The Fuller Number

Originally posted by blaze720 When both are planing it is a slightly different case but as was particularly obvious with boards when 'fullers roamed the earth' in pre-history waggling a long hull around with a (now un-needed) load of it in the air is simply not fast.

On our water in most winds the dinghies can cope with, a Raceboard has never been bested by a Dinghy, nor for that matter by a Formula board, even Olympic El would rather sit it out than get his trousers taken down by the old boys in the Raceboard Fleet. I put the typical race board details into my PeakyPunky Spreadsheet it came out similar to an EPS, 1019.62 against 1020.9, I know that's not right.

Edited by iGRF - 25 Apr 14 at 2:17pm

Mike makes a good point. Everyone assumes that numerical simulation can provide the answer but even the primary users of high fidelity numerical methods cannot get close to real world conditions. In the aerospace industry CFD is certainly used a lot but has to be validated and tweaked by using empirical physical test techniques. The day CFD replaces wind tunnel testing has for many years been predicted as being 10yrs away, in reality is more like 70-90yrs. The problem is that the codes have been developed based on existing configurations. As soon as you get a novel change which in sailing terms we see fairly regularly then the existing codes simply don't cut it. Take for example the Americas Cup, those guys have incredibly powerful CFD systems running but they are calibrated against as similar config as possible before they start their optimisation. In short they are defining their vpp for a small variation from a known baseline.
So back to general handicaps, in order to get a good numeric computational method, you'd first need to validate a hugely processor hungry code against a series of known baseline configs and then evaluate the impacts of the subtle changes needed for each class. When you consider that a single data point (one speed, one wave pattern, one wind speed, one wind direction) would probably cost you in excess of €100k and take a few days, you can hopefully see that to define numbers for every class for all conditions is nearly impossible with current computational power. Now if you partially validate your codes by running models of all classes in a tow tank, you may be able to cut a few corners and make some assumptions but to do a proper job is both financially impractical and you simply wouldn't have time to do it.
This is why we have the current system!

TV editing sadly portrays the impression that hi-tech activities like CFD and ALM are instant activities and the current marketing would also have you believe it. Sadly it's largely just a pipedream at the moment.

I think you miss the point ... the variance is crew skill factor is huge compared to variance between peakys numbers and the PYs if you want a simple measurement system then you have it here ... the 1 -2% differences are nothing compared to driver skill issues.

Originally posted by 2547 I think you miss the point ... the variance is crew skill factor is huge compared to variance between peakys numbers and the PYs if you want a simple measurement system then you have it here ... the 1 -2% differences are nothing compared to driver skill issues.

Edited by Punky - 25 Apr 14 at 6:28pm