Loads for swept spreader rig

Steve,
The deflection analysis was eye opening. Thanks for the advice. The intermediate is now 8mm, same as upper (was 7mm) That produces a nice even curve. Accounting only for the stretch in the shrouds, the mast falls off about 10" at the hounds! That would allow a lot of forestay sag. This really points to the performance advantage of running backstays.

 

shu,
Yes, it can be scary. I would recommend trynig to reduce masthead sag to less than 6" if you can. Anything more feels mushy.
Steve

 

Spreader Rake

Dear shu,
Just wondered if you've ever run across this website of Applied Engineering Services Limited of New Zealand, <http://ourworld.compuserve.com/homepages/Chris_AES/homepage.htm>

Here is what Chris had to say on the subject;

"Spreader rake is an interesting subject. If you have read the above and determined that you are interested in an inline rig; skip this. Once a swept rig is decided on the next question is what angle of spreader rake. If you have a spreader rake of 30 degrees on a fractionally rigged yacht, the chances are that you will need little or no runner loads. Indeed many yachts with 30 degree rake have no runners what-so-ever. The cap shrouds are effective at tensioning the forestay and provided J the jib foot is not too long considerable forestay tension is developed without runners. On a masthead rigged yacht the backstay is effective at tensioning the forestay and again without any runners you can tension the forestay using only 20 degrees of spreader rake. As you reduce rake to only 15 degrees the need for checkstays and runners starts to eventuate, sooner if you have an inner forestay/staysail of course. At spreader rakes of less that 10 degrees our best advice is DON'T. We say this because you obtain the worst of both worlds with spreader rakes between 5 and 10 degrees. You need runners and checks because if you gybe with 5 degrees spreader rake or fly a spinnaker you might snap the sidestays if you have no runners. At the same time you have a swept rig and can not modify mast bend / luff curve as in an inline rig. So, you have not been able to get the benefit of either configuration. The cliche 'in for a penny, in for a pound' holds true albeit poorly paraphrased. If spreader rake is between 0 and 5 degrees this is to all intents and purposes inline and would normally just be treated as 0 degrees/ inline."

 

Determining Loads in Sailboat Rigging

Sail Loading on the Rig, Rig Loading on the Vessel

Let me first look back at some of the previous quotes on this thread:

shu:
"they (Larsson and Eliasson) don't provide a means of determining the shroud loads due to headstay tension"

terhohalme:
"Exact headstay tension is impossible to determine."
"Multiplier 15 includes all guestimated loads...."

shu:
'The tricky part in Larsson and Eliasson is that you calculate "real" athwartships static loads on the rig ..., then apply separate safety factors for the other components based on some multiplier. What was the assumed factor of safety for the forestay?' (Ed:not the exact quote)

terhohalme:
"Impossible to know exactly..." (Ed: about the Nordic Boat Std)
" Where do you really need exact numbers? Tightening sagging of foresail ...will mix the whole calculating process"

shu:
"However, this all assumes that you have a permanent backstay.... I see no accounting for the additional loads the shrouds must take to oppose the forward component of the load in the forestay."

"...changing the sag of the forestay can make large differences in the resulting shroud tension. I don't know if I can predict the sag as a function of the foresail loads."

tspeer:
"So it's not as simple as figuring the component of the righting moment that is borne by the forestay. I suspect the factor of 15 applied to the forestay tension was intended...."

shu:
"assuming half the propulsive force acts at the hounds"

SailDesign:
"I have seen sticks that were designed to death from a safety factor point of view, but were pretzels when you sight up them under load."

"you just need to apply the sailing loads to each length of rigging, calculate the stretch, and plot it…" (Ed: how does one determine these loads with any exactness)

gonzo:
" I have observed that many of the rigging formulas don't take bending into consideration"

shu:
" had started looking at getting uniform unit stretch in all the shrouds to keep the mast more or less straight, but kinda gave up when the 3 shrouds all came out at very different ultimate diameters for safety, when their working loads were much closer”
_____________________________________________________________


What really surprises me about these few quotes, and many other discussions on engineering a sailing rig, is the total dependency on the use of ‘guesstimates’ and a variety of ‘multipliers’, some of unknown origin and application.

I recently purchased a copy of Larsson & Eliasson’s Principles of Yacht Design, specifically to investigate their analysis of these rigging loads. But what I found at the very opening paragraph of their chapter on Rig Construction, "in dealing with the dimensioning and construction of the rig, over the years different methods have evolved, ranging from old rules of thumb…to sophisticated computer models for exotic composite materials. We will take a middle line (approach) using accepted practices (old rules of thumb?).…..” Page two (text 202) of their chapter, "It is common practice that the transverse and longitudinal stability are studied separately”
And this is supposed to be a modern analysis? Later in the chapter (text222), “another factor which improves performance is the rake of the mast. Although not numerically proven…”

In this modern computer age why have they chosen to ignore the "sophisticated computer models"? Are sailboat rigs such a complicated structural problem to analyze? Even the more simplistic steady-state ones (minus some of the more complicated dynamic questions)?

I guess my frustrations with understanding and defining the actual true loads on the rigging of a sailboat is best summed up at this Classic Marine website,

http://www.classicmarine.co.uk/Articles/rigging_loads.htm
“Rigging Loads- a study in guess work, or a tale of scientific progress?"

I will quote a few of the more notable passages from his very interesting summation:
a) He opens with a quotation from Douglas Phillips-Birt, "Masts are tricky things. It is not for nothing that Lloyd's, which is ready to specify the scantlings of nearly every other part of a yacht, washes its hands of them altogether and plants the responsibility for their size and shape squarely on the designer's shoulders....suggesting that mast are perhaps a little beyond rational analysis."

b) For all its crudeness, this rule (a particular one) at least recognizes that the strength of the rigging relates more to the size of boat rather than the size of the rig.

c) You can see that these factors will bear on the issue, but the more you look at it, the less you can understand why they are combining in the way they are.

d) With a method so opaque in its assumptions, you never know what the range of validity is in terms of rig type, or arrangement of stays.

e) From Skene's book, the ‘long method’ is based on SAIL LOADING. Good Heavens! That is the first time it has been mentioned, which considering that it is the sails that load the rig, must be an improvement. Don't get too excited though. How much is the mast loaded and where? The answer is that nobody really (seems to) knows.

f) It provides not a real life start point for some rigorous analysis, but a common assumption that can be used to compare craft with each other, and/or with empirical data. To try to rationalize an assumption like this is at best pretentious—an attempt to ennoble guess-work, at worst dangerous—someone might believe it.

g) You may be getting the impression that this is not much advanced on earlier efforts.

h) Interestingly that the NBS method does not specifically relate mast loads to shroud tensions, but starts again with the righting moment.

i) Secondly, however numerate the rules appear, in practice they are all founded on empirical data.

j) So we are a long way from a complete picture of the loads in a rig, particularly tradition rigs. Why? For a start, the more sophisticated approaches have developed during the age of the BERMUDAN rig. The usual assumption that shrouds can be analyzed separately from fore/backstays probably holds better for Bermudan, than for gaffers, where there will be a complex interplay between peak halyard, runners, mainsheet, bowsprit, shrouds and so on.

k) And finally and MOST SIGNIFICANTLY, none of the methods derive loads from the force of the sails, which is after all what is loading the rig!!! Such an analysis could be fiendishly complex, but with ever more powerful tools and computers, I think it is not an unrealistic thing to attempt.

Brian notes, maybe I am being a little naive here, but I find it hard to believe in this computer era that we can't set up a three dimensional 'map' of a sailing rig and be able to analyze the forces in the individual components, and how they interact, and how changing one component's size, strength, geometry, etc affects the other components at least in a steady-state environment

I would imagine that we must first redefine the actual load paths that the forces of the sails use to transmit their power to the rigging. And then how and where do the rigging loads get transmitted to the vessel itself? I propose to start a new tread on this subject,"Determining Loads in Sailboat Rigging" or "Sail Loading on the Rig, Rig Loading on the Vessel". And I think it might better be placed under the “Sailboat” general heading as it particularly applies to sailboats

We have previously bunched all the sail loads together and assumed they acted thru the sail's CE. Granted this might yet prove to be a reasonable assumption, but I'm not convinced we have included all components of this summation of force (are there some vertical components we have ignored, etc?). Certainly this summation force is not necessarily acting at a perpendicular direction to the sail surface at this CE point, and its not necessarily at a horizontal direction parallel to the water’s surface. And remember the sail cloth itself can not exert a forward force on either the mast nor the forestay, at least not in an upwind situation. So how are these ‘sail forces’ getting physically transmitted to the vessel?

I'm sure there will arise considerable discussions about the magnitude of these sail loads, but at least this could be dealt with as a variable aside from the question of load path. If we have the load paths defined, then we can play around with a variety of different load magnitudes and look at those new consequences. And then consider how the load paths can deform in direction under different loads.

I don't pretend to be any kind of an expert in these engineering/computer structural analyses. I would just like to get a clearer picture of how the sails actually transmit their forces to the vessel; at what points, and in what path(s)??

P.S. Finally, maybe someone could answer a particular question I have concerning the Larsson book. On page 149 there is presented a chart of lift coefficients for 5 sails as functions of the apparent wind angle. How is that the jib's coefficient is only equal to, or considerable less than that of the mainsail at the two upwind conditions? (and particularly in the light of acknowledgement that the jib is helped into a superior condition by the upwash of the mainsail).

 

brian says (in part):
"What really surprises me about these few quotes, and many other discussions on engineering a sailing rig, is the total dependency on the use of ‘guesstimates’ and a variety of ‘multipliers’, some of unknown origin and application."

My personal rig spreadsheet does not use guestimates or multipliers. It considers the sails, the RM, and the bend, and puts them together in a form that <gasp> forces the operator to think, and tweak, until the desired result is achieved. Not "push-button" design by any means, but I know what my masts will look like when bent, and how close the UBS each chunk of rigging will be at max RM. As far as dynamic loading is concerned, well..... there are "multipliers for that. Can you think of a method to calculate (accurately) dynamic stresses for a rig that would not take the national supercomputer network a month to solve for each boat?
THat being asked, the only way to assess these loads is by the "well, that stick didn't fail, so let's try one size down" method.
I don't liek it, but it's the best game in town these days, unless the owner wants to pay more for the design of the rig than the boat costs to build.
Steve "just ask Hall to design it......"

 

Brian,
I think you have voiced alot of my frustration with the available methods. I suppose that once someone like Steve has developed a program or spreadsheet to deal realistically with the problem, they want people to come to them and pay for their expertise (and deservedly so). That leaves the rest of us to flounder with half-baked methods not much better than the old rules of thumb.

Steve,
The problem with letting Hall Spars design it is, I have this structural engineering degree, darn it, and I've got to figure it out myself. So, as you apparently have done, I am putting the puzzle together so I can at least figure out the "real" static loads (as if any loading on a boat is really static). Now, if I can just get some believable rules of thumb for dynamic multipliers...;-)

 

shu,
Try this for a dynamic loading...
Assume the boat is upright, all sail set, no wind at all. Now assume the boat gets hit by a zero-warning gust, and capsizes in a very small amount of time. What (using generalities) would the load on the sails have to be? You can use the stability curve from 0 to 90 as a basis for calculating the work required to 90 degrees, and the time factor will give the forces required to achieve it.
That said, you have to assume at some point that it will be preferable for the rig to be ripped right out of the boat before something happens - better to lose the rig than the whole boat....
Steve

 

Static vs Dynamic Loads, & Load Paths

Hello Shu,
You wrote, "I have this structural engineering degree, darn it, and I've got to figure it out myself. I am putting the puzzle together so I can at least figure out the "real" static loads (as if any loading on a boat is really static).."

Brian replied:
I was going to suggest that we begin with the most basic structural analysis, that of the static case. And then consider the more complicated dynamic case (and safety multipliers) at a later date.

When I speak of basic, I really mean basic. Lets say we take the headsail alone of a bermuda rig . If we remove the sail from the vessel and replace it with the loads it excerts on the vessel, how would you define those load paths?? Woops I think we have to consider at least three different conditions here, upwind, reaching, and downwind. So to begin with lets just say upwind.
Does any of the forward drive imparted to the vessel get transmitted thru the forestay??

Interesting that you have a structural engineering degree. There is another naval/computer engr from India who has expressed a great interest in working on this analysis with me. Would you be interested in participating in a group effort?

 

Brian,
Yes, I would be interested. I must warn you though, I am quite rusty. It has been 15 years since I earned my degree, and I have solved only simple structural problems since then.

As regards the forestay. The sheet tension on a jib is quite high when going to windward, so I think the forestay will actually sag to leeward and aft. This is borne out by photos that I have seen in C. A. Marchaj's books. Next time I'm out I'll take a trip to the bow to verify this. I will also take a straightedge to photos in my favorite sailing magazines.

I would like to avoid getting overly detailed in the analysis though. You mentioned taking sail construction, seam location, etc. into account. I think that is unnecessary, since once you take the uncertain dynamic and safety factors into account, any precision you gained from such a fine-tuned analysis would be lost.

Steve,
That would be quite a horrific dynamic analyis to determine the time required for a knockdown. Perhaps you could assume the natural roll period for the complete boat? That will take some thought.

 

Re: Determining Loads in Sailboat Rigging

"In Hazen's model the lift and viscous drag of each sail are prescribed as functions of the apparent wind angle." (Larsson & Elliason) [emphasis added] Take a look at the angle - 50 degrees. The jib is either stalled or has been sheeted out and is not that effective. The drag coefficient is way higher than the main's, also indicating the jib is at least partially stalled. This condition is representative of a reach, not a beat. Even the spinaker is more effective than the jib under these conditions.

 

shu wrote to me:
I would like to avoid getting overly detailed in the analysis though. You mentioned taking sail construction, seam location, etc. into account. I think that is unnecessary, since once you take the uncertain dynamic and safety factors into account, any precision you gained from such a fine-tuned analysis would be lost.

brian replied:
I think you have me confused with another responder. I did not suggest this amount of detail. I was in fact trying to take the discussions back to the basics....forget the magnitudes of the loads at first and just concentrate on their paths. I think I will expand upon this approach on the other related thread, Sail Loading on the Rig, Rig Loading on the Vessel

guest wrote:
"In Hazen's model the lift and viscous drag of each sail are prescribed as functions of the apparent wind angle." (Larsson & Elliason) [emphasis added] Take a look at the angle - 50 degrees. The jib is either stalled or has been sheeted out and is not that effective. The drag coefficient is way higher than the main's, also indicating the jib is at least partially stalled. This condition is representative of a reach, not a beat. Even the spinaker is more effective than the jib under these conditions.

brian replied:
The problem I am having with this table presented in the book is that it is being presented as a "model" to predict with, to build rules of thumb upon. They do Velocity Prediction Programs with this standard! I don't know the details of its origin, but I would think that if they intended to utilize it as some sort of standard then they would seek to have each element acting at its best potential.

So I don't buy your argument that in this particular case the jib was just not sheeted properly. That mistake would flaw their base model. And BTW I did notice that it was apparent wind that the table was based on.

I trim my sails to the apparent wind, not true wind. So in opinion 50 degrees is not really yet a reach. But besides this I believe most modern text have come to agree that in an upwind situation the flow of air around the sails is diverted from the slot and around the outside of the main and the jib. This results in the jib receiving an upwash from the main that increases its efficieny and the jib downwashes the main which decreases its efficiency. This table would have me believe that the jib is MUCH MUCH less efficient than the main at 50 degrees apparent wind. Sorry I 'm not ready to buy this argument. And thus I may not be able to accept much of the follow on discussions that might include this as a founding basis. See my problem??

 

BTW, I sent an eMail to the author of the book inquiring of my perceived inaccuracy of this table, but have not received an answer yet.

 

Hazen's model came from an extensive series of wind tunnel tests conducted at MIT with the object of applying contemporary engineering to sailing yacht performance. It's the original basis for the IMS velocity prediction program.

Campbell compares the Hazen sail model with other wind tunnel tests for a single configuration, and the Hazen model agrees reasonably well with the data collected at the University of Southampton.

Since its inception, the IMS model has undergone continuous improvement, with US Sailing funded wind tunnel tests conducted specficially to address weak points in the model, plus practical experience. Claughton presents a more up-to-date version of the sail model, and it's clear the main has a higher lift coefficient at all apparent wind angles. The jib also has higher drag in the model.

So these data do not bear out your contention that the jib should be more effective per square foot than the main.

However, one should remember that these coefficients are intended for a component build-up model of a complete rig, and are further intended for handicapping purposes, not necessarily for design purposes. So the discrepancy may be one of bookkeeping, not accuracy. It's rare that a boat would be raced with jib alone, but sailing with main alone would be more likel. Plus the model has to handle cat rigs as well as sloops. So the coefficients for the main may well represent the performance without the jib, and the jib coefficients intended to give the proper total when added to the coefficients for the mainsail.

Physically, the loading on the main may be diminished in the presence of the jib and the loading on the jib increased in the presence of the main. But it's not necessary to explicitly show these effects when the object of the model is to predict the overall performance of the rig. All that's needed is to get the right total.

I suspect you are reading more into the Hazen model than is warranted. You can't divorce the model from its intended purpose or the bookkeeping inherent in its aero buildup equations.

If you really want to calculate the loading on individual sails in a multi-sail rig you need to employ computaional fluid dynamics (CFD). The simplest model is provided by the vortex lattice method. VLM will not give you the maximum lift, nor will it predict the viscous drag. But it does give reasonable estimates of the loading and interation between the sails as well as the induced drag. Predicting maximum lift requires either empirical corrections to the CFD results or more advanced prgrams, like Navier Stokes methods, that really require a specialist in the field.

The papers by Thrasher, Greeley, Masuyama and Courser illustrate the use of VLM to predict sail rig performance. Masuyama's paper is particularly useful because it includes full-scale sailing data for comparison.

There are many CFD codes that could be used for sail rig predictions. One commercial code is CMARC from Aerologic ( http://www.aerologic.com ). It's a full Morino-method panel code, originally developed at SDSU for NASA, but also has vortex lattice panels that can be used to model sails. It also has the ability to allow the wake to roll up for a nonlinear analysis. The granddaddy of VLM codes, Lamar's, is freely available from Prof. Mason's site at Virginia Tech ( http://www.aoe.vt.edu/~mason/Mason_f/CAtxtAppD.html ). He also has an optimization code ( LAMDES ) that might be useful for main/jib design. Other codes that might be suitable include the public domain version of Boeing's Panair code ( http://www.pdas.com/ ), Desktop Aeronautic's Linair ( http://www.desktopaero.com/ ), and VSAero ( http://www.am-inc.com/ ).

I suggest you might find it useful to become more familiar with the available literature on rig design and aerodynamics. Acquiring the proceedings from past Chesapeake Sailing Yacht Symposia would be a good place to start.

See:

Hazen, G. S., "A Model of Sail Aerodynamics for Diverse Rig Types", SNAME New England Sailing Yacht Symposium, New London, Conn, March 1980.

Kerwin, Justin E., Newman, J. N. , "A Summary of the H. Irving Pratt Ocean Race Handicapping Project," The Fourth CSYS - January, 1979. ( http://wseweb.ew.usna.edu/nahl/csys/default.htm )

Campbell, I. M. C., "Optimization of a Sailing Rig using Wind Tunnel Data", The Thirteenth CSYS - January, 1997.

Claughton, Andrew, "Developments in the IMS VPP Formulations", The Fourteenth CSYS - January, 1999.

Thrasher, D. F., Mook, D. T., and Nayfeh, A. H., "A Computer-Based Method for Analyzing the Flow over Sails", The Fourth CSYS - January, 1979.

Greeley, David S., Kirkman, Karl L., Drew, Alan L., and Cross-Whiter, John, "Scientific Sail Shape Design". The Ninth CSYS, March, 1989.

Masuyama, Yutaka and Fukasawa, Toichi, "Full Scale Measurement of Sail Force and the Validation of Numerical Calculation Method", The Thirteenth CSYS - January, 1997.

Couser, Patrick and Deane, Norm, "Use of CFD Techniques in the Preliminary Design of Upwind Sails", The Fourteenth CSYS - January, 1999.

 

Nbs

I am trying to find the Nordic Boat Standard.

Could you please tell me where I can find them in English?

 

B&R rigging discussions

...over on another subject thread:

Shallow-Angled Backstay of B&R rig
http://www.boatdesign.net/forums/showpost.php?p=223293&postcount=18

Shallow-Angle Backstays in several rigs
http://www.boatdesign.net/forums/showpost.php?p=223297&postcount=19

Cap-Shroud Loadings on B&R Rigs
http://www.boatdesign.net/forums/showpost.php?p=223299&postcount=20