The creek is maybe fifteen feet wide and it was enough to cut the park in half.
Now something pale and tapered lies across it, thick at one end and slimmer at the other.
Timber decking runs along it. Not on top of it, which is what you would expect.
The boards run down both sides, at the same level, one walkway split either side of the shell.
That arrangement looks like a design choice.
It is closer to a requirement.
Why a blade wants to twist
A turbine blade is not a beam. It is an aerofoil, and every property that makes it good at one job makes it awkward at the other.
Its cross section is asymmetric. The whole length is twisted, so the angle at the root is not the angle further out. And it tapers, so the depth changes as you walk along it.
Put a load on a section like that anywhere other than the line it wants to be loaded on, and it does not simply sag. It rotates.
That rotation is torsion, and a hollow shell resists it far less willingly than it resists bending.
Hang the deck on one side only and every footstep applies a twisting moment along the span.
Split the deck to both sides and the two loads balance about the middle, so what reaches the wind blade is bending, which is the thing a blade is genuinely good at.
What is actually lying over the creek
The park is Beaverbrook Park on the west side of Atlanta, and the water crossing it had left one corner unreachable on foot.
The span is a 49 foot cut taken from a much longer blade, a 161 foot blade built by a Danish manufacturer and retired from a Colorado wind farm.
The section weighs about 7,000 pounds and crosses roughly 50 feet of ground.
Which end you cut is not a detail. A blade is deep at the root and thin at the tip, and a simply supported span wants its depth at midspan.
Take your section from the wrong part of the blade and the stiffest material sits where the bending moment is smallest.
It went in during June of last year and opened about 2 acres of park that the neighborhood could previously only look at.
The evidence this rests on
The design came out of a research network spanning six universities in the United States and Ireland, working with a small Irish company, funded by the American and Irish science agencies and a Northern Irish department.
A research engineer at the university institute drove the Atlanta project, and an architecture professor there ran the design work with students.
The harder evidence is older and it is Irish. Two of these bridges went in during 2022, one in Cork and one in Draperstown.
In 2023 a team loaded the Draperstown span until they had stacked 36 tons on it, instrumented with strain gauges, drone video and digital image correlation.
Midspan deflection at full load was about a third of an inch.
Nothing failed, which is a serious result for a beam nobody designed to be a beam.
Why the fifth one is harder than the fourth
Every blade is a single item with no paperwork. Different maker, different length, different internal layout, and no way to see inside it without instruments.
There is no design code for a structural member with no mill certificate and a fatigue history nobody recorded.
A load test proves one bridge. It does not become a standard, and the next authority starts from scratch.
There is also no market. You cannot order a retired blade, so every project begins with somebody agreeing to give one away.
The alternatives are not idle either. Blades get cut in three and buried whole, or shredded into kiln fuel.
And a plant in Tennessee now bakes them down to recover fiber at high purity, which is a real answer at industrial volume.
What one footbridge is worth
Set against a waste stream measured in millions of tons, two acres of park is nothing.
What it changes is the category. A decommissioned blade stops being a disposal cost and becomes a component looking for a job.
Grinding one destroys a precision engineered structure to make low value filler. Baking one recovers the glass and destroys the shape along with it.
Keeping it whole keeps the engineering that was already paid for, which the first American example was built to demonstrate.
The number that matters is not the span. It is whether a fifth authority signs off without commissioning its own test.
Until one does, every bridge like this is a demonstration rather than a method.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
