When doing scientific reconstructions of ancient artillery, it's important to try to think like an ancient engineer, instead of a modern reconstructor. In the following chapters we discuss the various notions modern reconstructors tend to have, and which can lead us to making false conclusions about the ancient ballistas.
Performance at all cost Edit
Modern reconstructors generally start with a machine with (partially) known dimensions, such as the Orsova ballista or the cheirobalistra. Then they tend to try to squeeze every last bit of performance from that particular machine and either disregard or play down other concerns which were valid for the ancient engineers. The logic behind this reasoning is this:
Ancient artillery was designed to kill people and destroy structures from a long distance
From this we conclude that
The more powerful a piece of artillery is, and the farther it shoots, the better
And from this we reason that
I need to make my own ballista as powerful as possible.
While the chain of reasoning is basically valid, there's one important difference between us and the ancient engineer: the ancient engineer was not given some random machine and told to maximize it's potential at any cost. Instead he started from scratch, and also had other concerns besides performance.
Increasing washer size Edit
The most common modern pitfall is trying to artificially increase the washer diameter of a ballista with known dimensions to increase the power output. For example, Wilkins (1995: 21) and Marsden (1971: 223-224) artificially enlarged the washers of the cheiroballistra, which quickly led them into serious problems with dimensions of the other parts; this is discussed in detail in the personal torsion weapons article.
Similarly, Nick Watts' Orsova ballista reconstruction uses vernier plates invented to allow use of larger washers which in turn allow more energy to be stored, An ancient engineer did not need to do this, as he could simply enlarge the field-frame rings slightly to achieve the same effect and thus avoid having to make 4 vernier plates. That said, the vernier plates do serve another purpose (see below).
Inventing new parts Edit
It is very tempting to invent new parts to fix an engineering problem. The basic problem with inventing parts is that the fact that they do something useful does not prove they existed. It is possible that the invented part did indeed exist, but without proof (i.e. new archaeological finds) we can't tell with certainty. At best, we can tell that some solution is needed to an engineering problem we encountered, and our solution may or may not be the historically accurate (i.e. correct) one. In many cases there have probably been many solutions to the same problem.
To give a concrete example, there may be an alternative solution to Nick's vernier plates, which allow adjusting the washer rotation in 7,5 degree intervals in a reconstruction based on the Orsova artifact (see Baatz 1978: 10) which has only 4 holes in each field-frame ring; see this article for more details.
Sometimes there isn't a problem, unless our preconceived ideas of what the ballista should look like create one. For example, Wilkins' massive bronze locking rings (Wilkins 1995: 36-38; 2003: 49-50) are entirely useless if the cheiroballistra is reconstructed as an inswinger.
Focus on bolt velocity Edit
Nick's Orsova reconstruction has managed to produce very impressive bolt velocities in the 350-400 fps (105-120m/s) class and ranges of 800-1000 yards. These results have been achieved with as little as 45 degrees of arm rotation and lots of pretensioning. The reason such a high velocities are achievable with this little arm rotation is the geometry of bowstring and arm movement at the end of the shot, which is also visible in the cheiroballistra force-draw curves.
All of this could lead us to conclude that there was no reason for the ancients to use more than 45 degrees arm rotation. After all, if such a good velocities are possible with such a little arm rotation, why bother rotating the arms, say, 90 or 110 degrees? Now, if the Romans had been training for a flight shooting competition and had unlimited resources at disposal, our modern 45 degree approach would have made sense to them. However, we can fairly safely assume that the Greek and Romans wanted to get maximum amount of energy from minimum amount of sinew cord; this means, as discussed below, that a strong case can be made for maximizing arm rotation.
While using less pretension and more arm rotation might not increase maximum bolt velocity significantly, as Nick has told - and which is almost certainly true - it does increase the amount of energy that can be fed into the torsion spring safely, if a suitable amount of pretensioning is used. Essentially, with longer draw we can use heavier bolts than with a shorter draw without losing any bolt velocity. As an extreme example, we could pretension the torsion bundle so that the cords are near breaking point and draw the arm 10 degrees and probably still get fairly good velocities with light bolts. However, the energy stored in the springs during draw would not be nearly as high as with lesser pretension and longer arm rotation. The reason for this is that all of the energy we've spent pretensioning the spring has been lost, i.e. won't get transferred into the bolt. It only benefits us indirectly by making the force-draw curve climb up more aggressively.
Some of the side-benefits of a short draw are actually symptoms of lessened energy storage. For example, shorter draw probably makes the cocking require less effort, i.e. either strength or time. Should this happens, it indicates that less energy is stored in the torsion springs.
All this said, different engines need different amount of arm rotation. The cheiroballistra, which is cocked without a winch, will need lots of arm rotation to produce long draw, which is necessary to store maximum amount of energy during cocking. Due to it's winch an Orsova class machine can have a long draw or a short draw, depending on the situation: for example, the rate of fire can be temporarily increased by reducing the draw, at the cost of having to use lighter bolts.