Sails are foils that generate lift as a result of air flowing across them. Because foils can generate more lift with air flowing across them than the maximum drag would be for the same area there is a huge incentive to sail across the wind instead of sailing dead down wind. What sail shape is ideal for any particular conditions depends on a large number of factors, and sail shape is also constrained by inherent limitations in the construction of a sailing rig. A good source of information about wind tunnel testing of models and the aerodynamics of sailing is "Sail Performance: Techniques to Maximize Sail Power" by C.A. Marchaje, 2nd edition published in 2002 by International Marine.
Two is More than Twice as Good
Sail Shape
Sail Trim
Stability and Comfort
Aspect Ratio and Roach
Roller Furling
A common, and very significant, piece of information about sailing is the fact that two sails work a whole lot better than just one sail. The reasons that two sails are dramatically more effective than one sail are numerous, but one really big factor is the interaction of multiple foils. Basically two interacting foils are more efficient, generating more lift with less drag, than any single foil. This can be seen in the fact that a sailboat will be much faster, and able to point higher on the wind, with a small head sail and a reefed mainsail than it would be with either a large headsail alone or the full mainsail alone. The reasons that adding a headsail makes a boat so much faster though are not limited only to the increased efficiency of two interacting foils. Another big factor is the interference of the mast with the mainsail. The mast is placed in precisely the wrong place just ahead of the mainsail, and it really quite horrible messes up the flow of air down the leeward side of the mainsail. Adding a headsail makes the entry of the mainsail much less significant in the overall foil system, and even more dramatically increases overall sailing performance than would be expected simply because of the dual interacting foil boost in efficiency.
The other main reason that two sails is much better than one is that it gives a whole lot more flexibility to the sailing rig, both in terms of getting just the right amount of sail area up for the conditions as well as in terms of placing the center of effort of that sail area at just the right position fore and aft on the vessel. Two smaller sails is also always much easier to handle than one larger sail, and this becomes even more significant on large vessels.
The basic concept in sail shape is that the most efficient foil has more curve in the forward section and then becomes much flatter in the aft section. Just what the ideal sail shape is though gets muddled and confused by the fact that the interfering mast means that the mainsail ends up with quite a bizarre stepped shape. The origin of strangely shaped forward sections of mainsails has to do with the fact that the mast is less of a problem if it is off to the windward side of the sail as opposed to being off to the leeward side of the sail. Rotating masts are able to get the mast off of the leeward side of the mainsail, but with a fixed mast the best solution is a stepped sail shape where the forward part of the mainsail hangs off to leeward of the mast Getting the mainsail to hang off to leeward of the mast is however not so easy to attain, especially with heavily loaded lightweight sails.
The compromise most often employed is simply to put more curve (more camber) in the mainsail. More camber makes the mast problems less sever for a variety of reasons. One of course is that the more heavily curved sail tends to dwarf the mast, in other words the mast is less of a problem when the mainsail has a significantly larger camber dimension than the diameter of the mast. The other reason that more camber tends to make the mast interference problem less severe is that more heavily curved sails can be used with much tighter sheeting angles. That is a sail with more camber will remain decently efficient when it is sheeted to have a larger angle of incidence with the wind, and that large angle of incidence with the wind effectively moves the mast toward the windward side of the sail. Flatter sails can generally attain higher maximum efficiencies, especially in strong wind, but attaining this higher efficiency from flatter sails requires shallower angles of incidence and the range of angles of incidence that will work well also becomes narrower with flatter sails.
Because fuller sails will work over a wider range of angles of incidence to the wind they often seem to make boats sail faster in real conditions. There are however some sever problems with fuller sails. The main problem of course is that the lower peak efficiency of fuller sails means larger healing forces will be generated, and the vessel will tend to heel over more severely. Fuller sails also cannot be made to point as high on the wind, somewhat compromising upwind performance on many vessels. Sails with a large amount of camber also flop around more in light wind, and this can be a huge problem for ocean sailing on small vessels. Using flatter sails makes a small sailboat immensely more enjoyable and easier to sail in challenging conditions.
All sails generally benefit from extremely flat aft sections, with almost no curve to the aft third of the sail. Attaining these very flat aft sections can however be quite difficult. If sails are made exesively flat on the aft sections problems with puckering and curling back at the leech can be a problem. The high performance composite ultra low stretch sail materials are generally able to attain smother and flatter aft sections, and this is one of the main advantages of the high tech sail materials. The oft quoted advantages of lighter weight and higher load carrying capabilities are also significant advantages of the high tech sail materials.
Sails are constructed with a certain shape built into them, but the actual shape that the sails take when flown on a sailing vessel can vary quite a bit depending on how they are trimmed (trimmed with lines not trimmed with scissors). The extent to which sails can take on different shapes based on sail trim depends on the material the sails are made out of and also depends on how flexible (adjustable) the sailing rig of the vessel is.
The main thing that can be done to change the shape of sails is to move the clew (lower aft corner) of the sail forward towards the tack (lower forward corner) for a fuller sail or to move the clew back away from the tack to pull the sail flatter. Moving the clew forward pulls more camber into the sail, but this is not a perfect adjustment. Moving the clew forward also tends to move the point of maximum camber aft in the sail, which is generally not desirable for good sail performance. Some less heavily loaded main sails constructed from slightly stretchy material can also be adjusted by pulling on the leech (the forward side of a sail) or the foot (the lower side of a sail) to larger or smaller degrees. The way that this is done is to remove the lower mainsail slides and pull the tack back away from the mast to flatten the sail or to use a Cunningham to pull the tack closer to the mast for a fuller sail shape. Removing the foot of the sail from the boom (converting the sail to be a loose footed sail) can also allow for a slightly fuller shape. When a mainsail is reefed all of these same adjustments are also available by moving the reef tack and reef clew in and out along the boom. High performance low stretch sail materials generally are not nearly as responsive to being shaped by different sail trim strategies. Allowing the clue to come closer to the tack can however nearly always be used to get a bit more shape in the sails, regardless of what they are made out of.
In practice the largest determinant of how fast and how well a vessel will sail has to do with how well the sails can be placed to catch the wind. When sailing close hauled hard on the wind there is generally no problem with pulling the sails any which way that is required to get just the right angle of incidence to the wind. Likewise falling off and sailing on a close reach usually still causes no sail handling problems as it is easy to get the sails placed just where they need to be. When reaching or running though getting the sails where they need to be for best performance can be much more difficult.
The main problem with sail handling when on a reach or a broad reach is that the sails tend to twist a whole lot. When the sheets are eased to catch more wind the top parts of the sails fall off much more than the lower parts of the sails, and this dramatically hurts performance. It is often pointed out that a small amount of sail twist is desirable because the wind speed is higher up at higher altitudes and lower down close to the wave tops. This is absolutely true, but the reality is that the ideal amount of sail twist is very small, and larger amounts of sail twist hurt performance and make for a rougher ride as well.
Preventing sail twist means moving the clew of the sails out farther instead of simply easing the sheets, and this is not always easy or straightforward to accomplish. For the head sail there are two things that can be done to reduce sail twist. The main strategy has been to use a genoa pole to force the clew out away from the boat, and this is extremely effective if not entirely desirable for ocean sailing. Since doing the pole dance on the foredeck in rough ocean conditions is not the most fun part of sailing genoa poles are often dispensed with for casual ocean sailing. The other options for broad reaching headsails have to do with getting the wind to pull the sail out away from the vessel. This is usually in the form of a cruising spinnaker that can easily be pulled in when the wind gets too strong, but other head sails can also be made to work in the same general way. A high cut headsail for example is much more flexible when sailing off the wind with no pole than is a deck scraper. The high cut sail can be made to fill out away from the boat to a certain extent simply by moving the sheeting position forward when broad reaching.
If no genoa pole or spinnaker is used then sailing off of the wind tends to rely much more heavily on the mainsail, and getting the mainsail clew farther out away from the center of the boat can be even more difficult. A boom vang can be used to keep the boom down, but on most vessels this only works in light wind. Supper vanging of a boom to broad reach in enough wind to sail fast can be made to work on some smaller boats, particularly on an open boat where the boom vang is led all the way down to mast step at keel height. On boats where the boom is only a small distance above the cabin top the boom vang is only of very limited use. The most talked about way of getting the mainsail to sheet out farther for reaching in substantial wind is the mainsheet traveler. By moving the mainsheet out to the lee rail of the boat sail twist can be kept to a reasonable amount while getting the boom out a bit for reaching. The amount that the boom can be let out with a traveler depends on how far forward on the boom the sheet is located. With the sheet all the way at the aft end of the boom even a mainsheet traveler only allows the boom to be let out a small amount. Mid boom sheeting moves the sheet farther forward on the boom, allowing to boom to be let out much farther without the sail twist becoming excessive.
There are however some problems with mid boom sheeting. The obvious one is that the farther forward that the sheeting position is moved the stronger the boom is going to have to be to support the large sheeting loads. Also of concern though is the instability of a boom sheeted far out to leeward. If it is only the sheet holding the boom then an accidental jibe can put absolutely immense shock loads on the boom, sheeting equipment and the sail itself. For security when sailing shorthanded offshore something has to be done about shock loads in an accidental jibe. The traditional solution was a preventer from forward on the boat out to the end of the boom. This did the trick for preventing the initial shock loads of an accidental jibe, but the difficulty in smoothly releasing the preventer while hauling in on the mainsheet meant that additional crashing and high shock loads were likely before the boat was brought back under control. Somewhat unexpectedly a boom vang is a great advantage in jibing maneuvers because it keeps the boom from bouncing up and down so severely as the mainsail luffs. A boom vang however does little to prevent the huge shock loads associated with crash jibes in substantial wind.
The solution is a mid boom sheeting system using two sheets instead of just one sheet. Each of the sheets are attached farther forward on the rails so that when the boom is sheeted down for reaching it is securely held forward and cannot crash during an accidental jibe. The reason that the dual sheet system works so much better is that both sheets can then be operated from the cockpit to smoothly haul the boom back to the center of the boat after an accidental jibe. This improved mid boom sheeting system still requires a substantial boom capable of being sheeted in the middle, but removing the shock loads of cash jibes dramatically improves the reliability and longevity of the boom and the entire sheeting system.
The challenge of ocean sailing is not all about just getting the boat to move quickly in the desired direction, but rather has a great deal to do with trimming the sails for a stable and comfortable ride as well. The basic idea of sail stabilization is to keep a rather large amount of sail up. The most useful thing that sail can generally do is to prevent rolling (the side to side motion of a boat), which requires that at least some of the sail be sheeted in rather close. Close sheeting of sails is contrary to good driving force when sailing off the wind, and insufficient driving force tends also to cause stability and comfort problems. The solution that has been used for the past several hundred years is to keep just some of the sails sheeted in close for stability while using the remainder of the sails let far out for driving force generation. This remains the best way to sail most displacement speed vessels, but it is by no means the only strategy that can be employed. If sufficient driving force can be provided with a spinnaker then the rolling of a vessel can also be controlled by sheer force. Obviously a sailboat that is capable of coming up on a plane will strongly resist rolling once it is up on the plane, but even heavier boats that are limited to hull speed or just over hull speed can use sheer force to limit roll. On heavier boats it is the wave generation that inhibits roll, if the boat is throwing a large enough wake it will strongly resist rolling from side to side. This is also the reason why so many displacement speed power boats have been fited with engines large enough to throw a giant wake. Burning twice as much fuel to go 10% faster seems pointless, but that extra power to the water is often considered indispensable for providing a smoother ride.
It is not just rolling that a large driving force is good for controlling, as pitching and yawing generally also diminish with large driving force. On a sailboat it is not just about the driving force either, as sufficient sail also tends to damp the pitching and yawing motions of the vessel. More driving force leads to a more stable and more comfortable ride up to a certain point. If the vessel is capable of speeds of more than about 12 knots then the most comfortable ride is a compromise between damping wave generated rolling and pitching versus minimizing the heaving associated with traveling fast over the swells. Going too fast over waves can make for an extremely bumpy ride. Just how fast a vessel can go over the waves depends mostly on the size of the vessel, bigger ships can go faster over the same waves. Because the wind waves and swell in the open ocean are made up of waves of many different sizes though, it ends up being a rather complex set of variables that determines what speed is possible without excessive heaving, surging and slamming. Certainly a vessel cannot go so fast that it flies off of the crests and then comes crashing down in the troughs, as this is the roughest type of ride imaginable. Even before a vessel actually becomes airborne off of the wave crests there are milder sorts of problems that can crop up. If the waves are steep from a recent large increase in wind speed or from the wind opposing the current then a situation is often created where the vessel ends up burying it's bow in waves. This generally makes for a very wet ride, and also can create a quite uncomfortable harsh surging motion on lighter vessels.
It almost goes without saying that going into the wind and waves at high speed makes for a much rougher ride than going with the wind and the waves. Reaching across the waves can usually be done at rather high speeds, which is convenient since reaching is where most vessels attain their highest sailing speeds. Reaching and broad reaching also is generally where rolling is the largest problem, making sail trim decisions quite difficult at times. When broad reaching letting the sails out as far as they will go is fastest, but letting all the sails all the way out also removes the roll damping of close sheeted sail. Often reaching or broad reaching in a moderate breeze is a matter of getting the sails trimmed for the highest driving force generation possible so that either planing or wave generation can be used to damp the rolling.
Keeping a head sail sheeted in close is usually a much better strategy for a stable ride than trying to use the mainsail for stabilizing. Even on boats with much larger mainsails than headsails the ride is usually a lot better with the mainsail let far out and the headsail sheeted in close. There are a couple of reasons why this is the case. One is that the vee shaped sailplane with the point of the vee pointing into the wind is an inherently more stable shape than other sail plans. This vee shape is particularly good for resisting yawing, helping to keep the boat on course and reducing stress on steering systems. The other reason why a close sheeted head sail is better for stability is that it allows both sails to work towards roll damping. The close sheeted headsail acts to damp the roll of the vessel even when it is fully blocked from the wind by the mainsail. The mainsail let far out to leeward acts to damp rolling when it is "filled with wind", to quote sailors of the 19th century. A sail let far out does little to damp roll if it is not "filled with wind". If a poled out genoa is used for driving force generation and the main is used for stabilizing then the genoa is inhibited from being able to stabilize rolling whenever it is blanketed by the mainsail. And of course frequent jibing to keep the wind angle as far off the stern as possible when running is also a whole lot easier with the head sail sheeted in close than when running on a poled out genoa.
Flatter sails are usually better for maintaining a stable and comfortable ride in the open ocean, but this is not universally true. A flatter headsail is of course far better for sheeting in tight as a stabilizer. A flatter mainsail can be let out farther for running dead down wind, which generates more driving force for a faster and smoother ride in light wind. When sailing upwind flatter sails allow more sail to be carried, which is often huge advantage. A larger area of flatter sails generally allows a vessel to point higher on the wind, resulting in better velocity made good with lower sailing speeds. The advantages of a flat main sail for motor sailing cannot be overlooked either. An overly full bodied mainsail crashing back and forth after the wind has died is not only loud and hard on the entire rig, but also makes for a rougher ride until either the wind comes back or the waves subside.
Sails with more camber can however be an advantage in some reaching situations. The most obvious advantage of fuller sails is when a vessel is not rigged to be able to let the sails far enough out to leeward for reaching. In this case the fuller sails can generate much more lift when sheeted in rather close. Sheeting fuller sails in closer can also be used to give a more stable ride when reaching in rough conditions. Flatter sails need to be more precisely trimmed, meaning that they have to be let out farther for reaching. There are many situations where the ride is much better reaching with close sheeted full sails than flatter sails let far out.
The most efficient foil shape is a rectangle much longer than it is wide. For sailboats though the ideal sail aspect is considerably different for a wide variety of reasons. First and foremost sailboats have triangular sails to fit under the stays holding the mast up. An unstated mast or running back stays removes this constraint of a triangular mainsail, but there are still substantial advantages to roughly triangular shaped sails mostly having to do with weight aloft and heeling moments. However the mast is held up it is going to be more difficult to support sail area higher up. An unstated mast needs to be heavier to support more sail area up higher, and a spar for supporting a rectangular shaped sail adds weight higher up. More sail area higher up also tips the vessel over more, requiring more ballast placed deeper down under the water.
The compromise between the aerodynamic advantages of tall rectangular sails and the weight and heeling moment problems of sail area placed high up means that the best shape for sails is a long tapered shape similar to the shape of a bird's wing. This ideal tapered shape has a convex curve along one or more edges so that the inner (lower) part of the foil has edges closer to parallel. Attaining this shape on a sailboat requires the use of battens to support roach sticking out the back of the sail. Head sails can use battens to support roach, but this is not really necessary because the lower part of the headsail can simply overlap the mainsail. Mainsails usually use battens to support a substantial roach, and roach certainly does increase the performance of mainsails considerably.
Partial battens are a bit lighter and allow a greater level of flexibility in sail shape but are not able to support as much roach as full length battens. Even though full length battens are heavier and impose more sever constraints on the shape of the sail there are other significant advantages of full length battens in addition to the ability to support much higher levels of roach. One really big advantage of full batten mainsails is that they do not flog around as severely when raising, lowering or reefing the sail. On some boats the extra time for sail handling permitted by the full battens can make a world of difference. In very light wind the full battens can also hold a bit of shape in the sail when a partial batten mainsail would hang uselessly in a wrinkled mess.
The reason that full length battens can support so much more roach is that the batten load is transferred directly to the mast, as opposed to the batten load being taken by the center of the sail. This more substantial handling of the batten load also tends to make full batten sails more stable and longer lasting under heavy use. Even with a rather modest level of roach a sail designed for the higher batten loads possible with full length battens tends to be smoother and is able to have a flatter aft section.
The ideal amount of roach is a very contentious subject. Increasing roach radically increases sailing performance, but large amounts of roach create their own problems. More roach means higher batten loads, which requires stronger and heavier battens as well as much more substantial batten car systems. More roach also puts much higher tensile loads on the leech of the sail, and the heavier sail cloth required in the area of heavy roach means more weight aloft. Heavily roached mainsails create much higher halyard loads, which increases the compressive loading of the mast. On racing boats more roach translates into a faster sailing boat with the same height mast, so very large amounts of roach are used. Even if the stresses on the sail and rigging are not considered there are some other reasons that excessive roach is undesirable. A huge amount of roach moves the center of effort of the mainsail up higher, which then requires a deeper and/or heavier ballast keel to provide sufficient righting moment. Large amounts of roach also require running back stays instead of a fixed backstay. Extremely heavily roached mainsails are fast, but they are not necessarily friendly.
The big disadvantage of full length battens is simply the weight. It is usually said that full batten mainsails are not as severely stressed, and can therefore use lighter sail cloth to make up for the weight difference. The reality though is that for extended use in challenging conditions the sail cloth still needs to be just as sturdy, and full batten mainsails really are heavier. The weight difference is mostly just the weight of the battens themselves, although batten car systems can also add some substantial extra weight. The extra weight of batten cars comes into play because it is so important for them to smoothly move up and down the mast if sail handling is to be easy. It is easy to get batten cars to move smoothly when the sail is unloaded and hanging straight back in the center of the boat, but getting a batten car system to work for reefing while sailing down wind is a much larger challenge. Getting a full batten main to go up and down while it is under a load can mean extensive rail and car systems with multiple rollers, and these certainly can add substantial weight aloft.
Sometimes it is considered acceptable for the vessel to have to be brought into the wind under power to reef or shake out a reef in a full batten mainsail, but this is really incorrect thinking. A mainsail must be able to be reefed while the vessel is sailing downwind for sail handling to be easy. If the batten cars jam and will not go up or down when they are loaded by the sail having wind in it then sail handling will be all but impossible and broken gear or a torn sail is a likely outcome. On very large sloops with immense main sails this may play out differently, but reefing a two thousand square foot triangular mainsail would rarely be considered an easy task. Very large sloops, cutters and ketches quite often use a roller furling mainsail, and this is a whole different can of worms.
Being able to quickly and easily make sails get smaller or disappear is a compellingly desirable dream of all sailors. The manifestation of this dream is the modern roller furler head sail. Roller furling is not always entirely trouble free ease of sailing, but the advantage of being able to just whisk the sail away is in fact extremely useful. Simply not having to go up on the foredeck for headsail changes often is the difference between an enjoyable sail and an arduous task.
The major drawbacks to roller furling head sails are weight and expense, but there are many other consequences of roller furling head sails as well. One big difference is being limited to using just one headsail most of the time. It is a whole lot more difficult to change the sail on a roller furler drum than to hank on a different head sail. This is actually a very significant limitation, but it is one that most casual sailors are only too glad to live with. The key to successfully using just that one roller furler head sail is for it to be sized small enough to work for a wide range of conditions and for there to be some other type of larger headsail that can be used when the going gets really light. The usual additional headsail is a spinnaker used either with or without a pole. As easy and fun as cruising spinnakers are to use in settled conditions they are not appropriate for many types of shorthanded sailing in challenging conditions. Dual head stays is a more flexible solution.
Dual head stays is not the same thing as a cutter rig. A true cutter does have some of the same advantages, but the fractional inner stay requires some form of lower backstays if the rig is to be strong and stable. Fixed lower aft stays generally means that the mainsail cannot be let out nearly as far when running, so the added complexity of running lower back stays is really the only option for a functional cutter. Without lower backstays a cutter becomes a mast cutter, that is it cuts the mast in half.
Dual head stays retains the simplicity of a sloop because both head stays attach close to each other at the top of the mast. Usually the forward head stay has a larger light wind genoa on a roller furler or sometimes a furling nylon "code zero". The forward head stay can also use a hank on light wind genoa instead of a furling sail, with the substantial advantage of lower windage in heavier wind. Hank on light wind genoas work quite well because the extra labor of using a hank on sail tends to seem much less onerous on a beautiful sunny day with light wind than in stormy and changeable conditions. As long as the boat sails well enough for most normal use on just the 110% jib, the hank on drifter can be reserved for special use in very light wind or occasional spirited sailing in a bit more wind.
It is often said that roller furler head sails should not be used partially rolled up, and there is substantial reason to take this advice seriously. Using a headsail partially rolled up does put stress on it that is much more severe than normal use as a full sail. Sails that are often used partially rolled up tend to get pulled out of shape and distorted quite horribly. The reality though is that a roller furler headsail does not really work unless it can sometimes be used rolled up smaller. A cutter rigged ketch with two roller furling head sails might be able to be sailed without using the headsails partially furled, but a sloop has a very hard time managing a wide range of conditions unless the head sail is sometimes used partially rolled up.
The key to successfully using a roller furler sail partially rolled up is to always roll it up sufficiently to take a large amount of the stress off of it. Once the wind is too strong for the full headsail it is necessary to roll it up down to no more than two thirds of it's full area, and keeping rolled up headsails smaller than half of the full area is even better. This does make for a sort of a dead spot where a certain amount of wind makes the boat slower. The dispensation is that this is not an absolute limit, but rather an imposed "soft" limit used to prolong sail life.
Another trick to getting roller furler head sails to work well is to use a rather flat cut sail with no additional UV stip. The UV cover is usually considered absolutely necessary, but it does add a whole lot of bulk to the rolled up sail and severely interferes with the performance of the sail when partially rolled up. A flat sail with no UV strip rolls up much tighter and actually does not tend to get pulled out of shape as easily when used partially furled. The disadvantage of using no UV cover is that headsails then last only about two years. One way that the sun damage can be mitigated is to reserve a headsail with no UV cover for passage making, and install another sail with a UV cover for more casual use during in between times. Anytime that some serious sailing over long distances or in changeable and challenging conditions is expected the flat sail with no UV cover is installed, and languishing in port is handled by any old sail with a UV cover. This way the vessel is always ready for a casual sail, but expensive head sails are not uselessly rotting off of roller furlers every year.
Roller furling mainsails are tricky business, entirely outside of this authors experience, but a few general points should be made anyway. The main problem with roller furling mainsails is that they are not able to be rolled in or rolled out when they are under a load. The reason for this has to do with the way that the luff loads are supported by the mast. The roller furler mast has a drum inside for the sail to roll up on, and a rather narrow slot in the back of the mast for the sail to come out of. When the sail is pulled outward the drum rests against the inside of the mast. With the drum resting on the inside of the mast there is no way for the sail to be rolled in or out under a load. When the load is removed from the sail though the drum is pulled back away from the slot in the mast by the tension in the cable that it rides on. With the drum free to rotate inside the mast the sail can easily be rolled in or rolled out. The tricky part is maintaining just the right amount of tension on the sail when rolling it up so that it rolls tightly around it's drum, but does not pull the drum so far back that it jams against the mast. If the sail is rolled up too loosely it will fill the available space before it is fully retracted. Increasing the tension in the cable that the furler drum rides on increases the amount of force that can be applied to the sail while rolling it in or out, but the increased tension in the cable also increases the compressive loading of the mast.
Most roller furler mains can be rolled in or out quite easily in light wind without having to bring the bow into the wind. When the wind picks up a bit though and a smaller sail size is required it is necessary to power into the wind and roll the sail up as it flogs freely in the wind. When this strict procedure is followed roller furler mainsails seem to work just fine, but there are substantial performance penalties associated with furling main sails as well. Most roller furler mains are battenless sails, meaning that they require substantial hollow in the leech to prevent leech flutter. This hollow in the leech both reduces the sail area and also severely compromises sailing performance for the same reasons that roached mainsails so dramatically improve sailing performance. A rig designed specifically for a battenless roller furling mainsail can however be a bit taller with the same keel and ballast. Still though battenless mainsails are a large hindrance to performance.
To get back some of the lost performance furling mainsails are sometimes built with vertical battens to support a small amount of roach. The extra weight of the battens is however very unappealing because they always stay high up on the mast. The battens can also interfere substantially with furling performance because they tend to get caught on the slot in the mast. A longer and rounded slot can allow vertical battens to work acceptably, but vertical battens have not been widely used.
In boom furling mainsails seem to be a bad idea because they are prone to rolling up crooked if the boom is not constantly maintained at just the right angle to the mast while the sail is being rolled up. In boom furling has however been quite popular at some times, probably because of the perceived advantage of being able to use horizontal battens.