Why do anchors break free ?
If you, like me, enjoy anchoring in
those small, remote coves, you are sure to have dragged your anchor at some
time : no matter what anchor you use, poor holding or strong winds will have
made sure of that.
I've certainly had my share of this
kind of drama, and I have many memories of long nights of darkness and fear,
checking my anchors and my position. And praying! .... Will I drag? Or am I
safe enough to go to bed?
But such problems, rather than putting
me off sailing, made me think seriously about anchors. I began my study of
how anchors work - and of why they sometimes fail to work. Are all anchors
equal, dragging with the same load on the same bottom? Will they drag with a
constant resistance ? And if they do let go, will they dig in again? In a
word, how do anchors behave?
I also studied my fellow sailors, each
with his own relationship to his chosen, favourite anchor. How well did they
understand their anchors? Were they really wise old "sea dogs", or were they
simply acting with blind faith, blind also to the faults of their partner in
anchoring? I soon learned that surprisingly few sailors ever really
understand what is going on, on the sea floor under their boats.
I investigated the logic of anchor
design, and began the extensive series of tests and experiments that led me
to create the entirely new SPADE anchor. It is, I am totally confident, the
best anchor that has ever existed.
This is the anchor that we can
now offer you !
POIRAUD - SPADE chairman
The interplay of boat, sea and weather is
already a complex dance. Attach that boat to the sea floor, itself an
enormously variable material which is out of sight below your boat, and you
have an interaction of variables that is very difficult to predict.
For a scientific study to be useful, its
testing conditions must be reproducible. In order to obtain constant testing
conditions, and to allow sensible comparisons between tests, we have
deliberately avoided testing in real situations, from a boat.
We had several reasons for doing this :
had selected the smallest available sizes of anchors to test, the pulling
effort required to drag some of them would have called for a boat's engine
of more than 150 HP We didn't have such a boat nor the adapted equipment,
heavy rope and shackles, strain gauges...
we wished to
be able to observe the anchor's behaviour during the tests, and preferably
in very shallow water, in which a boat of sufficient power could not
operate. To reduce the number of variables we needed to pull the anchors
at constant speed, which a boat under full power cannot do.
would always try to launch the anchor in the same place, the natural
variation of any sea bottom gave unacceptably large variations, even for
repetitions of tests with the same anchor. We have assumed that the large
flat surface of sand uniformly subject to the action of waves ensure that
sea floor is of fairly constant form along a long beach.
In Press Reports
you can find the results of numerous tests carried out in real situations.
The fact that many of these tests contradict each other totally can be put
down to this difficulty alone - without even considering the pressures put
on the testers by sponsors, or advertisers.
To better understand anchor behaviour, we
sought to reduce the variables, and yet remain in a real marine situation.
We decided to carry the anchors out by hand, into water about 50 cm deep (knee
depth) off a beach. We chose a nearly horizontal bottom of good uniform sand,
with a shore area of sufficient space to allow us to use an agricultural
tractor to apply the load. Our beach was long, and uniform, long enough so
that each test could be carried out on undisturbed bottom.
The mooring rode was made of 8 metres of
8 mm chain attached to the anchor; and a long rope that attached this to the
tractor through a strain gauge, with input to a laptop computer. We could
thus measure, graph, and most importantly, record the load profiles of each
An enormous advantage of these shallow
water tests was that an observer could walk beside the anchor as it moved,
observing through the water exactly how its behaviour related to graphs
produced by the strain gauge. While this would be possible in deeper water,
using scuba gear, it would be much less convenient.
Our test set-up also allowed us to
closely observe anchors under conditions which are the most critical for a
sailor, and yet which the sailor can never observe - under extreme load.
Such conditions happen only rarely, perhaps once a year, and when they do
arrive the sailor is too busy trying to save his boat to observe how the
anchor is behaving. Likewise with tests from a boat, these conditions are
hard to produce, or to observe. Our tests required two observers, one
walking in the water beside the anchor, the other walking beside the tractor,
with direct comparisons being made between actual behaviour and measured
Not surprisingly this study allowed us to
confirm the well-known behaviour of some models of anchors, but we also
observed some rather surprising behaviour. We found that the results of our
tests allowed us to classify the various anchors into two main categories :
Those anchors that, once set, remained
buried under increasing load, but only up to a certain load. They then broke
free from the bottom, and dragged across the bottom surface. Some were able
to re-set, repeating a cyclical, or cork-screwing pattern of digging in and
holding, then releasing, and so on, as they moved across the bottom. Others
were unable to re-set at all, and dragged along the surface with minimal
Those anchors that, once set, remained
buried under increasing load. Once loads became too great for the resistance
of the sea floor, these anchors moved through the bottom but did not break
out of it.
It is not hard to decide which is the
best type of anchor to have! Even if your anchor does move under extreme
load, if it remains buried, it will only move during the moments of most
extreme load. Your boat will only move when the worst gusts hit, and then
only a short distance - a few metres, perhaps. You will have plenty of time
to consider moving to another anchorage, or setting further ground tackle.
But if your anchor lets go completely, the boat comes side-on to the wind (creating
much greater windage), and may accelerate to a speed where the anchor just
cannot re-set. At best your boat may move tens of metres with each gust; and
at worst ...... how far away were those rocks?
Let's look a little further at these two
families of anchors, starting with the ones you don't want (but which, sadly,
you may already own !).
That an anchor is unstable is not an
accident : it's a result of its original design. At the time it was first
designed, the detail of anchor behaviour was not well enough understood. Or
perhaps the designer was forced to compromise between what would work in
reality, and what would be convenient for the buyer : he was compromised by
In our studies, the unstable anchors fell
within two categories. They were either of the type we can call "flat
anchors" (we don't wish to use particular brand names !), which are often
called fluke anchors, or lightweight anchors ; or they were of the most
common "Plough" type, where there is a hinge between the shank and the body.
Let's look a little further at these :
Instalility of flat anchors
These anchors all (and there are many
different versions) have two flukes attached to a common axis which acts as
a hinge through the shank, with a fluke each side of the shank. The angle
that these flukes can turn, relative to the shank, is normally fixed at
about 32 degrees, but some versions allow the user to change this angle,
depending on the nature of the bottom expected.
There are two categories of these flat
anchors. The first generation, common with French sailors, are as described
above (i.e. they lack a stabilising strut). The second generation of flat
anchors, more common with American sailors, have the hinge axis extended
beyond the flukes on each side of the shank, this extension being a
stabilising strut designed to prevent the incorrect behaviour described
During our tests we observed both of
these types of flat anchors, and describe below how they behaved.
(a) - first generation flat anchors (without
Contrary to what one might expect, by
their symmetrical design these anchors are fundamentally unstable. Numerous
studies have already shown this, notably that conducted by Beg Rohu, of the
French Ecole Nationale de Voile (National Sailing School) in 1980.
Through these studies, and our own, we
see that when a flat anchor sits on the bottom, only rarely will the two
flukes dig in at the same moment. Depending on the density of the bottom, or
on the angle between the load (the chain) and the shank, one of the flukes
will almost always dig in before the other. Under load, this first fluke
will be pulled down into the sea floor, causing the anchor to begin rotating
around its shank, and thus causing the other fluke to move away from the
bottom. Under sufficient load this process will continue until the first
fluke breaks out of the bottom, on the other side from where it first
penetrated. The anchor again lies flat on the surface. It will then drag
some distance before the process begins again. Perhaps ! it may well become
choked with weed or rubbish before it can dig in again.
These flat anchors, thus, "corkscrew"
across the sea floor, alternately holding, then releasing, then holding
In our tests, we obtained many graphs of
"load versus distance" that demonstrated this perfectly (please see Figure
1.) The anchor first slides a small distance, offering minimal resistance,
before one fluke digs in. The load curve then rises to a maximum resistance
of nearly 200 daN (deca-Newton, the metric unit of force), before dropping
rapidly to nearly zero again. The process is repeated continuously, as the
anchor is pulled across the bottom.
Figure 1: Holding curve of flat anchors
This is a very dangerous behaviour for an
anchor. When you first anchor, and test the set with your engine in reverse,
you get the illusion of good holding. Many sailors convince themselves (wishful
thinking!) that their anchor is well set, and that it will settle in further
over time (pure faith!). The sailor then relaxes into the pleasures he feels
he deserves after a passage : time for a drink, a meal, a sleep .....
But remember that your motor, pulling in
reverse, pulls with a traction equivalent only to about 20 knots of wind !
Testing with this kind of force gives a false sense of security : just when
you've taken your corkscrew to your second bottle, your anchor may get the
same idea. The winds may increase, and at 25 knots of wind, your anchor may
begin to cork-screw ! Your dream can rapidly turn to a nightmare.....
Anchor manufacturers are fully aware of
this problem, but - like you - they are forced to make compromises : they
try to survive, and choose to produce a product they think will sell. Armand
Colin, designer of one of the French types of flat anchors, tried to prevent
this corkscrewing behaviour by incorporating, on both sides, folding
stabilising struts. The manufacturer chose not to follow the design ! A
second French manufacturer of flat anchors has added, on his aluminium
models, a strut that aims to reduce this instability. Unfortunately, in a
compromise between efficiency and size, the strut has been made so short as
to be totally ineffective.
The French use another type of flat
anchor, without a strut, and with a large heel between the two flukes. They
flukes open to a larger angle than usual, about 45 degrees to the shank. In
our tests, these anchors never dug in properly. The best they could do was
to rise up on the tips of the flukes, and scratch the bottom, skidding
across the sea floor with very low resistance. Such an anchor is to be
avoided at all costs - or it will cost you all !
(b) - second generation flat anchors (with
Tests of American fluke (flat) anchors,
in steel or aluminium, have shown a slightly different behaviour. Their
designers have incorporated an extension to the hinge axis, creating a long
strut each side of the two flukes, designed to prevent the corkscrew
behaviour. Sadly, the strut not only fails to prevent the fault, but it also
forms a trap on which the chain can readily entangle during major wind or
Even worse, when these anchors drag on
their sides they tend to stand up on three points: the end of the shank, the
point of the lower fluke, and the end of the strut. In this stable position,
they will slide under load, with almost no holding resistance, and - in
every case where we observed this behaviour - they will never dig in again.
(Please see Figure 2.)
Figure 2 : Holding power of flat anchors
with a stabilising strut
Note how the curve shows the anchor
sliding on the bottom for some distance before digging in. The holding power
then increases rapidly until, at a force of about 800 daN the anchor
releases completely. The holding power drops suddenly to almost zero, and
stays there as the anchor drags freely across the bottom surface. It does
not dig in again, not even with the cork-screwing behaviour of the previous
This pattern of behaviour is, in my view,
even more dangerous than that shown in Figure 1. I leave it for you to
imagine how it would be, on your boat, some dark night when the wind picks
up to 35 knots .... and there's a rocky shore 100 m downwind ...
Instability of plow anchors with hinged shanks
One might expect that the plow family of
anchors would be found in the stable group, but to our surprise we found a
pattern of behaviour quite similar to that of the flat anchors. We studied
these, too, in detail. Figure 3 shows a typical holding power graph for a
traditional plow type anchor, i.e. one with a hinge in the shank.
Figure 3 : Holding power of a plow anchor
with hinged shank
The curve shows how the anchor digs in
readily, and presents an increasing holding resistance, up to about 600 daN.
This holding might be adequate for most conditions, and you will come to
feel secure with such an anchor. But as the load increases further - a
squall comes through, for example - suddenly the resistance drops to about
one third of the peak - 200 daN; and then the cycle is repeated. The anchor
has released momentarily, and the boat will have moved - but at least it
hasn't let go completely!
Unless, that is, the squall endures, or
worsens : in that case the anchor may break out for a longer period, and the
boat will rapidly accelerate, moving sideways and gaining momentum beyond
the power of the anchor to control. Remember, in our tests the velocity of
the tractor was constant : in reality, this is far from the case !
But how to explain this behaviour ?
Please now consider Table 1, which shows how weight is expressed onto the
bottom by a plough anchor at its various points of contact.
Table 1 : How weight is distributed on a typical
At end of shank
A surprising proportion - 47% - of the
anchor's weight is on the hinge. Once the anchor has dug in, and a large
load is applied, the hinge portion tends to go down into the bottom, and the
anchor begins cork-screwing. Walking beside these anchors in the shallow
water was very valuable in understanding their behaviour. We often saw them
dragging through the bottom completely inverted, with the tip poking up
through the surface - like a submarine's periscope ! Perhaps this is due to
the hinge-weight relative to the point-weight, and you cannot avoid the laws
What appears to happen with these hinged
anchors is that once dug in and as the load comes on, the body of the anchor
is thrown to one side of the shank, perhaps by areas of uneven density
within the bottom. With the centre of resistance thus thrown off to one side,
the body begins to act like a rudder hard over, causing the anchor to begin
its rotation. Once in every full rotation it will break out, partially or
totally, before digging deeper again. The holding resistance thus cycles
abruptly from high to low as the body digs in, and breaks out - (see Figure
Many sailors have total trust in this
type of anchor, and will tell you they "have never been let down" by their
favourite plough.... most of us prefer to forget our own occasional lapses
from perfect behaviour ! so, too, is the nature of sailors in relation to
their anchors. If you dig deep enough, they will remember nights of terror
when the anchor wouldn't hold, or when they juggled two anchors on one
chain. But rather than choosing to learn from these experiences, the average
sailor will often excuse the problems as aberrations. It is such moments,
when loads are extreme, that our boats are most at risk, and when we need
the very best anchor. Most sailors are traditionalists, and traditions
resist change !
Remember how the famous solo navigator
Moitessier lost his equally famous yacht, JOSHUA, off the coast of Mexico ?
When the squall came through, his well tested plough broke free from the
sand bottom, and the yacht was buried in the surf. My personal experience of
this type of anchor is negative, often having found them very difficult to
set in the hard sand, or weedy, bottoms of the Mediterranean. I also found
their holding power to be unreliable
Strangely enough my bad experiences were
fortunate : they set me to searching for the best possible anchor !
If you are at all unsure as to how your
anchors will behave, try taking them onto the beach, or into shallow water
beside the shore. Try pulling with a rope : you may be very surprised by
what you find with this anchor you have trusted your safety to so often !
As before with the unstable anchors,
these stable anchors also owe their behaviour to design, not to chance. But
by behaviour, as by design, they are the opposites of the unstable anchors.
In our tests we found three types of stable anchor : the new generation
plough anchor, redesigned without the hinge in its shank ; a new German
anchor ; and our newly designed SPADE anchor.
All three of these types shared some
common design elements :
effective resisting surfaces lead to a single point, which is in the same
plane as the shank
structures are rigid, without hinges, so that the symmetry of the
effective resisting surfaces can not change through articulation
effective, resisting surface areas lie around a plane set at a fixed angle
of about 35 degrees from the line of the shank.
they all fall
onto the sea floor on their sides, and commence digging in from that side
once the load comes on
The resisting surface areas of all these
anchors pull them into the sea floor, starting from the side on which they
fall. The forces that pull them into the floor are created like those on a
kite, or a wing, by the resistance of the medium - the bottom - they are
moving through. As they bury into the floor, they settle to a position
within the floor in which the plane through the shank and the point is at
right angles to the floor. There is no asymmetry around this plane, and
therefore there exists nothing to cause them to start cork-screwing. They
remain buried, no matter what the load, even if they do start to move
through the sea floor under extreme load. Because they remain buried, they
must offer a fairly constant, and high, resistance to the loads applied. (Please
see Figure 4 : Holding Power of a typical stable anchor.)
Figure 4 : Holding power of a typical
Note how the first part of the curve is
similar to that in Figure 3, both curves showing how the anchors dig in
rapidly, and offer a holding resistance that increases quickly to about 600
daN. From this point on the difference between these two types is very
important : while the unstable anchor in Figure 3 begins its cyclical
behaviour, offering alternately high and low resistance, the stable anchor
continues to offer roughly constant, and high resistance to load, even as it
moves through the bottom.
Stable anchors do not break out of the
bottom : they may move through it under high loads, but they remain deeply
buried. They continue to offer a reliable, high and fairly constant
resistance to the loads the vessel applies onto them through the rode. The
level of that resistance depends, of course, on the size and type of anchor
used, and on the nature of the bottom.
Compare again the graphs shown in Figures
3 and 4, and ask yourself which type of anchor you'd rather have when the
wind's getting up, and the night is dark !
Comparisons of holding resistance of various
We assembled the results of all our
tests, and compared the various types of anchors tested. We choose not to
name the types of anchors tested here because we prefer not attack the
performance of other makes, but rather to show how performances vary. We
are, however, proud to name the SPADE anchor in these results, as you will
understand when you consider the following. Please see Table 2, which
compares the holding resistance of various anchors tested.
Table 2 : Holding resistance of various anchors
Weight in kg
surface area (cm²)
Max holding (daN)
Mean holding (daN)
Figure 6 - Holding comparison between unstable
and auto stable anchors
Comparisons are usually made between the
weights of anchors and the holding resistance they offer to load. We
consider it is much more valid to compare anchors on the basis of their
effective surface areas, because it is these surfaces that do the holding,
not the weight. Some anchors, by unintelligent design carry a lot of useless
weight. Others are made of different materials, such as the aluminium
anchors which have about half the weight of steel anchors of the exact same
dimensions. Comparisons based on effective surface areas are much more fair
to all anchors.
In Table 2, the most important line (other
than the first two) is the last line. The data here relates the holding
power of each anchor tested divided by its effective surface area, in units
of force/area : we have used the metric units, daN/cm²
Table 2 summarises our findings, and can
be put into words that clearly show which anchors to choose for peaceful
Unstable anchors all
let go of the
bottom under high load
fail to give
constant holding resistance under load
holding, offer rather low holding resistance / unit surface area
Stable anchors all
never let go
of the bottom under high load
constant, and high, holding resistance under high load
holding resistance/unit surface area at least double - and up to six times
that of the unstable anchors.
SPADE anchors all
resistance/unit surface area much higher than that of other stable anchors
; and four to six times that of the unstable anchors.
Our tests have shown two families of
anchors, based on behaviour under high load.
Unstable anchors give a cyclical pattern
of holding resistance, and at best a relatively low average holding
resistance. At worst, they may break out of the bottom and completely fail
to engage again. These anchors are unreliable and will always put a vessel
at risk when weather conditions apply high loads to ground tackle.
Stable anchors never break out of the sea
floor, and even if forced through the sea floor under adverse conditions of
high load, they continue to offer a high and fairly constant holding
resistance. These anchors offer greatly increased safety to vessels using
them, and are certain to become the standard type of anchor of the future.
Anchors are best compared on the basis of
units of effective surface area, rather than on weight. Even within the
family of stable anchors, average holding resistance per unit of surface
area can vary by a factor up to 2.