When rotating the washers one has basically two options:
- Rotate opposing washers in a field-frame to opposite direction (spiral pretensioning)
- Rotate opposing washers to a field-frame to the same direction
This article discusses both of these methods with their pros and cons.
Rotating washers to the opposite direction (spiral pretensioning) Edit
If one rotates the opposing washers of a field-frame to opposite direction the spring bundle becomes a spiral. Interestingly there is a relief in Pergamum (see Marsden 1967/1971: plate 3) where the torsion springs of a ballista are clearly depicted as spirals. Also, Philon mentions the following:
"The engine loses its springiness because the strands are huddled up into a thick spiral and the spring, becoming askew, is robbed of its natural force and liveliness through the excessive extra-twisting." (Marsden 1971: 121)
These two sources suggest that rotating the washers in the opposite direction would be the correct way to apply pretensioning to the spring.
However, in practice this method has many issues. The most important problem is that when rotating an arm, only the upper (or lower) half of the spring tightens and stores energy while the other one actually loosens. This means that much of the energy of the spring is wasted. Also, the arms do not return to their exact starting position after a shot, probably because the half of the spring that was loosened during cocking has to readapt into the higher tension levels.
When pulling back the arm by hand the torsion bundle clearly lacks the necessary springiness even when washers are rotated a significant amount. This is in part because the halves of the springs "meld" into each other thus increasing the mechanical leverage of the arm.
Rotating washers to the same direction Edit
When using this method the washers are always rotated to the same direction, so that the arms lean against the curved field-frame bars. As a side-effect the halves of the spring bundle move away from each other instead of melding into each other (see above).
When the arm is rotated, the spring bundle halves are much farther away from the imaginary center of the bundle than when the spring is a spiral (see above), thus forcing the spring cord to stretch a lot more. This, in turn, means that more energy is fed into the springs during cocking, which is exactly what we want:
- The energy spent pretensioning the springs is mostly lost: it only benefits us indirectly (for details look here).
- The more the spring cord stretches during cocking the more energy we can feed into the springs without tripping over the limits inherent in the materials.
- Less pretensioning means less strain on the spring cord when the weapon is not in use (i.e. cocked).
So far (~150 shots) there have not been any issues with using this method alongside cheiroballistra-style arms. My initial suspicions about arms moving up or down or getting pulled away from the bundle have not materialized, possibly because I've always pretensioned the springs enough.