sail aerodynamics

[Mikko Brummer]

Thanks Mikko, nice picture!
Does it include vertical gradient of TWS?
What are the total drive and lateral forces?
Interesting to see the low contribution of the bottom part of jib and top part of mainsail.

Vertical wind gradient is included, yes, as well as the crew sitting on the rail, although it doesn t show. The plot is from one frame with motions in the waves, so forces vary a lot from one instance to another - I don't have the total forces here, but you could integrate them from the curve. The plot is from this simulation

 

[tspeer]

The caption in the movie says the turbulence is due to separation on the windward side. But maybe there is more is going on than just that.

The windward separation occurs when the angle of attack is reduced by the head moving forward as the boat pitches down. But the turbulence on the lee side is shed when the head is moving aft as the boat pitches up. I'm not sure the time difference is due to the convection of the turbulence from the windward side of the jib to the leeward side of the main.

It looks to me like there's a starting vortex being shed from the leech of the jib as the lift on the jib changes due to the boat's pitching.

 

[P Flados]

I noticed the same sequence as Tom.

Regardless of the exact mechanism of the turbulence, a sail with lots of pitching motion does not have seem to have much chance for maximum efficiency.

For a long time, I have felt that there can be a big overall impact of pitching motion. It looks like optimum trim would be a trim that is not optimum for the median condition, but is instead geared more toward maximize the percentage of time in the "good region" of performance.

One of the significant advantages when cats & tris go for "foil assist" is probably the reduction in pitching motion.

 

[brian eiland]

The caption in the movie says the turbulence is due to separation on the windward side. But maybe there is more is going on than just that.

The windward separation occurs when the angle of attack is reduced by the head moving forward as the boat pitches down. But the turbulence on the lee side is shed when the head is moving aft as the boat pitches up. I'm not sure the time difference is due to the convection of the turbulence from the windward side of the jib to the leeward side of the main.

For clarification Tom, what are you using as your definition for 'head' in this posting?

 

[tspeer]

For clarification Tom, what are you using as your definition for 'head' in this posting?

Top of the mainsail. The corner opposite the tack and clew.

 

[brian eiland]

Thanks. I wanted to make sure it was not the 'leading edge' of a sail, as some folks refer to the head of the sail, nor the headsail, etc.

(PS:I should have known you would be using the correct terminology. Perhaps it was just the manner in which I read it at first that I a little confused)

 

[brian eiland]

Interesting to see the low contribution of the bottom part of jib and top part of mainsail.

In other words the most productive portion of the 'rig' as a whole is where to two sails work together ?

With that in mind I would ask Mikko if the top portion of the mainsail could be improved by having a mastheaded jib?

 

[Jamie Kennedy]

Interesting. When you look at the colour graph the low velocity blue indicating separation occurs up high on the windward side, mostly on the mainsail, when it is pitching forward. The pitching forward creates an apparent wind shift forward so you lose pressure on the windward side and separation occurs, but it does seem to reattach as the boat pitches back again. However when you look at the smoke simulation on the leeward side that is where you see trouble on the leeward side, as it pitches back. The apparent wind shift back again and, perhaps because of the immediately preceeding loss of pressure on the windward side as fell, some turbulence or vortex shedding occurs on the leeward side, but only on the mainsail not the headsail.

This raises some interesting questions. Assuming you maintain optimal heel at least on average, how does it affect boatspeed and pointing ability compared to if it was flat water with less pitching. Also, the mainsail leech is fairly tight, perhaps like it should be if the boat doesn't have to depower. How would things be different if the rig was able to work a little bit more with the waves? Would it be better or worse?

Here is a simulation of a Star, in windier conditions, and with a more fractional and considerably more bendy rig...
https://www.youtube.com/watch?v=zCpVfFDYfWE

It's not a fair comparison and I don't know which is better. The star is designed as a one design class with a HUGE amount of sail area and a slender rig that has a very wide range of ability to power up and depower, at the expense of no headsail downwind. The more conventional racing boat has a less fractional and less slender and depowerable rig, so less range of draft in the mainsail, but the ability to carry spinnakers and such downwind. Both boats appear to be handing the waves pretty good in my opinion, The Star rig is really working, but it has lots of extra sail area to work with and so it san probably have the leech open and close quite effectively in these conditions without worrying about losing too much power. The more conventional racing sloop is sailing with a tighter leech and trying to maintain more power and pointing ability through the waves without losing too much, because it has less to work with relative to its ability to carry sail. So that is my take anyway. Interesting simulation to see what is going on, and where it is going on. If I sailed either of these boats it would make more sense though. Love these simulations. Would like to see one of a Laser sailed upwind with and without vang sheeting, and with body kinetics. There are some Finn simulations but it is quite a different boat I think. Great stuff.

 

[brian eiland]

Vertical wind gradient is included, yes, as well as the crew sitting on the rail, although it doesn't show. The plot is from one frame with motions in the waves, so forces vary a lot from one instance to another - I don't have the total forces here, but you could integrate them from the curve. The plot is from this simulation

So the vertical wind gradient can be accounted for with modern CDF methods?
The reason I ask is I went back to one of Arvel Gentry's documents and read this in the last paragraph.

The Application of CDF to Sails.
The basic problems in applying CDF to sails are about what one would expect: the tools were developed for aircraft and sometimes do not contain certain capabilities needed in the analysis of sails. For example, the standard programs do not provide a means for changing the freestream velocity vector to represent wind shear from the deck to the top of the sail. A basic problem exists in that the sailor is able to change the shape of his sails to match the real flow conditions and the computer cannot.

 

[Jamie Kennedy]

CFD. Computational Fluid Dynamics.
Gentry was a wizard. CFD is getting better and better though as processing power improves, now with large array parallel processing. There will always be limits though, with things like turbulence, and when dealing with significant things happening on boat a large scale and a small scale, and how nothing is ever steady state in sailing. It's impressive how far we have come though. It does provide a lot of insight.

I would like to see some simulations of sailing laser, finn, optimist etc. by-the-lee. Lots of three dimensional flow I think, big vortex over the top as well as around the sides, and lots of dynamics, but I think it could be handled by CFD. Still thinking about trying a smoke bomb some day, or sailing in a snow storm.

 

[brian eiland]

It looks to me like there's a starting vortex being shed from the leech of the jib as the lift on the jib changes due to the boat's pitching.

I see now more clearly what you were saying.

And that image arose on its own,...as a consequence of flow, not as a directed input.

This CDF stuff is quite amazing. I just don't know much about it from a technical standpoint.

 

[Jamie Kennedy]

I see now more clearly what you were saying.

And that image arose on its own,...as a consequence of flow, not as a directed input.

This CDF stuff is quite amazing. I just don't know much about it from a technical standpoint.

In CFD they break up the volume of fluid into a very large but finite number of small elements. They don't have to all be the same size and shape now. You can squeeze more elements in where there is more going on and you need more accuracy. Then for each of these elements, for all of these elements at once, you solve the Navier-Stokes Equations for fluid flow. That includes things like conservation of mass, conservation of energy, viscosity, pressure, density if not constant, temperature if not constant, etc. To solve these equations required a number of iterations to converge on a solution of pre-determined accuracy, and that is just for one finite segment of time. From there you walk forward in time and iterate all over again, although it is a little quicker now that you already have the previous time segment, which is pretty close. Turbulence complicates things immensely. There are different ways of dealing with that depending on the type of flow and what you are after. There is a bunch of other stuff I am forgetting, like boundary layer theory and so on, but basically it is brute force iteration, and you end up with a pretty decent approximation of pressure distribution on your sails and mast and so forth, and the hull also of you are solving that at the same time. If there are waves then you have gravity waves and buoyant forces in addition to the hydrodynamic and aerodynamic forces, and if the hull and rig is dynamic then you have all those inertial forces as well. But that is basically it. I don't know what simplifications and assumptions they make but they always have to make some. They usually calibrate the solution using results from the real world or wind tunnel tests. That's all I got. Hope it helps.

 

[brian eiland]

Slot Effect again

The slot effect is a myth. Don't get confused by all the aerodynamic models people try explain to you - most of them are degree level or beyond. If you want to sail fast then get a fast sailor to show you how to trim sails and if you really want to learn about aerodynamics or CFD then talk to students/graduates/lecturers from a university that does aeronautical engineering. Lastly if you want great advice on sail set up, talk to a competitive sail loft like North or Hyde.

Huw

Sorry if this 'slot' subject seems repetitious, but I've battled naysayers for a long time. As I was looking back thru some of Mikko's archived articles I ran across a pretty good clear explanation he had,....and interestingly he makes a distinction between 2 different 'slots':

http://cdn2.hubspot.net/hub/209338/news/Ad_aerodynamics/index2.htm

Air in the slot - accelerated or not?

The old question about the slot effect seems to have a different answer depending on how you define "slot", and also how you define to accelerate:

If you consider the slot as the area between the mast and forestay, in front of the mast, then the air is slowing down. But if you define the slot as the area between the mainsail and the genoa behind the mast (like I would), then air is definitely accelerating in the slot, as witnessed by the colors turning from yellow to orange and red, when the air moves from the mast towards the leech of the genoa. You can try this yourself: When you stand of the fore-deck, the wind is lame, but when you walk back into the slot between the main an the genoa, you really feel the wind blowing in your face by the time you reach the genoa leech.

This acceleration is relative, however, the airspeed is not faster in the slot area than in case the mainsail were alone, without any jib at all. The main effect of the genoa or jib is to slow down the air speed in the front part of the main, no Venturi-effect there.

 

[brian eiland]

On that same page of his I found this:

http://cdn2.hubspot.net/hub/209338/news/Ad_aerodynamics/index2.htm

Jib head vortex on the main

As always at the tips of airfoils, there is a vortex forming behind the leech of the jib towards the head. When this vortex travels behind the leeward side of the main, it bends streamlines above itself away from the surface of the main, effectively "ripping off" the flow from the main surface at the hounds level and a little above it. This can be seen on the simulation as a triangular, disturbed air area. Above this area, closer to the top the flow can still be attached.

Lower fractional rigs (7/8-rigs) suffer more from this than masthead or high fractional rigs. This could be one reason that 9/10-rigs have grown in popularity lately.

Naturally I would say, why not masthead in lieu of 9/10 ?

 

[CT249]

For a long time, I have felt that there can be a big overall impact of pitching motion. It looks like optimum trim would be a trim that is not optimum for the median condition, but is instead geared more toward maximize the percentage of time in the "good region" of performance.
.

Yes, that's been explicitly noted in trim and tune articles by racing sailors for decades. For example, it's a factor that one looks at when determining headstay tension in many classes.