The Ballista

The webbing and “sinew” lashing seen in the photo below is what I came up with three shots ago to stop the string from flying off the limb.   Power levels have increased to the point that the  string feels little obligation to hang around if it ain’t lashed in place.

The gold coloured nylon binding around the limb is controlled by grooves in the back and belly of the limb.  All of this grooviness is far enough out on the lower stressed end of the limb,  that the grooves don’t threaten the integrity of the tip.  Note how the webbing is doubled back on itself under the binding.  It’s firmly secured.

This next video shows a longer stretch of the cocking sequence.  Click here, 20110429170026(2)

Because operating this winch is a full on chore for an old fat guy with a sore back, it follows that the mechanism is probably perfect for a couple of stout young lads with muscles.

Firefly has always aspired to be a plausibly authentic design  based on the Orsova artifacts.    Clearly a machine this large would have been a crew served affair.   If I am ever going to get to the bottom of all this,  I am  going to have to recruit a crew or attempt to deduce the probable man/seconds it would take a crew of necessary size to cycle between shots.  Until I can locate a crew,   I fear that photo ops and philosophical meanderings will be the limit of the hand winch’s usefulness.

The hand winch requires a force of about 93 lbs. on the end of it’s 67″ lever to rotate it’s 2 1/2″ drum against 5000 lbs of draw weight.  In the sense that the hand winch was intended to be operated by at least two individuals,  I am very pleased with it’s current performance level and have no plans to make it more amenable for one person to operate.   That task can be better accomplished by the miracle of electricity.  Sustained testing by yours truly,  will require the 12 volt Warn winch.

The shot seen in the video below depicts rope bundles that are working through a severe case of the slushy’s.  This is the segment in the adjustment schedule  when the washers are still being rotated  to torsion-up the bundles, but before they hit a definable wall that will only give way in a unified and controlled manner,  i.e.  like a spring.  This stage of indifference in the springs is always a bother.   It pays to go carefully now and make washer adjustments slowly.  I can hear acts of slippage going on in the bundles when she is being cocked.

……… That sounded……. never mind…..

Here’s the vid.       20110429170026(1)

The new and improved interface between  the hook and sideplate teeth can be seen in the bottom of these next two photos.

Neighbor Richard says that these more acutely angled teeth look kind of “flamey”.

New and improved, with flames coming out.  Sounds like a ballista salesman’s wet dream, doesn’t it?

In no particular order we observe the following:

(1)   The new blocks that pack out the crossbars are holding position nicely.  Sinew binding may be unnecessary.

(2)   As we have noted before in this blog,   it seems suspiciously fortuitous that after the springs have been  stretched to the maximum both linearly and torsionally,  that the crossbars always seem to end up directly over the vertical stanchions in the field frame.  Is this deliberate?   Would a sinew based torsion spring be so obliging as to place the area of maximum stress directly over the area of maximum support?  Are we seeing the original design express itself here?

(3)  The teeth in the side plates that are bolted to the sides of the stock, do not have an ideal angle for trapping the dual hooks that are attached to the trigger assembly.   It seems that the new hand winch is pulling at a slightly different vector than the old electric winch.  Because the locking angle is too shallow on the tooth and hook interface,  there have been repeated instances of the hook riding out of the teeth on the sideplate.  The fix is to recut the teeth and hooks.    This is such a different part of the machine for me work on,  this is actually going to feel like a vacation.  Which leads us to……

(4)  It’s only work.   Never give in.

(5)   I am finally putting some sense to an observation I’ve had before:   there must be a certain amount of slippage of the rope over the crossbar, especially because all of the materials doing the slipping are waxed nylon.  (The nylon rope is bearing down on a piece of nylon seat belt webbing that covers the crossbar, and the whole area has been smothered in gasket wax.)    By prodding with a stick,  I am able to feel that the areas of slackness in the spring that were not touched by our linear stretching, are quickly firmed up by the torsional twisting and are indistinguishable from the areas that originally held more firmness.  In short,  I have come to see that the spring has such powerful on-going forces at work that tend towards normalizing the length of all the ropes,  I suspect that the special, four point winding jig,  mentioned the other day, will be superfluous.  Maybe such a jig will make more sense if I start making springs from nylon thread.  With thread it seems there might be less tendency for slippage over the crossbar.

(6)  Never ever,  give in.     ……. And also,  in answer to Bonnie’s question of what is my technical definition of “finished” regarding this project:   for now, a working definition might be, “When I’ve run out of stuff to fix.”

The washers have been rotated to start the torsioning procedure on the springs.  They are still pretty slushy because I intend to work up their torque levels very slowly.    This is quite time consuming because I need to shoot the machine several times between the washer adjustments.   The object is to progressively wring out  any pockets of slackness while at the same time moving the two spring bundles closer to a state of balance.

I only had a couple of hours to fuss with it today and I will have to put it all aside for a few days because other duties are clamoring away at me.   The three anemic shots I did get off answered one question at least:   at 45 degrees of draw the minimal nocks for the bowstring seem to work very well.  I didn’t  have any instances of the string flying off the limbs at the end of the power stroke.  This is surprising considering that the string loops were not lashed to the limb as I had planned.  I am grateful for this development.  Hopefully things will stay like this when the power gets cranked up.

The new hardened support blocks for the crossbars have been installed.  It took 4400 lbs of linear stretch on the port side, and 4200 lbs on the starboard side, to pull the crossbars apart enough to slip in the blocks.   Actually those forces were applied to just one corner of the crossbar,  (i.e.  three blocks were installed and the final one could only be popped into place when the aforementioned forces were applied to the end of the crossbar directly over the last block.)  I had taken the time to carefully adjust the thickness of the blocks to get both  springs this close in balance to one another:  an exercise that was complicated by the fact that, in terms of overall length,  the starboard spring is 2/10″ longer than the port side.  This discrepancy naturally required two sets of differently dimensioned blocks.

I will not bind on the blocks with sinew yet.  I’m curious to see if the sloped bottoms on the new blocks will curtail their outboard migration.    Below, Firefly has been fitted with a low grade test string just to draw the limbs back far enough to extract the bronze hardstops for further fit up work.

Now that the linear stretching of the springs is complete,  tomorrow I will start the torsional tightening of the washers.  After a round or two of gradually rotating the washers and pulling back the string,  the static position of the limbs will settle down and I should be able to get a pretty good fix on how much to relieve the bronze hardstops to get them fit to the new limbs as sensibly as possible.

A good fit up between the thickened impact area on the limb irons and the detachable bronze hardstops,  is strongly recommended  by my little friends.   They say everything needs to be peachy-keen in the limb impact department before we advance to any live fire exercises.

The retention notches on these new blocks have been heavily chamfered  with a round file.  They are smooth enough for “sinew” now.

On a completely different note:  In further studying how the spring is deformed when the limb  is installed inside it, we see the probable cause why the inner strands of the spring, the ones up close and touching the captured limb, are strained tauter than those on the outside of the bundle.   These inner strands are forced, by the limb, into a more acute angle than the outer ones and are thus stretched tighter.  The current springs were pre-wound on a simple jig consisting of two pegs with a straight line lay-up of windings between the two poles.   This two pole jig does not really represent the final shape the spring must take.   If the jig had a set of additional pegs to represent the  limb when it is inserted into the bundle, then the lay-up of the windings would take much less exception to any bend in their middle.   They would find the final insertion of the limb into their midst a fairly normal state of affairs as they would have been, quite literally,  born to it.  With such an improved jig, the outside and inside of the spring would be more evenly strained.  All the fibers would have an equal say,  it’d make ’em more democratic.

In my experience any braided form of rope  always performs poorly when used as a spring,  at least when compared to a similar volume of three-strand.  I have always suspected this is because the fibers in braided rope crisscross back and forth at a multitude of link points, and this self-intersecting disposition of the fibers dampens their linear  extension to a level unacceptable for use as a spring.  Three strand rope has its fibers aligned more in a linear direction and  even though they  spiral,  there are no pesky changes of direction to ruin their elasticity.  Because this somewhat straight line disposition of the fibers in three strand rope seems well suited to spring making, we might conclude that an even straighter line disposition of the fibers would be better yet.

Perhaps some kind of thread, not unlike nylon artificial sinew,  would be a good bet when we get around to winding up  new springs.   In a way,  Sir Ralph suggested as much himself when he  claimed that a great number of finer, individual strands were better suited to making a torsion spring,  than a limited number of whacking big ropes.  Like I said,  good springs are largely democratic affairs.

The limbs seen here are ready for “final” installation.   The fit up tests I have done so far indicate that the detached  bottom irons I’d been considering a few days ago,  no longer seem necessary for a robust capture of the spring bundle by the limb.    That sounds backwards, I know.  It is sometimes useful to reverse the usual perspective and remember that the limbs  actually do have a controlling effect on the bundle,  just as much as the bundle obviously has a controlling effect on them.    Even though the limbs are not supposed to change shape,  they are in some ways,  independent cusses.  It is important to remember that they are under a lot of stress and can have a mind of their own when it comes to doing what is expected.

In the photo below, these metal blocks that hold up the crossbar have their down slope side facing in towards the center of the spring bundle when they are installed in the machine.    The inboard half of their concave channel has been crushed by the flexing of the spring steel crossbar.    These are the perils of using mild steel in such a high stress application.

After a while this damage to the concave channel  becomes sufficient to form it into a shallow graded slope.    Eventually the angle on this slope becomes enough to cause the block to shoot out of the machine much like a bar of soap  squeezed  by slippery hands.  It seems these blocks can have minds of their own too.

The solution is to make up a new set of blocks from hardened steel.   At 60 RC,  they would not crush under load like mild steel blocks.  I will also include a deliberate back draft angle  (i.e. the bottoms of the new blocks will have a  slope cut on them such that the inboard end  is thicker than the outboard end.)  This will cause the softer bottom of the u-notch, in the ears on the washer, to deform into a sloped surface.   The sliding movement this creates will propel the blocks inboard, thus locking them in a sensible direction.

Further capture of the block can be achieved by including a large notch on the outside of the block to allow a stout whipping of artificial sinew to bind it into the ear.   It is imperative to make these blocks listen to reason.

For general reference, the  photo below is an oldie from last season and shows the blocks  fitted under the crossbars.

Nosing the air on a fine spring morning,  Firefly wonders when she is going to get up and running.

It’s all very well  pushing for completion,  but, as a general rule, it doesn’t do to try and meet arbitrary deadlines when you’re building catapults.

Why?

It can form clots in the sequencing.  Horses pushing carts and all that.

Well,  no clots allowed then.

Thank you.