loop. Without the limb inserted into the bundle, there is just the one loop per side, but two per side after limb inserted. A round limb would probly add another two inches of loop to both top and bottom loop – a good thing in this particular situation.

    Too much Grist?

    There is a lot of really good information in the above post. The brick experiment is excellent to see exactly how twist or torque is created, what factors effect it and how they effect end performance.

    In simplest of term my findings are that in order to produce the highest speed and performance, the crossbar could do with a larger width while the limb itself(where it contacts the rope) undergo some major surgury.

    On another day or another post perhaps, I’ll share my theory on why the curved stanchions(purpose and importance of) – theory still a little rough around edges. I would edit this to make sure it does indeed make sense, but its now 1:30 am…Please email to discuss further as nec.

    Captn Harpoon
    (Imperial Whaling Fleet of Japan LOL).

Leave a Reply