Anti-Roll Bag ver. 2.1

 Here is what my off-the-shelf Anti-Roll Bag (ARB) version 2.1 looks like.

Folded up.

Unfolded in front of the seat boxes on the flying bridge.

It was intended for use as a "water diversion tube" and cost $34, including shipping.  When full, it is supposed to be a 12" tube that is 12' long.  Laying partially filled on the deck, it is about 20 inches wide.  The problem is that the construction material is about the same as a child's inflatable pool.  It leaked right from the start and I had to take it home to patch it.  While not a long-term solution for an ARB, it was valuable for experimenting.

Because it is several feet longer than my home-made bag (ARB ver. 1.0), I was curious if the water would, when passing side-to-side, inflate the excess tube at each end.  By "inflate," I don't mean with air.  All of the air has been bled out.  But the extra length would allow a more effective total transfer of the water and possibly a slight slowing of the release when the boat rocked back, as was discussed in ARB ver. 1.  At the end of ARB ver. 1 is a video that shows the effect of a flat seam at the both ends of that bag.  That restricts the amount and timing of the water.  The amount and timing of the water back and forth is the key to maximizing effect, i.e., comfort.

I partially filled the tube with about 3 inches of water.  It can't be filled much more than that and work properly.  At minimal rolls, the bag needs to be loose enough to take advantage of the "free surface" effect, although it is not quite "free" here.  There also needs to be enough "head space" in the tube to allow all of the water to transfer to the ends during a larger roll.  The tube could likely hold 70 gallons in theory (590 pounds).  I don't need or want that much shifting weight.

I put approximately 20 gallons (180 pounds) of water in the bag, which limits both the amount of roll attenuation and any possible deleterious effect to vessel stability.  The bag has the effective nature of a 180# person running back and forth on your flying bridge.  Some have expressed concern that this would collapse their flying bridge or overturn their boat.  The solution?  Do not board any vessel for which this would be a possible problem!!

Actually, I've found that the biggest problem with any roll reduction system is testing it, i.e., getting real numbers.  People who have added some type of roll attenuation systems (sometimes erroneously called "stability systems") are often pleased, despite having spent upwards of $40,000.  Yes, they report that their vessel is more comfortable.  Okay, but the question arises, at least for me, how much of that improvement is because of roll reduction and how much is confirmation bias after having dropped a pile of money?  What empirical testing was done before the modification?  Without that, and further testing after modification, how would one know whether simply filling the galley lockers with canned goods would have accomplished a similar roll reduction?  Or 500 pounds ($1,200) of lead in the bilge?  Or just spend the money to buy a lifetime supply of Dramamine for $40 (also available for dogs) and forget about your boat rocking?

What is needed is both a way to actually measure the roll reduction (if any) and then quantifying the effectiveness of a roll reduction system (maybe as dollars per degree of roll reduction??).¹  I'm not sure if there would be any agreement on the latter, although it seems like common sense to examine the cost/benefit issue.  If $40K stabilizer fins reduce a roll by 90% and $200 reduces the roll by 50%, which is the better system?  It could be that some will pay any price for the same feeling of stability as their "dirt home,"  i.e., they will pay any amount to have a "dirt boat."  Solid as a rock (which in a boat usually means you are hard aground.)  Seriously, I don't know if actual numbers comparing cost with effectiveness will change any opinions and would only cause arguments.  It seems that roll reduction has something in common with politics.

To further complicate things, roll reduction consists of two measurements, amplitude and acceleration.  A 10 degree roll can seem fairly minimal if the "snap back" is slowed down because of an increased roll period.  And it will depend on each individual's "sea legs."  ("Step right up and get your sea legs: $40,000 for a pair!")  But how does one test a system to get an accurate before and after picture?  One traditional way is to pre-guesstimate by doing complex calculations when the vessel is designed.  The other is to hire a naval architect after the fact and hope for the best.  As noted by the naval architect in the prior blog (ARB ver. 2.0), a naval architect can be expensive and still get the estimate all wrong.  

I think that what might be better for most recreational boaters is an old-school "tank test."  I'll just put my boat in a giant tank, generate a big wave of known repeatable amplitude, and see how the hull reacts given different adjustments to a roll reduction system.  Fortunately, I happen to have a huge tank and a giant wave generator available to me.  The Washington State ferry Walla Walla runs close by my marina several times a day.  At a length of 440' and a beam of 81', travelling at its standard 17.5 knots, the Walla Walla generates quite a wave.  It runs in a charted ferry lane in the calm Port Orchard Canal and its course, speed, and distance from me show up when I monitor its AIS broadcast.  That can be my tank with a wave generator producing a standard amplitude wake which I can monitor. 

As it turns out, the smaller Kitsap County "fast foot ferry," a catamaran running between Seattle and Bremerton, puts out a nastier wake for my boat when I'm running in the same direction.  These catamarans were designed to be "low wake" ferries (really?), but that is the wave generator I used for most of my experiments.  The sacrifices I make in the name of science.

Running my boat on repeated parallel courses at a standard speed gives me the required benchmark to accurately determine the effectiveness of changes to my ARB roll reduction system (or anybody's system, for that matter).  I can get the ferry SOG and COG from AIS.  For instance, when the ferry is running in the channel at its normal 271 degrees true at 17.5 knots (30 knots for the fast foot ferry), I can run parallel at 271 degrees, 5 knots, and exactly 300 yards away.  Having had the ferry pass me many times in the past, I know that within a minute or two this will give any roll reduction system a good workout.  And because of the ferry schedule, I can repeat the exact thing more than 12 times a day if I choose, or a couple times a day for a week, or try it again next month with a different system.  

It turns out that having the ferries approach from behind gives me the greater rolling.  If I'm going in the opposite direction, I would effectively be crossing the wake at a large angle.  When the ferry (or any large wake making vessel) sneaks up behind me, that is when I encounter a generated wave that most closely harmonizes with my boat's natural roll period (approximately 3.55 seconds).  At one point, I was lucky enough to get both the big Washington ferry and the little Kitsap ferry wakes at the same time.  I recorded a 12.5 degree roll (without the ARB).  We call those a "bell ringer" because my boat rocks enough to ring our bell.  I'm not going to do that on purpose again.  Although for big rolls, check out my post on Anti-Roll Bag ver. 2.2.

To measure the effectiveness, if any, of my roll reduction system, I use my inclinometer phone application.  I needed a standard roll test without any roll reduction system in operation to see just how much my boat rocks from that wake.  My experience has been that a 10 degree roll from the ferry wake is "normal."  Note that a 10 degree roll measured accurately with an inclinometer is what most boaters claim to be a 20 degree roll.  If one is not looking at an inclinometer, it just isn't possible to claim accuracy and, of course, the guess is always extremely high.  I introduce my inclinometer in Anti-Roll Bag ver. 2.2.

I also have an accelerometer phone app.  I have found that the "sensation" of a gentler roll is more easily felt than trying to analyze the sine wave recordings on the phone.  The resolution just isn't sufficient to really see what's going on.  Measurements to date indicate that the ARB has a significant effect on roll degree and, therefore, acceleration.   And what appears to be the biggest benefit is the reduction in a synchronous roll.  It's that old feeling, especially with a wake, where it starts small with back, Forth, BACK, FORTH . . . and then you know what's coming.  With the ARB (as likely with other systems), there is much less tendency to build.  Therefore, it is less likely that "the big one" will ever come.

The few detailed studies that I have found on anti-roll tanks (which have field-tested measurements) show that there tends to be certain amplitudes that are really reduced.  It sort of makes sense being that we are essentially "tuning" to offset frequencies and neither my roll period or the water transfer in the ARB remains exactly the same for different amplitudes.  The studies showed that, for instance, a tank may have sort of a "standard reduction" of 40%, but at several certain rolls the reduction can be 60%.  So the sine wave reduction pattern is not likely to be mathematically perfect. Kind of odd to think about, but one ends up with a 50% reduction overall.  All of the studies I have read were primarily concerned with reductions in amplitude with little addressed as to acceleration. It might be more complex to examine what an ARB does (and where it does it) to the vessel's roll acceleration.

With accurate measurements using my Washington State Ferry "tank and wave generator" and the accurate monitoring of my vessel's speed, direction, and distance off, I can generate accurate and quantifiable roll results, something which I have never heard any recreational boater mention when talking about their new "stability system."  

For the first numbers, I was surprised.  I had rolled my boat at the marina and timed 10 complete rolls.  It is actually quite complicated to do this without the boat sooner or later tugging on a mooring line.  But I got some clean "rocks" and came up with the 3.55 seconds per period.  Those tests usually started by building to a 7 degree roll, then I let go and measured.  With the ARB filled up, I had a very difficult time getting the boat to rock 6 degrees.  I was winded by that time and just let go and recorded the rocking.  Unfortunately, the rocking died down so that I couldn't get an accurate 10 rolls.  And what I did get surprised me.  The roll period averaged to 3.45 seconds.  That's not much of a change, but it is a faster roll period!  Just the opposite of what I was hoping to accomplish!  

But the benefit of the ARB can definitely be felt on board, so what's happening?  It appears that a reduction in roll angle/acceleration is more "sea kindly" than a reduction in roll period.  And one way to reduce acceleration is to reduce the arc, i.e., the distance traveled in a given period of time.  Getting the boat to rock was more difficult with the ARB filled, and the ability to stay rocking was greatly reduced, so much so that the additional 1/10 of one second in the roll period was immaterial.

Y

X

Above is a typical damping sequence.  The X and Y axis are not indexed, but we can see what's going on.  We will call the first roll (X₁) a "1," and that's basically the max we can get at the dock.  The boat then flops back to the other side to about an .8 (X₂), then back to .6, back to .5, back to .4, etc.  Not much damping going on and, looking at the entire graph, we could probably measure out to 10 complete rolls.  We then divide the time by 10 and get our seconds per roll measurement.  For my boat, that averaged 3.55 seconds.

Here is another damped cosine graph with greater damping.  The first roll at X₁ is a 1 again.  The second roll, X₂, is only a .6, followed by X₃ (a .36), then .24, .18, .11 and basically gone.  We can't measure a sequence of 10 rolls here and it's unlikely we could rock the boat enough.  Maybe 5 complete rolls is all we can get.  When the time is divided by 5, my boat averaged 3.44 seconds with the ARB filled (increased damping).  That is a 10th of a second faster.  But ultimately much more comfortable.

I think that some of the improved comfort when field testing was because of the "other side" of the graph that isn't taken into consideration here.  How does a boat generally get to a roll of "1"?  It is usually because of synchronous rolling.  Basically, what I was trying to do by running parallel to a ferry wake.  A graph for the building of a roll would look similar to a reflection of the above graphs because the same amount of damping effect would take place at the start of a synchronous roll sequence as well.  First port, then further to starboard, then even further back to port.  We've all felt that.  

But look at the difference increased damping would make in graph #2.  The first beginning roll would be choked more by the increased damping, making it unlikely that we ever get to a "1" based on the same wake height and frequency.  We would begin with a smaller roll because the boat is rocking less, a synchronous roll would then not build as fast (therefore as great), and the rolling stops sooner.  What more could we ask for for a $34 roll attenuation system?

Which brings me back to my $34 ARB.  It worked, but it leaked.  Bad.  I took it home and patched 4 leaks and still didn't get them all.  So I went with a slightly different system.  To be continued at Anti-Roll Bag ver. 2.2.

¹   Naval architects are very fond of abbreviations and acronyms for various formulas.  LOA for length overall, S/L for speed/length, etc.  I propose a new one for roll reduction systems (aka stabilization): Dollars per degree damping (abbreviated $/°d).  Capt. Beebe, in his book Voyaging Under Power, recommends that an ocean cruising powerboat should have both active fins (+$50K) and paravanes (+$20K).  But wait, there's more.  Operating both systems will increase fuel consumption and neither will be very effective at anchor.  I'll leave it up to the NAs to figure out how to calculate the actual ongoing $/°d of a +$70K stabilization system.  For that kind of money, the system better also keep my boat washed and my beer cold.

Anti-Roll Bag ver. 2.0

 Although my first experiment with the anti-roll bag (ARB ver. 1.0) was successful, I wanted to try a different bag. To accompany a blog about a brand new bag, you may want to open Papa's Got a Brand New Bag in a separate window for a musical accompaniment.    

I wanted to try some kind of a bag that was an "off-the-shelf" product.  I also wanted to get some actual numbers so that I could compare one version to another, both as to cost and effectiveness.  The cost for my first DIY bag, ARB ver. 1.0, was about $200.  The cost numbers are easy to calculate for version 2.0.  I found a "water diversion tube" on Amazon for $34 (free shipping).  These are intended to be filled with water and placed across on opening to a building in order to stop water intrusion.  In my case, the tube was 12 feet long (probably to protect a garage door opening.)  

At 12 feet long, the new bag is longer than what fits across my flying bridge.  Both ends of the tube have to be pointed up in the air, which makes them empty.  But that is not a bad thing.  First, all the air is bled out of the bag.  When my boat tips, the water rushes to one side and sort of fills up what was the empty space at the end.  Thus, unlike my first version, more water can "slosh" up into the empty tube.  This changes the timing a little and provides a little more bang for the buck than my original flat bag.  In the video of ARB Ver. 1.0, the end of the bag is constricting and doesn't allow the slosh to run up the side.  I thought I would see if that makes a difference in stability.

Here is an article by a naval architect that briefly compares the use of an anti-roll tank to other forms of vessel roll attenuation:

"Passive Anti-Roll Tanks . . .

Space requirements are very difficult for small pleasure vessels (say below 60 feet). Possible undesirable effects on stability, depending on the vessel (large free surface effect). Very unlikely as a retro-fit. Possibly noisy. Relatively complex to design correctly (therefore relatively expensive to design). Relatively inexpensive to build. Relatively simple in use."

Here are the claims broken down numerically:

Con

1.  Space requirements are very difficult for small pleasure vessels (say below 60 feet). 

2.  Possible undesirable effects on stability.

3.  Very unlikely as a retro-fit. 

4.  Possibly noisy. 

5.  Relatively complex to design correctly.

6.  Relatively expensive to design. 

Pro

1.  Relatively inexpensive to build.

2.  Relatively simple in use.

I'll use this NA's Pro and Con assessment of an anti-roll tank in analyzing my ARB.  I've already passed the first hurdle (Con #1, i.e., "space requirements are very difficult for small pleasure vessels say below 60 feet)."  I'm 30 feet LOA and space isn't an issue.  The "space" issue is simply one of personal preference.  Which would you rather have on your flying bridge, a freezer, two kayaks, or a roll attenuation system.*  As we will see, a roll attenuation system can be by far the cheapest to install, maintain, and operate.

Con #3 ("very unlikely as a retro-fit") also hasn't been a problem.  I guess it depends on if one is willing to sacrifice some space for comfort, so this is simply a version of Con #1.  I don't know of a roll attenuation system that does not require any space, so this isn't really a Con for an anti-roll tank, it is a Con for every system.  In fact, I can't think of any roll attenuation systems that require less space than an ARB.  I'm not sure why this is included as a Con only for this particular system.

Con #2 ("possible undesirable effects on stability") is also true for all systems.  Take, for instance, the new concept of hydraulic operated fins on the outside of the hull.  I say "new" because anti-roll tanks of various types have been around for over 150 years.  And, yes, there have been a few incidents of undesirable stability (can't find them on Google, but can imagine it happening).  But when hydraulic operated fins have been around for 150 years, there will be the same or maybe more incidents.  Why?  Hydraulic gizmos on boats (just like electric gizmos) fail given time.  If the fins freeze in the wrong position, or get 180 degrees out of sync, there will be an incident.  So Con #2 also applies to all systems.

Con #4 (the "possibly noisy" claim) is likely based on someone's imaginary perception of what a partially filled tank of water would sound like (in a thin steel tank?)  First, when a powerboat is in operation the engine noise generally covers up the sound of liquid moving in a remote tank, at least in boats under 60 feet.  To the operator, the "noise" would be like that generated by the bow wake or water tanks.  Have you ever heard anyone say that their boat is really noisy because of the splashing water in their water tank?  Based on the experience of millions of boaters, I would change "possibly noisy" to "implausibly noisy."  Certainly not a noisy as running a generator to operate a gyro stabilizer system.  (The NA ignores the noise issue for the gyro system).  And the majority of those dragging paravanes also note that there is a humming from the cable that it transferred throughout the boat.  The "cure" seems to be having a section of chain connected to the paravane, thereby reducing harmonics by increasing drag and fuel usage.  (The NA also ignores the noise and snag issues for the paravane system).  

I should speak a little bit about "sloshing" because an anti-roll bag differs from a tank. The "slosh effect" is a real thing and can be a problem, though not as big of a problem as is often stated.  Here is some info if you want to nerd out about the slosh effect.  It should be noted that "slosh" is both a noun and a verb.  We can have the verb (i.e., the slosh effect) without the noun (i.e., the slosh sound).  How?  The slosh sound is generated by the interaction of the liquid with the atmosphere.  If there is no interaction, there is no sound.  Think about trying to hold on to a big water balloon as it squirms around in your hand.  The water moves back and forth inside, but there is no slosh sound because the water balloon contains no air.   That is one advantage of an antiroll bag over an antiroll tank.  Of course, if one can soundproof an engine compartment one can soundproof an antiroll tank or bag.

On the Pro side in the above article, the NA notes that an antiroll tank is "relatively inexpensive to build and relatively simple to use."  As already noted, my ARB ver. 2.0 cost $34 and it is definitely "simple to use" (it required filling up with a hose.)  Filling it probably took 5 minutes and cost nothing at my marina.  There is no hydraulic system to activate, gyro to engage, generator to fire up, or paravanes to drop overboard.  Not even a button to push.  And the article calls that relatively simple?  What could be simpler?  Certainly no other roll attenuation system.

Which brings me to another plus that isn't included in the NA's critique.  Maintenance.  My $34 bag had a little leak even when new (grrrr).  It needed to be topped off every couple of weeks.  I have tried to fix the leak, but I expect that the bag will wear out over time from motion in rough seas.  I'll guesstimate that it will need to be replaced every other year.  So the annual maintenance cost for this ARB stabilization system is equal to one-half the installation cost.  $17 per year and hopefully the next one won't leak.  Check out the maintenance schedule, cost, and down time for a gyro system.  Hint: it's more than 5 minutes and $17/year.

And the NA forgot Pro #4, the ARB works at anchor (without running a generator).  Is there another roll attenuation system that works as effectively at anchor as underway?  The NA forgot to include that one, likely because none exists.

But according to the NA we still have two remaining minuses for the use of antiroll tanks.  The article claims:

5.  Relatively complex to design correctly.

6.  Relatively expensive to design.

These two Cons are really just one claim.  In fact, all of the negative claims against the use of antiroll tanks for roll attenuation are just Con #5 restated.  Yes, an antiroll tank system can be "relatively complex to design correctly."  That's the only downside and term "correct design" subsumes all of the other Cons. 

"Correct design" requires the intelligent use of some space (Con #1).  If not designed correctly, it could have "undesirable effects on stability" (Con #2). Skilled designing could be required for a retrofit (Con #3).  An improperly designed system "could be noisy" (Con #4).  Designs created by an expensive naval architect could be expensive (Con #6).  Even more expensive if designed incorrectly and needing redesign (duh).  Of course, all of these difficulties also pertain to every other type of roll attenuation system available.  And all of the other systems are going to cost more than $34.  And all other types do not allow for inexpensive incremental experimentation in tuning the system as allowed with an ARB.  It turns out that "correct design" of an ARB can be a DIY project for the small recreational vessel.  A gigantic Pro only available for the ARB.

For instance, the installation process for some forms of recreational vessel roll attenuation systems begin with 1) "haul out your boat" and 2) "cut a hole in your hull."  Sounds costly and makes me more nervous than a bag of water on my flying bridge deck.  If you don't want to haul out and drill holes, then "install a $50,000 gyroscope and the required $15,000 generator."  That's more than I paid for my boat.  Or "fabricate 20 foot masts operated by electronic winches" and then "drag stuff through the water" to make it work.  Really?  Is there is someplace that doesn't have low bridges, crab pots, abandoned gill nets, submerged rocks and floating snags?  What fun is that?

I recently saw all of the popular roll attenuation systems at the Seattle Boat Show (excepting the ARB, of course).  The booths reminded me of "Psst, hey buddy, I have a bridge to sell you."  Commercial systems have to be really, really expensive.  How could a business make money selling a $34 vessel stabilization system?  It is actually the $34 cost that is the biggest Con of an anti-roll bag to a naval architect, a commercial business, and a boat yard.  Where's the profit?  An ARB business couldn't even afford a booth at a boat show.

Next, I need to experiment with what a $34 roll attenuation system can do.  Specifically, what must it need to accomplish to be worth the time and expense of "installation?"  We could look at it mathematically.  If the installation of a $52,000 hydraulic fin roll attenuation system reduces an 11 degree roll to a one decree roll (a delta of 10 degrees), what must a $34 ARB system do for an 11 degree roll?  If I've done the math correctly, a $34 stabilization system should reduce an 11 degree roll to a 10.9925 degree roll based on a simple cost/benefit analysis. Any additional roll attenuation is a win for the ARB system.  I should have some test results in Anti-Roll Bag ver. 2.1.  Anti-Roll Bag ver. 2.1.

  *  Roll attenuation systems are often called stabilization systems.  I'll try to stay away from that misnomer.  The stabilization of a vessel generally remains the same and a roll attenuation system merely makes things more comfortable by damping the vessel's roll.  Still, "stabilization system" is so commonly used that I'll likely use the term inadvertently.