Part 1
The 1992 and 1994 Sea Trials and Other Excursions
1. Olympias 1992 Trials Report
Edited by Paul Lipke
Acknowledgements
Paul Lipke
My thanks go to the Hellenic Navy, the entire Olympias trials staff and crew, and particularly to Ben Brungraber, Denis Chagnon, John Coates, Ben Fuller Jr., John Howarth, John Morrison, SeĂĄn McGrail, and Ford Weiskittel for their expert advice and support.
Perhaps the most important âLesson Learnedâ in the course of the Trireme Project is that such a huge project only moves forward on the patience, determination and support of the families of the participants. It would be impossible to count their hours of clerical and logistical support, waiting, heat prostration and the many meals they lovingly prepared, and that then grew cold while they waited for their trireme-addicted relative. They did all this with the utmost good nature, in support of obscure historical research and their loved ones having a great deal of fun.
Thus the author has special debts of gratitude to pay to Jane Coates, Mary Morrison, Kati Rankov, Ann Roberts, Nan Shaw, Harriot, Elisabeth and Charlotte Weiskittel, and the mostly-unknown-but-very-appreciated families of the rowers with whom Iâve been honoured to work.
These chapters are dedicated to my wife Marcelle Lipke, in deep appreciation of her constant good humour and support through 24 years (and counting) of my fascination with the trireme project.
1.1. Introduction: salvaging value from a failed effort to publish a 1992 Olympias sea trials report
Paul Lipke and Ford Weiskittel
Nearly two decades ago, we were asked to write and edit the 1992 Olympias sea trials report. This remains unfinished due to various complications, other obligations (such as earning a living), and an overly ambitious outline for the work. The envisioned publication was to be written to be valuable to both specialist and non-specialist audiences, and include new data and perspectives in a number of areas.
One advantage of having developed such an expansive outline is that it offers insights into new areas of study. Therefore, in order to advance any future sea trials, historical research, and publications, and to inform readers of some broad concepts worth consideration, below are brief descriptions of the major areas that were to be included in the 1992 report. Some of these topics are considered briefly in the context of other chapters within the publication you have before you.
I. 1992 Aims and performance on the water:
a) | Ship position and speed measurements: quality of data from the shipâs log and the global positioning system |
b) | Details of rowing performance in daily outings: Minute-by-minute logs, highlights, including 1 hour of firm, longest row, fastest speed, turning, rowing astern, etc. |
c) | Details of rowing performance on voyage to Corinth and Salamina: route, wind, duration of rows, speed made good, sailing performance |
d) | Performance comparison with other trials: from sprints to voyages, addressing speed, power and pacing, including tables of adjusted data for earlier trials |
e) | Operation with partial crews |
II. The thesis that Olympiasâ previously published performance data warrants even further caveats than those presented in the Log Summary (see Chapter 1.2: Some Results of Olympiasâ 1992 trials), due primarily to lack of data and errors in performance of the measuring equipment. This makes suspect any interpretations and debates about Olympiasâ viability based largely on her top speed. There are many other, far more valid, reasons to value Olympias, her performance and the entire project.
III. Brief commentaries by veteran naval architects and shipwrights who have worked on other historic vessels, replicas and reconstructions. We wanted their insights on how much Olympias design âpushes the envelopeâ in terms of strength, safety, and hull durability. This proposed chapter was in no way intended to call into question John Coatesâ extraordinary work designing Olympias. Rather we sought to provide non-specialists with a relative sense of how extreme Olympias is from an engineering point of view, and how far and in what ways more or less conservative safety and performance standards might affect the design. Could a âmore riskyâ ship gain materially better performance, and in what parts of the ship might the most effective risks be taken?
a) | Useful questions for evaluating the design of any trireme reconstruction in the context of what we have learned. |
b) | Ship construction and repair: i) | Daily maintenance needs of triremes: frequency of repairs, the tools and materials likely carried on board | ii) | Possible battle preparations and likely repair strategies | |
IV. The evolution of our understanding of the optimum trireme stroke, âHow to row Olympiasâ, and how the latter might differ from rowing in a trireme where stroke length is completely unrestricted.
V. The human engine:
a) | Rower physiques and the necessary interscalmium (the âroomâ or space required for a fixed-seat rower to pull effectively) |
b) | Rower physiology: i) | The performance of our rowers in comparison to modern athletic performance | ii) | How fit Olympiasâ oarcrews have been | iii) | Using collected volume of oxygen uptake data and pulse monitors to relate the performance of rowers in the ship to their physiological capacity | iv) | Measuring effective power delivered to the ship | v) | Predicting shipâs performance and crew power output on the basis of this data for a given crew size, gender and fitness | |
VI. Suggested protocols for future trials
1.2. Some results of Olympiasâ 1992 trials and log summary
Paul Lipke, Andrew Ruddle and Ford Weiskittel, with assistance from Charles Hirschler
The principal efforts of the 1992 sea trials were to explore operations with a reduced crew, crew performance during longer voyages/hours at the oar, and to improve the accuracy of our speed and position data using Global Positioning System (GPS) technology.
1.2.1 Reduced numbers of crew and crew performance
The oarcrew in 1992 numbered approximately 154 (out of a possible complement of 170) at the start of the trials, and reached a low of 121 on August 8, 1992 (the low numbers resulted in part from a late decision to conduct trials, and therefore a late recruiting effort.) Despite the low numbers, the ship and crew performed well. In fact the reduced 1992 crew rowed better and faster more quickly than the full 1990 crew, especially during the first four training days (this period was followed by a physically demanding voyage leading to cumulative fatigue). Since the 1992 crew was no more fit physically than earlier crews, we believe the improved performance reflects continuing advances in training and coaching methods.
A full hour of non-stop, âfirmâ rowing showed what a small crew might do under short-term pressure to perform, i.e. in battle. A 156 kilometre (112 nautical mile) voyage to Aegina, Corinth, Salamina and return to Poros tested the small crewâs stamina, especially during an 11-hour, non-stop row into headwinds reaching 20 knots with higher gusts.
During much of this long day the crew rowed in rotations of 40 minutes on, 20 minutes off, the thalamian seats being occupied by those who were resting. In such a headwind it was very important to maintain the shipâs headway (and thereby her heading) while the oarcrew were swapping seats. This was achieved by reducing the time needed to complete a rotation to well under two minutes (sometimes as little as 80 seconds) and/or by keeping the bow or stern rowing while the balance of the crew changed seats and started up again.
1.2.2 Global Positioning System: accuracy of trials data
Researched by Charles Hirschler, written by Paul Lipke
Previous trials relied primarily on a somewhat inaccurate shipâs log for speed measurements (Morrison and Coates 1989, 44â5; Coates, Platis and Shaw 1990, 23â4). Measurements by Dutch log (which involves dropping a buoyant object, such as a wooden block, off the shipâs bow and counting the seconds needed for the vesselâs length to travel past the block.) and timed runs past measured markers on shore were used to develop an adjusting factor which reduced any recorded reading to 89% of the value displayed. It must be said that almost everyone involved lacked confidence in both the log and the adjusting factor. The 1992 results call for further modest adjustments, but overall greatly increased confidence in the data.
In 1992 Global Positioning Systems used signals from 3â7 orbiting satellites to provide highly accurate measurements of position, speed and distance traveled anywhere on the surface of the earth. It must be said here that GPS accuracy claims fuel stiff competition between manufacturers. Furthermore, there are tensions between users, manufacturers and the military because the latter intentionally introduces random error in the signals in the interests of national security. The introduction of error is called âselective availabilityâ and is measured by the Horizontal Dilution of Precision (HDOP).
The manufacturer of the Trimble Ensign hand-held GPS used in the 1992 trials claims in their literature that under the best conditions it will determine your two-dimensional position on the globe to within 10 metres (vertical position accuracy is not considered here since Olympias is always at sea level). When the signals are being degraded, as they have been (until recently) at virtually all times except during the 1990 Gulf War, GPS accuracy is limited to twice this distance, or 20 metres multiplied by the HDOP.
Typical HDOPs during the 1992 trials consisted of:
Lows of 1.4, i.e. accuracy was 20 m x 1.4 = 28 m (92 ft) Highs of around 3.0, i.e. = 60 m (198 ft) For entire outings, the HDOP averaged 2.1, giving an accuracy of 42 m (138 ft).
An average accuracy of ±138 feet of the position displayed seems realistic. Over distances of a few miles or more such an error is small. Over short distances, i.e. for a 2000 metre (1.25 mile) row, assuming an error of up to 42 meters (138 feet) seems reasonable and reduces the usefulness of the GPS for establishing distance traveled.
1.2.3 Speed
Speed resolution is ± one unit of the smallest units displayed per second (i.e. if the GPS displays 5.1 knots, actual speed could be 5.2 or 5.0 knots). Speed readings are more accurate under âselective availabilityâ than position readings because the error factor in the satellitesâ signals changes gradually over time. This means the built-in error factor does not affect the speed readings which are based on changes in relative positions taken within a few tenths of a second of each other. One of the major factors that typically has significant negative impact on GPS speed accuracy, i.e. blockage of the signal by buildings, bridges, and mountains is clearly not a problem on the water.
The published maximum speed record of 8.9 knots (Shaw 1993, 43) achieved during the morning outing on 9th August, 1990 deserves some discussion. The figure of 8.9 knots has since been widely published and quoted as Olympiasâs top speed. It should be said however, that this run had an average speed of 8.3 knots (adjusted) for the last half of the run with a single reading of 8.9 knots at the very end of the run (Table 1.2.1).
Clearly the 8.9 knot (corrected) reading was not sustained for any appreciable period, whereas the 8.3 knot adjusted average is solid.
In 1992, we sought to get a better sense of the accuracy of the 1990 (and earlier) readings. We received some reassurance, but the peak of 8.9 knots remains a little suspect. For example, the GPS consistently displayed 7.8â7.9 knots for a 2-minute speed run with a reduced oarcrew of about 135. This is consistent with an 8.3 knot average achieved with a full crew in 1990.
The next day, with a reduced crew of 121, a brief peak of 8.2 knots was recorded by the GPS. Given this and other runs with a small crew at speeds well above seven knots, the authors believe a brief peak speed of about 8.5 knots and more sustained speeds of 8.3 knots can be claimed for Olympias with confidence. Given the previous uncertainty about the accuracy of the shipâs log and the correction factors used in 1990, the GPS readings are reassuring.
This lower number is further strengthened by the speeds recorded in 1988 with a laser tracking system called a geodimeter, which showed that a less well-trained crew was capable of producing a burst of 7.9 knots, with most of the acceleration runs producing speeds from 7.3â7.5 knots (Lowry and Squire 1988, 53â60).
1.2.4 Summary of results of the 1992 trials of Olympias
Based on Andrew Ruddleâs log
Note: The early days of each set of trials have always focused primarily on crew training and adjustment. This means getting the international crew understand the command language and process on board, learning to row in unison with 169 other people, moving rowers around the ship to find levels and triads within which they row and mesh well, etc.
Trials day 1: (22/7/92): max speed of 5.8 Nautical Miles/ hour (NM/Hr)
Trials day 2: (23/7/92): max speed of 6.0 NM/Hr
Trials day 3: (24/7/92): max speed of 6.9 NM/Hr, distance covered 6.78 NM
Trials day 4: (25/7/92): max speed of 6.3 NM/Hr
Overview of the Voyage
This was four-day voyage totaling 111.85 NM: 67.8 NM rowing, 7.7 NM rowing/sailing, 24.3 NM sailing; 4.73 NM under tow (through the Corinth Canal). Some long passages were m...