Principal Particulars were as follows:
Waterline Length = 750mm
Waterline Beam = 150mm
Displacement = 500g
Rowing machine uses a 3v electric motor to drive a crank. The rotation of the crank is reduced to approx 45rpm by a pair of elastic bands on wooden pulley wheels. The crank axle is fitted with a sprocket gear to drive a matching crank axle at the forward end of the galley via a lightweight plastic chain drive. The cranks are connected to a pair of beams (port and starboard) which follow a rotational motion. Holes are drilled in the webs of the beams to take the oar “handles”. Hence, in this simple model the port and starboard oars both work together and follow a circular path.
A fairly basic numerical spreadsheet model was produced to predict the performance of the 12 oar galley. This model was tested against the measured performance of the galley for the following parameters (both at 45 strokes/min):
1. Bollard pull (thrust at zero speed)
2. Steady state speed
Output for steady state speed shown below:
Trials results on the actual model were as follows (spreadsheet predictions in brackets):
1. Bollard pull = 0.0123N (0.0125N)
2. Steady State Speed = 91 mm/sec (90 mm/sec)
The model was further developed to cover the trireme rowing machine describes elsewhere on this site.
The following link covers the calculations behind the spreadsheet model and the results of the two trials:
Click on the following link for a drawing of the rowing machine:
The following video shows the galley on "basin trials"!
It is worth noting that when a scale model of a rowing machine is operated at the same stroke rate as the full size vessel the speed will be approximately scale in proportion to the length (i.e. a 1:24th scale model will travel at approximately 1:24th the speed of the full size vessel).
However, when model tests are conducted as part of a ship design process (either to predict resistance or manoeuvring characteristics) they are generally run at corresponding Froude Numbers. This is because it has been demonstrated that geometrically similar vessels at different scales will generate the same wave pattern (and hence comparable wave making resistance) if run at the same Froude number. Froude number scales inversely in proportion to the square root of the waterline length. In the case of a 1:24th scale model this means that the model speed would be equal to the full size speed divided by the square root of 24. This would be nearly 5 times as fast as the current speed of the model and would require the stroke rate to be increased approximately 5 times (assuming the same slip). The implication of this (and other scale effects) is that the models performance is not likely to be a very good indicator of the performance of the full size vessel.
It would, however be interesting to see how the working model of Olympias performs, as full size trials were conducted with this design and so there will be a real basis for comparison.