Ask Josh Anything Episode #33 – Wind Tunnel Secrets – PART 1

In this riveting behind-the-scenes filming of the Marginal Gains podcast episode, our hosts, Hottie, Fatty & Josh, dive into the world of wind tunnel testing, covering everything (for part 1 anyway). Join us as we explore the importance of tire air pressure, the wonders of the SILCA Terra floor pump, and a mention of the recent UCI rules revealed regarding road bike brake levers. Get ready for a fascinating journey through preparation, experiences, and the critical differences between closed loop and open return wind tunnels. The conversation peaks as Josh unveils the secrets behind computational models, and some jaw-dropping aha moments in the wind tunnel.

Takeaways:
• Directionality of prediction is key in working towards performance improvements.
• Aha moments can lead to unexpected discoveries and insights.
• Wind tunnel operations involve various roles and meticulous data analysis.
• Avoid groupthink, encourage individual ideas for more insightful conclusions.
• Implementing changes requires buy-in and support from a broader team.

Chapters:
00:00 – Introduction and Hottie doing a bit of Cyclocross Racing
02:21 – Fat Biking and the Terra Floor Pump
05:26 – Importance of Knowing Tire Air Pressure
09:29 – Wind Tunnel Testing and Controversial UCI Rule
11:52 – Preparing for Wind Tunnel Testing
15:30 – Why Go to a Wind Tunnel
18:17 – Closed Loop & Open Return Wind Tunnels
25:17 – Preparation, Experience, and Variables in Wind Tunnel Testing
41:43 – Directionality of Prediction
49:24 – Aha Moments in the Wind Tunnel
01:19:23 – Day in the Life of the Wind Tunnel
01:23:04 – Analyzing Wind Tunnel Data
01:23:59 – Data Collection and Interpretations
01:25:51 – Post-Processing and Analysis
01:31:10 – Successful Implementation and Buy-In
01:32:10 – Continued Iteration and Planning

Don’t miss this exclusive look into some of the secrets of wind tunnel testing! Subscribe, like, and hit the bell to stay tuned for Part 2!

#AskJoshAnything #WindTunnelSecrets #CyclingTech #SILCA #silcapursuitofperfection #DylanJohnsonCycling

I am also punching record on audition okay I’m recording in so many ways now guys okay let myself squared up here either are you going anywhere for for Christmas I have a cycle cross race this weekend be my third of the season and then yes off to see family awesome so

Cyclocross that’s awesome Coss how are you doing I’m pretty awful but I’m having a lot of fun actually work so you’re doing great yeah well my Red Zone power is just is dissipated um because of lack of racing for for some time now

So and I was in no mood to get on the trainer and do anything about it but what I did do is go to the park a couple times a week and work on my Dismount remount skills which it paid off I was always terrible at it just you know I

Would just to show up the cross races and add Li it right but this time around I went you know what I’ll at least work on something and I’ll work on an absolute weakness which is that and a couple a Sven nice video and some other

Kind of tips here and there and um that skill has picked up nicely um that’s awesome yeah and uh I haven’t Red Cross in since before the chains spaw day so having a wax chain and cross rcing has been Brant love revelatory huh yeah yeah um been experimenting with tires a

Little bit too so and then Josh the the Sila B tape the kushino am I pronouncing that right kushino yep kushino yeah kushino yeah perfect it is it is terrific so that wrapped around a set of carbon bars and it’s great so so I think cross is the one discipline where the

Worse you are at it the more fun you will have is my my like hypothesis about cross like it’s like when you start getting good at it you know it’s a little bit it’s like like golf right you know you go hit some golf balls and like you know

When you hit the good one you’re like oh that was awesome and then when you’re actually good at golf it’s like every ball that’s like you know less than perfect God damn it for cross for that for me is like that like oh I only fell eight times per

Lap on this one and uh then look how muddy I am let’s have some beer so for me it’s fat biking season it’s yeah uh couple of feet of snow in American for canyon and I am digging it I actually had a question for you Josh

On uh a new pump that I saw this Sila has and I’m wondering if it might be a good pump for my fat bike because with my fat bike I knowing your air pressure is more important than with any other bike I have but none of my pumps will

Gauge it properly because I generally at six to between three and six PSI and it looks like you’ve got a new pump that will actually give me a reasonable measurement on that yeah uh the te pump I can’t remember what its lowest graduation is but I think it’s like

Three three and a half I know we we really had fatbike in mind as we we working on that and you get it’s nonlinear it’s actually really cool Tech it’s like a you know gauges are the is driven by sort of like a rack and pinion type mechanism and so it’s it’s a

Nonlinear uh mechanism so like the first what like essentially the first half of the gauge travel goes zero to 30 and you get quarter PSI increments and then from there to the other half you go all the way up to 120 um so it’s a you know much get gets

Much more compressed as you get higher and by the time you get to 120 you’re at you know basically like two PSI increments um yeah course that’s a pressure where that matters a whole lot less so yeah that might be a good a good option for your fat bike I would say it

Is time for me to make another order with Sila I it it is I mean forever I have had to do essentially the the test for pressure with the fat bike is just basically how hard am I having to push before you know I feel the rim you

Because you you are riding with that low of pressure uh in soft snow and it is nice to know I mean because there’s a lot of chatter in the Utah fatbike Forum on Facebook you know it’s like well what pressure are yall running right now

Because there’s a lot there is a lot of dialogue about what is the right pressure both for a good quality ride and to not ruin the grooming of the trails which we are all very precious about here in Utah is if you ride with a

Hard Tire you are likely to dig a trench in the trail and ruin it for everyone for the rest of the day right so we are all very we’re very snobby about when it’s too late in the day to ride and when it is too hard to ride and you guys

On ebikes you lay off the throttle but but yeah no it’s knowing the pressures is probably more important on my fat bike than anything else just in terms of can I stay on the bike so yeah I got to I gotta I I gotta make an order man that’s

Fun all right so let’s I got a Bonus question for the YouTu I want to start I start with this so as we lead into the theme music and everything we’ll we’ll find out a little bit more but this is going to be a our wind tunnel Ask Josh

Anything and by who is asking Josh anything in this one it’s just hottie and me man we we want to know more about wind tunnels I I don’t think either of us have ever been in one but this is mostly going to be about things that you

Have learned and things that you can’t learn but I want to know what is the worst the most hairbrain hypothesis that you have ever tested or seen tested or been told about testing in a wind tunnel oh God worst as in like worst idea or worst

Outcome I I was thinking worst idea but I love the other one as well so God one of my first times ever at the wind tunnel Texas A&M in like 99 give or take um in in the industry and there was a guy uh Paulo who I don’t

Think he’s still around in the industry but he had L composits for a while he did a wheel called the black black hole which you remember was like a like a fork that was kind of like a had a circular carbon thing and then the rim

Ran in like bearings in that so it was a hubless spokeless wheel um but he I think the company was called like wear and tear but he did um arm and calf fairings they’re like foam blocks that you could like velcro to yourself um and like they worked but they were just

Horrible and and looked Terri and and and so it’s timely because you I think Escape Collective did a a thing about the UK TT scene a couple months ago and like people are like stuffing um like cut up pool noodles in the back of their really tall socks and things to

Try to make body fairings again I’m like oh my God we’re coming back to this like like oh yeah that that I just think from a I get it I get it it most like works it most likely works well but oh it’s just it’s just terrible right everything

About it like please don’t turn our sport into that we already wear socks that are too tall and are turning our brake levers Inward and you know doing all this other stuff for gains that I get on the gains perspective but like it it’s just not a good look for it’s not a

Good look for our sport cosmetically but um the I would say funnily enough if you think of it from a uh performance I will say like the probably the worst performance product you can think of to to put on there it is like a handlebar

Bag and and that I just think is a little bit silly you know you go to these races you know you go to Steamboat and there’s a you know $144,000 you know Factor semi arrow with you know 58 MM Wheels and it’s like you know somebody looks like they’re

Contending to race and they’ve got like a little burrito bag on the handlebar and you’re like you know that thing is like eight watts of penalty um and so that’s what like you know I I’ve used them in the past I totally get why you you want one but

It’s like wow it’s amazing that that something that that big when you put it in that spot can have just that significant of a negative impact uh performance-wise all right should we do the actual show then let’s do it I uh yeah by the way we’re

Recording the show I think on the day that news broke that the UCI is going after more um Arrow advantages by restricting the turned in levers yeah yeah I just I just saw that earlier today on on escape and um and yeah they they did it and they provided like like

I don’t know who did that CAD work but man they they need some help with their CAD work so you know UCI you oh call [Laughter] me I kind of want to dig deeper into this no tell me what you’re talking about I mean what was wrong with the

Catw work oh just they built this model and they made kind of a jig with the model and there’s just I mean just from somebody who’s catted for a long long time you know we opened it and my Engineers here and we’re all like like

Oh man you know like they went and got like some high school kid to like do free cadw work for them it’s just kind everything about it is it’s just off and um so sorry for I’m clearly offending somebody somebody did this work and they’re going to be super offended to

Hear it but it it just is a little bit um I don’t know I feel like it it just detracts from the professionalism of the governing body of sport is like laying down the law on this topic for this particular reason and then they’re doing it with you know like very

Very amateur supporting materials I let’s say that or put it in those terms so um again I’m sorry I’m sure I just totally totally pissed somebody off Merry Christmas use the person we will no doubt get get angry comments from somebody there but um but anyway we’ve remained Scandal free for 100 plus

Episodes it is about time we cause some controversy that’s ging about yeah yeah of all all right let’s do a countdown H you want to lead it okay okay sure uh this is episode 108 wind tunnels in five four three two two one one terrific okay

Um this is the margin that was nice little rougher on The Voice something extra for the YouTubers this is the margel greens podcast Ask Josh anything number 33 this ajaa is going to be different as we have an episode dedicated to the wind tunnel that’s right a full episode’s worth of

Questions all about wind tunnels we’ve got Josh with us who has spent well actually that’s a good first question Josh how much time have you spent in Wind tunnels in your career oh it’s uh 1,00 Plus hours give or take holy cats a lot a

Lot so 30 working weeks of time of your life has been spent in a wind tunnel give give or take yeah yeah which is crazy and which if for those out there who’ve done it you know it it is very much like watching paint dry like the actual operating of it and

So I yeah last we we were there a few months ago I took my son with me so he could see it he’s interested in engineering and he lasted about two hours he’s like Dad this is miserable like hey I’ve done like 1,00 maybe I don’t know a lot I stopped

Counting at some point like in the 1100s but um but yeah it it it can be fun but generally yeah it’s like watching the grass grow yeah the margin L podcast is presented by Sila Exquisite pumps wax chains Aeros stocks titanium dropper hangers bik computer mounts and cleats bik cleaning

Products and more I’m Michael hotton hied most and marginal gains is also supported by Shimano known mostly for its outstanding group sets Shimano also makes an extensive set of wheels let me say that again marginal gains is also supported by Shimano known mostly for its outstanding group set

Shimano also makes an extensive line of Wheels the set is most relevant to this episode hold on let me add this line what do you say and I am fatty I forgot your name here and I’m am fatty I’m super famous I people know my voice

Okay do you need help with a salign fatty I I mean apparently it’s it’s just like it’s uh he uses too many two syllable words you know that’s I have it so much better than you guys like you guys have like words to read and all of mine are just like three pink

Question marks like like that is so much harder to screw up okay deep breath Center yourself laughing the margin no where am I and I am fatty marginal gains is also supported by Shimano known mostly for its outstanding group sets Shimano also makes an extensive line of Wheels the

Set most relevant to this episode would be the shimano’s deepest carbon Arrow wheel the dura a C60 shimano’s road wheel lineup is huge with models in the tgra and 105 lines and offerings in tubless tubular and rim brake if you need a new set of wheels be sure to check out

Shimano bike. shimano.com all right we want to get into the mysteries of wind tunnels here but we should probably start with a foundation the why the how the what and whether we should be in a wind tunnel at all now Josh you taught us all about the Chung method how it can

Be completely useful to us and it’s certainly a lot cheaper than the wind tunnel why go to a wind tunnel at all yeah I’ll say you know that the Chung method has changed a lot and has really you know democracy ized uh arot testing and to some extent crr testing

Um for all of us but if you really want to try to understand sort of the not just the correlation but the causation of certain things particular on the equipment design side of things the wind tunnel is really just the best tool for testing your various hypotheses from a product design

Perspective um it also from a writer perspective can be a much much faster way of refining a position than than what you’re going to get with with something like the Chong method you know a typical um wind tunnel you know yaw sweep of a position we can do in you

Know five to seven minutes uh and then maybe a quick change over in between and then you’re back testing and so you can really get a lot of you know move through quite a lot of variation in say a day of testing whereas uh you know to

To Chung that you really you’ve got to do your your laps or your your saddles or you know kind of whatever method you’re using and you’ve got a act you know you’re hoping for not too much variance through the day and and wind speed and direction or or air

Temperature or that no weather fronts come through and you know the wind tunnel really kind of takes a lot of those out and kind of lets you just get down to business and you know I’d say typical like World Tour riter I mean we in three hours you can

Do you can make some pretty significant uh significant gains and then you you got to put the rider in the re world and see if they can actually you know ride that position yeah and uh make make sure that it works at the back end so you’re never

Going to get away uh from from the field testing part long term but yeah the the tunnel as expensive as it is you can really get a lot of bang for your buck if you know how to operate in there with a high level of efficiency um now not all wind tunnels

Are created equal I’ve heard um in researching this topic something about closed loop wind tunnels and open return wind tunnels you have tunnels that kind of suck in air across an object or blow air at an object so what are the differences here and and what do you

Prefer oh that’s that’s good yeah so there’s a ton of variation um we start with so a closed loop tunnel you know if you were looking at it from you know over top of the building is is sort of like a big doughnut um most of them are

Probably more rectangular some of them are actually kind of curved in the corners but um you know closed return is essentially like you have one volume of air and you are running that air around this Loop over and over again um one of the things that’s really nice about that

System is that you can almost all of them um you can air condition uh or control control the temperature and humidity of that air really very tightly um independent of what’s Happening you know kind of outside the system and so you you can run through an entire day with very minimal uh pressure

Temperature variation and so the reason that’s important is that you know as uh air gets warmer it becomes less dense and your drag numbers go down if you’re not accounting for that or if humidity goes up uh you know your drag numbers are going to come down uh which might be

A little counterintuitive but that’s why uh you know like for your ideal hour record right you want to pick a a hot humid day at high altitude with low barometric pressure um and so with with a closed loop tunnel you just have much much greater control and you know some

Tunnels very few and far between um you think of like the the tunnel at MIT it’s actually able to be pressurized and that is generally done so that you can take out like the scaling effects of using scale models um typically if we use a scale model

You within the Reynolds number you have this length piece and so if you make your model half the size the Reynolds numbers are all haved but there’s like a pressure um uh or air like kinematic viscosity component in there that you can kind of offset that scaling by

Pressurizing the tunnel um and then there are some tunnels that they just use that to control the pressure you know kind of as a very fine level just to take that out of uh the variability between tunnels but of course you can imagine pressurizing this thing with

This huge volume of air um it’s not cheap the tunnel is more expensive uh there’s more power required to run it um yeah there’s it it’s pretty exciting when you think of all the things you know essentially the the from the engineering perspective Engineers love to take variables out of

Equations right you know we we get a bad W for you know solving like liking to solve these huge mathematical equations but really deep down what most Engineers want to do is make those equations as simple as possible and the more of those variables you can take out the easier

The math becomes but also just the the less variability you have to you know contend with when you’re post-processing the data and so um you know for every hour trying to take the it depends out of the yeah yeah or or just the the random uncertainty or the the you know

One of the big challenges with testing like this is you know for every hour in the tunnel you probably have two hours of post-processing and and kind of work with the data to try to really understand it but when you’re in there you know it’s a thousand bucks an hour

So you’re really full tilt on just trying to get your runs completed and buttoned up and make sure that you’re not making any m mistakes well you know good good example actually with a test I did at MIT with uh some of the CSC guys ion boso and and caros sastra and like

Way way way back in the day and um we started to see like some really weird drift during the day it but it took us because you had humans in there and humans mess up the air right and they’re not very repeatable and they throw a whole bunch of uncertainty into the

System I mean it took us a couple of runs to find it and then we found out that there was some um some issue with the heater that was supposed to be controlling the temperature and it was actually causing it to warm and that was throwing all of our other data off um

And that’s really hard to find in the moment because you know you’re also trying to think like hey this is a thousand bucks an hour I’ve got to get these spacers ready so that as soon as the run is over we can open the door and

Run in there and you know like a F1 Pit Stop like get the position changed and then get the door closed as quick as possible to get the air spun up again um and so there’s really these two uh well there’s many but two big competing

Um like factors of your kind of like mental capacity right you want to be as quick and efficient as possible with your runs to minimize the expense but you also really need to be on top of all your variables as much as possible to ensure that your data is as clean as

Possible so um in my experience a closed loop tunnel uh generally gives you better repeatability especially if you’re running through like an entire day or over multiple uh multiple multiple days because they’re less sensitive to external temperature pressure weather type stuff um San Diego innel is a great example of this you

Know it was kind of built like World War II era and it’s concrete and so it’s this huge kind of thermal Mass um that the air flows around and it’s in a place that has pretty consistent weather dayto day and so you know we can

Go to San Diego and actually it zip we twice a year go to San Diego and do five 10hour days right and so that’s where a lot of my 1,00 Plus hours come from you know you’re doing it 50 hours at a time um

And go ahead that kind of gets to the qu I guess the picture that I’m trying to develop in my mind and that is you know is the day before or you know the days before A Day In The Wind Tunnel what is the day in the Wind Tunnel like and then

You know what is are the days after you’ve been in the one tunnel like I mean there’s got to be a lot of preparation of developing the questions you want to answer there’s the day when you’re actually in there and trying to collect the data you need and then

There’s the days after I assume where you are processing and analyzing and and I I’m really interested I think especially in the days before how do you develop a question that you know the questions you want to answer and then the day of I mean it’s I think all of my

Understanding or picture of wind tunnel experience is like from a few photos and a minute or two a video where you know which are you know control room shots right over the shoulders of the people who are at computers looking at a guy in an arrow position on a bike and it

Can’t be like that what is it actually you know what is the prep what is the experience what is the post experience oh I love that yeah one minute or less there’s so many more people doing it wrong than right on this one and I I

Think you you kind of nailed I mean that the preparation is is way more important than the rest of it uh to some extent I mean you need you need to know what what you’re going to test build a run list you know have talked to the tunnel like what

What’s a typical runtime look like um you know what’s our what’s our setup are we going to sweep are we going fixed angle you know are we time averaging our data points um you know it’s pretty common to like if you have a riter you

Know you test at zero and you take a 30- second you know time average of of uh you know all your data points and then you yaw well you know you say stop at five degrees well your balance has mass to it and so now as soon as you stop

There’s like a harmonic wobble in the balance for maybe 30 seconds um that needs to settle out and so you have to wait that time and then you maybe take your 30 seconds of data and then you do it again um and so there’s tons of stuff

Like that that if you aren’t asking the question or don’t know to ask the question you can really get caught out by you know like oh we’ve yawed the table take take the data well okay there’s going to be all this weird we harmonic stuff from the moving of the

Table um you know things like uh a lot of closed loop tunnels uh have doors that magnetically lock so you can’t enter them while they’re running because that would be super dangerous um and so you have the time to spin up the air volume and then the time for that air

Volume to Coast down to a safe velocity and so you know you might have a a 10-minute run time but it’s a minute and a half to spin up and a minute and a half to spin down and so you have to factor that stuff in um and then you

Have to yeah think you know what is it that we’re testing is it rer positions okay what’s the best order to do this is it is it faster to add spacers or to take them out you know is it faster to leave the arrow bar loose and let the

Riter kind of you know bend it or not bend it but you know shift it from say flat to 5 degrees up to 10 you know degrees up or whatever you’re testing so that we don’t don’t have to open the door and let let the air volume spin

Down and spin up again now there’s a ton of strategy that that can go into that and then the the last thing and we actually got caught out a little bit uh in the test that we did with Dylan Johnson is to really just run through your numbers and kind of remember all

Your your sanity check stuff you know right like what what cdas are we expecting um you know different tunnels talk different ly some use different conversions you know what what’s the a that we’re agreeing upon uh or or or that we’ve measured and that we’re

Bringing it to to the table in our CDA because most tunnels will have like an assumed a for you know if you put a wheel in there a bike in there a bike in a riter well you know the the a is going to be X and so then we’re measuring the

CD and you know all of a sudden you’re talking CDA and you think oh these numbers don’t make sense oh well you know maybe the tunnel this tunnel uses a different a than the other tunnel did or or they have a a generic a that’s not what you calculated for that writer um

And so you can get yourself lost really quickly there um and then for me honestly one one of the big things I have to do is is actually run through like tables to reorient myself and this sounds so stupid but I I worked for so

Much of my career in grams that I still think in grams and not in CDA and so it’s just kind of like getting the cheat sheet out and having that ready and having your brain in a right spot to really again sanity check yourself you know there’s certain things like oh wow

That was a one of my challenge with CDA you know cdas are point you know that was a 0. Z3 Improvement like that’s a that’s a hard number in my brain but if you tell me oh that was you know 12 grams like that that clicks straight up

Um and so simple things like that you know like that one’s particular to me but I think we all have our own baggage and and uh you know having all of that sorted and rehearsed ahead of time will save you a fortune uh an absolute Fortune going in there and um you know

The the wind tunnel folks like they’re awesome and they’re every tunnel I ever have been in you know they’re willing to help and they’re willing to do what you want but you know they’re not there to tell you how to do it and they’re you know they’re not there to do

It for you I mean they’re really just there to operate the equipment and make sure their stuff doesn’t get damaged and you know I think quite frankly at some level too like if you go in thinking you’re going to do 30 runs in a day and

You do nine like that for them hopefully you just come back another time um and and I’ve have seen that from so many um I’ve actually walked out of a couple of wind tunnel tests before with uh with pro teams because they’re just so badly managed and run and planned that it’s

It’s a bunch of people sitting around doing all the wrong stuff and just burning a giant pile of money and and and making everybody myself included very frustrated and so yeah I’ve I’ve actually walked out on two different uh tunnel tests over my 25 years or whatever of doing this um just from

Frustration and and feeling that total inability to to help or to be listened to um they’re you know like so many things in life right they’re they’re 10,000 ways to do it wrong and there’s probably only like a handful of ways to really do it

Right and so you know if you don’t have the experience um or the plan in place you’re almost certainly doing it wrong so tell us about open return wind tunnels how do they work and are they any good in your mind yeah so open return tunnel is just kind of like a

Line um and so this would be like the A2 wind tunnel uh the specialized wind tunnel and they are much much less expensive to build which is a huge um a huge plus you know when I when I was actually about the year before I left zip actually a little bit

Played into my leaving you know we’d run a project where Michael Hall and I hired a guy uh Bruce Ashcraft who designed the um the auto research wind tunnel here in Indianapolis uh we’d actually hired him as a consultant to design a wind tunnel that we were that saram was planning to

Build for zip and uh it was pretty fascinating at the time to get through you know we were probably $3 and A5 million dollars to build an open return pass through style tunnel we were probably n to 12 million to build a closed loop uh tunnel and so

That put some scale on it um the challenge with open return tunnels is that they are pulling in new and different in random air from whatever building they’re in all the time and so um anyone who’s tested at A2 will tell you you know there’s there’s like a

Garage door and a couple of man doors that you know if if it’s lunchtime and four employees go through the man doors like you can change the temperature in that building that the the tunnel is by you know well over a degree and and now you have to chase that through your data

Right or it’s North Carolina and it’s hot and sunny and if you know you start in the morning and it’s it’s you know 68 degrees inside that building and even with the HV fact that they’re trying to keep on top of it with you could be 74

75 by the end of the day um and so you’re just you’re it’s constantly becomes a moving Target yeah and uh one of the ways you can offset that is to have you know essentially a control item uh and so you like at zip we had this control 404 it was a

2000 uh model year 20044 with a 21 millimeter uh Victoria cors CX Tire on it and you know we called it the Golden Wheel and it just it you know had silver Sharpie on the side indicating the direction it had to spin and you know the air pressure that it had to be

Pumped to and kind of all the stuff and you would take it to the the tunnel and every first run was with that wheel right and um if you were at a tunnel like A2 probably every six or seven runs you’d put that wheel back in and run it to make sure that

It was repeatable to what you had run you know two hours prior um and so things like that you know especially if you’re working with people things like temperature drift um are just almost impossible to find right because you know a human is variable they’re never quite still they never exactly repeat

The position I mean you you know we have tools and techniques we really work on but um very subtle things can make differences and so you’re always already chasing that if you have a human and then if you’re now dealing with like a temperature drift um you start dealing

With this combined factor of okay well we’ve got to compensate for that in our data but now we also have to compensate for that in in the rider you know the rider is going to feel different and and ride differently at 70 than at 75 um

Especially if you’re in there for for a few hours right that’s gonna g to change things there so yeah I I would much prefer um to work in a Clos Loop uh tunnels and if you’re really lucky you work in tunnels with open uh what we

Call open nozzle and so you know the nozzle is like where the air is narrowed down to to pass over uh whatever you’re testing and so you think of what’s Happening typically there’s a fan on the opposite side of the loop um and it’s sucking air so no tunnel really

Blows um because your putting too much turbulence and mess into the air by when you blow it but if you’re sucking it from somewhere behind the rider um you can actually use the design of the tunnel and those like the curvature of the tunnel and the what they call

Turning veins and flow straighteners in other parts of the tunnel to clean up that uh that dirty flow from the fan and so uh and so if you have a a closed nozzle or a closed kind of working environment tunnel that would be like

What you see at uh the lswt at San Diego right it’s the rider is kind of sitting in this tunnel and it’s rigid walls with you know recessed lights and all the air is flowing through that and and one of the problems you have with closed nozzle

Tunnels is uh this thing called a blockage effect and that’s that you’re the thing that you put in there can only be up to a certain percentage of the cross-sectional area of the tunnel before you start almost acting like a valve um and like San Diego’s a

Fantastic tunnel but it starts to flirt with this as you yaw the rider that the the kind of flow and the wash over the rider begins to kick out into the walls and kind of be reflected back um and that has all sorts of potential to mess

With your data in ways that you don’t understand um and you know this is one of the problems with you see a lot of small companies um develop a wheel or a bike or something and they’ll take it to like a University Research tunnel that

Might have like a a 1 by one or or a you know one and a half by one and a half meter cross-section and they’ll put like a disc wheel in there or a whole bike and do a big study on it and you’ll get all these crazy uh data and essentially what’s

Happened is you know they they’ve basically you know made like a butterfly valve out of that wind tunnel um and so you just at that point you kind of can’t trust anything um open nozzle tunnels like the arc uh like the Silverstone Research Center they actually take those

Walls away and allow like a room kind of around where that would be um and and these are really almost a necessity if you’re doing race cars because you particularly with like rear Wings on race cars and uh under trays you could have so much upwash coming off the back

Of the car that it’s otherwise going to be kicked up into and then reflected off a ceiling and so the way you can get around that is take the ceiling away um and because the air is being sucked in behind the car um we’ve actually done

This with smoke like at the arc I mean like the the flow can kick out you know uh half meter Beyond where the ceiling would be and cleanly just gets pulled back under I mean it’s hard to believe that it works but it it actually works really well um the other huge benefit

With the open musle tunnel is that you don’t you aren’t constrained by the door open to door close timing issue that we we talked about earlier so you know you you can be in the test environment just outside the flow that the rider the the

Product is in and you can have your um you know your next run or your part for your next run or your helmet change or whatever and all you really have to do now is wait for the safe velocity of the air to come down uh and that’s typically

Higher than what the door open velocity would be um and so those tunnels if you if you’re really in the game of maximizing your your tunnel time you can probably squeeze an extra 10% give orake runs in a day if you’ve got an open nozzle and you’re not having

To wait for this door hold Time stuff so what’s kind of amazing to me is the whole under underlining value proposition of a wind tunnel is that you are controlling your variables you can control variables in a way that you can’t in an open environment and yet

There remain so many variables that seem to be at least on the threshold of being beyond your control or easy to miss or if you weren’t paying attention you get a result that incorporates some variable that you weren’t even thinking about yeah totally yeah totally yeah and and

That’s I mean before you you you know it’s the complexity right I mean you you see those pictures of the Wind Tunnel control room that you talked about and you know me look at most all of it’s now on a computer screen but um you know all

Those gate like San Diego right I mean it was again built like World War II era so you know it’s I mean there’s probably 200 analog gauges in there that that they’re all doing something and and every one of them is connected to some sort of measurement device which is

Itself um you know has some uncertainty right or some error um that is subject to Breaking uh you know all the so yeah it really um it’s a challenging environment to work in and it’s if you’re doing it right you can really glean unbelievable um insight into to to how

Things are are working or or you know I think to me too maybe change the way I’m saying that one one of the the sort of joke about the wind tunnels you know like you know for every every time you go into the wind tunnel you know you go

In with 10 questions and you come out with a h 100 and and that is just so true right you think okay if I can just figure these things out or understand or or give some sort of directionality you know I feel like like that’s probably

The best you can really hope for is directionality like oh if if I have this this shape and I start making it bigger or deeper or that like what’s the directionality is it improving is it not is it um I think you know almost from like a scientific research perspective that’s

About as good as you’re going to get you know the a lot of it past that is setting you up and this is where like the computational stuff now is so exciting um a a lot of what we’re doing in the tunnel now especially on the product side of things is trying to

Validate what’s being done computationally and prove that our computational models have the the proper directionality of prediction um you know meaning that if the computational model says this is faster that it’s actually faster and and if it says it’s 2% faster that it it’s roughly 2% faster um you know because

I’ve certainly done things where like oh this should be 3% faster and you’re like well it’s 5% slower so something’s broken um and that was if you go back there’s amazing oh go ahead oh I’m sorry I I was wondering both in that direction where things seem much worse than what you

Expected vers or when they’re much better than you expected when you go in with an idea of how things are going to go and then you get something that is wildly different that could mean either that you know you misunderstood in the first place or there was something that you weren’t

Anticipating that that is giving you basically useless data how do you plan for and accommodate that well yeah I mean like so getting a negative result that was totally unexpected is like exactly the time where you break out you know the the the standard right right like the Golden

Wheel you know we we thought this thing would be better here and it was worse let’s make sure it’s not the equipment put the thing you know put the thing in that we know what the drag is and and you know there there’s some tunnels that

You know one of the popular uh standards is just like a cylinder you know looks like a a Ballard in a parking lot you know like you know so we always joke we you know call it you put the Ballard in and and if and if that repeats then you

You can pretty safely you know maybe you’ll repeat the thing that was unexpected again to make sure that it did it again but yeah it gives you insight’s not quite the right word but it it at least gives you some insight into the fact that you’re you know

You’re going down the wrong path with that thinking and and I certainly in the early days of the cfd work that I was doing back 2008 9 10 um you know we had a number of those I mean I remember a couple of couple of things that we built

Into our CF and our cfd model was pretty predictive of um of both wheel and wheel and bike um and then we started playing with some different surface textures and things and we’re finding these huge improvements and so we we really LED ourselves to believe that oh that this

Is you know the model’s really good at predicting Rim shape and and you know tube shape uh performance related changes like these surface textures must be magic and of course we took them to the windall and they were all they were all worse than standard and so that gave

Us some idea of like oh our our model is much better at predicting kind of the bulk flow around the thing but it’s actually you know our um and God there’s so many different ways of trying to model the flow interaction on the surface of things like clearly we’re

Just using a wrong um model when it comes to like our transition to turbulence um or our skin friction um you know pieces within the model and and so you know the the tunnel is a great a great way in an environment like that to keep you from going down the wrong path

For too long um but it can also be just a maddening tool in that it it it doesn’t tell you why anything right and so you know there really does just have to be this you know it’ll only tell you information about things that you physically made that you can put in

It and it will only give you the number that is the outcome and so every time something unexpected happens you instantly start going well why did that happen the tunnel will give you absolutely zero insight as to the why that happened um and there there are

Tools that people have come up with the smoke uh smoke and there’s these like um uh kind of like micro bubbles that you can put in the end things like that but I gotta be honest I 90 n% of all the smoke flow and flow visualization stuff that I’ve ever seen

And have ever done have been about marketing and not about I don’t think I’ve ever actually figured anything out using a flow visualization technique in an actual wind tunnel um but that’s where it it is gets super exciting to think about the computational tools because they allow

You um they can actually tell you why something is happening you just have to get the tool to a spot that it’s actually matching what’s happening in the real world right so it’s they need to be used in conjunction and they’re both extremely good and capable liers I would

Say well I think next let’s get to marginal gains that can be found in the Wind Tunnel but first the three of us I think deserve a little coffee break and we are going to be able to do that thanks to our sponsor Perk coffee yeah more than just marginal

Gains here fatty perk has been a solid upgrade thanks to their attention to detail on all things coffee for me I got to have single origin light roast beans I love the Ethiopians and most coffees from Central America perk knows exactly how to pull out the subtleties of a

Coffee region with its roasting techniques yeah the Ethiopians are on the sweeter side and Central American coffees tend to emphasize more citrus notes and you get all of that when you brew with a bag of perk and I’ll say someone who has had to learn a lot about packaging and labeling

Really appreciate how perk has colorcoded their bags of beans to match the roast level and the flavor profiles of the coffee um and they’re just cute and clever like man it being out of Savannah they got the art college right there like every if you see the holiday

Box it’s the poor tortured gingerbread men um it just it just makes me smile if I’m gonna be completely honest so you know I have to give it uh really give it to the team at Perk coffee is a crowded space uh and they have done an amazing

Job of standing out from the crowd both in the quality of the coffee but as well as in the the packaging and labeling and understanding of what you’re getting our listeners can get a 15% discount off their first order by using the code mg15 for 15% off go to perk coffee that’s PC

Coffee.com mg15 and remember to use that code mg15 at checkout all right let’s spend a few minutes looking at real gains that you have uncovered or at least gotten the inspiration or data for in a wind tunnel Josh tell us about a couple of aha moments that started in the Wind

Tunnel oh gosh um you know that was one of the earliest for me was the effect that the tires play and you know I think I’ve told at least some of the story before but you know 200 uh God two three somewhere in there um I was at the Texas tunnel with Lance

Armstrong and that whole you know cohort and Bernal and all those guys and um watching them test the wheels they were using the the head three which you know Lance was old friends with Steve and and had been sponsored by him since he was a junior

And he loved that threes spoke wheel but watching them test it with like a 19mm continental time trial TI um and you know I’m sitting there eating a Subway sandwich never forget we having Subway sandwiches and Yan berneil was telling me how you know he’s only getting 21 millimeter tires from the

Sponsor from their forward um to really cut down on the flats and the um you know they handled better and the rid it’s just maybe one of the first moments I’m like oh the tires are going to get wider um but he’s telling me this whole

Thing about how he’s all in on 21s and they’re not doing any TT specific tires anymore and um saying oh well why are we testing that wheel with this 19 like oh that’s the wheel we always test with and uh you know it’s fine it’s no big deal

You’re like oh no I’ve tested that wheel a ton like that wheel is fabulous with a 19 and it’s totally average with a 21 like you need to think about that and they were not interested in and and of course right I was I was the zip guy we

Were making the disc wheels and you know they’re looking at me like of course the zip guy is gonna you know trash the head wheel but uh you know which was also true right that was also part of my job but um but I really came out of there

Thinking oh God like that’s a real that could be a real actual advantage in the field if we started making a wheel Around The Wider tire that they’re going to use you know we we had always when tunnel tested with 19s and 20s um at that time and and that’s where we set

The the Golden Wheel actually got set with that 21mm Victoria Tire because that’s what so many of the teams were using um but most people weren’t time traling on them at least at that point and so uh from there I really got obsessed with tires and then uh if you

Think of the the original zip Arrow tire that the tangente was kind of came from our noticing that wow smooth tires aren’t great like rough tires you too much tread Is Not Great there’s seems like there’s this sweet spot of tread and so uh we actually bought a bunch of um

Michelin tires and and Michael Hall who was with me he came from car racing and had like a basically made a miniature tire groover out of a soldering iron um because he had familiar with that from car racing where they hand Groove tires and we started playing with different roughness coefficients and found

Something and then then we noticed and Michael made this thing uh that we all called the magic tire and he he made this tire and I won’t give away how he did it because it was so so brilliant and clever um but it had a bit of a more

Of a uh oval kind of Arrow shape to the tire and holy smokes that thing was fast and so that’s where the second generation um tangente tire came from and and then of course we all got schooled on that and that we made what was really Far and Away the world’s most

Arrow Tire um I think even more so at that time than the GP 4000 which we found kind of on accident and then Continental turned out they had made it on accident um but by making the tire more profiled and shaped we’d actually given it much worse rolling resistance

And so that ended up being a total trade-off where like oh we made the world’s most Arrow tire but we’ve offset all of those gains with this increased rolling resistance um and so that product uh ended up going away and being replaced with a a much more rolling

Resistance focused product but um yeah I I will say having having left um that world and and moving into the silker world I was pretty quick to seek out the Continental guys to say like just you know like just tell me like how how did you guys do it with

That gp4000 like it’s got good rolling it’s got just the right amount of roughness the arrow is good the shape is good and never forget he just looked at me goes honestly total accident it was like you know like I tell my team here the time like you know

Sometimes you’re lucky and sometimes you’re good and sometimes you’re both and uh you know one of those were just like oh man you you guys you got so lucky but hey congrats on the way they capitalized on it because they you know they really took it and ran with it um

Once that became known that that that was such a good tire so I’d rather be lucky than good that’s yeah yeah yeah in that case but better lucky than good although then you got to run your business and and actually be good and and deliver right

Because we’ve seen a whole lot of a whole lot of products out there that that were good or or somebody got lucky but man if you can’t fulfill um or the quality is not there it doesn’t matter yeah Luck’s not repeatable that’s the problem yeah yeah like if yeah luck is

Is like that uh that bowl that you got to grab by the horns right go so that uh and then more recently Josh you were in the you were in the Wind Tunnel like you mentioned with Dylan Johnson and and you folks found that Not only was a Hydra

Pack or at least some hydration packs not bad but they they could be good they could help you yeah cheat the wind a little bit that was I think a lot of people were like wait what I mean those straps aren’t hurting me all that weird funkiness out front doesn’t mess things

Up yeah that that shocked me I mean my my prediction had been that the bag as is would be would be bad but that if we really cleaned it up all the straps de wrinkled stuffed it just right like you could make it into something

And I I was shocked I was totally wrong I mean even with all the Flappy strappy and the sticky Audy bits and all the the nonsense you kind of associate with the bag yeah it was good uh you know and it was what what is it an USI or whatever

The it is a it is a better than most a like just looking at it better than most um we did test a couple other brands that did not perform as well they were not better although even they were not much worse sure um um but yeah I was

Really surprised about that and then if you do clean up all the Flappy strappy and the other nonsense um it it got even better so that yeah that that was a good one I mean the other one and I I think he covered it in his video but and uh

This is one where you you just don’t know exactly although we have theories but you know putting a big water bottle in the middle back pocket was a huge savings for him and the question sort of became you know how much of that is that it’s further dewrinkling and and you

Know he’s wearing a pretty tight like racing suit but to put that bottle in the back pulls everything Tighter and you know probably reduces your a a little bit by kind of squishing you inwards uh even a super fit guy like Dylan but you are also filling in that

Dead space behind the rider’s body with a you know a shape that’s worse certainly you know better than nothing um and so that’s one that you know I’d like to look at further from a number of of angles of like what what are the contributing factors here and what what is the

Causation um of bottle in the pocket is it that improves the clothing or is it that having something in that space um actually helps clean clean flow up a little bit um there’s a ton of ton of stuff like that when you get a result like that that is kind of

Unexpected does that generally send you back to retest and see if there was something that is wrong or do you accept that it was like oh well I’ve just learned something new it seems like there’s a fine line that you’re going to have to walk with

That yeah so we over the years I’ve sort of developed this um kind of this schematic of you know we you always go in with like you know I think like I think we took Dylan in with I think 25 runs that we wanted to get

Done knowing that there’s no way to get 25 runs done um but you know as I told him before I said you know the thing with the you the tunnel it’s like you know what’s the the World War II general you know like you know planning is everything but plans become useless

Quickly right or you know it’s like Mike Tyson’s like you know everybody’s got a plan do you get punched in the face um and so we knew like hey this is all the stuff that we want to do and we’re going to be prepared to do it but you’re gonna

Find things or run into things that are going to shock us and then we need some sort of decision criteria on what to do next and so you know I tend to like put a number out there that’s like okay if it’s if it’s an improvement you know

Greater than this amount right and we can all choose between us to agree on the amount maybe we break the plan and we go chase it or if it’s some shocking result in a negative way maybe we stop to figure you know you’re going to instantly you know push in the the

Control product to make sure that the tunnel is not bro broken or something um and then what’s the decision criteria for like do we try to figure out why that’s super bad or do we just say it’s bad don’t ever do that and we’re g to go

To you know let’s get back to what what we’re here for um so yeah something like like that water bottle was was interesting um it it made a notably positive result above our decision criteria and we said oh well if you know if if a standard water bottle is good

You know Dylan had these like I think they’re zepl or something these like oversized they’re extra long and they’re a little bit like larger diameter bottles and um he had one of those like well okay if if that’s good is a bigger one better and we put it in I think it

Was maybe slightly better still it still didn’t really answer the question but at least now we know you know it wasn’t it wasn’t a one-off um but yeah there are times where where you know you’ll certainly do something uh you know you you think might be good and it’s

Terrible and you have to go okay do we really want to do we really want to sink time and energy into figuring out why this is terrible or do we just keep with the original plan you know strike it off as don’t do it um and and keep with the

Plan so you know and that like back to your planning question of of earlier fatty you know I think it’s things having things like that in place before you walk in the door probably get you another three to four runs out of a day at the tunnel MH um which certainly

Helps maximize your spend but you know as as you’re doing stuff you’re learning and you’re you know again every every run is kicking up 10 new run ideas oh well this you know this thing did that I what if we made it bigger what if we

Made it smaller what if we rotate it what if we you know it there’s this like terrible like exponential kind of like effect uh to every run and you can really get yourself lost uh in the excitement of that um and so yeah you you need to have that

Decision criteria of what do we Chase what do we not Chase and and then I mean really the most important job there is who like the data collection person because you need them collecting observing and collecting the data and making sure that like everything looks you know all of your

Your other data points look you know nominal right all your pressures and temperatures and and that stuff um but you then also need them taking notes on you know what happened what did didn’t happen what idea came out of this you know how you might want to test that

Idea in the future and so that person through a day at the tunnel if you’re doing it right that person actually generally stays really quite busy um with you know essentially this massive bookkeeping um and then everyone else is running around pretending to be F1 pit crew I’m so intrigued by the planning

Piece of this and now we’re kind of in the execution piece and you’ve mentioned a few of the roles but kind of fill that picture in what are the roles being filled in the actual wind tunnel you you’re talking a little bit about I mean I assume there’s someone who essentially

Is the crew chief the customer who who is there to learn something but you’ve got someone who is taking notes someone who is observing and it sounds like there are people who are actually moving things tell us what the I mean what are the roles being played and filled um

Both by the wind tunnel crew and by the customer so I mean typically there’s like one to three wind tunnel technicians you people who work there and they know the tunnel and they’re working a control board you know full of uh you know feedback and and numbers and

Gauges you know some analog mostly digital anymore and they’re really just there making sure that the tunnel is performing in the way that the tunnel should be performing and that nothing is getting broken or being pushed too hard or um you know any of that stuff they’re

There to troubleshoot if things do start to go wrong um and and also I’d say they’re generally playing some sort of safety role because most of the things that you as the customer want to do to maximize the runs that you’re going to get are are

Going to be in some way less safe than doing it the way they want you to do it and so I I I don’t think I’ve ever been to a tunnel where I haven’t had some sort of pretty serious negotiation with the tunnel manager of like okay no like

My people are are pro and we do this a lot and like I need you to you know kind of we talk about like the the safe speed for Crossing into the air flow in an open nozzle tunnel like that’s relatively negotiable as far as I’m

Concerned you know and like my my people aren’t you know high school kids on a field trip they’re gonna gon to walk across the threshold and get thrown down into the tunnel by a you know 15 M hour residual wind speed that’s that still exist and and you know there can be real

Danger there right because you think of like in an open nozzle tunnel you have essentially this this Giant volume of air that’s moving through the open nozzle and there is a wall I mean you it’s really cool you can stand there and take your finger and just start moving

It forward and all of a sudden I mean it’s even more pronounced than like sticking it out the window of a car because it’s so laminer and it’s so clean and smooth I mean you literally go from the front half of your finger in a

30 mph wind and the rest of your hand is in still air um and so you know there’s a huge safety component especially you know they’re running race cars in there it’s you know 120 M hour uh air speed you know if you accidentally walk into that threshold you’re going to get torn

Up um in a big way but if the speeds were running um you know they don’t want you to cross it until it’s like eight and I’m more like you know we’ve done this a lot I I need that to be 15 because that’s going to save me 30

Seconds on every run and um you know things like that and so you know we do things like like if we have tools um you know we tend to have them like like like coat we use like a lot I would say we use a lot of electrical um tooling that is like

Rubber dipped or coated for insulation but for us we use that as a we have a wind tunnel toolkit and all of those tools are essentially um much safer and less damaging if you drop them right and so like that’s a thing I use you know

Like if if I’m talking with you know Henry at the at the arc you know Henry you know us you know we can cross it at 15 because we’re a good safe crew you know we’ve got drop safe tooling um so the the risk is down and you know and

Then I think once they see you operate for a while you can start doing more stuff like that so um but you know like Arc actually has a a a moving floor it’s called Rolling Road and so it’s basically like the biggest treadmill belt you’ve ever seen um and you can’t

Use it for for bicycles because it would just put way too much friction and noise and everything else in the in the data um and so it’s bridged with this sort of bridge that has the the load cell balance in it but it’s still exposed everywhere and so you know like dropping

A a you know even something like an adjustable wrench and having it land on on that treadmill belt I I mean that’s it’s probably a $10,000 belt um and so you know they’re there really thinking about stuff like that right all you know the balance uh that lives inside that

Bridge piece it’s a quarter million dollar balance and so you know if a sticker says Don’t Step here like you better not step there so so that’s what the tunnel staff are doing the you know you’ve generally got someone like me who’s doing a lot of work but also

Negotiating things like that like that you know let me cross when I ask you to cross um you know if it’s a closed tunnel they they generally are much harder to negotiate with because there’s some automated system of like nope you know San Diego has an automated system that

The door is magnetically locked down to a certain wind velocity and it’s all automated and so you know you can negotiate all you want and they just like like tough um you know you’ll you’ll have typically one to two technicians who are the ones doing the adjustments or the movements

Or you know some sometimes it’s as simple as you know they’re standing there holding the helmet that you’re going to run in the next one and they run in there and they give the person a helmet and they take the old helmet and they get the hell out of there as quick

As possible um you know sometimes they need to be quite mechanical because you’re you know testing risers in a an arobar setup and you know a lot of these modern arobar setups like they’re not easy to get apart and put back together quickly and they’ve got lots of small

Pieces that you really can’t afford to drop in that environment and so um you know I’d say I’ve seen even as much as we did a test some years back when the um original shiv was out there and we had all these different machined angled risers in these bolts and would

Literally went out there with like a like a big beach towel and would like kind of unfurl it under the bar and then have somebody basically ready to catch what dropped um because you just can’t you know it’s a thousand dollar an hour um and so you know you don’t want to

Spend half an hour looking for that you know that screw that dropped and fell into that crack oh yeah um or have that shut you down all together so and then yeah you’ve got your uh your athlete right and and it is a hard day for an

Athlete to be in the tunnel and you know it’s funny you got a guy like Dylan like oh Dylan like this this is hard it gets old you’re only running you know 200 watts of pedaling but it’s it just just beats the crap out of you and you know

He’s kind of like oh man you know I’ve done Leadville and Unbound and you know of course by by like four o’clock on that day he’s like man I feel like I’ve like lost a fight against a gang and uh it’s just it’s just hard and and then the other

Thing that uh you don’t always think about but you know it’s typically 70 like you go for like 72 74 degrees controlled um but that’s cold old and so you know generally whatever you’re like if you’re able to leave the athlete out there while you’re making changes you’re

Also generally running out with like a jacket or some you know like a a throw that they can wrap around themselves um because they get they get quite cold through the day of doing that but you don’t want to test them with any sort of warmer clothing because that’s that’s

Generally not what’s going to be used for their event you know like um that was a big that was a big Lance Armstrong thing in my experience you know he he really hated the cold and so he would really fight to like like no like let me wear the leg warmers like dude

The leg warmers change a lot and and the problem is we don’t know what all they change and so you know if you’re not racing with the leg warmers you shouldn’t be wind tunneling with the leg warmers um but I I’ve seen him in more than a number of instances get get to

Keep the leg warmers and the arm warmers through the test you just go well I that’s some level of uncertainty that you know hopefully it’s not doing anything but we have we have no idea and we have to admit that so i’ would say that’s that’s your general wind tunnel

Crew and then probably the most important person there for the day is the person that’s getting the lunch if we’re really honest um you got to have some good snacks and food and coffee at lunch because because the the crew that’s really doing the work I mean it if it’s

Happening properly it’s probably every 10 to 13 minutes it’s like a fire drill you know I mean it’s just chaos for five minutes and then it’s 10 to 13 more minutes of waiting and so it would be tough to be that focused knowing that if you if you

Drift if you sort of you know check out for a second you know that that cost you a couple hundred bucks right I mean that that is rough for sure yeah for sure I mean it’s uh cuz cuz other things too like you know you lose a little bit of

Focus and get a little bit sloppy M and and I would say this probably happens at least once at every test and if if there’s humans involved it is guaranteed that you get sloppy and you know everybody runs out of there and you close the door and you start the wind

And you look up and go he’s still wearing the jacket you know you got to hit the kill let it spin that right and you like oh we just burned half a run because we forgot the jacket or and the meter doesn’t stop running right I mean and

The meter doesn’t stop running yeah yeah and and you know too there’s always a guy like we always run a board and the board has the run and you know you write the Run number and you write the change or whatever the the test is and I would

Say that’s another one that um is just so easy to screw up right because you know the person doing that needs to be really consistent like okay I come out of there and I update the Run number and the run to the next run and you know

I’ve seen it so many times where like they forget to do it and you look up and you take the picture and go oh oh wait I I thought the last run was run 36 and then everybody’s going what Run is it you know and the guy with the boards

Going oh I think I changed it but I we’ve done 36 of these I you know like everybody’s like panicked and then you’re like back to the the data guy who’s running the the the spreadsheet you know what run was the Run we just ran and but but they’re usually you know

They’re in their world of taking notes on you know what to look for in that run when you post-process the data after the fact and so it uh yeah wow it can it can be a good time but it really I mean it I probably make it sound more chaotic

And interesting than it is it really it’s it is long days of standing there staring through a window at a person riding a trainer for 13 minutes at a time and that’s the tension right I mean it’s intense and expensive and also tedious and trying to trying to I guess keep

That tension in Balance has got to be a tough thing to do you know staying focused when as you say you were just watching someone ride nowhere for a full day yeah at an incredibly High rate of money and if it’s it’s even worse right it it the changeovers and stuff are

Easier but with products I mean you think it’s like a a bunch of people standing there watching a a a bicycle through a window like oh yeah and there’s not even anyone there yeah here oh and then I yeah oh yeah go ahead no I just we’ve

Been recording for an hour and 14 minutes we’re B maybe halfway through our questions I think you were absolutely correct that we’ve got two episodes worth of questions here and this is as far as I’m concerned like one of the most interesting episodes we’ve ever done I’m loving this I don’t I

Don’t want to just give it short drift I’m trying to figure out I don’t know hotti what what are you thinking do we just keep I I I I have day job stuff I got to go back to it some point but I don’t know if we like say you know try

To get to some good break point and say we’re going to make a part two of this or I don’t know first we got to do bicycle we have to get bicy done right now we should just jump to that right now and then I can and then what we do is

Um weord move that up I’m sorry go ahead yeah I no I can find some place to put it yeah that’s not that’s not a problem I could find somewhere to to cram that and actually on that one could could we could we get him moved up earlier in the

Episode ahead of perk because that was he was the one that yeah we’ve done that the last two times did we do that on the last one at least the last one I think as long as we’re alternating them right exactly I know we’ve done it we have

Done it at least once Josh and if not maybe twice I’ve moved him up to yeah yep yeah I caught your email on that so yeah so that has been done awesome but they’re they’re fairly movable usually so oh yeah so fatty I would say we jumped um sorry bicycle and

Then record that in the close saying you know with a caveat saying hey we’ve we’re going to do more of this we’re GNA have more wind tunnel this is turning into our you know our our record part do and and in a way I mean we we sort of

Started this part two here you know what are the marginal gains to be found and I’m afraid that I drifted back into Day in the Life In The Wind Tunnel because I I find that incredibly intriguing and can’t I thought that was awesome I love the yeah I love the direction that took

I thought that was that was perfect so um yeah you know what I think we should do go ahead go ahead let’s let’s run the bicycle I think announced like the change over um you know now it’s mythbusting wind tunnels maybe do we ask the question and

Then before I can answer go like guys this show ran two and a half hours so you’re going to have to get the answer in the ne you know part two or whatever um you know sort of like Cliffhanger it and you know and we can you know like topics

In part two how do Liars lie with wind tunnel data and what about the marke you know the tension between marketing and Engineering I I think that would be a heck of a way to end the episode and yeah and give us a a great setup for the next one so

Do maybe do I don’t know do we have time to maybe just try to finalize or finish this what’s the day in the life of the wind tunnel and then say you know we’re going to come back to more more about marginal gains in the Wind Tunnel what you know let’s

Do that and so we’ll do what you’re saying do Post yeah we can do like the post-processing part of the day in in the life and then end the episode with a setup for the next one I I think that’s perfect this is going to be sophisticated edit for you hottie um

Yeah you’ve lost me there I don’t know where you guys just went so we’re going to record what next so we’re going to do bicycle then we’re going to go into the the post-processing part of of Fatty’s day in the life like so you know we’ve

Come out of the Wind Tunnel you said there’s two hours for every one hour in the tunnel what’s that look like we can talk through that that’s five minutes seven minutes it’s me I talk a lot so it’s maybe 10 and then we close um and

Then we close and with a setup that there’s going to be a we we’ve gone way over we have this and this that we want to talk about but we’re going to close now and th we’ll pick this up in part two right yeah okay

Okay so did I break in in a place where yeah it’s always hard to say because I I I thought we I know I I sort of broke into mid-sentence I didn’t mean to do that you know lagginess but no it’s good I I think I think we were right at ready to

Read the mid roll okay well let’s go ahead and let’s go ahead and get to the bicycle then well this is going to go in sounds like I’m going to be moving this but maybe not okay I think you’re right think I’m moving I don’t know what I’m

Doing if you take bicycle and move it up higher I I think that I mean that’ll make them happy and we can do that so this is this piece is modular yeah and then okay I better edit this tomorrow because otherwise yeah this will be easy easy to get confused on

Yep okay here we go mid roll bicycle well wind tunnels and time trial bikes they kind of go together don’t they you know I’ve not owned a time trial bike and I don’t know a few years at least but if I needed one the first place I would check is bicycle bicycle

Is the largest global Marketplace for pre-owned bikes you know I do have a time trial like I have a first generation shiv we’ve talked about that a little bit I had to sell that on bicycle cuz it’s just sitting there in the garage I have not ridden it in a

While I someone would want it right bicycle has over 15,000 pre-owned and refurbished bikes available they connect buyers and sellers on a global scale hottie if you want a time trial bike unfortunately I cannot sell you mine because you I what a foot and a half

Taller than I am but bicycle has almost 400 triathlon bikes and surely someone out there is going to be able to help you out with something in that category in addition to Triathlon B oh God first generation shift with first generation di2 I think I think I got the charger

Somewhere and addition to triathlon bikes bicycle has hundreds of road gravel and mountain bikes there’s also an other category where they have ebikes City bikes and comfort bikes yeah bicycle they have it all with guaranteed buyer protection secure payment first class services from carbon checks to refurbishment Bicycle

Believes that your next dream bike does not have to break the bank and can still provide the quality and performance desired for years to come buy cycle is also sell cycle when you sell a bike bicycle provides a bike box and packaging materials directly to your door pickup times are coordinated with

Your schedule so you can hand off your bike in person and for a limited time you can enter the code mg23 at checkout for free shipping on your next bike P purchase so if you are considering pre-owned check out bicycle.com that is spelled buy cycl e.com and find your next dream bike for

Up to 70% off retail all right and you will just pick that up and place it where somewhere where it can fit and so we’re going to just go ahead and say okay now we’re done or I can I think I can probably just go ahead and

Do this we’re going to do a little bit about after you know the day after you’ve been in the wind tunnel and what you’re doing with that and then we’re going to say hey we’ve got a lot more and we’re going to do a close and we’ll

Do part two that sound good okay I think we can go from there sounds good okay Josh we’ve planned our wind tunnel experience we’ve collected you know 10 zillion points of data and now you are back at Silka HQ or wherever and you are looking at you I don’t know how

Many terabytes data that you’ve collected how do you start to analyze that oh man it yeah this really is the the hardest part and the the least fun part by far um like I said it’s it easily two hours of of post-processing for every hour you were actually in the

Tunnel um and that’s where you need somebody with the right the right mindset to to really do this I am not good at this job because I you know I’m like oh squirrel no I just don’t have the uh the focus to to do this job but

Fortunately I have people who do and um yeah you you think you know each each run is essentially a line in a spreadsheet that is giving you something like like 256 give or take data points um from The Tunnel right and so from that you’re manipulating that and

Calculating um you know to get to your your CDA um you know your side force your you know your lift force your yaw Force depends on exactly what you’re looking at um but then you know we also in a lot of cases like if we’re working with

Athletes you know you bring some other equipment to the table you know repeatable body position um is is one of the the biggest hardest ones when you’ve got athletes involved and so you know there are sensors out there and my my friend pork mlen is you know probably

One of the best in the world this of you know using sensors that you know essentially give kind of like live readings of uh you know the angularity of of your torso um you know looking at at the writer’s breathing um you know we’ve we moved the elbows in like what

Did that do to his the respiration rate um you know looking at uh in every run you’re taking photographs generally from sort of a top front and side uh and then you are able to you try to line up with that Row in the data set from a

Standpoint of um you know if you you did have something weird like okay is there anything in the photos that we weren’t catching in real time um you know typically one of the people there if you’ve got athletes is you’re projecting The Athlete on a screen and you’ll

Actually use a like a dry erase marker on the screen to kind of hit a couple of key body body parts to make sure that those aren’t moving during the run but then also like as you’re changing things you can use the prior Mark to kind of

Show you the difference um and so we always take a picture of that screen as well um but you know you in there you might find things like like oh this run was an improvement from that run but look the rider changed the way the hands were positioned or the way the fingers

Were on the front of the bar and so now you’ve got to mark that run is is possibly suspect or possibly something for a future test you know was was it the extra spacer we put in there or was it the hands that actually made the difference

Um or or was it somehow the combination of the two um that’s in there and so this post-processing piece is really just tedious work where you’re ultimately trying to get to you know what are my key variables you know if it’s a a product uh we’re looking at

It’s you know like a wheel for example they’ve got to be where you know obviously drag is important but you’re also plotting and graphing side force uh drag side force and um steering torque values uh against wind angle right and so you know somebody’s making a every

Run is you know maybe you know six or eight plot points that get plotted out you know make sure the Run num is in there make sure that accurately connects back to what the change was look at the picture to make sure that that’s really

What that was you know and and so on and you know I mean some of these you know 50 hours at uh you know the lswt you we might get aund 30 to 150 runs and so all of that has to be plotted and accounted for and then you know you typically go

Into you know sort of a that person is going to come out of that that person is going to come out of working with the data with a lot of ideas and thoughts and and sort of hypothesis about what’s happening in different things but you’re going to go into a meeting and you’re

All going to kind of look at the data and see you know ideally that person’s not saying anything other people are going to look oh that’s interesting that’s weird well you know what do you think is happening and everybody kind of forms uh or the way I like to run it you

All kind of form your own ideas and then we start to compare them and that avoids that sort of like like overly powerful group think that happens where you know somebody sees something and then somebody else kind of agrees that they see it too and then other people start feeling pressure

To see the same thing even if maybe they didn’t actually think they’re seeing that or feeling it um and then from there a lot of times you can go back in and try to test other things in the data uh or or look at other elements within

The data um to see if that looks looks like that may or may not be true you know or or in the case of weird outliers sometimes there are other things that um you know like like oh wow we had a a really low drag at high yaw with um with

A low side force and a low torque that that doesn’t really make sense right right like to to have low drag at high yaw you have to have high lift um on the leeward side of the wheel and because the wheels are never perfectly symmetric you you typically have both high side

Force like a sailing effect and some sort of you know either torque variability or just high torque in a direction and like if you’re not seeing that maybe you go back into the data see like is you know am I missing something is it is it acting on a different axis

Um you know was there some sort of data drop uh you know you do get aile that’s I mean like gigabytes that has like the literal raw data in it and so you can uh in most cases at most tunnels you can trace things back to almost the sensor

Level and look at oh well here’s our 30 second average at you know 60 Herz or whatever we were taking at but for some reason there’s like seven data points in this run so like data got dropped and we probably can’t put too much faith in in this particular number and so there’s

Just a ton of stuff like that um and so yeah you you generally by the end of all that you come up with like okay here’s the two or three things that we think we’ve learned and we hope they repeat the next time we go back here’s the

Handful of things that we think we see some directionality or we think we see some you know potential improvements that we would like to test in our next run and and maybe the best thing you you can take away from some of these is to say wow here’s some things that we

Thought would work that really didn’t work at all and were terrible and so now we can kind of put those you know over here on this list of stuff to not do again because we don’t want to pay money again um to see you know the thing that

Didn’t work um and and yeah from there you’re you’re either you know particularly with athletes okay well now we’ve got to take the stuff that worked and the stuff we liked and agreed upon and we’ve got to make that positional change with the athlete and they need to

They need to tr train with it for a month um and then that really kind of goes on and becomes the the problem of the coach uh you know or or somebody generally within the team um if it’s World Tour you know there’s a lot of the

Teams we work with now have technical directors oftentimes that falls on them to to Really push the writer to train the position to try to be able to sustain it um and and appreciate it and make it work um because I I cannot tell you how many times

We’ve done all this all the work all the thing and you know a week later like ah the writer really didn’t like it and so he’s changed back you like guys we just we just set this you know again this Briefcase full of cash on fire to

Make this happen and and you let them change it back in four days like that that’s crazy um and so you know there really does have to be a broader system in place around you know probably that final step is the actual like successful implementation and belief and buyin uh

Of the team and the rider uh to to ultimately get the thing to the literal Finish Line uh at the end man and then the season ends and then you start planning how you’re going to do it all over again next season time in the Wind Tunnel begets more time in the wind

Tunnel is kind of what what I’m hearing it is so true it is so true that has uh something in common with this episode of the podcast uh hotti I don’t know if you’ve looked at the runtime but I have we have been recording for I don’t even know how long

And we’ve gotten through I think I don’t know around a third of the questions that we had put down so uh what do you say let’s pick this up in uh the next episode of the show okay because there’s a lot of ones and zeros here that I get to you know

Comb over and make a show out of absolutely lot data it’s wind tunnel data just in audio form you’re the it I think I think it’s very analogous to the the post-processing you you do here right I I get to show up and blab for an hour

And typically you guys have spent like I don’t know how much time and couple hours writing the script and what do you spend couple I mean easily it’s got to be three or four hours doing the editing and the post-processing um it depends when it getss out of this out of control yeah

Then it can’t be it depends though like like it did here today yeah our pre-planning wasn’t the best here I guess we we didn’t quite understand how long this this topic could go on but what the hell we’ll we’ll we’ll wrap this thing up somehow

We’ll put a bow on all this I would say that uh having an uh a topic that is just too interesting to be contained in a single episode is a good problem to have um we’ve got a lot more questions and soon we will have a regular Ask Josh anything

Episode for your questions here’s how to submit them text or call the marginal gains hotline 317 343 4506 you can also leave a comment on this episode or any episode at marginal gains podcast.com for listening to the show and we will be back soon with more marginal gains questions and answers so

Thank you all for Listening