He Tinkers With Anchors as a Fun ‘Science Project’

Roll Bar or No? That Is the Author’s Question

READ IN APP

The author is a cruiser, boating writer and retired chemical engineer who describes his life as “one big science project.” He lives and tinkers in Virginia’s Delmarva Peninsula and sails an F-24 trimaran.

By DREW FRYE

Winter is boring so I decided to play with anchors. I built a bunch of miniature fluke anchors and played with them in a tub filled with saturated sand from one of my favorite day anchorages. Clean, fun for testing and easy. A nice low standard deviation.¹ I learned a few things.

The idea was spawned by having a number of Mantus M1 dinghy anchors left over from testing. They are modular, coming apart quickly for storage under the seat of your jet ski. But even larger anchors by many brands have transitioned to bolt-together constructions to reduce shipping costs.

Interestingly, the 4-ounce anchor holding in test sand scaled within about 10 percent of testing I have done with full-scale anchors on the same sand, when using a formula of hold = constant x mass^0.9. That’s a scale factor of 160:1. That won’t hold for some crusty bottoms, but for uniform sand and mud, it’s pretty amazing. But 4 ounces is too small for real testing on real bottoms. Good for trialing and eliminating bad ideas, though; a few styles did not move through to the next scale.

That Formula

For any given anchor design, constructed to proportional strength (metal thickness increases with load) in a consistent, saturated soil (no layering):

Holding capacity = constant x mass^0.9, with the exponent varying between 0.85 and 0.92 depending on the soil. With fine sand 0.9 is typically about right.

A 10-pound anchor will hold about 1/2 what a 20-pound anchor will hold in the same sand. Really. For example, the 4-ounce anchors held 18 pounds where the 2.5-pound anchor held 150 pounds. (2.5/0.25)^0.9=143 pounds. Pretty darn close, and close enough to test some trends. I’m trying to design a 12-pound anchor for my F-24. I need about 500 pounds of hold to be safe in thunderstorms, so I need a 2.5-pound anchor that will hold (2.5/12)^0.9=122 pounds in soft mud. Since most only hold about 80 pounds in the local soup, I need to find some improvements. We’ll see.

Also, the wind load on a boat is load = constant x wind^2. This includes waves too. If the load is 70 pounds at 20 knots, it will be 70 x (50/20)^2 = 438 pounds at 50 knots. I’ve tested this from 5-40 knots, and this is very accurate. However, if the yawing of the boat increases (chain lifts off the bottom and no longer drags, for example), then the wind load can go up far higher, nearly double.

Sometimes people say “boats yaw more in storms” without realizing that it is the chain lifting off the bottom that changed. It’s not some change in the aerodynamics. For example, a multihull on a bridle yaws the same amount at 5 knots and 60 knots, because it isn’t the chain holding the bow steady, it’s the bridle.

This formula holds spookily accurate from 4 ounces to 1.000 pounds. It is often said that larger anchors are disproportionately better, but there are only two ways that is true:

There is a hard layer that the heavier anchor can more easily push through. Weeds and shells in a layer between mud and clay, are two examples.

The 25 pound anchor drags and the 30 anchor holds. It seems MUCH better to the owner, but it’s only 20 percent better. Like the difference between a weight you can lift and weight you can’t lift. All the difference in the world.

In fact, the biggest differences in anchor holding are the bottom and how much the boat yaws (assuming enough scope and enough chain).

Back to the experiment: I cut and welded five flukes, three shanks, two roll bars and some wings, all designed to be fully interchangeable. The thickness and weight scale accurately to my Mantus M1 reference anchor. Combined with variable crown attachments, shims to adjust fluke angles, and bottoms ranging from fine sand to super-soft trashy mud, this gave me more than 100 possible combinations.

Share

Considering it takes at least five pulls and a few veers to develop any statistics on a combination, testing everything could take 1,000 pulls or more. I’ve only explored a small corner of the possibilities, but as the weather warms, I will test more, and I’ll probably add a few more components. Oh dear, I already have:

• High, medium and low shanks
• Mid and aft crowns
• Five flukes, including split toe, concave and convex.
• Three roll bar/wing options

(I have no bias toward or away from the Mantus M1. I chose it as the reference because it is well-known. I had several. And it is modular, so that I could mix and match components with ease. In fact, most of the combinations tested contain no actual Mantus parts. I do not, in fact, use a Mantus on my boat. No particular reason, it just didn’t end up that way. I have lots of very good anchors.)

I’ve learned a few things.

The standard deviation of anchor testing is huge, typically 15-60 percent, depending on how homogeneous the bottom is. Thirty-five percent SD is about average, with a 50-70 percent range. I knew this from prior testing. If you look at other test programs, you will see how true this is, and that publishing the “max hold” is a bad joke. The low end of holding that felt like a set is more relevant, but it’s still all over the place.

• All of the flukes work about the same. I have trends and comments, but I will hold them.
• Both crown positions work about the same.
• Fluke angle is critical to the bottom type, but it is always a compromise. In fact, all of the flukes work best at about the same angle. This is potentially the biggest difference between models and the one that is always statistically significant at just 2-3 degrees of change. It is that important.
• Medium and low shanks work about the same. There would be some advantage to a higher shank in weeds and cobbles. But not in sand or mud.
• With or without roll bar does not have a significant effect on holding. This is before subtracting area for any toe weight. All of the flukes set fine without a rollbar, when equipped with the correct hank and wings to roll them upright. The basic fluke design factors that cause the fluke to roll upright and dig are the same, with or without a roll bar.

(I have not tested toe ballast. Maybe later.)

BTW, I’ve cruised with high-end roll bar and non-roll bar anchors. Once they disappeared below the water, honestly, it was hard to guess what was on the chain by behavior. They were very good, and much better than their pivoting fluke or plow predecessors, which I have also used.

There are several arguments against roll bars. They’re ugly (IMO function is beautiful). They collect trash (no, I have not seen this—the fouling was always on the toe). They don’t fit (if that is your case, good point). If I fit the high shank to any test fluke and mount wings on the heel of the fluke, they all roll over and set fine (some variability—I’m not pretending there was not—but that will take more testing) without adding toe ballast. Which performed better? I don’t have enough data yet, but I’ve seen examples go both ways, both in this testing program and with full scale anchors.

The CQR Ain’t That Bad After All, Tests Show Peter Swanson   ·   February 5, 2025 Read full story

What was the biggest problem, across all anchors? Clogging with sticky mud near the toe. The huge ball would not release and could inhibit resetting.

My question: Is the elimination of the roll bar a holy grail of anchor design, or do people favor the robust obviousness of roll bar function, even if it feels a bit like a cludge?

Chesapeake Bay

A few thoughts specific to Chesapeake soft mud. It is layered.

• Lower the anchor, stretch out the rode, pull just enough to get it aligned and the tip started (very light set), tie it off … and then wait 10 minutes. It’s known locally as soaking the anchor. This allows it to sink through the top compost/soup layer to the real bottom. Then set slowly. There is a firm layer under the soup, but there can be oyster shells at the interface, so slow is better. Too fast just plows furrows, with any anchor.
• A light set on short scope helps with pivoting fluke anchors because it prevents the shank/chain sinking and the flukes floating. It helps the flukes drop into position. That’s the one exception. Other than that, every anchor I have tested likes long scope just fine, and I’ve never heard a logical explanation otherwise.

The reason for the weird Chesapeake bottom is the detritus that comes from the leaf fall from the woods and the marshes. There is an upper layers that is a very light soup of super fine compost, that is too light to consolidate into anything. Under that is firm clay, with oyster shells and sticks in between. It can be challenging, but if you get the anchor into the clay, it is actually good most places.

Want More Science?