The weight she had on board was for her 31inches (79cms) length was 13.75lb (6.25kg). The weight of the model's steel shell was 6.05 lb (2.75 kg). an all up weight of 19.8 lbs (9kg). At these weights she sat in the water just over half way to the gunwales amidships and slightly less aft.
In extrapolating those weights to T3's length, some 7 times that of the model, I think the weight hike will need to be like that of the volume hike ie the cube of that ratio or 1/7 scale.
The cube of 7 is 343 so if this idea is correct that would mean T3's all up weight should be 6791.4 lb (3087kg) or 3 tonnes.
If that calculation is to be believed and as her actual all-up weight, engine, cabin, lead filled keel pod and accoutrements was 1155lbs (525 kg), one imperial ton or half a tonne, it is small wonder that her footprint on the water was that of an 18' canoe.
Back when I was playing with the model I rang the test tank people at Southampton and they quoted some runs there and calcs against established data for some £2500 pounds.
That seemed a lot. I thought for that much I could knock up a 'life size' model I could actually test with me in it and in many way this, in material terms, was not that wide of the mark.
What was wide of the mark was the other big costs - time - I was working part-time as an obesity adviser - and rent and rates of the space I had here in London to do it - this by far the biggest killer financially.
Using Xray film with its wangy or elastic spring steel quality very much like that of the marine aluminium sheeting of the definitive boat, I played around with various cuts into rectangles of it to produce nice shapes letting the cut surfaces slide over each other.
Having played around with xray film I moved on to using thin gauge steel.
The pattern was simply this: the length of the sheet is twice the width - it can be whatever you like - it's what I chose as it seemed to work out on xray film - the cuts in the bow end (the blunt end) allow the sheeting to slide over itself to a point such that the bow profile from the side was 45degress, a concession to internal space and an appropriate attitude to oncoming water.
Since the sheet is truly rectangular it is easy to be bilaterally very accurate in one's markings. The excess of the overlap that lies above the gunwale is trimmed off forward in the line of it aft. This produces a cusp or hollow at the top of the bow which is conveniently filled with a symetrical bi-ellipse to form a supra-prow. For geometrical reasons I never tried to fathom the top of this ellipse is a splendid starting point for decking of the model again forming along the length decking which is at right angles to the gunwhale side along the whole length. Thus the forward decking runs upwards quite markedly and and progressing aft become less so and eventually flat.
The working model was made from a single sheet of thin steel in a few hours.
Naturally I feel flat at the moment and feel I want to be shot of it all. But I have to look back at what I wanted when I started out and that was:
1. to determine that which is not intuitive being whether the tear drop profile so ever present in nature is useful in boat design and
2. whether origami like simplicity in metal bending can offer a easy-build yaght embracing this non-intuitive concept.
I read somewhere that the ultimate in boat design is one that leaves no wake - allowing that a wake represents wasted energy no one would disagree with that though the parameters of performance need defining. Yet in evolutionary terms the tell-tale nature of a wake giving the game away either for the prey or the predator-hunter would be an action packed situation and would suggest the reverse tear drop shape is also meant for speed. Of course how this speed will be affected by the shape will be a function of the fluid's viscosity, which should be low, and its density, which should be high, so as to allow the pressure head developed to be usefully, and that means quickly, redistributed in the post head section of the form to 'squidge' it forward and thus retrieve the initial energy expended.
A bath of mercury might be an ideal liquid satisfying these criteria if one's model was heavy enough not to be displaced onto the surface!
To stand by a pond and watch a large koi near the surface give a single swish of its tail and no more than this to shoot it forward with not an iota of visible disturbance is exhilarating.
One might think that I should have been building a submarine but the physical mores apply to any fluid, liquid or gas and one would think therefore to an body that is part immersed. The immediate responder might be the termoil at the liquid/gas, water/air interface would be better negotiated by a sharp leading edge - well this remains to be seen.
One of the outstanding features I found when testing the model is incomparably better directional stability when the craft was pushed - no great shakes of course when the motion comes from pulling - though - when towing any conventional craft through the water the tendency for it to rapidly drift from the line of pull is well known.
Perhaps mentioned earlier, the buoyancy of my craft being well forward will prevent or lessen the likelihood of it being a pitch poled from a following sea by this acting on the wider and larger surface and much more buoyant forward section. Also the fine entry of the stern and its lesser buoyancy must offer an easier helm to control.
Well here I| am with a 'too floaty boaty'. What to do? Well, one positive need is re-ballasting and one positive outcome from that will be to bring the water level in the engine well dangerously close to filling her with water from the inside. So, the engine and its well will have to go. That can be done quickly. The well can be removed completely or cut just above its junction with the hull thereby offering some purchase for future bracketing etc. Of course the central hole that provided the entry point for the propeller drive will have to be plated unless option* below is taken.
Then what are my other options for propulsion for the craft? They have to be compatible with adding lots more ballast.
- put an inboard engine in and run a shaft aft and provide a propeller in a cutaway section just forward of the rudder.
- think about an electric motor/battery/generator set up inside the boat.
- attach an electric long shaft engine alongside(outside) aft and accept a bias whilst under power.
-* using the existing hole in the bottom of the hull for a long vertical well to accommodate a very long shaft electric 'outboard'.
No comments:
Post a Comment