Real World Aero Testing via Chung Method - Data Thread

Back by popular demand, the general all-things Road forum!

Moderator: robbosmans

User avatar
kgt
Posts: 8749
Joined: Sun Jun 18, 2006 10:29 am
Location: Athens, Greece

by kgt

WMW wrote:
kgt wrote:It is expected of course that in real conditions (wind, slopes, other vehicles, uneven tarmac etc. etc.) the pure effect of aerodynamics becomes much more complicated if not immeasurable.

The only reason it is "immeasurable" is because the results are obfuscated by too many shifting variables that aren't measured and accounted for.


I agree. That's why aero tests in labs and wind tunnels can never tell you if A is faster than B in real conditions.

User avatar
Sacke
Posts: 643
Joined: Fri Sep 26, 2008 2:33 pm
Location: South of France

by Sacke

wingguy wrote:
RyanH wrote:True except we have to keep in mind there's a difference in fit between the Foil and Cervelo so it's not apples and apples. If I got my hands on an S5 to test, it would be an apples and apples comparison.


The new S5 is longer and lower than the R series/California/S3 geometry.


Only because the head tube is shorter. The geometry has stayed exactly the same apart from that. I had an S3 before, and "dared" to take the S5 because I knew I could just put roughly 1cm of spacers under the stem and get the same fit.

I would love to see an S5 in action as well. Wind tunnels are great, but this real world testing is really tickling my curiosity.

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



RyanH
Moderator
Posts: 3185
Joined: Tue Jan 17, 2012 4:01 pm
Location: Los Angeles, CA
Contact:

by RyanH

In case some missed it since it was on the previous page, I did a run with the Fuji Altamira. The estimated CdA from that run should be in the center of the bell curve of future data points since the run was so clean. The curious thing is that despite it having tube shapes that are almost twice as large as the Cervelo, it's practically matching the Cervelo CdA.

If someone wants to lend me a size 51 S5 for two weeks, I'd be more than happy to test it. I'm sure we could crowdsource funds to cover shipping both ways.

Sent from my SM-G930V using Tapatalk

wingguy
Posts: 4318
Joined: Thu Mar 08, 2012 11:43 pm

by wingguy

Sacke wrote:
wingguy wrote:
RyanH wrote:True except we have to keep in mind there's a difference in fit between the Foil and Cervelo so it's not apples and apples. If I got my hands on an S5 to test, it would be an apples and apples comparison.


The new S5 is longer and lower than the R series/California/S3 geometry.


Only because the head tube is shorter. The geometry has stayed exactly the same apart from that.


Not true. The front centre is also lengthened, especially in the smaller sizes. The S5 is not just lower, it is truly a little longer.

User avatar
WMW
in the industry
Posts: 893
Joined: Sun Jul 29, 2012 2:59 pm
Location: Ruidoso, NM

by WMW

kgt wrote:
WMW wrote:
kgt wrote:It is expected of course that in real conditions (wind, slopes, other vehicles, uneven tarmac etc. etc.) the pure effect of aerodynamics becomes much more complicated if not immeasurable.

The only reason it is "immeasurable" is because the results are obfuscated by too many shifting variables that aren't measured and accounted for.


I agree. That's why aero tests in labs and wind tunnels can never tell you if A is faster than B in real conditions.


Actually you don't agree at all. The best test for aerodynamics is the one where all other variables are eliminated or accurately counted. "Real conditions" is a random number generator. Just because there are many shifting variables in the real world doesn't mean that the aero effect you measured in the lab isn't the same in the real world.
formerly rruff...

User avatar
kgt
Posts: 8749
Joined: Sun Jun 18, 2006 10:29 am
Location: Athens, Greece

by kgt

Read my post again. I am talking about testing what is faster, not testing what is more aero. My objection is against those who claim that a more aero frame is always a faster one. What I am saying is that in the real world that is simply not true. The reason is that in real conditions there are "many shifting variables", as you say, that one eliminates in a lab or wind tunnel.

Anyway, I will wait for RyanH's conclusions. He already found that "despite it [Fuji Altamira] having tube shapes that are almost twice as large as the Cervelo, it's practically matching the Cervelo CdA." Interesting, don't you think?

User avatar
Sacke
Posts: 643
Joined: Fri Sep 26, 2008 2:33 pm
Location: South of France

by Sacke

wingguy wrote:
Sacke wrote:
wingguy wrote:
RyanH wrote:True except we have to keep in mind there's a difference in fit between the Foil and Cervelo so it's not apples and apples. If I got my hands on an S5 to test, it would be an apples and apples comparison.


The new S5 is longer and lower than the R series/California/S3 geometry.


Only because the head tube is shorter. The geometry has stayed exactly the same apart from that.


Not true. The front centre is also lengthened, especially in the smaller sizes. The S5 is not just lower, it is truly a little longer.


Sorry for the off-topic, and yes, you are right about the front center difference. It is likely due to the different forks.

The rear-center, BB-drop, seat post angle, fork angle, all remain identical. The front-center difference of 2-3mm will not have any difference on rider position on the bike. End of off-topic. If you want to debate further, shoot me a PM.

RyanH
Moderator
Posts: 3185
Joined: Tue Jan 17, 2012 4:01 pm
Location: Los Angeles, CA
Contact:

by RyanH

Real world is the real world and that's why we have statistics. Eventually, we can calculate the confidence interval, p value and such if necessary given a sufficient number of no wind day runs.

As kgt mentioned, we already have wind tunnel data for a lot of this equipment, but one of the interesting take aways from this test is that the most aero may not be the fastest. I dislike the shape of the Foil's bars and its stiffness, which means I probably ride on it slightly differently than the Cervelo where I am one with that bike. Just a small change in shifter placement and removing 4mm of spacers resulted in a significant drop in CdA, which leaves me curious what the CdA would be with Ergonovas and Campy SR.

I may continue pursuing testing of the Foil even though it's currently sitting as a bare frame right now and it's on the classifieds. I'm in SF/Napa for the week and will be racing Winters road race Saturday. I brought the Fuji for that, which was a bit of a surprise for even myself. I'm mulling over the fate of my bike collection this week and trying to vette the Fuji's full character. There's a lot of things I like about it (it's an amazing sprinting platform) and if it's not coming at an speed penalty to the Cervelo then it's only disadvantages are slightly less rear end comfort and being a "budget" bike in name.

So, bringing this back around, if a bike rattles the crap out of you and you're riding it "stiff" then even though the frame could have a 20w aero benefit, you'd maybe end up with the same overall CdA as the non aero but more comfortable bike.

wingguy
Posts: 4318
Joined: Thu Mar 08, 2012 11:43 pm

by wingguy

Sacke wrote:Sorry for the off-topic, and yes, you are right about the front center difference. It is likely due to the different forks.

It isn't. The fork offsets are identical.

The rear-center, BB-drop, seat post angle, fork angle, all remain identical.

So? Rear centre, BB drop and (within reason) seatpost angle have no effect on rider positioning at all.

The front-center difference of 2-3mm will not have any difference on rider position on the bike.

Except for moving the handlebars 2-3 mm further forwards or further back at a given stack. Which is a difference. End of story.

RyanH
Moderator
Posts: 3185
Joined: Tue Jan 17, 2012 4:01 pm
Location: Los Angeles, CA
Contact:

by RyanH

No update for today, too windy.

Can someone elaborate on the half-pipe testing protocol? I'm having trouble wrapping my head around how to analyze the runs. I need to test a handful of bar setups so it'd be useful to do it this way. What do I do? Choose a spot on a hill, coast down and come to a stop (do I pedal to start or just let go of the brakes)? How do I return back up without affecting the data? Is this method even possible if the interval markers are not lining up properly in GC?

User avatar
WMW
in the industry
Posts: 893
Joined: Sun Jul 29, 2012 2:59 pm
Location: Ruidoso, NM

by WMW

kgt wrote:Read my post again. I am talking about testing what is faster, not testing what is more aero. My objection is against those who claim that a more aero frame is always a faster one. What I am saying is that in the real world that is simply not true. The reason is that in real conditions there are "many shifting variables", as you say, that one eliminates in a lab or wind tunnel.

Anyway, I will wait for RyanH's conclusions. He already found that "despite it [Fuji Altamira] having tube shapes that are almost twice as large as the Cervelo, it's practically matching the Cervelo CdA." Interesting, don't you think?


The shifting variables just mess up the data. They certainly don't invalidate what was measured in the wind tunnel.

He isn't measuring CdA, he is using it as a proxy for drag. Wind, rolling resistance, drivetrain resistance, etc will vary also "in the real world". A huge wild card in this is position. Your body is by far the largest source of drag. If body position isn't identical, you certainly can't tell which frame is fastest.
formerly rruff...

User avatar
zappafile123
Posts: 655
Joined: Sun Jun 01, 2014 8:24 am

by zappafile123

RyanH could you do us a favour and create a spreadsheet listing each bike, the results of each trial and a summary of your main findings? Thanks!
FactorOne|SartoSeta|StorckF.3|Enigma EliteHSS|SworksSL6|ColnagoC60/V3Rs|DeRosa Protos|BianchiXR4/2|BMCSLR01 16|Cdale EvoHM|Focus IzMax|RidleyHe SL/SLX|BH Ulight|BassoDiamante|Scapin DyseysS8|TimeRXRS/ZXRS|TCR SL|RidleyNoah|Look585|CerveloS5/S3

RyanH
Moderator
Posts: 3185
Joined: Tue Jan 17, 2012 4:01 pm
Location: Los Angeles, CA
Contact:

by RyanH

GSC10 Run 14: Fuji Altamira on 404s with 23mm Veloflex Criteriums
Recording Device: SRM PC8
Speed Sensor: GSC-10
Tire Pressure: 90F/100R
Beginning Temp: 67*F
Ending Temp: 69*F
Wind: 0-3 mph N (weird on and off wind)
Avg HR: 144 bpm
Avg Speed: 22.3 mph
Avg Power: 241 watts
Distance Covered: 15.657
Duration: 42:03
Estimated CdA: 0.3224
Quality Rating (out of 3): 1 (below average)

Ride Data (Strava)
Google Sheets Aero Data - ***All Test Runs are in this Spreadsheet***

Notes: Rapha 2015 Pro Team Jersey

Test Setup:
Bottle front, fred can with tire in the rear cage.

Chung Method Aero Analysis
Image

Comments:
Sorry, I accidentally overwrote this.
Last edited by RyanH on Fri Sep 02, 2016 7:09 pm, edited 3 times in total.

RyanH
Moderator
Posts: 3185
Joined: Tue Jan 17, 2012 4:01 pm
Location: Los Angeles, CA
Contact:

by RyanH

zappafile123 wrote:RyanH could you do us a favour and create a spreadsheet listing each bike, the results of each trial and a summary of your main findings? Thanks!


I'm not there yet but I do have this which is a link in each run as the Google Docs link:

Google Docs Spreadsheet

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



justkeepedaling
Posts: 1712
Joined: Thu Jan 03, 2008 6:14 am

by justkeepedaling

Not enough variables are tied down to make such a simple assertion that the Fuji is as aero as the R5CA.

At a minimum, you'd need to use the same brand/model of stem, seatpost, seat, bars, groupo, crank, and EXACTLY the same position. This test has good info, but without having the same constants, you can't make any meaningful conclusions between bikes.

You really need to test one variable at a time. Then you know how each one affects aero or speed.

Post Reply