« Previous Page

For a couple of days now I have been attempting to get a nice long series of shots to gage the velocity consistency of this machine. It seems like we get two or three data points and the way forward is blocked by something that needs adjustment or repair. Here are the shots in the order they were done.

Shot.     Velocity in fps.           Comments.

1             329.8                  First shot after some unspecified washer rotation.

2             326.4                  I’m liking it.

3             307.5                  Not so much.

And then we tighten the washers 15 degrees all round.

4             351.6                  That added some zip.

5             352.8                  Very good .

6             349.0                  Loving it …… but….

The bowstring is tightened and the serving is replaced.

7             365.0                  Okay, a bit more zip … but…

We add 7 ½ degrees more rotation to the starboard washer for better limb balance, and….

8            394.4                   Hoopla! And it sounded fast too….. but…

One of the leather buffers on the port side falls off. This minor detail altered the angle of the at rest limb position for the next shot, (consequently the string does not get picked up in the center, all of which throws out the limb balance). Stop, breathe , repair.

9          368.5                     About as expected.

10        362.5                     Not horrible.

11        125.3                     Twilight zone.

Mildly disbelieving of this last number, I stared at the chronograph for a good long while. That ugly little “125.3” stared back at me.  I reminded myself that we should be running things through two chronographs just to verify the instrumentation.  It looked like another mystery…… until Rebecca poked her head from around the corner,  “Could that have anything to do with it?”,  she said, pointing  at the bowstring that had  snapped 12″  from the starboard end.   The pathetic remains of the string drooped from either limb like a pair of dead snakes.   Where it had failed looked as if a firecracker had exploded inside the braid.


Despite its dramatic failure, whatever actually happened to the bowstring occured without any noticeable noise or flapping around  Probably it succumbed  to one of its own dance moves. Too bad we didn’t have video rolling at the time.  Looks like we need a protocol for that too.  Clearly we must lower expectations on how quickly we can get stuck into accuracy testing.  And that is where things stand today.   Stop, breathe, figure it out. ……..perhaps a hawser with double braid eye splices is what we need.   Too thrilling.

Dear anomynous person that asked about teddy from the photo posted in “The terror of the double braid eye splice”,  here is our earliest record of him.  He thanks you for your kind inquiry, and yes, he is in the best of health thank you.

Go ahead and dance silly string, you’ve already shot your bolt.  By the time you straighten up, your enemy will be falling down.  Click for video,  string


Sometimes it is interesting to compare utterly different things.  For example, 2885 grain Dura bolt @ 394.4 fps has twice the penetration in newspaper as a .44 mag, 240 grain JHP out of a 4″ Smith & Wesson.  The bolt made it through 12″ of baled newspaper and appeared to be stopped only by the pinch on the swelling at the rear of the shaft.  The .44 slug, being a hollow point and not designed to maximize penetration, barely made it to 6″ when fired into the above bale.   Totally different things of course.  Interesting though.  Yesterday’s bolt made it through 12″ of newsprint plus 1 1/8″ of plywood.  And that at a slower velocity too.  We can only think that the density of the newsprint must have been different in the bale we used yesterday.  Perhaps we hit the financial section or something.

Spent most of the day tuning the bundles.  Pretty tired now.   I’ll make this short.  Last shot of the day was a stunner.  2885 grain Dura bolt got up to 394.4 fps,  (120.21 m/s,  268.91 mph).  Energy was 985.93 ft/lbs. (1337 joules).  We are still only on notch #17, with 50″ of draw length.  The final notch, if we dare take it back that far, is #23 with 63″ of draw.  Various snicks and pops were heard around mid-draw @ 3000 lb draw weight.  Final draw weight for this shot was 3600 lbs.   The shot before this got up to 365 fps, 844.42 ft/lbs.  Power settings were changing each time I tightened a bundle attempting to get the limbs balanced.  The difference between 365 fps & 394.4 fps is a mere 7.5 degrees extra rotation  applied to both the top and bottom washers of one bundle.  The limbs are almost balanced now, good enough anyway for the next phase,  consistent velocity tests.  What fun,  we are moving into the extreme sport aspect of this project now.

Here’s a quick vid of that 394.4 fps shot.  I was a bit stunned by the increase in velocity, given that the only change made was the extra 7 1/2 degrees of twist put into one of the bundles (around 3/8″ movement on the outside of the washer)  Click here for video, 394-fps

The previous shot @ 365 fps could have used some more newspaper backstop.  Click here for video, 220090313182427

Or should that be “screamin’ Dura Europos bolts”.  I couldn’t resist making a few shots today with the new 2855 grain bolts.  As soon as I get more time, I will make a chart with all of today’s data and its implications.  For now, suffice it to say, the best shot of the day showed a velocity of 352.8 fps (107.53 m/s), with an energy of 788.91 ft/lbs (1096 joules).  That particular shot was one from a string of three made at the same power setting. All three shots were within 1.5 fps from one another.  This bodes very well for the accuracy tests that will be coming up in a month or so.  If the powerplants of this machine can drive out consistent velocities, all the other variables that might adversely affect accuracy are essentially fixable with good tuning.  Today has been a profound high point for this project.  I am pretty elated by it all. 

We managed to capture that 352.8 foot per second shot with the slow motion feature on the new Sony SR-11.  Slowed down 4 times, it looks like this. Click here for video: 20090311170012

And the last shot of the day at 349 fps, was recorded in real time.  Click here for video, 20090311170640

I sense that these new bolts are a  bit on the light side.  We will bump up their weight by 500 grains and see what that does.  So many things to try.


Ten years ago splicing the eyes that form the loops in the bowstring for the Gallwey ballista, was a  challenging affair. The passage of time has not made the task any easier.   Still, I got stuck into it this afternoon and after half a dozen tries managed to spice a pair of loops the correct distance apart (that’s the real trick).  Now, with the limbs reinforced with more steel, Dura Europos bolts ready to go, and a new camera to record all the action, it is time for the games to begin.  High power testing will commence this Friday and end when………….. I guess it will end when something breaks or the springs show signs of losing their memory………… kinda  sounds likes a metaphor, doesn’t it?









Whipping over the splice is an effective lock to help prevent slippage.  The above string will eventually receive a full whipping so all traces of the modern dacron yacht braid disappear.  Cordage is one area where this project makes no pretense at authenticity.

Grappling with the dreaded double-braid eye splice.  Colored felt pens are a handy way to mark the entry and exit points for the fid on the core and outer braid.











A full wrap of spring steel strapping around the heel of the limbs and larger chafing gear to protect the spring bundles.  Just about ready to rock.

We had hoped to be into some souped up shooting this week.  As it turned out, what free time I have for this project was spent trying to forge out a complicated pair of spring steel straps to wrap around the butt of the limbs.  Quite a bit of end grain is exposed by the concave depressions for the bundles, it is going to need a little help from this 3/16″ strapping if it is ever going to survive full draw.  For all their structural disadvantages, the depressions do a good job of forming the cross section of the  spring into a circle.  This not only keeps the individual cords from spreading out and interfering with the field frame, but also prevents any lateral shifting of the limb.  There are of course other ways to accomplish this,  but my intuition keeps hammering at me that this design is the most stable way to grip the limbs.  I also suspect there may be performance reasons why having the cross section of the spring form into a circle at this location, may be important.   More on this later.

Also, we have ebayed our way into a better camera for the various video footage that is coming up.  For camera buffs, “HDR-SR11″ probably means something.  Sony’s “smooth, slow record” feature should come in handy for slow motion analysis of any more mysteries like the wobbly string.


The trimmed down proportions of the new 18″ long Dura Europos style bolts, compared to the clunkers I have been shooting, can be seen in this photo.



The beastie on the top is left over from ten years ago when the Gallwey project was going full steam.  Clearly I had no idea of what a decent ballista bolt should look like back then.  Even so, these 2 1/2 foot long, large winged projectiles proved to be both formidable and accurate when shot out of the Gallwey (1150 foot pounds, 6 moa @ 50 yds.).  Weighing in at around 5000 grains, they  did a good job of soaking up the energy from that machine’s heavy limbs.  The question of ideal bolt weight is vital to optimum performance.  Because the bowstring will only move at a certain speed, and will exhibit only so much tendency to be slowed down by increased bolt weight, it follows that too light a bolt may not have enough mass to take full advantage of a power stroke that would be unfazed by using a heavier bolt.   In other words, the highest velocity bolt is not the same as the highest power bolt.   Apart from the frontal area and sharpness and construction of the tip, the deepest penetration will be achieved with a bolt that balances its mass against the machine’s ability to drive a heavier projectile without being unduly slowed down.  Too heavy a bolt will lumber out of the machine with such an arcing trajectory it will have little real world utility.  Bolt weight has a sweet spot for any particular class of machine,  something between a needle and a log is what we are looking for.  As a starting point, we will assume that the Dura Europos bolt at around 3,000 grains is just about right for the Orsova ballista.  From there we can test to see if that premise is correct.  The relationship between any given bolt’s foot pounds of energy and it’s actual drop in trajectory will pretty well tell us what is ideal.  One thing is for sure,  the Romans knew where the sweet spot was for the type of bolt used in their Orsova machine.  Perhaps we can reverse engineer the type of fodder this machine liked best.  That might give the archaeologists a baseline of comparison for any new finds.  Maybe not though.  The imponderables of matching ancient projectiles to ancient ballistas,  does seem daunting.     ……..Unless of course we could prove that there are certain characteristics about a style of bolt that make it particularly suited for use with a particular type of machine.  For example, bolts for inswingers benefit from having small fins that do not interfere with the acutely angled string at full draw.  Because of this, the Dura Europos bolt seems to be a perfect candidate for an inswinger.

The mildly comic effect of fin interference can be seen in this video as the string fails to pick up the old Gallwey style bolt,  oops

And here returned to its usual vigour after the fins have been radically cut back,  trimmed

And in other news: To test its innate stability our new version of the Dura Europos bolt has deliberately not had its fins added yet.   It was fired into the catalpa log at a low power setting, something like 26″ of draw with bundles that are in dire need of tightening.  Even without fins it flew quite straight,  no doubt the swelling at the rear of the shaft has a stabilizing influence.    We didn’t have the chrono set up, but just eyeballing it the bolt couldn’t have been going over 100 fps.  Very wimpy.   Amazingly, the new style tip buried itself over 3″ into the wood.  This much penetration perhaps says as much about the relative softness of catalpa, as it does the new sharp tipped bolts.  If we had being firing at full power the darn thing would probably have sunk in seven or eight inches.  It certainly indicates the need to come up with a better backstop medium if we don’t want to be chiseling bolts out of catalpa logs for the rest of the year.  Compressed newspaper maybe the answer. I am wondering how well a box of those rubber pellets they use as kid catchers under playground equipment might work.  Right now it is a bit of a problem.  We need something that the bolt will stick in, yet be brought to a stop within six inches or so,  and then be fairly easy to remove. Any suggestions would be very welcome at this stage…….


The following photos of some original quadrobate heads were downloaded from the website for “Ancient Caesar”, found on ebay.  Some serious penetration going on with these babies if they can be kicked up to 300 fps or so.

Circa 1st-3rd century AD.  Length: 7.8″ (198 mm),  weight 1527  grains (99 gm).  Probably compatible with an Orsova sized machine.


Circa 1st-3rd century AD.  Length:  9.250″ (235 mm),  weight  2932 grains (190 gm).  This would be a dandy for the Gallwey.

« Previous Page