Factor Ostro

Discuss light weight issues concerning road bikes & parts.

Moderator: robbosmans

Post Reply
ericlambi
Posts: 139
Joined: Mon Apr 18, 2005 6:16 pm

by ericlambi

wheelsONfire wrote:
Wed Sep 09, 2020 9:38 pm
Wind tunnel time seem to cost alot of money, so that's why most seem to rely on CFD only?

It would be interesting to hear an honest answer if bikes really need to be designed in a wind tunnel.
The problem is that people don't seem to care too much about optimizing the last ~5 aero watts anyway, as evidenced by the fact that Specialized is getting rid of the Venge, people are putting disc brakes on their bikes, etc. So it is a significant expense, and also a delay to the manufacturing process as you test different ideas/iterations, you can do relatively well without it using some generally accepted principals and CFD modeling. I'd say the return on wind tunnel testing for road bikes must be negative.

Skillgannon
Posts: 3631
Joined: Sat Apr 01, 2006 12:17 am
Location: A bigger rock in the Pacific (AUS)

by Skillgannon

wheelsONfire wrote:
Wed Sep 09, 2020 9:38 pm
Wind tunnel time seem to cost alot of money, so that's why most seem to rely on CFD only?

It would be interesting to hear an honest answer if bikes really need to be designed in a wind tunnel.
It's not just cost, but also time. With a wind tunnel model, you take a model (or several models) to the wind tunnel. You can test different configurations (things like bars / stems), and possibly look at making minor changes to tube profiles by adding fairings or bondo. At the end of the process you end up with some measurements, and (hopefully) some flow visualisation to help guide your thinking on where you should be looking to improve. Then, if you have the budget, you build more models, go back to the wind tunnel, repeat. But that's all pretty expensive, and worst case scenario you just get a number for the marketing department to use.

Through CFD, the iterative design process is both faster and cheaper. You can make a change to your design, remesh it, and rerun the simulation multiple times in the time frame that it would take to set up a new wind tunnel model and test it, with far less cost.

As to if you need a wind tunnel - no, not really. Yes, a wind tunnel may theoretically be more accurate, but with good CFD design that accuracy should be small. Furthermore, with the exception of meshing difficulties, your CFD should be consistent - so the changes in drag seen in the CFD model should reflect changes in real world drag, even if the absolute amount is slightly different from the real world. Furthermore, a wind tunnel is only as good as its design and operators - there's no guarantee of perfect accuracy in a wind tunnel either, it's a tool, and use that tool incorrectly and you've just spent a lot of money for nothing.
This board would be a nicer place if everyone would take themselves less seriously.

I almost miss Mr Search...

by Weenie


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

www.starbike.com



User avatar
wheelsONfire
Posts: 5060
Joined: Mon Jul 07, 2014 8:15 am
Location: NorthEU

by wheelsONfire

Well, i guess NASA doesn't use a wind tunnel so perhaps a slow bike doesn't need one either (haha...)
Bikes:

Ax Lightness Vial EVO Race (2018.12.21)
viewtopic.php?f=10&t=156137
Paduano Racing Fidia (kind of shelved)
Open *UP* (2016.04.14)


Ex bike; Vial EVO D

madik
Posts: 309
Joined: Sun Jan 08, 2017 12:44 pm

by madik

I would argue its a way cheaper to test bike in wind tunnel than to run CFD in house for a smaller brands If its done just few times a year. As other mentioned you get some data from wind tunnel very quickly If you bring few models and see what is faster (in tunnel anyway) and its done within a few days of testing. It surely cost a lot per hour but for a bike testing just 1-3 days must be enough and it probably wont bancrupt the bike company.

In other hand CFD is not a magic button, it requires to pay for expensive licenses, have an experiance specialists for FE/CFD computation that have a good engineering and fluid dynamics background from phd Studies, auto or aerospace industry to actually give some insight and genuin new approaches to the design of bike thats is not just a copy of the other brands. Also it takes months or more likely years of FE/CFD work to learn how to analyse the bike properly to take the advantage of those tools. Software licences, FE/CFD specialist that have to be payed whole time will cost more than some testing now and then in wind tunnel.
Functionality > Performance > Weight

robeambro
Posts: 1513
Joined: Sat Jul 07, 2018 6:21 pm

by robeambro

madik wrote:
Thu Sep 10, 2020 9:09 am
I would argue its a way cheaper to test bike in wind tunnel than to run CFD in house for a smaller brands If its done just few times a year. As other mentioned you get some data from wind tunnel very quickly If you bring few models and see what is faster (in tunnel anyway) and its done within a few days of testing. It surely cost a lot per hour but for a bike testing just 1-3 days must be enough and it probably wont bancrupt the bike company.

In other hand CFD is not a magic button, it requires to pay for expensive licenses, have an experiance specialists for FE/CFD computation that have a good engineering and fluid dynamics background from phd Studies, auto or aerospace industry to actually give some insight and genuin new approaches to the design of bike thats is not just a copy of the other brands. Also it takes months or more likely years of FE/CFD work to learn how to analyse the bike properly to take the advantage of those tools. Software licences, FE/CFD specialist that have to be payed whole time will cost more than some testing now and then in wind tunnel.
I don't think that CFD and wind tunnel testing are necessarily fulfilling the exact same purpose.
As I see it, CFD (if done properly and with the right competences) allows you to try and screen loads (dozens, hundreds, thousands? I don't know) of different combinations for each component (ie fork, headtube, bars, etc) and see in theory how they perform. If you develop exclusively using a wind tunnel as a small brand, you can maybe try a handful of molds that may or may not be aero at all to see which one's the fastest, but that's hardly something I'd call "aero optimisation"..

Sure, as a small brand, wind tunnel can be more accessible, but I'd be wary of something developed by a small brand exclusively in a wind tunnel. You may be getting a frame that has all of the compromises of an (old generation) aero frame such as excessive stiffness, excessive weight, without any of the benefits (ie the frame may be only as fast, if not slower, than aero climbing frames such as SS Evo, SL7, new Emonda, etc). All of the compromises with potentially none of the benefits? And, at the same price (usually on grounds of "boutique feel" :roll: ) of a well-established brand? Nah - I'll pass, thanks.
Last edited by robeambro on Thu Sep 10, 2020 11:43 am, edited 2 times in total.

Stueys
Posts: 641
Joined: Sat Nov 22, 2014 1:12 pm

by Stueys

I'm not an expert by any means but I would have thought that a wind tunnel provides a known baseline, concious that each tunnel is different but ultimately it's the same technology so there is a level of consistency across the different tunnels. Whereas CFD software is multiple different models and software programmes, we've no way of knowing whether software version x is equivelent with software version y.

User avatar
wheelsONfire
Posts: 5060
Joined: Mon Jul 07, 2014 8:15 am
Location: NorthEU

by wheelsONfire

If different wind tunnels shows different data under same conditions, it's not 100% reliable is it?
So the miniscule gains we see might not be correct?
Bikes:

Ax Lightness Vial EVO Race (2018.12.21)
viewtopic.php?f=10&t=156137
Paduano Racing Fidia (kind of shelved)
Open *UP* (2016.04.14)


Ex bike; Vial EVO D

Skillgannon
Posts: 3631
Joined: Sat Apr 01, 2006 12:17 am
Location: A bigger rock in the Pacific (AUS)

by Skillgannon

Stueys wrote:
Thu Sep 10, 2020 10:37 am
I'm not an expert by any means but I would have thought that a wind tunnel provides a known baseline, concious that each tunnel is different but ultimately it's the same technology so there is a level of consistency across the different tunnels. Whereas CFD software is multiple different models and software programmes, we've no way of knowing whether software version x is equivelent with software version y.
There can be significant differences across wind tunnels. Sensitivity to inlet turbulence, how the bike is mounted on the wind tunnel, etc. etc. can all cause issues. Ferari Formula 1 a few years ago realised their cars were underperforming, because their wind tunnel set up was wrong. This is a billion dollar company with a large team of dedicated aerodynamicists.

Both with CFD and Wind Tunnel work, you need to pay for experts who understand what they're doing. With CFD, poorly understood/chosen turbulence models that are innapropriate to real world conditions, or meshing issues, can both cause problems. These are less software problems though (at least not from software X to software package Y), and more methodological problems. With a wind tunnel, you have less access to full flow field data, so when you move beyond fixed metrics (which are only of limited value when trying to improve a design) there's more interpretation required, and this is the type of thing which can be gotten very, very wrong.

None the less though, I would 100% be more likely to trust a small-mid scale manufacturer who has worked mainly with CFD, as compared to mainly with a wind tunnel. Because with CFD, you can only screw things up so much, and you can iterate far more. But in a wind tunnel, are they using the right tunnel? Not every tunnel is appropriate for bike applications. How are they using the tunnel, is it just for a marketing number? Are they paying external experts (because they won't have this knowledge in house) to understand how best to use and exploit the tunnel test environment?
wheelsONfire wrote:
Thu Sep 10, 2020 10:53 am
If different wind tunnels shows different data under same conditions, it's not 100% reliable is it?
So the miniscule gains we see might not be correct?
Correct. A great example with this can be seen in wheel drag data. Every brand will come out and say "our wheel is better than our competitors by Y watts". Of course it's Y watts at a certain speed, and under a specific yaw angle (or a specific weighted average of yaw angles). But those yaw test protocols can be different.

Now with wheels, a lot of the white papers at least will have measurements at different yaw angles, so you can compare a little bit more, but even still, there can be differences in relative performance across the wheels that are driven by the test protocol. And when it comes to frames, there's usually no information about the yaws or anything usually.
This board would be a nicer place if everyone would take themselves less seriously.

I almost miss Mr Search...

User avatar
wheelsONfire
Posts: 5060
Joined: Mon Jul 07, 2014 8:15 am
Location: NorthEU

by wheelsONfire

I'd like to set up a few aero bikes appropriate for how i like to be positioned and see if i really notice a difference.
Both in speed, freshness and comfort.
I has to be a clear benefit or in all fairness it lean to much to a placebo effect status.
I'd like to see how typical aero bike as Cervelo S5D would look.
Bikes:

Ax Lightness Vial EVO Race (2018.12.21)
viewtopic.php?f=10&t=156137
Paduano Racing Fidia (kind of shelved)
Open *UP* (2016.04.14)


Ex bike; Vial EVO D

ericlambi
Posts: 139
Joined: Mon Apr 18, 2005 6:16 pm

by ericlambi

How much do you suppose you'd be behind if you get the Force build for a little more than $8k then upgrade to the new DA when it comes out and has availability? The DA bike will not be less than $12k if the red is $11k, so you'd have ~$4k to spend before you even sweat it. A quick google search shows 9170 complete groupset is $3300 at Colorado Cyclist, $4100 with a power meter in the crank. Obviously, the new one will be more, but seems like you'll be in the ballpark. Will the wheels you get on the Factor be compatible though? The Force build does come with 30mm wheels in stead of 45mm, which is a slight bummer as well.

Geraint
Posts: 144
Joined: Thu Aug 06, 2020 7:46 am

by Geraint

It's hard to say given that we don't know what 'standards' the new DA will use. 24mm BB is probably a reasonable bet but will it be microspline (MTB 12sp) hubshells or the current Shimano 11sp road type?

It comes down to how much of a rush you are in and how much you're prepared to gamble. I personally would either buy the frameset or a custom 'rolling chassis', or just wait for new DA to come out and then buy the complete bike with new DA build. My 2p.

azaggie
Posts: 22
Joined: Thu Aug 20, 2020 7:08 pm

by azaggie

ericlambi wrote:
Thu Sep 10, 2020 7:27 pm
How much do you suppose you'd be behind if you get the Force build for a little more than $8k then upgrade to the new DA when it comes out and has availability? The DA bike will not be less than $12k if the red is $11k, so you'd have ~$4k to spend before you even sweat it. A quick google search shows 9170 complete groupset is $3300 at Colorado Cyclist, $4100 with a power meter in the crank. Obviously, the new one will be more, but seems like you'll be in the ballpark. Will the wheels you get on the Factor be compatible though? The Force build does come with 30mm wheels in stead of 45mm, which is a slight bummer as well.
Looks like you can swap for different wheel sets, no problem. I think you'd be ahead if you sold the Force for maybe $1500. Free OSPW can be sold for a few hundred bucks as well.
Really hope new DA is lighter, not heavier like new Red was.

Scott2017
Posts: 148
Joined: Sat May 20, 2017 11:40 am

by Scott2017

FlatlandClimber wrote:
Wed Sep 09, 2020 8:44 pm
wheelsONfire wrote:
Wed Sep 09, 2020 7:42 pm
FlatlandClimber wrote:
Wed Sep 09, 2020 3:39 pm
What is the Ostro comparable to in another brands portfolio? The One being even more aero and the O2 VAM being super light, is this like the Domane Koppenberg and Roubaix Team Edition?
Back off on the Emonda and go for Ostro?
Nah, although I really like the bare carbon paintjob. I like the more classic look of the Emonda better. Also, Factor have never been to a wind tunnel with the bike, which is kind of questionable. Probably still a very good bike.
Really can't wait for Tour to get their hands on both of these bikes, the wind tunnel debate raises some good points but it will still be good to have one consistent test to compare.

bluebottle81
Posts: 67
Joined: Mon Feb 10, 2020 1:23 pm

by bluebottle81

I've ordered a frameset. Pretty happy with the deal given bars, BB and OSPW included...

Went with Sicilian Peach which I think looks awesome and was a big drawcard for me. Though I will be interested to see if that changes after the ribbing I get from my mates for having a partially pink bike. Should be an interesting experiment for me and can always get paint redone down the line.

Been working through the build components which I'll be buying new - so far reckon I might be able to get weight including pedals, cages, garmin mount down to ~7.2kg (noting total cost needs to be A$10-11k max)

azaggie
Posts: 22
Joined: Thu Aug 20, 2020 7:08 pm

by azaggie

Sounds awesome. Every Factor I've seen had a cool paint job, betting the peach will look rad. And who cares what your buddies think- if you're excited to ride it, that's all that matters. Looking forward to hearing some real world reviews!

by Weenie


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

www.starbike.com



Post Reply