Wind Tunnel Report
In addition to its numerous practical and functional features, Omni was also designed to for superlative aerodynamic performance. From the head tube back, there are virtually no trailing edges on the frame. Instead, Omni guides air smoothly back towards the rear wheel in order to avoid air detachment and reduce drag. Of course, that was just the design theory. Making that concept work in practice required iterative design including multiple trips to the San Diego Low Speed Wind Tunnel, and many rounds of Computation Fluid Dynamics (CFD) throughout the design process. How does the final product measure up? Let's take a look.
Bike + Rider drag
Omni in the tunnel
Control room at the San Diego LSWT
Compared to our baseline rig, Omni saves approximately 10 Watts with a pedaling rider on board. This translates to three minutes saved over the course of an Iron-distance bike leg. These savings come from approximately 8 Watts of frameset savings shown in the wind tunnel, plus a CFD-estimated 2 Watts of savings from Omni's aero-matched skewers (skewers aren't testable in the wind tunnel, since tunnel fixtures require skewers to be removed).
In analyzing the data, we compute weighted-average drag figures twice: once using low-yaw weights that simulate less wind and/or a faster rider, and once using high-yaw weights that simulate more wind and/or a slower rider. In the testing presented here, weighted average results were similar for both weight sets. So for simplicity on this page, we show only the low-yaw averages. However, both sets of averages, the weight sets, and the raw data are all available on the next page for readers to interpret themselves.
Our baseline bicycle was a 2014 Cervelo P3, equipped with Alpha X aerobars and dual Omega X brakes. In our previous testing, we found this configuration to be as good or better than even the higher-priced P5, so we consider it representative of the fastest bikes currently on the market. Of course, it would be great to have specific comparisons against every other bicycle in the industry, but time and resources make that impossible. Moreover, none of the data we have seen suggest there is anything currently on the market that is categorically superior to our baseline testing rig. Again, our working assumption is that the baseline rig represents the best bicycles currently available. As always, the reader is encouraged to draw their own conclusions.
Baseline Testing Rig
Visual position confirmation
As mentioned, two bicycles were used in comparison testing. The Cervelo P3 with TriRig Alpha X aerobars and dual Omega X brakes, and the TriRig Omni. Both bicycles were equipped with the same drivetrain, a SRAM 1x setup, FLO Carbon Clincher 60/90 wheelset, Dash Stage saddle, TriRig Beta bottle carrier, and two bottles (one in the Beta, and one in BTA position in the Alpha X aerobars). The bikes were set up in precisely the same position, with identical saddle heights and setbacks, and identical stack/reach to the front end. Because the Cervelo P3 has approximately 5mm more reach than the Omni, the pad XY was slightly different. However, the shifter terminus was placed in the exact same spot, 860mm from the saddle tip. This results in the exact same rider position, even though the arm pads hit the rider's arm approximately 5mm further rearward while riding the Omni. In addition to double- and triple-checking the bicycle setup positions manually, rider position was confirmed visually in the San Diego LSWT control room. The tunnel technicians have a camera pointed squarely at the rider, and draw an outline in marker of both the bike and rider during the first test run. On subsequent runs, the position is confirmed visually to ensure the rider's body and head are precisely the same. Note that the rider's head position during the test wasn't meant to represent one that would be ridden at all times on the road, but rather was chosen because it could be held consistently throughout the duration of testing.
For rider-on tests, the wind was set to 30mph, and the rider held a constant 30mph wheel speed. Each testing sweep sampled from -15 degrees to +15 degrees, in 5 degree increments. Upon setting the tunnel fixture at a particular yaw angle, the wind was allowed to "settle" for approximately 30 seconds, followed by a testing sample of 45 seconds. The data is continuously monitored to ensure there are no wild spikes or troughs - basically confirming that the rider is in a steady position, pedaling smoothly.
We also ran a number of tests with Matt Russell, which largely corroborated the earlier findings of the studies shown here. You can read all about Matt's tests, here. Those tests also illustrate the incredible value of Alpha One and its unique ability to fine-tune aero positions with a minimum of fuss in the tunnel. And Matt's testing also showed us how Omni's unique shape can cut down side-forces and handling torque down significantly, offering a ride that is much easier to handle in the wind!
Bike Only Testing
Simplified front end for bike-only tests
The massive fan at the San Diego LSWT
Testing of the bikes alone was conducted between the same two bicycles (TriRig Omni and Cervelo P3), although with a slightly different setup in order to reduce signal noise as much as possible and isolate testing to the framesets themselves. The bicycles had no drivetrains (no cranks, derailleurs, shifters, or chain), but both had dual Omega X brakes and the same Alpha X cockpit as with the rider-on testing. The front end was simplified, with shifters and brake levers removed. Again, the goal with this setup was to test only the differences in framesets, hence the reduced component count. Both wind and wheel speed were set to 30mph. Testing was conducted from 0 to +17.5 degrees, at 2.5 degree increments. Why the different protocol from the rider-on tests? This is because a frameset without drivetrain is very nearly symmetrical, so there is little need to look at both sides. Instead, we trade one side's data for increased granularity on the side we do test, sampling at every 2.5 degrees rather than every 5. We have used similar protocol for equipment-only tests in the past with the Alpha X testing.
Interestingly, the delta between bikes is slightly smaller here than for rider-on testing. Said differently, Omni shows even greater drag savings with a rider on board than when comparing the bare framesets. This may suggest that the design goal of moving air down towards the rear wheel (and away from the rider) works as intended.
Omni in the tunnel
Omni in the tunnel
There is been recent discussion in the industry about the aerodynamic characteristics of rim brakes. To that end, we wanted to study how the Omega X brake works on the Omni. So for our bike-only runs, we tested Omni in three configurations. First, with the brake completely removed, leaving only the fork in place. Because the Omni's fork and head tube are not designed with any odd cutouts or strange angles, this would seem to be a best-case scenario for testing a no-brake configuration. The second test setup was with an Omega X brake in its stock configuration, and its stock Front Plate cover. This configuration is basically drag-neutral or better (showing a 2-5g savings, but basically within the tunnel's margin of error). Finally, we tested Omni's ordinary configuration of removing the stock Front Plate on the Omega X and replacing it with the custom Omni cover that snaps into the brake and the Alpha X aerobar, forming the final intended shape of the Omni's front end. This configuration saved approximately 20g of drag over the baseline.
In summary, Omni's front brake configuration actually saves drag compared to having no brakes at all. Moreover, Omni did not make any design concessions to accommodate its rim brake. The fork legs were designed with a wide stance width to allow more space between the inner surface of the legs and the spinning wheel, in order to reduce drag. The Omega X brake fit perfectly onto the resulting shape without the need to alter the design to accommodate the rim brake hardware.
Click to the next page to see all wind tunnel data and images from the tunnel: