TT bike fit progress



frenchyge said:
... Anyone have a good technique for measuring frontal area from a digital photo?
Yeah, it's not too hard in Photoshop, you use Levels, Curves or the Threshold tool to drive all pixels to either black or white and then bring up the large pop-up version on the histogram and you get percentage of pixels which works out to percentage of area. The tougher part is setting up the shot with a good known reference area and minimizing parallax distortion. Since we can't get object plane telecentric lenses big enough for this application the best bet is to use a longer telephoto lens placed ten to twenty feet in front of your trainer but you'd still want the reference plane close to the plane of interest (the rider's head and shoulders) so you don't get a lot of geometric distortion like you would if you drew a square on the wall way behind your trainer. You'll also want to dress in dark colors and try to create a mostly bright to white background.

But even then you'd get A and although Cd is often swagged for modeling purposes off some known shape like a cylinder it's really a catch all for all the complexities of fluid flow. IOW, the frontal image is interesting and as a first estimate will probably help you reduce CdA but it won't really tell the whole story. Kirk and Andy went round and round on this a few times on the wattage lists. In the end it's not clear that the frontal silhouettes really have that much value beyond what you'd see in stills, a video clip or even a mirror and that it's probably not worth the time to carefully set everything up for a good capture or to process the area results.

-Dave
 
daveryanwyoming said:
Yeah, it's not too hard in Photoshop, you use Levels, Curves or the Threshold tool to drive all pixels to either black or white and then bring up the large pop-up version on the histogram and you get percentage of pixels which works out to percentage of area. The tougher part is setting up the shot with a good known reference area and minimizing parallax distortion.

-Dave

Or, he could just snap a pic of his shadow on a wall at sunrise or sunset for which Kansas is particularly well suited. :D
 
Piotr said:
Or, he could just snap a pic of his shadow on a wall at sunrise or sunset for which Kansas is particularly well suited. :D

Ugh... I had enough of the air density correction factors between apparent and actual sunset back in my celestial navigation coursework. :(

I think I'm just gonna go for something like this [ame]http://www.youtube.com/watch?v=xrixWmMoNR0[/ame]
 
daveryanwyoming said:
I haven't found a good long 'Chunging' venue since moving to Seattle but have used the parking lots and loop roads in the industrial park near my work for some position comparison testing.
You don't really need a long course -- in fact, there're advantages to using a short one.
 
Is this legit, how do you do it and can it be done on a PC?

[ame]http://www.youtube.com/watch?v=hql-CdbQF-s&feature=related[/ame]
 
flapsupcleanup said:
And I can't imagine doing a TT without a powermeter. The temptation to go out too hard is overwhelming... speaking from experience!

It's not actually that hard - IF you get enough experience. The big variables are:

Warm up.
It varies from person to person but I used to find for a 25 mile TT that 20 minutes on a little portable tacx trainer would get the job done. 10 minutes slow ramp to race pace, 5 minutes alternating above and below race pace and the remaining time. People used to look at me funny when I got off the bike - given that early season events in England could be held in 30 to 40F weather the "steam plume" could be 'interesting'

The start.
Don't go hard for the first 4 minutes. It took a couple of years to dial this one in with heart rate back in the day but 3.5 to 4 minutes or holding in the reigns before the legs would realise what was going on... From then on in it was a gradual sufferfest.

... of course with a powermeter the only big variable you have to account for is the difference between training and racing. Will power can only take you so far, adrenaline is a wonderful thing - it'll take a fair few events to figure out what you can do in races. Chances are, done properly, you'll be exceeding what you did in training by a fair bit. Chances are, done properly, that 25 mile time trial will leave you feeling like a truck ran over you, stopped and reversed over your for a second pass...


flapsupcleanup said:
And then I have to actually put it down on race day. Another non-trivial thing!

If you have a power meter, do your first couple of events at the power you can sustain during training for the first half of the event and take it from there.

I know people get worked up over time trials - the moniker "the race of truth" doesn't really help.

Think of it another way. It's you, the bike and the course. You get to use that extra incentive to do well to put out that extra bit of power. You don't get that in training - use it to have fun.

If you plan on riding a good number of time trials, ride some just for fun/experience. Mentally, it's just wrong to give them all the same level of importance. You need some events where you're going harder than training but it's not going to fry your brain and leave your breakfast somewhere past the finish line...

When climbing hills look at the highest point of the road and become transfixed with it. Unless you have a technical descent don't go over your threshold. On the downhills revel in the big gear work (do your math before hand to make sure you have a gear big enough to cover the cadence/rpms at that same when you're on the aero bars if there are any long straightish downhills - people may snicker when you show up with 56x11 on a hillish event with non-technical downhills. Laugh at them when that long slight downhill with a tailwind means that you're doing 45+mph on the aerobars and they're reving their spuds off on the drops.). On the flat, relax. Concentrate on the furthest point up the road and maintain your best effort. Forget everyone else - they're irrelevant. Everyone has pretty much the same challenges during the event unless you have a mystical change in weather during the event.

You can never gain time in a time trial but you can lose lots. Sure you might gain time relative to others but you're always losing time against the clock.

The good thing about time trials is that a rider who does "their homework" can seemingly ride much better than a similarly fit rider. It's all about the homework...
 
frenchyge said:
Ugh... I had enough of the air density correction factors between apparent and actual sunset back in my celestial navigation coursework. :(

I think I'm just gonna go for something like this http://www.youtube.com/watch?v=xrixWmMoNR0

... you'll might want to reconsider that as the pain of a time trial will seem irrelevant compared to the pain induced by your better half at the mess that's left. ;)
 
fergie said:
Is this legit, how do you do it and can it be done on a PC?

That's cool! Thanks. :)

Makes me wonder if anyone has done a profile version of this with angles calculated for each horizontal line to estimate Cd? :confused:
 
I understand that one separates the foreground and whites out the background but how this is done and if I can do it on a PC escapes me. Any computer geeks wish to show us how?
 
Piotr said:
FWIW, I've noticed that adjusting riding position indoors may not be optimal, since the forward momentum when riding outside plays a role in how balanced one feels on the bike.

I have a feeling that might be true as well.

For an experiment try riding in a normal position with the hands on top of bars and then while keeping your torso still remove your hands from the bars. With a decent fit, you should be able to stay in that position fairly comfortably outside, but indoors you'll likely start falling forward (this may very between individuals depending on their legs to torso ratio). Forward momentum should be taken into consideration when making bars and saddle adjustments. This is probably one of the reasons why riding indoors feels different. I suspect that it also applies to riding a TT bike indoors.

I just thought I throw it out there.

+1 I experience this very thing every winter when I start riding the trainer. At first I thought it was because the rear and front wheels weren't level...but that wasn't the case. In the past, when I've adjusted saddle fore-aft and stem height for my winter trainer position, I would then have to re-adjust everything back to the way it was (my usual road position) the first few weeks of spring training.
So yes, there definitely seems to be a discrepency between indoor and outdoor positions. Which makes me not want to ride the TT bike indoors, as I had it perfectly adjusted last summer:(
 
I found a big difference in riding my track bike on a windtrainer in pursuit position and on an indoor track to riding on the road and our bumpy outdoor track
 
Maybe this Q is for Alex but i ought to put it out to all & see if one of you has experience with this...

anyone had luck with the Alex Simmons aero spreadsheet using a course that is not perfectly flat??? ok ok. what do i mean 'not perfectly flat'...
- best i can come up with is a 1000m run with approx 4-4.5 meters of elev gain.
(thats about a 0.4% slope (assuming it is a constant slope - but the above course i found is not a constant slope even as slight as the elev. change is. so i cant simply add a .04 slope in analyticcycling to correct i think?).

i guess you could average the runs in both directions but i know speed/power is not equivalent even for perfectly paced runs on a slight uphill/downhill course.

am i better off using the Chung VE model?
anyone have experience w/the VE spreadsheet?

i assume that with a decent elevation profile, you can ride almost any course, download the data into the VE spreadsheet & get a much better than SWAG cda estimate?
 
DancenMacabre said:
anyone had luck with the Alex Simmons aero spreadsheet using a course that is not perfectly flat??? ok ok. what do i mean 'not perfectly flat'...
- best i can come up with is a 1000m run with approx 4-4.5 meters of elev gain.
(thats about a 0.4% slope (assuming it is a constant slope - but the above course i found is not a constant slope even as slight as the elev. change is. so i cant simply add a .04 slope in analyticcycling to correct i think?).
That's correct.
i guess you could average the runs in both directions but i know speed/power is not equivalent even for perfectly paced runs on a slight uphill/downhill course.

am i better off using the Chung VE model?
Depends on how much precision you need, and how non-constant the slope is. The key with the regression approach is that speed and power should be constant during each run. If the slope isn't *too* far from constant then you might come close to that so it's possible to try the regression approach.

Alternatively, you could try the VE spreadsheet and fiddle with the parameters until you get 4 to 4.5 m of elevation gain over the course. One virtue of VE is that you don't need to keep power or speed constant -- try doing one run while holding speed and power constant (as for the regression approach) and you'll see that it can be kinda hard.

I encourage you to try both.
 
DancenMacabre said:
...anyone have experience w/the VE..
It sounds like you might be able to get reasonable data on your course, especially if you can find at least 500 meters or so in the middle that's flat or steady grade. IOW, use the whole run to get up to and settle into as steady a power/speed pace as you can but trap the steady portion for your data.

Alex's spreadsheet is really good because it takes into account both kinetic and potential energy changes (as long as they're both steady gradients) so you don't have to maintain a perfectly constant speed as long as the speed change is modest and in one direction. Same for elevation changes. But if your course is rolling over the portion where you capture data you'll introduce some error. How much error, that depends on how much it rolls in the data trap section.

Some of my best venues for regression testing were on low traffic access or residential roads near small regional airports. Airports pretty much have to be flat and the roads outside their fences are often quite flat as well. Plus you can go home and download the airport weather data history for the time during which you tested to double check wind and to derive air density.

FWIW, I got my best results with the regression method when I was able to settle into a steady pace for at least a minute and a half but I'd usually clip the center 30 to 45 seconds as the most stable and constant speed/power data section to drive the spreadsheet. Slightly longer runs seemed better and more stable, but you don't want them too long as you want to do a lot of them before the wind picks up and you want to do some of them at or above race speed so they can get a bit tiring and you want to be able to hold fairly steady speed.
...am i better off using the Chung VE model?...
Well you've already heard from the man when it comes to VE testing. Personally I like the VE approach but I've done both so that I could nail down Crr vs. CdA. Elevation changes (small enough that you can stay in your aero position and not have to brake) are a good thing during VE testing so the only trick is finding loops or hammerhead out and back courses that let you ride in the same position and stay off your brakes and hopefully also have some elevation changes.

On perfectly flat courses VE becomes really sensitive to wind, although ideally even with some elevation change you want a very low wind to calm day.

-Dave
 
yes i can say i have heard from the man himself now :D:D:D

thanks for the tips Robert and for your suggestions also Dave.

holding power constant is doable indoors but outdoors especially in the flats is very tough i know. regardless i'll give it a go to see what i come up with.

-i'll try Alex's spreadsheet on the course i described & will run the data through the VE spreadsheet too to see how it compares.
-i'll also do a dedicated VE course/test run but will use an out/back or loop course with more elevation gain/loss. i figure that makes it easier to identify the amplitude & crest height of the 'virtual elevation' profile, hence better estimate of Crr & Cda.
-will have to practically tape a sign onto the aerobars during the VE run saying 'no brakes, no position changes! to remind myself not to blow it!

looks like i've got some work to do. much appreciate the help ;)
 
After a battle to clamp these aero's on, I finally got on to test ride. Still not 100% secure, but ok for indoors for now. They definitely help out with a variety of positions and are pretty comfortable, but I do feel like I'm sacrificing some of the basic/normal grips (top bar mostly).

Do any of you use the clip-on aero's for regular road rides ? I like them, but I'm not sure if I will on the road.

I know they are waay too far apart for any TT's, but for now I'm just looking for comfort and change in position.
 
frenchyge said:
My question is whether anyone has tried to evaluate saddle position fore/aft versus sustainable power? The biggest issue for my pedalling seems to be an inability to powerfully kick forward at the top of the stroke because of the sharp knee angle. I'm thinking of pushing the saddle back to help this, but that will likely close my hip angle somewhat. I was wondering if anyone else had solved that issue differently?
Well that hip angle issue is probably one big aero positioning paradox (one of them).

I think that one's ability to bring the saddle back for easing the application of power, thus closing this angle is largely dependent on one's lower back flexibility.

Over the years, I've been following this simple and non scientific rule of thumb. If one's aero positioning impairs his ability to sit on the Ischiums (is it how those little butt bones are called in English?) then I know that this angle is probably somewhat too closed. My business though involves preparing cyclists for the 40k, 90k, and 180k time trial events. No one should ride a 180k sitting on the "Jewellery box".