Yacht Velocity Polar diagrams

Does anyone know of a library of yacht Velocity polars for various yachts.

I want a very general reference for comparing average speed/heading data for a variety of hullforms and rig deisigns in the ballpark 35 - 55, foot range. I want some from the heavier cruising configs as well as high performance flyers.

I expect that the racing yachts will be well documented,
but if anyone has a VP diagram of a heavier cruiser perhaps you could post here.


Heres one of a (well sailed) J-105
Cheers

 

This polar was from the Dave Pedrick designed Navy 44 footer. From memory the boat was medium displacement ~ 250 disp/length. I will see if I can come up with some more.

 

It’s very interesting to see that all the sources agree so closely on the speed. In the group of data points, IMS PCSail tank and splash, is splash the data from the actual boat and tank represents the data from the tank test of the boat.

 

Yes, it is encouraging to see that, even hand calculations can go a long way to predicting boat performance. CFD and tank investigations are good for that last ounce of performance.

 

Well, I have to reply since I'm the guy who did the VPP for the David Pedrick 44'. If you want more pictures, go to:

http://www.redcoopers.net/PA/navyvpp.php

What my research showed was that hydrodynamic sideforce and drag could be predicted with reasonable precision from many sources. However, I put all sets of data into a VPP linked to the Hazen method for aerodynamics.

In truth, I think that hazen aerodynamic method is a little too simple for really good calculations (i.e., no headsail-mainsail interactions). Furthermore, I had to apply my own center of effort approximations to yield an appropriate yaw balance.

Although this work of mine is complete, I was wondering if anyone else knows of a newer aerodynamic model of sails to incorporate into a VPP?

-Jon

 

Looking at the Navy 44 chart posted by dionysis, I note that the worst agreement between IMS and the others is around 45 degrees, i.e. the windward sailing angle. The difference is perhaps 0.3 kt, which is monumental in racing terms. This is not to imply that using the IMS model would guide a designer to an inferior boat, though it might. What is implied is that the IMS model is not especially accurate at the point of sail where a racing boat is going to spend 50% of its time (at least on hulls of this type). Perhaps the model works better on a more mainstream contemporary hull form.

The science of math modeling confronts this all the time: the model is good at predicting the middle range, but it is the extreme that is of interest.

 

If you want really "poor" results, look at my 6-knots wind polar. But this is understandable.

Even with sand strips along the appendages and bow, I had a sneaking suspicion that I was still in the transition zone (or had reattachment). Especially downwind, my tank predictions showed the effects of small model testing. In fact, I'm pursuing designing a 12-meter with the bustle cut off...

On the other hand, SPLASH shows reasonable accuracy: but it is a potential flow code with some calibrations. In any case, when viscous resistance dominates wavemaking, it's really hard to get good results.

-Jon

 

redcoopers, I have been thinking of writing a fully integrated (Hull+Sails --> CFD+Hazen ->>VPP output) for some time. The difficulty is in building good 3D grids for hulls in CFD (structured grids,naturally). I am slowly chipping away at this huge problem... see

http://forums.boatdesign.net/showthread.php?p=28063#post28063
for the basic hull-design system, and
www.MarineDesign.tk (and go to research)
for the (at present 2D) structured grid generation.

I suspect I will work on empirical formulae prior to adding the CFD. Details will be on my website when I start to progress any considerable way with it.

Cheers,

Tim B.

 

Hi Tim,

Yes, I think I went to your website a few days ago. The grids are impressive. I have three projects which are ongoing myself.

The first is a hull design system based on parametric terms mostly. What I do is construct a sectional area curve from a set of terms, and then automatically use NURBS to draw the hull. (this is mostly in waiting since I don't know enough about NURBS yet).

The second is a boundary integral potential flow simulation coupled with a hazen aerodynamic simulation. For sailboats, it's a necessity to remesh a grid for every iteration (to account for trim and sinkage), so I believe that a boundary grid is the most efficient way. The drawback is that this is limited only to potential flow - both for speed and because I don't know a boundary integral formulation capable of anything but potential flow.

My final project is a meshless method of the full navier stokes equations. The advantage of this is that it does not require a grid - and is very accurate. However, all simulations must be carried out in the time domain, and steady flow can only be achieved at a large computational expense.

In any case, if you have skills in grid generation, I'd be willing to couple some of my hydrodynamic codes with your grids. The first two projects are only hobbies of mine, but I do most of my work in C++ (or matlab for speed of coding).

By the way, does your code offer 3d capabilities yet? To me, this seems to be the most difficult portion of grid generation (and is why my thesis is a meshless method!).

-Jon