GhostRider62

He said the model nailed it and one could ignore MOI.

The guy is a scientist and professor. Doesn't his language obviously imply both.

With sufficient instrument resolution and replicates, one can ascertain the nit on a fly's ass from the moon. MOI is meaningless on a bike. Period.

PeteHski

Maybe we should move on to the aero vs MOI effects of dead flies stuck to the wheel rims.

Maelochs

One of my jobs is doing man-on-the-street interviews, so I talk to a wide range of people. I therefore have heard, over the last several years, pretty much every conspiracy theory, alternative theory, crackpot theory .... at some point one cannot continue a discussion .... the other party is saying, (figuratively) "Ignore the science, the facts, the video, the personal experience, and ignore the obvious inconsistencies and and logical and/or physical impossibilities .... my favorite internet guru has the Real 'inside information'."

This is why I remind myself not to argue with Flat-Earthers.

Possibly on a flat Earth pedaling a heavy wheel takes less effort than pedaling a light wheel because the power pulses are damped .... and possibly the aliens which made this planet out of cosmic Lego blocks and live under the ocean in invisible cities, using telepathy to control our leaders, and going to eliminate me for revealing their secret.

So you can see why I am glad I don't live on a flat Earth.

Thanks to all the actual knowledgeable people who have taken the time and energy to make the physics and math clear to even a simpleton like me.

chaadster

Well now, Chung did not say that…

GhostRider62

I said that. In a separate paragraph.

Where did I quote him on that. In fact, I merely summarized and gave my impression of his statement.

I'm not going to dig back but the F1 professional engineer sort of made a similar albeit less direct assertion WRT the importance of MOI on bike wheels. Trivial. Is that better for you

GhostRider62

JP says rotating weight actually makes "no difference at all"

Summary for those who did not watch

On long famous climb in italy, the 1160 gram wheels were 50 msec faster overall compared to the 1600 gram aero wheels

On a 1 hour criterium, the $6000 Mallenstein 1160 gram wonder wheels were 700 msec faster due to better MOI and all of the accelerations out of the corner but they lost 20.5 seconds overall to the aero wheel due to inferious aerodynamics.

I would consider 50 mSec to be meaningless. Period.

chaadster

Well I’m glad that’s clear, but it’s still an inaccurate summation/take-away of his comments.

GhostRider62

Bull

You said I quoted him

That is a lie

An appropriate response would have at least been an oops, sorry

This is my interpretation redone

I took the time to summarize and quote another scientist. That was for your education. Not mine. You are welcome

RChung

Functionally. I think statistical inference in this kind of situation, where we were more interested in a practical result than testing a hypothesis, is sometimes overkill. The power meters were reporting at 2 Hz, and the wheels had reed switch speed sensors which we knew were imperfect (though we also had installed high-precision timing strips at several places around the velodrome surface). At a speed of 60 km/h, you're traveling 16.66 m/s, so the wheels are going 'round about 8 times per second, or roughly 4x faster than your legs are going round. So there's not much "pulsing" going on. What's a much bigger effect is that when you dive into the turns (which are banked), you lean over so the wheels are traveling further than you so the wheels are turning faster. Since on the track you ride a fixed gear this is a very noticeable effect and novice track riders tend to "float" the turns--experienced track riders know to apply power all the way through the turns.

I think we may have been the first to actually measure differences in effective CdA according to position in the team (i.e., whether you were lead, 2nd, or last rider). There were no wind tunnels at the time appropriate for measuring a team of riders. On the Canadian team that Olympics, there was a big difference between the "largest" and "smallest" rider (largest is relative--she was around 172 cm) so if you put the largest rider directly behind the smallest she got not much draft, and if you put the smallest directly behind the largest she basically had a vacation. But women's team pursuit at that Olympics only had three riders (now it has four) so you had to make one of those two choices. Team pursuit is a very interesting optimization problem from a modeling perspective.

[Edited to add:] By "nail" I meant to the resolution of the measuring devices, though that included the high precision timing strips. During some test runs we were getting very good agreement between observed speed and power and predicted speed and power when suddenly the predictions and observations diverged, but not by a lot. A lot of commotion, we though one of the sensors was off or, worse, my model was off. Then someone looked up and realized that the velodrome staff had opened up the loading dock door and the air density had changed. So we all took a coffee break until they were done and waited a few more minutes for the air density to equilibrate, then we went back to testing and the predictions and observations went back to what they'd been. That was satisfying.

Maelochs

Science----we made a model, plugged in numbers, made some predictions, then we tested in the real world, and the results pretty much matched our models, so we believe the theory is valid.

W#hen I was young and infatuated with motorsports (as opposed to old and interested in motorsports) I read a magazine article about when Ferrari first attached electronic sensors to their car' suspension. .... in the early seventies, i guess, sensors small enough, precise enough, and tough enough for the application were brand new.

The designer, whose name escapes me, said something to the effect that he had had absolutely no idea what was really going on with his cars. The tires were flexing into abnormal shapes, the various suspension parts were twisting and bending, the wheels relative to the chassis, the wheels relative to the track, and the chassis relative to the track were moving in ways his mathematical models never came close to prediction.

Interesting part is, he had a history of designing winning cars, and that continued as the data collection grew to obscene levels. I guess the guy knew what6 changes to the design and adjustment would likely have what effect even though he actually didn't know why.

Whatever. As far as I know he never designed a winning bicycle, so I guess he wasn't all that.

Interesting that a change of air density could have so much impact on the bike racers.... the sensors, however primitive, must have been pretty accurate.

terrymorse 

Pro tip: Do your time trial in the afternoon, when temperature is higher and the air density is lower (assuming it's not windy).

Air density is a big deal to a pilot with a small plane at a high altitude airport. They might be fine taking off in the morning, when the temperature is lower and the air is denser. But try in the warmer afternoon, and they might not survive. That has happened before at Truckee Airport.

RChung

The test of a model is how well it predicts the data you are going to see but the purpose of models is help us ask new questions and generate new theories.
The way I estimate CdA and Crr was a by-product: I was trying to figure out a way to determine how accurate and precise the data from consumer-level power meters are.

PeteHski

We were still struggling with F1 tyre modelling in the 2000s. Michelin were trying to develop a thermal tyre model we could use in our F1 simulator, but it was far from perfect. The rest of the car is relatively easy to model - including suspension and chassis compliance, but tyre modelling is complicated in this application.

RChung

I've been fortunate to have worked with a handful of riders attempting the hour record. The first time I did that, we collected the weather reports for the same period of the year from previous years so we could see how much the air density varied across time of day, then I made a guess about what the density would be on the date the attempt would be made. The rider had some flexibility in when they could make the attempt so as the date and time of the attempt got closer we refined the guess. It turns out that for various scheduling reasons a different rider had to make an attempt at a less auspicious date and time of day, and fell short. The rider I was working with could wait until a better time, and broke the hour record.

asgelle

That's one purpose. In my former line of work, models were used to avoid tests which were difficult or in some cases impossible.

Maelochs

While trashing other posters is the real reason I come to BF, I have to admit, this "learning" stuff is pretty cool in its own right.

Thanks again, all.

terrymorse 

In my former line of work, we made models to see if the design would pass qualification tests after the hardware was built. Failing a qual test was very costly in money and schedule slippage.