Next Page »

It seems that every time I look at someone else’s vision of how an inswinger must have looked, the machine is invariably striking a pose with it’s limbs showing dramatic amounts of rotation and, presumably, extra long draw lengths.  This cool painting by Mr. Dominic Andrews shows what I’m getting at:

Dom 1


Have they never heard of stacking?   The leverage angle between the bowstring and the limb becomes so nearly like a straight line it’s just not possible to pull it back any more.  The springs are at maximum strain just when there is minimal draw back power to get them that way. There is a fundamental illogic about that arrangement.   Anyway, Firefly wasn’t having any of this over-rotated hype.  A short, intense, 45 degrees of limb rotation is where she found her power.  I’ve heard the theories about how, with inswingers, long draw lengths and big rotations are better for throwing heavier objects.  Maybe so, but there didn’t seem to be anything to it when Firefly was set up that way.  Of course, Pumkin Chunkin ain’t exactly her forte.

I don’t always believe things just because mathematicians tell me they are true.  Real world modelling has to count for something too. Maybe the math isn’t big enough.  Of course, in fairness, maybe the modelling isn’t big enough either. Be that as it may, with Phoenix, I’m sticking to Firefly’s proven strategy for nice flat trajectories with Dura bolts:  45 degree limb rotation with an extra-extra taut resting bowstring, along with all the rotational pre-tension that implies.  Locking in a couple of tons of linear stretching first, makes an excellent foundation for all the rotational stuff you are going to have to spend so much time foolin’ with, if you want to birth a balanced and powerful machine.

And if that all sounds a little bit serious, that’s because it is.  (Rant alert)  There is just too much work in this hobby to go down any more fruitless paths.  If someone from the school of super-duper rotations wants to show us how to shoot an authentically weighted Dura Europos bolt a thousand yards from an actual machine (as Firefly has done), then we are all ears.  In the meantime, my experience with Firefly (here we go again) warns that the stacking from big rotations would require a low rotational pre-tension to achieve full draw, which causes the bowstring to be loose when in the just fired, at rest position.  This lack of tension in a resting bowstring is coincidental with sluggish velocities — not at all the thing after so much work is put into dreamin’ big.

The particular game I’m playing does not favor the Hamish approach:



On the subject of accuracy and precision, this graphic just about says it all.


Firefly’s long range shooting has so far been typified by that sketch on the upper right.  While it is true that on a calm day she has enough precision to repeatedly hit something the size of a Volkswagen bus at half a mile,  it is doubtful she could do so unless the driver first parked his rig in the middle of an established pattern.  The Romans would have had to engage in some clever maneuvering to get their foes into a pre-determined kill zone for this type of long range ballista fire to be effective.  Pre-sighted artillery would be the trick.  High-precision machines like Firefly could act similar to long range booby traps, just waiting for someone to step into their “beaten” ground.  Half a dozen machines trained on the same spot might well nobble Otto….. if he could just be persuaded to pause by that big oak tree….

On our fifty yard shop range, where Firefly’s peep sight can be utilized, the performance is much like the sketch in the lower right.  Accuracy and precision makes all sorts of trick shots possible. For example, back in October of 2011,  we had good luck juicing apples perched atop pumpkins. (Unfortunately, Firefly remains hidden in the shop because we had no way to move her back then.)  Click for video:  2011102711243911


When she is dialed in, Firefly can do this kind of duty with monotonous regularity.  Such is the nature of precision and accuracy when they are combined.


And not to forget the coconut balanced on top of a wine bottle.   That is, a full wine bottle, certain to incur the wrath of the Rebecca if it gets broken.

Click for chunks of flying coconut:   201110221249222


Center shot,  just like in the movies.  Except, of course, this is real.


Firefly will very reliably toss all of her bolts into a 2″ circle at 50 yards.  When the wind is calm she can shoot 3 consecutive bolts into a 156″ circle at 800 yards.  Those two data points I have good confidence in because they have borne up under repeated testing. Other intermediate distances have been difficult to experiment with because the bolts keep shattering when they impact the available backstops.  But I am determined to get a decent 150 yard group before winter hits, even if I have to sacrifice the eight remaining projectiles I have left.

In the sketch below, I included a speculative 4″ group at 100 yards.  (Not enough data yet to be really sure. That’s why it’s marked in red. Possibly dodgy.  Probably should disregard.)

MOA = minute of angle.

Click to enlarge:



Just for fun, I thought I would put it to all the math people that are not reading this, how they would predict the group sizes at the other ranges based on the 50 yard and 800 yard results. Firefly will be filling in the blanks with more testing, so there will be real world corrections and corroborations.  (Yeah, I know, that’s all sounds a bit squeaky clean doesn’t it?)

And here’s a photo of that 156 inch group at 800 yards.  (Actually 790 yards, but round numbers are good enough at this stage.)

13 group

And here is one of the many 2 inch groups we knocked out on our small shop range.

2 inch group

That flyer was a sighting shot, while the ragged one hole group was made by 7 bolts fired consecutively at 50 yards.  The highly authentic Roman uniform and helmet were just a lucky find from Goodwill.

I’ve put the winch battery on to charge.  Which can mean only one thing.

Unless, of course, you have enemies.  In which case, it might be a good idea to skip the middle bit.

And the hero of the hour is ….. fanfare please …….


Today, Oona warned us of a big fat rattlesnake coiled up under the couch. (No, they don’t always rattle.)  The pesky serpent would have been about a foot away from my leg when I sat down. That would have been a drag, to have his boorish little fangs stuck in my leg. I like that leg too, it’s the only good one I’ve got left.

So, kudos to Oona!  …….And ticker-tape parades, and dog cookies, and ball games, and swims in the lake, and fetching ballista bolts…….. What an uber hound!

As a craftsman I am struck by the irregularity of the top loops on the Lyon field frame seen in yesterday’s post. (I don’t impute some special design strategy to their mismatched shape and placement.)

loopy 1

The unsymmetrical sizing is not the result of sitting underground for many centuries, it was made that way originally.  This is not what I would call a high order of forge work. Don’t get me wrong, I have no doubt it worked just fine.  But that is as much a testament to the adjustable nature of a design that uses wedges to lock it all together, as it is to the precision of it’s manufacture.  “Good enough”, I suppose we could call it, and think no more about it. However, to me it suggests that these machines were made ad hoc, by local talent, rather than at a centralized armory with presumably higher standards. If I were running the joint, I wouldn’t let that kind of thing out the door.

(Yeah, but your’e a machinist with a rectilinear little brain, of course you see it that way.)

Not just a machinist.  I’ve spent my time at the forge and know plenty of blacksmiths around here that could do a much better job of it than what I see coming through in the artifact.  Is this the kind of degraded product we might expect as things unravel at the end of Empire? Although, 197 A.D. is not exactly “late” is it?  Those goofball top loops just look like sloppy workmanship. Slave labor?  Lowest bidder syndrome? Hangover? There just doesn’t seem to be any pride in them.

Back in the day when I worked as a modelmaker at Boeing’s wind tunnel, we had a common expression to describe a lot of the work that came through the modelshop,  “quick and dirty” we called it. Management was almost exclusively concerned with how quickly a project could be completed.  So much so that we grunts that undertook their bidding had a saying,  “They never remember how quick, just how dirty.”   I strongly sense something like that is afoot with this field frame from Lyon.   “Bang me  out a dozen of these by the weekend and I’ll double yer grog ration for a month.”   Triple time, in Boeing speak.  You can just sense whoever was doing it wasn’t going to take the time to use some bending blocks and make a neat job of it.  It’s not difficult to do symmetrical work at the forge if you are not being harangued by upper management.  Very intimidating if they are packing swords, I’m sure.  But it’s really no excuse is it?  Sloppy work is sloppy work.  Not at all the thing for impressing the natives, fortifying the troops or keeping the Emp. off your back if he spots all that shoddiness out on parade.

Unless of course, it was just a decoy ballista. Like one of those rubber tanks from WWII.  There is usually a fairly pressing reason for management to go all quick and dirty like.

(No, serious people, I don’t really believe it was a decoy.  We’re just building up a head of steam here for what comes next.    …..Those loops are still some pretty sloppy work though. I remember seeing Roman garden implements with more symmetry)


Update:  According to the rant above, the basic design of iron frame ballistas must have made them the AK-47 of torsion engines. Regular blacksmiths could knock together a rough looking machine that no doubt worked tolerably well.  Loops and tangs and wedges allow some amazingly discordant geometries to come together with great strength and forgiveness.

Update of the update:  Coming back to this a couple of weeks later, I’m not so sure there isn’t something to the “decoy” idea.  There are only 4 holes in the washer’s rim for locking pins.  Also, there are only 4 holes in the end cap of the field frame. Unless we are to suppose the existence of another case of missing vernier plates (not likely in the Lyon case), then the holes in the artifacts don’t appear to have enough locking pin holes to accomplish the rotational tuning.

Update of the update of the update: Yup! Still convinced. There is something amiss here.  What was it Sun Tzu said?  “If you are big, appear small.  If you are small, appear big”.  Perhaps the hordes were better kept at bay by putting on a battlements show that used lots of lower quality, locally made machines, mixed in with real hot-shooting professionally made ballistas (i.e. the Elenovo, by the looks of her pretty field frames) in proportions the commander deemed sufficient to reinforce the bluff. Complicated theatrics have always played a part in warfare. There has to be some rational explanation for the obvious crudity of the Lyon artifacts.

Update of the update of the update of the update:  Unless, of course, they had a way of tuning their machines that did not rely on rotational adjustments.  A pure wedge machine perhaps? That might explain the non-radiused top surface of those crossbars on the Lyon machine.  For that machine, I’m envisioning some  nicely radiused steel spreaders atop those square-edged crossbars, driven by very hard, hardwood wedges to increase and decrease the spring tension. Must remember: there is more than one way to skin a ballista.  A fella would probably need several lifetimes to explore them all.

Update of the update of the update of the update of the update:  Amateur hour is over.

Samuli Seppanen, over at Greek and Roman Artillery Wiki, reminds me of this sketch of the Lyon artifact that I hadn’t seen in years.


What a find!   Frames, washers, crossbars!  but a resounding zero on my vaunted vernier plates.  With their three inch spring holes, and 12 3/4 inch tall field frames,  it is clear the Lyon machine was of the same size and class as the Orsova ballista.  Again with the big bottom loops?  (….. musings afoot).  And what is it with that skimpy curved stanchion compared with the bulbous overkill approach evident on the Elenovo stanchion? (….More musings afoot.)  Also the top surface of the crossbar is not radiused — that would be prime territory for the chaffing trolls. Maybe there was a wooden shoe there or something.

And clearly, no vernier plates …..

And yet, for reasons made abundantly apparent in previous posts (9/5/2015, 8/25/2015), the Elenovo artifact (not the one seen above) is practically begging for vernier plates.  Are we to suppose that the Elenovo had tiny little washers and springs just to accommodate that small diameter hole pattern in it’s end caps?  And if we do accept small washers and springs on the Elenovo, why would they build the stanchions so damn wide and imposing? Unless of course that was the device’s main function: to look cool and imposing while working with limited sinew resources.  Maybe it wasn’t military at all.  A target model perhaps from the local ballista club.  How about a rich man’s garden toy? with small springs.

I am reimagining a strategy to generate high velocities with less draw weight.  What a relief that would be. Two and a half tons of hoop-la is more on the plate than I want to chew right now.  Maybe getting older has it’s advantages in catapult work. Conservation of energy starts to have real meaning.

(Well that’s a big change from what you were saying yesterday Torsion man.)

Yeah, well yesterday my back hadn’t gone out.


Update:  I didn’t mention that the curved stanchion seen in the Lyon field frame would be too weak to handle the kind of loads I have been feeding into Firefly lately.  Unless we suppose that it’s curved stanchion was made out of a high grade of properly hardened carbon steel, we might reasonably conclude that the Lyon machine is not in the 5,000 lbs class. Given how Firefly’s curved stanchion bent at one point in her development, I’d say about 3,500 lbs tops for the Lyon machine if that curved stanchion was mild steel or iron.

Ran these welds a little hot to make sure they reached the root.


Nice clean puddles on both sides.  Phoenix’s field frames came out amazingly square.  Which is a handy feature if we ever suspect that they have become bent under our typically massive loadings.  A quick check with a small square will tell all.

I wouldn’t want anyone to think that what Firefly has achieved ballistically has been done with anything other than brute force.  Two and a half tons is a lot of draw weight to be messin’ with. On the other hand, the Romans were not unknown for using brute force when it suited them. What good is a torsion engine if it doesn’t excel at handling it’s most salient engineering concept — torsion .

A somewhat suitable hillside has been found with a 300 yard range attached.  More testing as soon as it rains a bit.


It is all range land, not yet despoiled with “no trespassing” signs.   I suppose I should track down the owner and ask for permission….


Respect born of fear.

Next Page »