Rough Draft

So I’ve finished my rough draft of wheel selector, it’s live here:

http://chrisuberti.com/wheelselector/

Mainly the biggest hurdle I’m having is coding the damn thing into a website.

So right now what it is essentially a power estimator given there’s zero cross wind, no rolling resistance, and you’re riding by yourself (aka Ironman). It’s estimating pretty high on the power/work numbers based on my own rides.

What it needs to be a viable wheel selector:

  • Wind included (with inclusion of atmospheric boundary layer effects)
  • Wheel drag across range of wind angles (currently only running for 0 degrees, which means all wheels are pretty much the same)
  • Add rolling resistance
  • Time spent Drafting?
  • TT position
  • Refined bicycle rider CdA (this is pretty rough currently, the topic of it’s own blog post)
  • Add more explanation to what the hell is going on

Other things I plan on doing

  • Some sort of submit to database, involving entering your ride metrics (from power meter). This could help refine CdA measurments, which are particularly difficult to estimate
  • Better interface (ajax implementation, oh the confusing computer languages, why cant everything be like matlab…easy)
  • Step 2:  ?????????
  • Step 3: Profit!

Ultimately I’d like to incorporate something like this into what this guy did here.

But that’s a LONG way off for a side project.

I don’t want a silly plain power estimator, that doesn’t really give you much (plus that’s kinda the reverse solving process from what you need to benefit you as a racing cyclist). With this tool, you could essentially calculate your power needed for say a mountain cilmb, pick which wheelset is best, etc.

StumbleUponRedditShare

A Hairy Vengeance

I like to go through and watch what other bike companies are doing in terms of Aero-ness. It’s hit and miss, largely miss. However the funniest one I’ve come across is this following Specialized #AeroisEverything video:

If you want the bit I’m talking about FF to 2:25. They propose that shaving your hairy trunks will save 70 seconds on average over a 40km TT. “To put that in perspective that’s like going from a traditional round tubed bike to something Aero like the Venge.”Capture

Think about that, you can spend $3500 on a bike frame….or just spend $3 on a razor to shave your legs.

StumbleUponRedditShare

I am done with power meters

I’m done. Power meters suck. Firstly, they dysfunction and break faster than any other product that charges a $2k+ premium aught to. Second they suck the soul out of cycling, just ask Froome dog:

Frooming

Does he look happy? No. An the fact of the matter is that no one’s happy watching their power numbers because they’re always too low. That’s the whole reason you’re looking at your power numbers in the first place: to get better numbers (or to brag about how many watts you did or didn’t put out). It’s all about Marginal Gains, #AeroisEverything, not having any fun, and whatever else it takes to get an edge.

I’m not that type of person. Sure having a power meter taught me a lot about what it really means to do a constant effort. But in all honesty, the only time I genuinely use my power meter now is the handful of times during the year I do intervals (I don’t do intervals), other than that they’re just numbers to tell you you’re going too hard.

I just don’t care enough, I enjoy beers and Krispy Kreme. I always knew that I didn’t want to be a pro cyclist forever, I didn’t want to end up 30 and realize my life was dependent on racing kids for a few hundred bucks (at best). That’s why I applied to grad school, I told my team that I was thinking of going back to grad school and unfortunately that’s why I didn’t get on a team.

I’m super bummed I didn’t get a contract for next year. I applied to grad schools for Engineering for the winter semester, it was sort of a long shot with not a lot of spots available…and I didn’t get in anywhere. Then just as I was wrapping up talking to schools finding out why I didn’t get in I found out that I didn’t get a contract because the possible scheduling conflicts from grad schools. #FirstWorldProblems

I love cycling, I tried to not care one year (I even had a full time big boy job) and I had my best season to date and got a pro contract from that year. Because cycling my passion I think that’s why it’s never really sat well being my job. You’re not really adding any value to anything or anyone’s lives (sponsorship struggles to make sense via dollars), it’s important to admit to yourself you’re a pro cyclist because you’re selfish.

We had long talks in Belgium about what’s the point of all this. Pipe dream for us older folks is to make it to DII teams like UHC. What does that mean? More travel, more obscure races, slightly more money…then if I’m successful there, what next? Getting to be the C team for a DI team? It doesn’t end, so I’m glad the choice was made out of my control because I’m not sure I could have made that choice myself. Fact of the matter is you’re fighting a big uphill battle if you didn’t get into the system when you were a junior. I’m a late entry and I’m done fighting that uphill battle, I’m done having to put the rest of my life on hold for a few bucks and a title that a lot of Cat 1 and some cat 2 riders misuse anyway (and even that doesn’t matter, because to an outside observer, you’re making money therefore you’re a professional).

So now I’m back to a clean slate with literally no commitments for next year or even next month.

StumbleUponRedditShare

Assumptions to Estimate Work for a Ride

So in this model I’m building I’m making A LOT of assumptions.

1. You don’t gain any energy back from going downhills. That is essentially only works when you’re going uphill. Potential energy is a pretty simple calculation:

\,W = -\Delta U

 U(\mathbf{r})=mgh,

(yes these are stolen from Wikipedia, as will most equations be)

The problem with factoring your descent into potential energy is that you would essentially end up with Zero additional energy for riding over a hill vs going the same distance on a flat. Anyone who’s ridden a bike knows it’s a hell of a lot harder to ride 10 miles up hill then 10 miles downhill than riding 20 miles anywhere in Indiana. Also, most of the time when you’re converting your PE to KE you’re coasting downhill, thus doing zero work, and most of that speed is quickly eaten up by drag forces.

2. There is no wind. This is important because a lot of wheels have nearly identical drag data with no cross winds(within error of measurement). This is an important exclusion that really places a handicap on all the wide rims out there.

I do plan on adding wind data by doing some weighted averages, but that’s not going to be until I2.

3. There’s no drafting (Triathletes REJOICE)

4. Assumming everyone’s riding GP4000 tires. These tires are the most areo and most wheel wind tunnel tests are done using GP4000.

5. No rolling resistance, this is pretty minimal and constant, neglecting for now.

6. the Drag Coefficient Area is applied via an addition/subtraction factor. That is when you make a more aerodynamic object the only thing that is effected is the CdA in the Drag equation.

F_d\, =\, \tfrac12\, \rho\, v^2\, c_d\, A

So what I did was take a basic CdA from a crappy wheel (most rider-bike systems that I’ve CdA values for are old i.e. Merckx, Moser hour records). I take an existing wheel (like 32 spoke wheel) that every wheel company uses as their baseline calculate CdA for the crappy wheel, then subtract that from the CdA of the fancy plastic wheel we want to analyze. So usually you’ll get a negative CdA that is applied to the overall rider-bike system CdA.

One mistake I already came across was trying to add Cd values directly. The problem is a rider had a Cd on the order of .6 -.7 and if you just add your wheel Cd (usually below 0.2) you get some huge Cd’s, which overestimate not just Cd but more importantly the drag effects of the wheel (which should be VERY small).

 

Now I’m kinda BS-ing all of this from my leftover college text books. If I’m doing anything wrong let me know, and I’ll beg forgiveness of the Engineering Gods (Newton, Euler, etc).

StumbleUponRedditShare

Wheel Selector

Since I’m currently a cyclist in winter (aka lots of free time), I’m trying to build a little app that shows you which wheels are actually fastest over an entire course. What am I talking about? Well there’s a lot of different depths of wheels, usually trading of aero advantage for weight.

Mostly I’m curious where that trade off point is, or even if there is one. It may be that a 90mm wheel is ALWAYS faster, or that a super shallow lightweight wheel is the way to go.

In the first iteration it’s just going to take in rider metrics, and general stats from the ride like distance, speed, and elevation gain…make a lot of assumptions, then give you an estimate of required work (kilo-jouiles) for each wheel selection.

This will be pretty crude at first, you know, no drafting, no cross winds.

So far the hardest portion of this has been trying to figure how to factor in changes in wheel drag to the overall drag of a cyclist.

To do this properly I figure you need to break the rider and wheel into your individual drag coefficient areas. This way you can do sort of a component addition of which wheel you’re using (or more specifically an addition or subtraction of CdA to your overall CdA)

    drag

drag

                drag

Like this, except with bicycles.

The difficulty is that there’s a lot of data on CdA of riders (especially with respect to hour records), but nothing that I’ve found so far on drag on individual components IN THE SYSTEM.

There’s a lot of wheel drag data out there (BS or otherwise) but the way the air flows around the wheel with a bicycle and rider present is difficult to distill.

 

 

StumbleUponRedditShare