Cycliq Fly12 3D printed camera mount

Since buying a 3D printer at the start of the year, I've spent lots of time playing around with it, producing various things for my bikes and home.  A lot of my other bike projects have been put on the back burning temporarily while I'm enjoying the world of 3D printing and creating my designs in CAD.

I've already designed and printed a flow diffuser for my indoor cycling winter setup (see blog post here) and a fan mounting bracket (see here), and I have number of other bike-related things I want to create using the 3D printer in the coming months.

My latest creation, shown in the photo above, is a handlebar mounting bracket for my Cycliq Fly12 bike camera.

The Cycliq camera came supplied with a handlebar mount, also shown in the photo above, but one of them.  Since I often want to put the camera on several different bikes, I need to have several camera mounts, to avoid the need to unscrew and move the mount from bike to bike. Unfortunately, the spare handlebar mounts that Cycliq sell (see here) are quite pricey, at £22.99 each.

I'm not prepared to spend almost £100 to buy several mounts for my spare bikes.  Instead, I’ve created a replica which can be 3D printed.

It works well and only uses 16g of plastic filament to produce (equivalent to about 20 pence).  It requires a couple of M3 machine bolts and nuts, but that's it.  It works well, and I've even created a version with a slightly ovalized radius that fits onto some of my stems.

It's free to download in case anybody has a 3D printer and wants to use it:

https://makerworld.com/en/models/1333580-cycliq-fly12-camera-handlebar-mount#profileId-1372448

Custom made fan diffuser

This trumpet-shaped object is the diffuser that I've designed and 3D printed for my indoor cycling setup.  This blog post describes my reasons for doing this, how I designed it, and how the diffuser alters the airflow characteristics of the fan.

In my previous blog post I explained how my new improved indoor cycling setup now includes two Cleva Vacmaster fans.  The fans are excellent, providing powerful jets of air to help keep me cool during hard workouts.

The fans have three speed settings, with the fastest #3 setting delivering a airspeed of 31 kph at the centre of the jet, at a distance of 1 metre from the nozzle.  Directly in front the nozzle, the airspeed is obviously faster, 55 kph, but the measurements at 1 metre are more appropriate to how I use the fan.  On its lowest #1 setting, the speeds at 1 metre are approximately 60-70% of the airspeeds at the highest setting.

Why a Diffuser?

The problem I've found is that during the colder winter months, when it's around 10 degrees Celsius in my garage, the airspeed at even the lowest fan setting is a bit too fast.  This is especially true when doing easier Zone 2 endurance rides.  I find that need some airflow to avoid getting sweaty, but I only need a very light breeze.  With the more powerful Cleva fans I found myself often getting too cold, even with the fans on their lowest setting.   I would get too sweaty when the fans were off, though, so found myself cycling them on and off.

I decided that I would try to make a diffuser.  A diffuser should slow the flow down by causing it to 'spread out'.  This would, if it works, also cause the jet to become broader, having the additional benefit that the flow would cover more of my body, which should be helpful at times when I wanted the fan to be on it's full setting, to keep me cool.  To do that though, the diffuser would have to work properly, meaning no flow separation within the diffuser, in order to maintain full aerodynamic efficiency.  If flow separations occur inside the diffuser, then aerodynamic losses occur (total pressure losses), the result of which is that the air would slow down somewhat, but the jet would not widen correctly.  In that case, the same result could be achieved simply just by restricting the flow with an object that partially blocks the nozzle (e.g. a grid, or a gauze).

Diffusers are also used in a couple of other applications that people may be familiar with:

• On the rear underside of Formula 1 cars, and other race cars (see here).  Race car diffusers achieve the same result, slowing down the air flow by causing it to spread out, to expand.  The objective is slightly different though - when the flow slows down, it increases in pressure, returning to the ambient pressure as it leaves the back of the car.  This in turn allows the flow under that car to be faster, operating like a venturi, which reduces the pressure below the floor of the car to a pressure below the ambient air pressure, 'sucking' the car downwards, thereby creating aerodynamic downforce.
• Diffusers are also used in wind tunnels, behind the working section, to slow down the flow, allowing slower moving airflow in the return loop of the wind tunnel, reducing losses in the wind tunnel.

Fan Diffuser Mk.1

My first diffuser design was a straight tapered device, shown in the screen shots on the left.  I used the Autodesk Fusion CAD package to create the diffuser in two parts.  The first part is permanently fixed to the fan and replaces the standard black nozzle on the fan, having a similar shape and fittings.

The photo below shows how the two parts fit together. The fixed part has a forward facing slot around the edge of the nozzle.  The diffuser has a tapered flange which then fits into that slot.

The slot and flange connection is tight enough to stay there by itself, but for extra security I drilled a couple of small holes to secure it with 3 mm wood screws.

The next step was to see whether the diffuser worked correctly.  I taped down a number of small ~30mm lengths of wool onto the inside of the diffuser to help determine the flow quality.

Wool tufts are a common type flow visualisation technique used by aerodynamicists to determine where the flow is attached to the surface or separated.  Attached flow means the air is moving in the intended direction, flowing smoothly across the surface.  Separated flow means the flow has broken away from the surface, causing regions of  recirculation where the flow can be moving in the opposite direction, creating eddies that restrict the flow and reduce the efficiency of the diffuser.

The video below shows the that some of the wool tufts are quite stable.  However, other tufts are quite active, showing that the flow is separated, or nearly separated.  One tuft on the lower surface is being blown backwards, indicating it's in a region of re-circulating air caused by flow separation further upstream inside the diffuser.

Generally, the result was a bit disappointing.  I measured the characteristics of the airflow downstream, just to check it (see plot on the left).

This plot confirms that the airflow speed was being slowed down, but the jet wasn't getting much wider.  This is another indication that the diffuser was lacking efficiency because the flow separation in the diffuser causes losses that slow the flow down but do not cause it to spread out, meaning the mass flow rate was not being conserved because the flow through the fan was being constricted.  My feeling is that the angle of divergence I chose for the nozzle was too severe, causing the adverse pressure gradients in the diffuser to be too high, prompting the flow separation.

Fan Diffuser Mk.2

As a second attempt, I decided to try a diffuser with gradually increasing divergence, combined with a lot more internal vanes to help guide the flow.

In my mind, these multiple guide vanes would operate in a similar way to how corner vane cascades work in the corners of closed return wind tunnels to efficiently turn the flow through 90 degrees (see diagram on the right).

The CAD screenshot to the left shows the external shape of the diffuser (left), along with the fixed nozzle that permanently attached to the fan.  The screenshot on the right shows a cut-away cross-section, showing the shape of internal guide vanes.  The diffuser required about 400-500 grams of plastic filament (PETG) to make, and it took about 12 hours to print on a Bambu Lab P1S 3D printer.The number of guide vanes made it impossible to attach fixed wool tufts inside the diffuser, so instead I used a wool tuft on the end of an old wheel spoke to inspect the flow quality.

It can be seen from the video above that the flow quality looks good, with some clear changes in flow angularity across the exit of the diffuser, showing that the flow is being turned.  This was encouraging.  The next step was to measure the airspeed profile.
Airspeed profile measurements

I used the anemometer that I described and calibrated in a previous blog post (here) to measure the airspeed at 1 metre from the fan.  I measured the airspeed at 50 mm intervals up and down the centreline of the fan jet, as shown in the photo to the left.  I also checked the velocity variation to the left and right of the centreline, to ensure I was measuring the centre of the jet where the peak velocities were occurring.  The airflow speeds for the three different nozzles, with the fan on the highest speed setting, are shown in the plot below.The airspeed profile below shows that the Mk.2 diffuser is doing a good job, not only in slowing the flow down, but also spreading it out too.

For the lowest fan speed setting, the airspeed profiles are progressively lower, as shown in the plot below, but the differences are otherwise more or less the same.

Conclusion

Overall, I'm really please with the diffuser.  It makes a noticeable difference to the airflow on colder days.  On milder winter days, I turn the fan up to setting 2 or 3 to get the necessary cooling.  When the weather warms up even more, in the spring, I'll remove the diffusers altogether to maximise the airspeed.  For now though, in the winter temperatures, they're working great with the fans.

Indoor cycling set-up v3

In a couple of previous blog posts I showed my indoor training setups, the first one in 2016 here, and then my improved one built in 2020 here.  In the last few days I've improved it further.

With hindsight, my cooling fan setup wasn't as good as it could be.  The cheap pedestal fan that I've used for the last four years couldn't provide enough flow and cooling for the warmest days or the hard workouts, although it was okay most of the time.

That has changed recently though, after I bought two Cleva centrifugal fans, after seeing some good reviews from people on the TrainerRoad Forum, who discussed UK alternatives to the Lasko fan that was highly recommended by tghe guys on the TrainerRoad podcast.

Initially, I bought a single Cleva Air Vacmaster Air Mover fan, which about  £50 at the time.  This was great, but it has a manual switch, so I also bought a remoted controlled plug set from Amazon for £18.99, to allow me to turn the fan on, after I warm up, about 5 minutes into the cycle.  The fan speed was excellent, but the fast moving air stream was quite narrow, not broad enough to cool my whole body.  Therefore I decided to buy a second fan to give me more complete body coverage.

The second Cleva fan I bought was their more expensive Vacmaster Cardio54, which I bought for £79.99.  It's identical to the cheaper Air Mover, except that it has a built in remote control which is slightly better because it also allows the speed to be selected from the remote, rather than just on/off.

As shown in the photo at the top, I have mounted these two fans in two positions in front of my bike and trainer, one directed at my head/torso from above and the other directed at the lower half of my body.  After some experimenting, I found this to be the best setup.

The lower fan was easy to position, and required only a small block to get the angle just right.  Also, it could stay in that position and wouldn't get in the way when I'm not cycling.

The upper fan was a bit more tricky though.  The fan is capable of being mounted to a standard tripod threaded attachment, or alternatively mounted onto a TV wall mount that has a VESA mount.  However I didn't really want to spend more money on a TV mount, so I re-purposed the fan swing arm from my old pedestal fan, as shown in the photo on the left, using a cylindrical wooden rod as the arm.  To attach the fan, I made a 3D-printed backet, using the 3D printer that I brought late last year.

I designed the bracket in Autodesk Fusion 360, which is free CAD software. I incorporated a quick release mechanism into the bracket, re-using an old QR seat post clamp that I had spare.

This allows the angle to be tilted to the best angle to direct air on my head and upper body.  The swing arm allows the upper fan to fold out of the way, against the wall, when not in use.

I'm really pleased with the new setup.  The amount of cooling is excellent, and in the colder winter months it is occasionally even too much, even on the lowest setting.  On those colder days, I only turn the lower fan on.  The photos below show the setup with everything in place (left), and on the right it's hwo it looks when everything is folded away to make more space in my garage.

 

Indoor cycling fan upgrade and speed check

I recently bought a Cleva Air Mover centrifugal fan as an upgrade to my indoor cycling setup. 

It came highly recommended on several TrainerRoad forum posts (see here).  Compared with my old fan, it seems to be a lot better.  I wanted to measure exactly how much better it is, and this blog post describes my attempts to measure the air speed that the fan generates.  As usual, it's all a bit nerdy and over-the-top, so sorry about that.

Airspeed measurement methods

I bought a cheap anemometer (an anemometer is air speed measurement device) from AliExpress for an incredibly cheap £13.05.  It seemed good, but given that it was so cheap, I wanted to check its accuracy with a second source of airspeed measurement.

Having recently bought a 3D printer, I decide to make a pitot tube that I could use as a 2nd instrument to measure the air speed generated by the fan.  It was quick to create in CAD (<1 hour) and printed in less than an hour.

A pitot tube is a device that measures the total pressure of the flow, by having tube that is orientated in the flow direction.  When the flow enters the pitot tube is comes to rest within the tube, and when it does so the pressure will be the total pressure. Total air pressure is the ambient (static) pressure of the air plus the dynamic pressure of the moving air.

Hence, the difference between the static pressure and the total pressure is the dynamic pressure, which is a function of the air speed and the air density.  Hence, but knowing the air density and the by measuring the dynamic pressure (via the difference of total and static pressures) the air speed can be obtained.

I measured the difference between static and total pressure using a device called a manometer, which is the U-shaped device shown in the sketch on the left.  The manometer tube is filled with a liquid, shown in purple in the sketch, and could be any liquid such as water or mercury.  I used water because it was easy and its relatively low density helped with measurement precision.

I bought a manometer from Amazon for about £20.  A quick calculation in Excel allowed me to calculate the airspeed that corresponds to various manometer water height differences.  I also checked how sensitive the results would be to small changes on pressure and temperature, which affects the air density.

Results

I measured the airspeed at 0.5 metres from the fan nozzle and also at 1.0 metres from the nozzle.  These are distances that are representative of how far away the fan would be located from me when I'm cycling.

0.5 metre distance

The AliExpress anemometer gave a maximum airspeed, in the centre of the air stream, of 11.2 m/s.  The reading fluctuated by approximately +/-0.8 m/s.

My pitot tube gave a water height difference of 8 mm in the manometer, with a manometer measurement precision of about +/-1 mm.  This corresponds to an airspeed of 11.32 m/s +/- 0.8 m/s.  The nature of the pitot tube and manometer meant that any air speed fluctuations are damped, although the +/-1mm manometer measurement precision results in a +/-0.8 m/s uncertainty. The two measurement sources, the pitot and the AliExpress anemometer, therefore agree very well for this 0.5 metre case.

1.0 metre distance

At a distance of 1 metre, the AliExpress anemometer gave a maximum airspeed of 8.8 m/s, with +/-1.0 m/s fluctuations.

My pitot tube gave a water height difference of 4 mm in the manometer, +/-1 mm.  This corresponds to an airspeed of 8.0 m/s +/- 1.0 m/s.  Again, the two measurement sources therefore agree very well at this distance of 1m, and well within their measurement precision.

Speed directly at the nozzle

The Cleva website says that the maximum airspeed of their Cardio 54 fan, which is mechanically identical to their Air Mover Fan, is 54.0 kph, 45.0 and 34.2 kph for the three speed settings.  These speeds correspond to airspeeds of 15.0 m/s, 12.5 m/s and 9.5 m/s.  It's highly likely that these speed quoted by Cleva are the maximum speeds, directly in front of the nozzle.

I checked the speed using the AliExpress anemometer, and that device gave speed recordings at the fan nozzle of 15.4 m/s, 12.9 m/s and 10.7 m/s for the three settings, all with a variability of about +/-0.4 m/s.  These values are therefore consistent with Cleva's quoted airspeeds, and actually are slightly faster.

Back-to-back rolling resistance testing: MTB versus Cyclocross tyres

Last year, in October 2023, I published the results from some tyre rolling resistance tests (see blog post here) in which I tested my cyclocross bike with 33mm cyclocross tyres.

Those tests showed that my cyclocross (CX) tyres were much slower than my Schwalbe Thunder Burt mountain bike (MTB) tyres. The CX tyres had a CRR value nearly double the CRR of the MTB tyres, which was a huge difference.

However, the two tyres were tested on different days, with different bikes and different power meters, so there were a number of factors that cast doubt upon that comparison.  Since doing that test, and especially as the results were so surprising, I've wanted to repeat the test by testing both tyres on the same day, with same power meter, thereby improving the quality of the test.  This blog post describes my latest test results, and thereby allows me to conclude, with much more certainty, which of my two bikes is faster for dry cyclocross races.

Recap of previous test results

The previous results are shown in the plot to the left.

The data points shown in red are the CRR values obtained from my hardtail mountain bike tested in June 2022.  The CRR values reach impressively low values (~0.015), much lower than the two sets of cyclocross tyres plotted in blue and green.

As noted in the plot, the Challenge Grifo/Baby Limus tyre test was done a day when the ground was softer.  This will have certainly caused the CRR values to be higher, although it's not clear how much the softer ground accounted for the differences.  Note also that the softness of the ground was only determined qualitatively by me, with no measurements etc of the firmness.

New test setup

The new test is better in several ways:

• Three different sets of tyres, all tested back-to-back on the same day.
• Same power meter for all tests: Favero Assioma Pro MX-2 power meter pedals.
• The power meter pedals are a dual-sided power meter, rather than single sided power meter that I've used in the past for most of the previous testing.

The use of a dual-sided power meter ensures total power is measured, whereas single-sided power meters measure only one side (one leg) and then assume the other leg is producing the same power in order to infer total power. Dual-sided power meters are therefore superior, especially since I have discovered that my left leg/right leg balance is not 50/50 (it's more like 53% left, 47% right).  Also, and more importantly, I've also discovered that my left/right leg imbalance is not fixed, and it varies with power, so there is inconsistency and inaccuracy whenever I use my single-sided powered meter. 

Test order:

1. Drop Bar MTB with 2.35" (60 mm) Schwalbe Thunder Burt Super Race tyres @16 psi
2. Cyclocross bike with 33 mm Challenge Baby Limus HTLR tyres @20 psi
3. Drop Bar MTB with 2.35" (60 mm) Schwalbe Rocket Ron Super Race tyres @16 psi

The test was performed in my usual way (see Chung method description here), with the exception that I tested only one pressure per tyre/bike this time, using pressures that I know from previous testing give good rolling resistance results but without negative drawbacks like pinch flat risks.

Speeds were measured using a hub-based speed sensor, with the diameter in my Garmin head unit set to a fixed value of 2115 mm.  Measured speeds were then scaled in the analysis based on measured diameters to give the correct speeds:

  - Schwalbe Thunder Burt:  Speed scaling factor = 2228 / 2115
  - Challenge Baby Limus:  Speed scaling factor = 2110 / 2115
  - Schwalbe Rocket Ron:  Speed scaling factor = 2245 / 2115

Results

I performed the Chung method (virtual elevation) analysis using a fixed CdA or 0.40 m^2.  It's obviously wrong to use the same value for both bikes, because it's likely that the CX bike with a lower frontal area (narrower tyres, etc) will have a slightly lower CdA.  However, in the same way that I've done it previously (see here), by using this fixed value of CdA, any performance differences between the three setups, whether those performance differences comes from aero, rolling resistance or something else, will be attributed to CRR.  Hence, whichever setup has the lowest CRR using this style of analysis will be the fastest bike.

Results with Fixed CRR

Using a fixed CRR value of 0.0261, the virtual elevation plots show that the three setups are reasonably similar.

Note that the variations in the measured elevation (the green line), should be ignored because this is simply a drift caused by changes to ambient pressure during the test (note that a one metre elevation difference measured by the Garmin barometer is equivalent to just a 0.12 millibar change in ambient pressure).  The tests were done riding laps of a grass field, so the actual elevation always returned to the same value, in reality.

The variation in the orange virtual elevation profiles show which setup is fastest.  The CX bike with 33 mm Baby Limus tyres is fractionally slower than the drop bar MTB with Thunder Burts, because it's VE profile rises slightly.  The fastest setup is actually the third one tested, the MTB with Rocket Rons, since the VE profile falls slightly for that third test.

Results with adjusted CRR values

The CRR values can be adjusted to 'flatten' the VE profiles, as shown below.

This plot shows that the three setups have the following CRR values:

• Drop Bar MTB with 2.35" (60 mm) Schwalbe Thunder Burt tyres:  CRR = 0.0261
• Cyclocross bike with 33 mm Challenge Baby Limus HTLR tyres:  CRR = 0.0263
• Drop Bar MTB with 2.35" (60 mm) Schwalbe Rocket Ron tyres:    CRR = 0.0251

Note that difference between the Rocket Ron CRR of 0.0251 and the Baby Limus CRR of 0.0263 is equivalent to 6.7 Watts of rolling resistance for a 85 kg bike+rider travelling at 15 mph.  These differences aren't significant, but they aren't particularly large differences either.

It's also interesting to note the average speeds, although it's more difficult to conclude which setup is faster form these speed, because the average power are not exactly the same.

• Drop Bar MTB + Schwalbe Thunder Burt tyres:  Avg speed = 13.79 mph @203W avg
• Cyclocross bike + Challenge Baby Limus tyres:  Avg speed = 13.33 mph @193W avg
• Drop Bar MTB + Schwalbe Rocket Ron tyres:    Avg speed = 13.60 mph @195W avg

Incidentally, one of the benefits of the virtual elevation method is that it's not necessary, as a tester, to hold a fixed power.  That's a really inconvenient constraint that many testers put upon themselves if they don't use a virtual elevation method and try to hold a fixed power.

Finally, it's interesting to note that the final two laps for the Rocket Ron have a significantly rising profile, even though the first four laps were flat.  Those last two laps would have a measured CRR of 0.0271, which is strange.  I have no explanation for these weird results during the last couple of laps.

Conclusions

I set out to see repeat my previous tests but with a better standard of testing, to see whether my previous conclusion that MTB tyres are faster for cyclocross still holds true.

The conclusions from this test aren't quite so clear.  The MTB tyres are still slightly faster, but not significantly so this time.  The MTB tyres are only about 1-7 Watts faster.  Therefore, I'll likely use my cyclocross bike if there is any chance of mud, since my CX bike has much better mud clearance and is less likely to get jammed up from sticky mud.

Perhaps the most surprising result of this test is that the Rocket Ron tyres were actually faster than the Thunder Burt tyres, despite being how knobbly they are compared with the Burts, and considering the tyres are similar in every other respect.  This goes against what I'd expect and is contrary to Bicycle Rolling Resistance tests (although the Rocket Ron performs quite well on BRR).  This is a test result that would be worth trying to reproduce at some point.

Testing Schwalbe's Super Ground versus Super Race tyre versions

Last month I posted the results of the tyre rolling resistance testing I did, in which I compared the Schwalbe's Thunder Burt Super Race against the Continental Race King Protection.

Both are fast tyres, but the Super Race version of the Thunder Burt has never been tested by BRR (BicycleRollingResistance.com). Only the more robust Super Ground version of the Thunder Burt has been tested (see here).  Nevertheless, that Super Ground version of the Schwalbe Thunder Burt is so far the fasted MTB tyre tested by BRR, marginally faster than the Continental Race King Protection. 

In principle, the thinner carcass of the Super Race version of the Thunder Burt should be faster than the Super Ground version.  This is at least what Schwalbe claims.  However, some tests on BRR in the past gave some unexpected results, with the Super Ground version of the Racing Ralph testing a few Watts faster than the Super Race version.  This showed that it's not guaranteed that the Super Race version should be faster tyre, which is what motivated me to do this test.

Test details

For this test I was fortunate to have two new 2.35" Schwalbe tyres, that were brand new and unused until this test.  I had bought both the Super Ground and Super Race tyres in late 2023, so there was minimal uncertainty coming from the effect of any tyre ageing.

I tested the tyres using the same protocol used previously, which I explained in a previous post here.  As for my most recent test though, I used my MTB with its Power2Max spider-mounted power meter which measures total power.  As a consequence, this was one of the best executed and best controlled tests I've done.

I tested the two tyres in an A-B-A-B manner, so both tyres had a repeated test.  In addition, I also test my older 2022 Schwalbe Thunder Burt Super Race tyre first, to check that it gave similar results to last week.

Results

The results are shown again below.  The results are fairly conclusive, showing the Super Race version of the tyre (plotted with purple symbols/lines) being faster than the Super Ground version (plotted in red).

Validating my power delivery optimiser: Step 4

Step 4: Calculating the optimum power delivery profile using my optimiser.  

In the previous blog post (Step 3), I showed the BestBikeSplit recommended power profile and the results of my Zwift ride when I followed BestBikeSplit's recommended power profile.  It saved me 9 seconds compared to the constant power effort, for a similar 150W normalised power.

I will now see if my Excel-based power optimiser gives similar results when I model the Whole Lotta Lava segment using the same set of 29 intervals that BestBikeSplit used to model the segment.

Results of my power optimiser

The plot above, at the top of this blog post, shows the BestBikeSplit recommended power profile (dashed red line) and then the power profile recommended by my Excel-based optimiser (solid red line), when targeting a 150W normalised power. The agreement between the two plots is remarkably good.  In the past I've done a crude comparison between my optimiser and BestBikeSplit (see here), but I've not plotted the powers on the same graph before.  This really shows the two methods give the same results, and so it shows my Excel-based optimiser is a free alternative to BestBikeSplit.  It could actually be even better than BestBikeSplit if I'm able to model more intervals than BestBikeSplit.  I just need to write the script that can automate the creation of the intervals.

Predicted ride times from my power optimiser

For the constant power situation, with the power fixed at a constant 150 Watts, my optimiser predicts a time of 27 minutes and 17 seconds. Note that this is only 2 seconds different (0.1% different) to my actual Zwift ride time of 27 minutes and 15 seconds, which I've obtained on two occasions riding at a constant 150 Watts.

To achieve this constant power of 150 Watt simulation in my optimiser, I set-up a multiple goal-seek macro in Excel.  That was an interesting learning exercise in itself, because I have never programmed in Visual Basic before.

Running the optimiser to achieve an optimal power profile, my optimiser predicts a time of 26 minutes and 37 seconds, which is a fairly significant 40 seconds (2.4% saving).  That's very similar to the 2-3% savings that I was getting when I first created my power optimisation spreadsheet back in 2016 (see post here).  The speed and power differences between the constant power case and the optimised power case are shown in the plots above.

Summary

Riding at a constant 150 Watts, my optimiser predicts a time of 27:17.

(average power = 150W, normalised power = 150W)

That 27:17 time agrees very well with the actual ride time in Zwift of 27:15.

Riding at the optimum power, my optimiser predicts a time of 26:37 (40s faster)

(average power = 143W, normalised power = 150W)

I haven't yet tried following this power profile in Zwift yet, and although the BestBikeSplit power profile is similar, the speed modelling deficiencies in BestBikeSplit (described here) was insufficient to test this.

Next Steps

The next step is to create a script that can process the Whole Lotta Lava segment profile, and any other ride profile, in a more efficient way with less manual effort by me.

Validating my Power Delivery Optimiser: Step 3

Step 3: Using BestBikeSplit to create an optimum power delivery profile.  

In my previous blog post, I explained that the following values are needed to simulate the Zwift ride:

• Total weight (rider + bike) = 80 kg (176.4 lb)
• CdA (drag coefficient x Frontal area) = 0.2415
• Crr (coefficient of rolling resistance) = 0.004
• Air pressure = 101,250 Pa
• Air temperature = 15 degrees C
• Air density = 1.225 kg/m^3
• Drivetrain losses = 2.5%

BestBikeSplit modelling via ZwiftInsider

I first tried to use this BestBikeSplit link that is available via the ZwiftInsider Whole Lotta Lava route page to access Best Bike Split.  The page assumes a fairly light rider weighing 145lbs, and makes some other assumptions. However, the Time Analysis tab allows the user to change the value of various parameters using sliders to adjust the parameters in 1% increments to the necessary value.  Unfortunately, this method didn't allow me to enter the exact values above, so I used the closest values possible instead:

• CdA = 0.24 (i.e. 0.0015 too low)
• Average power = 144.5 Watts
• Normalised power = 149.9 Watts (0.1 Watts too low)
• Weight = 177 lb (i.e. 0.6 lbs too high)
• Crr (coefficient of rolling resistance) = 0.004 (exactly as required)

Note that I tried to get the normalised power as close as possible to 150 Watts.  It's unclear what assumptions BestBikeSplit has made for air density and drivetrain losses.  Nevertheless, with these values, BestBikeSplit predicts a route time of 27 minutes 42 seconds, which strangely is 27 seconds slower than my constant power time.  I didn't understand this, so I needed to investigate why.

BestBikeSplit dedicated race plan

Since the results above seemed to make no sense, I decided to create a dedicated BestBikeSplit race plan, which is possible even with BestBikeSplit's free account.

I was able to change the settings so that the parameters were exactly matching what I wanted:

• Rider weight = 160 lbs, Bike Weight = 16.4 lb -> Total weight = 80 kg (176.4 lb)
• CdA = 0.2415, for all yaw angles
• Crr = 0.004
• Drivetrain losses = 2.5%
• Air pressure = 101,250 Pa
• Air temperature = 15 degrees C
• Zero wind speed
• Drivetrain losses = 2.5%

These values give a route time of 27:47 for a normalised power of 149.94W and an average power of 141.93W, which is bizarrely still 32 seconds slower than my constant power Zwift time.

The final section of the BestBikeSplit Race plan above provides 'race intervals' which show the segment distances and gradients, then the power recommended by BestBikeSplit, together with the resulting speeds.  I copy and pasted those 29 race intervals into Excel and calculated segment distances in metres, instead of miles: 

I then copied the distance, gradient and speed values into my power optimiser spreadsheet to allow me to compare the power values calculated by BestBikeSplit and values calculated by my spreadsheet.  Surprisingly, the power values agreed for some of the route, but were significantly different for the downhill section, as shown in plot below for the solid red and dashed red lines:

As soon as the gradient becomes negative, BestBikeSplit (the red dashed curve) is showing that a higher power is needed to achieve the downhill speeds that BestBikeSplit has calculated.  I really don't understand this because, for example, it is calculating speeds of 23.08 mph on a -4.36% (downhill) gradient while applying 84 Watts of power.  That is clearly wrong and I didn't know why BestBikeSplit is doing that, unless it is somehow assuming a large headwind

To investigate this, I plotted the BestBikeSplit powers and speeds versus the gradient values, shown withe red symbols on the plot to the left.  The plot looks reasonably sensible, although the scatter seen in the corresponding speeds is slightly strange. Regardless of these doubts, I decided to try to ride the route using the BestBikeSplit power profile to see what time it would give.

Riding the BestBikeSplit power profile

To do this, I decided to use the BestBikeSpilt's feature where they allow the power profile to be exported as a TrainerRoad workout.  This is a nice feature, because it allowed me to ride exactly the prescribed powers by doing the workout in ERG mode, with my phone (for TrainerRoad) and also Zwift (via my iPad) connected to my Wahoo Kickr trainer.  The workout included the 29 intervals contained in BestBikeSplit's race intervals.  As shown above, I also added a long 150W interval afterwards to allow me to do a second lap at a constant power, if I wanted to.

I started the TrainerRoad workout as I got to the archway that marks the start of the Whole Lotta Lava segment.  As a reminder, the segment time I did previously with a constant 150W power output was 27 minutes and 15 seconds.  Using the BestBikeSplit power profile, I did a time of 27 minutes and 6 seconds, so 9 seconds faster than the constant 150W effort.  It's worth noting that although the normalised power for the BestBikeSplit run was about 150W (actually slightly lower, at 148W), the average power was substantially lower at 140W.  This is interesting in itself, and a positive results, that the average power was 10W lower than the constant 150W effort, but the time was 9 seconds faster.

However, something I noted when riding the segment with the BestBikeSplit power profile was that at several times the target power did not correlate well with the gradient.  At most times it was good.  For most of the uphill sections, the target power was higher than 150W, as  expected.  At some points, though, the target power didn't correlate with the gradient.  For example, the high power interval stopped about 10 or 20 seconds too soon, before I reached the top of the climb, on the steepest part, which seemed odd.  Therefore, I think that due to the BestBikeSplit speed issue mentioned previously, I think there was a synchronisation problem between the BestBikeSplit simulation and what actually happened in Zwift when those powers were ridden.  I have the feeling that the most likely explanation is that BestBikeSplit is assuming some wind that is not actually there in Zwift.

It's worth also be noting that the 27:06 time is significantly faster that the 27:47 time predicted by BestBikeSplit, indicating that their calculation of Zwift speeds are wrong, despite me entering the details such as Crr and CdA correctly. 

Finally, it's interesting and encouraging that my second lap of the Whole Lotta Lava segment was done again at 150W constant power, and again this gave exactly the same 27 minute 15 second segment time as I achieved before (described in my previous blog post).

Summary

This has been quite a long post, so let's recap:

Riding at a constant 150 Watts gives a time of 27:15.

(average power = 150W, normalised power = 150W)

Riding at the BestBikeSplit optimum power profile gives a time of 27:06 (9 second saving)

(average power = 140W, normalised power = 148W)

Next Steps

The next step is to see whether my own Excel-based power optimiser gives similar results, and to do that I will be to create a scripts that can process the segment profile in a more efficient way, with less manual effort by me.