The Doodson-Légé Tide Predicting Machine

Hi All,
I was just watching "Coast" on SBS, i don't usually but it just happened to be on. They were talking about the tides around Liverpool and how long it took to produce accurate tide charts for the mariners. It all culminated in the late 40's in a marvel of engineering that took over 120 variables to produce accurate tide charts. The small amount of footage shown displayed an amazing amount of work in brass, steel and cast iron.
Unfortunately after a google i came up with very little, this page here: The Universal Machine: The Doodson-Légé Tide Predicting Machine and the National Oceanography Center site here: Dr A.T. Doodson, CBE, DSc, FRS, Hon FRSE - Influential scientists | National Oceanography Centre | from coast to deep ocean

Does anyone know anything else about this rather amazing machine?

Hi Ewan

I watched that show . Yes what a mechanical marvel :2tsup:

They praised nylon but omitted to say that the worlds oceans are now plagued by bits of the stuff that will take centuries to break down .

The little horse drawn Victorian bathing boxes for women were rather novel

Thanks for posting that, a truly amazing bit of engineeering :2tsup:

The most complex real time software I ever wrote had zero inputs and only 1 output. It was to disable/enable a pump system based on tidal caclulations.

The algorithm was basically a calculation based on a large number of sine waves, with different phases and phaseoffsets.... , a little bit like a fourier series with some feedback components If I recall correctly, Just looking at the mechanicals I wonder if it's a mechanical implementation of the same algorithm?

I wish I could remember where the algorithm came from..

Regards
Ray

Hi Ewan,

Well there you go, it was using the Doodson Harmonic Analysis stuff after all...

Here is a link to the book...
http://archive.org/details/philtrans08044568

Regards
Ray

Thanks Ray, I just downloaded the book. I can't promise that I will read it (or possibly understand it) but I once wrote software to control an outfall valve (2 km out to sea) dependant on the tides.
It needed the tide tables downloaded from the maritime website and uploaded to the PLC every 3 months or maybe 6 months, I can't remember. A tide predicting algorithm would have been much better.