RAD

RESEARCH AND DESIGN

Hour Record Project

THE PROJECT

A UCI hour record attempt on a tandem. The aim is to break the standing tandem hour record, and to find out whether a tandem can also go farther in an hour than the best single rider. Pilot: Will Seitz. Status: in development.

THE BENCHMARKS

Tandem hour record: 49.625 km, set in 1997 by Herve Dechamp and Guy Rouchouze in Bordeaux.
Elite men's hour record: 56.792 km, set by Filippo Ganna in 2022.
4 km, tandem: 3:55.396 by Patrick Bos and Tristan Bangma (2024).
4 km, elite single bike: 3:59.153 by Jonathan Milan (2024).

WHY A TANDEM

Over 4 km the best tandem is already faster than the best single bike. Two riders put out much more power than one, while the second rider sits almost entirely in the first rider's wake, so drag grows far less than power does. If that gap carries over to an hour, a well built tandem should out-travel a single bike. Our job is to find out how far, and to build and test a bike that gets there.

HOW WE OPTIMISE THE POSITION

Test the front rider on a single bike. The front rider sets the position the rest of the bike is built around. About 1 km of field runs gives enough data to separate positions.
Transfer the components to the tandem. The rear rider's position is then built around the front rider's optimised position.
Test the tandem. The rear position is refined with multiple 4 km runs.
Correlate with CFD. Pressure fields and CdA from simulation are compared with field results. On earlier work the two agreed well between the single bike and the tandem.

A performance model then turns the measured CdA, power, and track conditions into an expected hour distance. Line choice is worth 1 to 2 km over an hour, and a tandem is hard to steer precisely at around 60 km/h, so it is part of the plan too.

TESTING A TANDEM

Fewer than five wind tunnels in the world can fit a tandem, and none of them spin the wheels and drivetrain. Without pedalling, tunnel data is off by about 10%.
So we test in the field. An Android data logger connects over Bluetooth to a Streamlines probe, a speed sensor, and both power meters.
Field variance is higher on a tandem, because there are two power meters and two riders who each have to hold position. We account for it with more runs and by stating the uncertainty.

THE FRAME

Our 2026 tandem was built by cutting two existing frames and bonding them back together. The method works: it raced at the Para road world championship. The full story, with photos, is on the Tandem project (2026) page. What we took from it:

Eccentric bottom bracket. A tandem needs one. We bored out the standard BB30 shell and bonded in a new oversized shell. The rear bottom bracket needs no changes.
Wall thickness. The hardest design problem is tuning wall thickness and layup for the higher loads, more than the outer shape.
Center section. The middle of the frame has to be built for the specific pair of riders, which sets the length.
Structural testing. There is no ISO standard for tandems. We load test, and an ISO-style test bench is being developed alongside our Bunch Bar project.

THE RULES

There are no ISO or UCI approval standards written for tandems. A tandem has to be commercially available to be raced.
The usual UCI limits on tube dimensions and rider position still apply, so building from parts of an already approved frame is the simplest route.
Prototype equipment cannot be authorised for a record attempt. The bike has to be on sale to the public.

back to home

rdand.net is a website of RAD Performance.

+1 [205] 902-5236
will@rdand.net