Even with doing VE before owning this thing I used to find that having a steady constant grade if you're going to have one worked better than having a route dotted with "pimples" along the way of little rises and falls of barely any consequence.

The out/back nature would make sense for a constant grade out and opposite of that back. The pimple or mole hill scenario mid-route I would think would be of more concern.

With these sensors, the wind/elevation "cancellation" for an avg shouldn't be too big a deal since it measures it. The issue with it with VE before owning a sensor was not knowing the instant wind scenarios. So I'm going to actually see if it can work for me to do a false flat downhill so I can do infinitely more testing at "10 mile race pace" without being limited to the 22ish minutes of work at that power I can handle. I'll just play with it and see what I can see. If not, just winds up being a good Z2/tempo workout.

I hate being negative on public forums as it is easy to be a naysayer without having knowledge enough to criticize. But....I read all these posts and go back to the original thread title "Aero sensors for dummies thread". The dialogue here is at a very high level and I can't use it, as I am as the title defines, a dummy about this stuff.

I still have the Notio, I have used it a few times this season. It really is useless to me if I am honest. I just don't have time (or the desire) to do hours and hours of testing. And then you have the hours and hours of computer time required after to get some use out the data. I'd rather just go ride my bike. It is not that the device doesn't work, it's just not practical for me. If I put a dollar figure on the cost of the device, the time it would take to do proper testing, I would be better off paying myself an hourly rate for that time and save up and go to "good" testing facility, either on a track or in a tunnel.

I am hoping something comes along that is almost as practical as a power meter, but I have to conclude as it is not possible. It's just too complicated apparently.

yes, some of this discussion is very "in the weeds". ST has a tendency to get in the weeds pretty quickly. Hopefully people will post high level questions that will get high level answers. Unfortunately sometimes the explanation to a high level question leads us right into the weeds. I apologize to my contribution to being in the weeds



Here is a fundamental difference between these devices and a powermeter. You go out and ride and you get 200w. It is really hard for you to question that 200w. Is it 196, 210, 180 ?.....you can' tell

But if you have a device, tthat gives you .282, repeat the test, get .264....repeat .272 and you know the numbers should all be the same but they aren't, then you are stumped. The possible inaccuracies of these devices are shown by these devices.

But I have high hope we will get to a point where we can just trust a number.

But I can guarantee you this, simplicity will not come until precision of these devices is there, because as long as imprecision exists, these devices will put it in your face.

Also remember, until Coggan wrote his book, how useful was a PM to most people ? We will get there.

Training and Racing With a Aerometer

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

IMO there has been way too *much* focus on getting accurate numbers while "just riding along". This would of course be nice and a manufacturer would have a large market for something like this and make big $$$; but we are a *long* way from achieving that.

Meanwhile we *could* have a device utilizing the best testing protocol, that generates and analyzes laps automatically, adjusts calibration automatically, measures and uses air density and temperature in real time, calculates CdA at the end of each lap (and will also record it as a variable with time), checks for anomalies, etc... making the process of field testing a whole lot easier and seamless.

Would it be worth your time to spend a couple hours solving the "A vs B" question with <1% error? Or is that still not worth it to you?

The CdACrr App by bugno (Pierre Ledac) came the closest to achieving this, and it's practically free and uses a $70 anemometer. I have to believe that an organization with resources, rather than one guy developing this part time as a hobby, could come up with something better.

I have seen you use this term "just riding along" and the acronym JRA. Where does that come from ? I have never seen a manufacturer claim that as their target but maybe I missed something

I think he’s talking about the difference between getting good numbers in a controlled environment, which we should be talking about, vs. JRA, which we are actually talking about.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

Sometimes people don't ask the right questions, and asking for "accurate numbers" for an "inaccurately defined" condition is basically one such instance. If you're not in a stable position, in stable winds, with consistent road quality, then even if you were to run an experiment long enough to where the estimate of the mean is very close to the "true mean," the variation is going to be so high that the number doesn't represent something very useful.

I think that engineers that focus on aerodynamics can be too focused on designing experiments as something of a black box from which they can get data, while the more meaningful improvements come from understanding why something "tests" better. Why does nobody ever talk about what it "feels like" to be aero? I think the answer is because people don't think about it enough, or maybe they don't know what to "feel for" besides something like "no wind on my belly." I can say that I actively seek the feeling of flow attachment on my lower back, and try to reduce the feeling of high shear on my upper back when I feel it, with the goal of achieving the former condition. This is not something that happens naturally or just by going to "position Y", I have to "hunt" for it with subtle adjustments. For that matter, most people could probably get faster improvements with traditional tufts and a camera looking at their back than they would with a ton more sensors.

That was kind of the situation with power meters 20 years ago. We rode with a speedometer, RPE, and sometimes a HRM but we didn't know what 250 watts or 350 watts or 150 watts felt like. One of the most important calibrations we had to make was to figure that out. I think there's some possibility that, over time, some riders will feel what 0.05 m^2 difference (either up or down) in drag area feels like. Maybe.

RPA ? Rating of perceived aero ?

For the distances I race (Ironman), 10w made all the difference... it was a 3min time savings riding at 200w, and I ran well. If I rode by RPE at 210w it would mean I walked the 2nd half of the run.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

I don't mean RPE or perceived drag, I mean that people don't seem to note, or at least share what the flow feels like when they are in an ideal position, as in my example. That the responses so far seem to ignore this point seems to confirm my suspicion.

This somewhat abstract notion can be captured visually with tufts.

For adjustments related to helmets, skinsuits, positions of shoulders/torso, etc. ultimately the goal is to achieve the best flow attachment, which can sometimes (more often than not, in my experience) be "felt" by the rider if they know what to look for.

Edit: maybe a few more examples would clarify.

If I feel high attachment in my shoulder blade area, that is good, (though wasteful since shear should be distributed as evenly as possible) but not as good as high shear on my lower back (or an even flow distribution on the back), so I'll adjust my position until I feel the high-shear flow as far down the back as possible. This may mean that I have to rotate my helmet according to where the wind is coming from, extend my neck, tighten my shoulders, some of which will have no effect, but it is a clear step change when it does occur, and it isn't something I could just do by "picking the best position I've ever held," it requires finesse.

I feel like you're going about it exactly the wrong way... instead, think of it as a series of binary decisions which you screen and evaluate thru testing. Eventually you'll land on what you're trying to engineer with your feelometer, which may or may not be accurate. Most likely not.

In other words, you have a hypothesis about attachment, shear, flow distribution, and your perceptions. You can test this hypothesis scientifically and definitively thru A/B testing using a power meter and an aero stick or the Chung method.

E

Edit: see my comment re: 200w and 210w above and then reread Robert's post in that light, a sort of horse to water situation (we're the horses).

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

I’m not suggesting to not test systematically, or without sensors. I’m saying that data which is not correlated with qualitative aspects and understanding of the flowfield is much (much) less valuable, and is sometimes misleading.

I’m suggesting the above not as a luddite who likes to feel things out without designing good experiments, but as someone with a lot of experience leading teams of aerodynamicists in wind tunnels and cfd :)

Although I personally didn’t know, I suspect that Dr. Chung took your experience into account. :)

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

He once responded to me that he prefers to be addressed as Robert here, lol, and we've had nothing but pleasant and interesting exchanges on this forum, but I don't know how much he knows about my background. I bring up my background only for context, and I wish I didn't feel it necessary. In a way, nothing is "necessary," but I enjoy the discussion and sharing the knowledge :)

Since I haven't really introduced myself, my first wind and water tunnel experiments using PIV were in 2009/2010 as a summer undergraduate researcher at Caltech. After graduating with an aerospace engineering degree and math minor from Cal Poly Pomona, I started and led the in-house CFD department at Team Penske,in collaboration with Ford, and stayed there for about 5 years before moving on. While there, we ran 50%-scale model wind tunnel sessions of 10-14hr shifts for a week at a time, every 2nd or 3rd week, plus full-scale testing, along with 1000's of annual CFD studies. I trained people to design for CFD and wind tunnel studies, and was the go-to "surfacing" guy that defined specs. Afterwards I was Sr. Aero (Co-lead) for Chevy's NASCAR Camaro ZL1 1LE, and led the initial studies for the 2022 NASCAR Camaro before returning to school for a PhD and becoming a broke (but happy) student again :) I had a life in art and product design before I went into engineering school too, which helps manifest the concepts. I haven't done any tunnel testing on bicycles, specifically, but my focus has always been on the theory and less so about the particulars of the test articles.

Cody

I didn't know the details of Codygo's background but it was obvious from his posts (and his questions) that he's had some experiences in either this or a related area.

I was agreeing that it would be good for us to learn how "fast" feels. I was speaking not just to him but, in part, to myself when I said that most riders haven't yet figured out "how it feels" when they're experiencing lower or higher drag. I can tell from testing certain things but not others (like, it turns out I can hear when my helmet is quieter). I think having a way to calibrate your senses with actual objective drag measurements can be an important and useful skill. That said, Josh Poertner has often mentioned that riders often equate noise or vibration as speed, so our senses can be fooled. Years ago, Tom A. said that had he gone by feelings alone he would have thought his P2K was faster than a borrowed P3C. In both cases, an objective measurement of drag helped them to figure out what cues and clues to rely on. Feeling or not feeling the air ripple across my back could just mean that I'm a perceptual klutz. Years ago before I had a power meter I had no idea what 10 watts under FTP felt like, nor 10 watts over FTP. I think having that display has helped me, over time, to calibrate that sensation. I hope that a real-time drag sensor will eventually train us to recognize "fast."

Thanks. I didn't mean to put words into your mouth, sorry if I did.

I feel like, while your comments re: 10w and threshold are spot on, might we be approaching Heisenberg's Uncertainty Principle with real time drag data calibrating our feelometers? Moving our heads to look is going to change things. I haven't played around with that feature admittedly so I'm a bit ignorant on how things play out with noise and feeling and drag. Obviously I'll have to change where my head unit is.

OTOH, I feel like I did have a different system, a kinesthetic one where if I "felt" fast, ie. I was feeling the feelings of shrugging and turtling and it tested fast in the tunnel, I remembered those feelings from the tunnel and looked for them in myself on the road in races.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

Sometimes it's harder to speak casually about technical things. It's also understandable that people will interpret a question or proposition (could be related to theory or experiments), colored by their own experience. Here, we're all chatting it up and sharing ideas and it's not always going to be clear, and that's ok, I think :)

In a more procedural manner, we could say the following:

• Using a power meter, speed sensor, a camera, tufts on jersey, Aerolab, etc.
• on a road with known elevation,
• on a day with approximately constant wind,
• record tufts during runs, and see if their behavior correlates with CdA.
• Take extensive notes on what shear feels like during each run; is the rider's sense (feeling pressure or aerodynamic cooling, sometimes sound) of "shear" reliable enough to "perceive" what the tufts indicate visually?

If the latter two points are true, then that is something one should focus on improving and building substantial notes. This does not have to be outdoors, and can obviously be done in a wind tunnel.


The difference between digesting information as outlined is that it is a theoretical model for low CdA: Flow attachment means increased stagnation pressures and therefore less drag, so we should seek flow attachment itself, and not to replicate a rote pose which is less likely to correlate with low drag in real-world conditions.


I think one of the main reasons I've excelled is that I avoid accepting numerical answers in either wind tunnels or CFD without a strong conceptual model of what is happening, particularly owing to the rather artificial conditions that are CFD and wind tunnel studies. There may be times when I must change my model based on new information, but I take the time to form a hypothesis and own it going into future experiments, rather than simply accept that "object/condition A" is some quality better/worse "than object/condition B". When I read about the voodoo that people consider helmet and skinsuit testing, I'm very much reminded how too many aerodynamics studies are agnostic of any underlying premise or theory, and are often binary results scrubbed of important contextual information that make them less useful for implementation.

Yeah. I haven't done real time yet, I only do post hoc, so I don't have to worry about spoiling my position by looking at a head unit.


Not only haven't I yet done real time, I do almost all my testing solo, at dawn. I'd never get my wife or my kid to record tufts while I was riding around.

If I had spent much more time in the tunnel, I'd probably be looking at tufts. I think that could be a good way to hone intuition. Maybe Marcag can get his drones to follow him around and do tuft recordings.

Been there, tried that, drone active track is limited in speed. Maybe DCR can tell us if anything better than my Mavic Air 2.

I have failed at communicating again :)

I mean to suggest that a camera (like a gopro style camera) can be mounted somewhere fairly benign, perhaps on a small boom rearward of the seat, recording the rider. I don't think recording tufts from a non-rider reference frame will be of much use. Wind tunnels have their own networks of cameras so this wouldn't be an issue there.

Regarding head movement, I would suggest that there are distinct flow sensations that aren't trivial to achieve... so riding around before a planned test run may help hone this sensation, which one would maintain constant during the run time as one traditional position without moving around mid-test, provided the usual test safety/sanity caveats :)

This drone seems to work rather nicely!

Yes, at that speed. When I go over 24ish mp/h sometimes I lose it. I do all my training and testing on rolling hills. Gravity is my friend on the downhills.

That being said, my drone is a year old now so I suspect it's time for an upgrade

Great video on their ability to track you https://www.youtube.com/watch?v=OHs9xgb9FeU

I chose the Mavic Air 2 based on it's ability to carry around an aerosensor, garmin and other baggage.

Mavic Air2 is said to be able to horizontally move at 19 m/s in "S mode" (that's 68.4 km/h). So it means if you want to film aero testing in the 40-50 km/h range you are able to film some of your runs, no issues ?

I understand you use the drone for more complicated stuff... but just curious about this simple application ;-) .