Trainer speed vs. TT speed

I've been training on my KK for the last 6 weeks or so and have the optional $50 power meter/cycle computer. Using that computer I monitor my progress with various short timed repeat intervals - e.g. 2 x 5" or 2 x 8" (a la the CTS training DVDs). I estimate my FTP this way at around 265 and rising (thankfully). In any case, while I perform these at max sustainable power (FTP?), my speeds are rarely greater than 20.5 mph. While I can certainly hit higher speeds on the trainer, I can't maintain them for very long. This past weekend I rode my first 9 mile TT. I used my road bike with no aerobars or disk wheel, unshaved, etc. and averaged 23.6. I'm fairly happy with my performance but I cannot reconcile that speed with my trainer speeds. I understood that the better trainers were fairly close in comparison to real road speed/power. Is my KK underestimating power/speed or am I simple able to find more power/speed under race conditions?

 

The KK has a power/speed curve that mimics riding up a slight grade so you should expect slower speeds than riding on the flat outside. IIRC, 300 watts is about 21 mph.

 

It sounds like this could have a lot to do with the course of your TT. Was there a tail wind? Was it a loop or just a single direction, where a slight grade could play a role?

 

Yep, about 21.5 mph according to the published KK curves: http://www.kurtkinetic.com/pdfs/Power_Curves419.pdf

Your point about the 1% grade is right on target, if you TT on a flat road with little wind you should go faster than your KK times. OTOH if you don't meet KK's assumptions: "..rider assumed to be 165 lbs, riding a 23 lb bike with 170mm crank arms up a 1% grade, at sea level with no wind on rough asphalt... " then your results may vary.

 

As pointed out by others, KK is making some assumptions about grade in their power/speed curve. Even if they weren't making these assumptions, the implications for speed of position, wheels, frame, weight, tires and the course and conditions (especially wind) are huge. But, the KK will help you increase the power of your engine and you can translate that into speed on the road.

 

I knew about the KK assumptions but must have missed the 1% grade factor. I guess that accounts for the difference, though on the road I don't seem to experience a 3 mph (~15%) speed loss on 1% grades - at least not over the 5" or 8" test periods I use on the KK (I occasionally ride a river trail that is exactly 1% grade out and back).

The TT course was out and back with no real wind but some hills.

 

Hi

The other posters (talking about the grade) are correct. I noticed the same thing with my KK trainer - it seemed much harder on the trainer than on the road.

When I investigated, I noticed the 1% grade thing. I then went to Analytic Cycling (they did the curvefit for KK) and played around with their stuff. I first reproduced the curvefit (pretty easy to do, since KK provides the necessary information), following which I turned the grade to 0% and redid a few points. The results were then pretty close to my on-road experiences.

 

Thanks, I'll check that out.

 

Nice work on using the KK and power training principles.

Just to illustrate RD's point:
In the last 6 weeks or so I have done two 16km TT tests using a power meter on the same 4km loop course with same equipment set up. Pavg varied by only 2 Watts but the time difference was 88 seconds! Weather conditions were quite different - and did not "even out" even though I was riding a circuit.

 

On a related topic, I've been doing longish intervals (20 - 40 mins) at a local velodrome - obviously a flat course - and am interested in what power level my average speeds correspond to. I've plugged the relevant data into two on-line speed-power calculators:

http://www.analyticcycling.com/ForcesPower_Page.html

and

http://www.kreuzotter.de/english/espeed.htm

and I get huge differences in calculated power, which on closer investigation is mostly related to the models coming up with very different values for the [Drag Coefficient]*[Frontal Area] (Cd*A) term in the power formula. Analyticcycling has a default Cd*A of 0.25 whilst the other site comes up with 0.45. At my usual speed of ~35 kph this translates to a power difference of 120 W (181 vs 304)!

Does anyone out there have any idea which is likely to be closest to correct? (I'm deliberately riding in a non-aero position - hands on bar tops - on a standard road bike - since my main aim is to increase threshold power for an endurance ride involving lots of climbing)

 

I'd go with AC. I get unverifiable results at the kreuzotter site.

 

Only problem with that is I'm sure that the AC number is too low, based on comparing perceived effort, HR etc for these intervals with reasonably steep climbs where you can ignore wind drag and get a reasonable estimate of power just from climb rate, mass, and gravity. This is probably because the AC site's default values for drag coefficient (0.5) and frontal area (0.5 m2) give too low a power estimate for flat roads where air drag is the main force to overcome. To get a sensible answer out of the AC site you need a better estimate of these things which is hard lacking access to a wind tunnel or backsolving from a power meter (which I don't possess).

So I guess the question I should have asked is does anyone out there with a PM have an idea of what _their_ output is riding at 35kph (call it 22 mph) on a smooth flat road with no wind, hands on bar tops?

 

Even this is going to be a pretty large range due to rolling resistance of tires and road surface. And, FWIW, I think the conditions characterized as "flat, no wind" is a null set.