Kontact 

Answer: How to design a test that favors the tester's product.


"Drivers who switched to Progressive saved an average of $550." Right. And who's going to switch if they aren't going to save money? Ask the question the right way and you'll get the answer you want.

chaadster

Well if they were laid by an older, more experienced machine...

KBentley57

It doesn't tell us anything.

That's a textbook example of what someone thinks is a scientific test, but is actually junk science. There's so many variables that aren't controlled for, that you can't make any useful conclusions.

woodcraft

"Tighter, please... fill my voids."

SethAZ 

That sounds like bulldoodoo to me. Pure, unadulterated bovine excreta.

woodcraft

My comment, or theirs?

SethAZ 

Theirs.

woodcraft

I was commenting on the fact that they wrote taught, when they meant taut.

The machine-laid carbon fibers does sound like marketing spin.

SethAZ 

Yeah, I smiled at your catch of their word mixup. I appreciated your joke. But yeah, whether their machines made a taut layup, or they created some AI to impart a little of its "machine learning" to their carbon, the idea that their tires conduct heat so much better because their machines produce a more taut layup just sounds like BS.

I'm no CF engineer, but about the only way I could see one layup conducting heat better than another is if one layup has voids in it and the other doesn't, or one layup uses less resin so that the CF fraction of the whole matrix is higher. Which I doubt.

woodcraft

One carbon fiber says to another "Wow, we got machine laid and now I'm going to have your baby."

The other replies "No way, I know you're pre-preg."

SethAZ 

The first carbon fiber knew it was just a ploy because he was vacuum-bagged for it all.

woodcraft

Let that be a lesson for us all.

chaadster

Check out FSE's filament winding construction, which both eliminates voids and reduces the amount of resin needed.

SethAZ 

Wow, that's very interesting. I knew that filament winding was good for round things like pressure tanks, tubes, things like that. I'm curious to see how they get the winding done on a closed loop like a wheel. Typical filament winding has the mandrel spinning, but you couldn't spin a closed loop mandrel this same way. Obviously they have something figured out. Maybe they wind it more like a tube and then join the ends together when they mate it up with the tire-seat section and join the primary wheel body and tire seat body into one piece? I'll be interested in seeing a cut-away cross section to see what shape the primary hollow rim body, and how they are joining it to the section that can't be wound that comprises the tire bed.

chaadster

I don’t think uncontrolled variables were the problem, rather the narrow design of the test. Which variables are you thinking of?

redlude97

total braking power/work would be the obvious one

KBentley57

There's a lot

1) The work done on the rim by the brake pad.
2) Disregarding the fact that while they applied the same force to each rim, the coefficient of friction between the rim and pad remains unknown. This directly effects (1).
3) pretty much everything in here: https://www.bikerumor.com/2017/12/14...im-brake-test/
4) As everyone else has mentioned, wrong brake pad for each rim.
5) Disregarding that while the rims may have melted, a person riding them down a mtn wouldn't have been on them as long, because they would have removed more energy, slowing them down more quickly. They 'overworked' the rim per se.
6) Using 1200W as they did in the study is outside the realm of reality for that length of time. Someone would have to be falling off a mountain to require so much work/power be dissipated.

There's also the lack of depth in the study. They need at least ~16 rim tests to be sure the failure isn't due to random chance, and that's putting a lot of trust in student's t stats.

There's the lack of cooling on the rim as well. Unless I glossed over it in the study, there's no active cooling as there would be on the road.

There's probably a million more, but for all of these reasons alone, it's bunk. Could it be improved to the point where their points are valid? Probably not, but they'd need a whole lot more carefully designed experiment if they were going to convince me.

chaadster

That’s far from obvious to me; I don’t even know what that means! “Total braking power”divided by “work” is variable, or the result yields an uncontrolled variable? I don’t understand.

chaadster

I don’t think anything there is an uncontrolled variable. There are unonowns and things outside the scope of the test, but not unknown variables.

KBentley57

I think what he/we are getting at is there is no actual measurement of how much energy was put into the rim, by the brake pads. It's clearly different for each rim, because the coefficient of friction between the pads and brake track are different for each. He stated the same force was applied to the lever for each, but that's only part of the work equation.

redlude97

Either total braking power or total braking work done. This controls the conversion of braking force to heat and actually gives you an idea of amount of heat dissipation that must be performed by each wheel.

Dean V

+1
If they had measured the braking force/torque at the caliper the amount of energy being dissipated could be calculated. They already had the rpm (the other part of the equation).
The 1200w motor thing was meaningless as it seemed to hardly vary in speed. I think it may not of been loaded enough to bring it down from its no load rpm.

Jiggle

I think the lab screwed up the power input measurement to the motor. It's the one factor that simply explains everything wrong with the test. And it fits that a bunch of mechanical engineers would bumble an electrical something or other.

Dean V

They didn't say the motor was drawing 1200w.
Just that it was a 1200w motor.

chaadster

That was not the point of the test. Would have doing so made it a better test? Maybe, but that’s a different test. It was a bad test for a variety of reasons, but not because of uncontrolled variables.