The Rope Effect Explained

The bit nobody quite explains

Open almost any timber design textbook and you will find the same sentence somewhere near the bolted-connection chapter: "the rope effect adds up to 25% to the lateral capacity." It is in the lectures, it is in the software, it is in the design notes - and it is almost never properly explained. Most engineers learn to take the 25% and move on.

That is a shame, because the rope effect is one of the more elegant ideas in Eurocode 5. Once you see what is actually happening to the bolt as the joint loads up, the 25% stops feeling like an arbitrary bonus and starts looking like something obvious you can almost derive on the back of a napkin. This sheet walks through the mechanics from first principles, dispels the two most common misconceptions, and shows where the 25% comes from.

It is pitched at students and architects, and at engineers who would like to understand the clause rather than just apply it. There is some algebra, but the equations are there to anchor the picture, not the other way round.

The wrong mental model: friction

The most common misconception is that the rope effect is friction. The story usually goes: when you tighten the bolt, the nut pulls the timber tight against the steel plate, and that clamping force gives you grip; the bolt then carries the lateral load partly through that friction, on top of whatever it bears with its shank.

This is wrong, and it is wrong in a way that matters.

First, EC5 does not assume bolts are pre-tensioned. The clause does not care whether the nut was made finger-tight or torqued to 100 Nm. The same 25% applies either way.

Second, even if you do tighten a bolt against timber, the clamping force does not stay. Timber creeps under compression, dries, and shrinks. Within months the nut is loose, and after a season of moisture cycling it is loose enough that you can turn it with your fingers. Anything that depends on that clamping force is gambling on something that disappears.

Third, the rope effect appears at the ultimate limit state, when the bolt is bending and the timber is yielding around it. By that point any friction that ever existed has long since been overwhelmed by gross plastic deformation. Friction is a small-deflection phenomenon. The rope effect is a large-deflection one.

If it is not friction, what is it?

The right mental model: a rope under tension

Imagine a bolt going through two pieces of timber, lying perfectly straight. Apply a sideways load. The bolt starts to bend - it has to, because the two timbers are sliding past each other and the bolt is the only thing connecting them.

Now look at the bolt's shape. It is no longer straight. It has gone into a shallow "S" or "Z", with the middle pulled sideways and the two ends pinned in their timbers by the washer and head. Geometrically, the bolt is now slightly longer than the straight distance between its end points - because curves are longer than the straight lines they replace.

But the bolt cannot actually stretch much - it is steel, and steel under modest stress hardly elongates. So if the bolt is geometrically longer than it was, and it has not stretched, what gives?

The ends must have moved inward.

That is the rope effect in one sentence. As the bolt bends sideways, it pulls its own ends - through the head and through the washer - inward toward each other. The head and washer dig into the timber face. The bolt is now in tension, like a rope being pulled tight, and that axial tension produces an inward clamping force that squeezes the timber from both sides.

That clamping force has two consequences. It pushes the timber inward, which means any lateral movement of the bolt now has to work against a normal compressive stress at the contact face - effectively, the bearing area is being pre-loaded. And by tightening the joint up, it forces a tighter contact between the steel plate (or middle timber) and the outer pieces. Both contribute to extra lateral capacity. The bolt has become its own clamp, not because someone tightened it on installation, but because it had no choice but to clamp once it started bending.

This is why dowels do not have a rope effect. A dowel has no head and no washer. When a dowel bends, there is nothing at the ends to anchor against - the dowel is free to slide through the timber as it deforms, so it cannot develop axial tension. The very feature that gives bolts the rope effect - the anchorage - is what dowels deliberately lack.

What Eurocode 5 says

EC5 captures this with a deliberately simple rule. The total lateral capacity of a bolt is the Johansen yield capacity (the bit from bending plus bearing) plus an axial contribution from the rope effect, with the rope effect capped:

F_{v,Rk} = F_{v,Rk,Johansen} + \min\left(\frac{F_{ax,Rk}}{4},\ 0.25 · F_{v,Rk,Johansen}\right)

Two things to notice. The rope-effect term is divided by four, because EC5 assumes only a quarter of the axial pull-out capacity ends up contributing usefully to lateral resistance. And it is capped at 25% of the Johansen capacity, not 25% of anything else - so if Johansen gives you 10 kN per shear plane, the rope effect can add at most 2.5 kN, even if the bolt has 50 kN of pull-out capacity available.

The pull-out capacity itself, for a bolt with a washer, comes from the washer's bearing against the timber face:

F_{ax,Rk} = f_{c,90,k} · A_{washer}

where A_washer is calculated using an effective washer diameter that EC5 caps:

d_w = \min(12 · t_1,\ 4 · d)

The cap matters because washers can be very large, but the timber can only spread the compressive stress so far before it ceases to be effective. The min(12·t1, 4·d) rule keeps the assumed bearing area realistic.

Why 25% and not more

The 25% cap is not a derivation - it is a calibration. EC5 sets it there because tests show the rope effect is real and significant, but it is also unreliable. Several things have to go right for the bolt to develop its full theoretical axial pull. The washer has to be properly seated against undamaged timber. The bolt has to bend without bursting out at the end (which is why it is rope effect, not splitting effect). The timber has to be intact around the head. The geometry has to be right for the bolt to develop a clear "S" shape rather than rotating bodily.

In a real connection some or all of these will be compromised in some bolts. Rather than try to model that variability, EC5 caps the contribution. The 25% is a reliability-adjusted estimate of what you can count on across a population of real connections. The mean test result is higher; the design value is deliberately lower.

A second cap is hidden in the F_ax,Rk/4 term. Even if the bolt could in principle develop its full pull-out, EC5 only credits a quarter of it to lateral capacity, because lateral and axial loading are not perfectly aligned, and the geometry of "S"-shaped bending only converts so much axial tension into useful lateral confinement.

Both caps - the F_ax,Rk/4 and the 25% absolute ceiling - are doing the same job: making sure the design value is something the engineer can rely on, not the best-case value seen in a single test.

When 25% is worth chasing, and when it is not

For a connection sitting well below its capacity - say, 60% utilisation - the rope effect is irrelevant. You have plenty of margin already, and the extra 25% just makes a comfortable answer slightly more comfortable.

For a connection at 90% utilisation, where the design is right on the edge, the 25% can be the difference between a working design and a redesign with bigger bolts. In those cases it is worth claiming - but it is also worth checking that the assumptions hold: washers properly sized, the bolt geometry sensible, the timber thick enough under the washer that the dw cap is not biting too hard.

For exposed connections where the washer is removed for appearance, or for connections using dowels, you do not get any of this. Plan for that at the design stage rather than discovering it when you flip to dowels late and lose the 25% you were counting on.

The rope effect is not a free lunch, but it is one of the better-justified design bonuses in EC5. The mechanics behind it are clean once you see them. The 25% is conservative. And the assumptions it depends on are exactly the assumptions you should be checking anyway - that your fastener has a head, a washer, and timber thick enough to do its job.