DrIsotope

My brain seems to unconsciously send me out on rides that I've not only done before, but have done before a year or two previously-- sometimes to the day. I have a ~3.5 mile lap around my neighborhood that I've done ~180 times. It's easy, all right turns, only 100ft of elevation change per lap, and I never get very far from the house, so I can quit whenever I want. The magic number seems to be seven-- 7 times around is just a smidge above 40km, so it's a nice length to go at a reasonable intensity and not render myself useless for the rest of the day. Not having to stop is probably the most important aspect-- that's an automatic average speed increase.

So as I alluded to, I had done this exact thing previously, and actually set the PR back on 8/25/17-- on my Cervelo. Today was on my Ritchey in "road format," shod in 700x35 Pirelli Cinturato Velo TLR.
This head to head is steel on 700x35s, 1X gearing 42T - 10/42 cassette vs. carbon on 700x25, mid-compact 52/36 w 11-25 cassette. The Cervelo is about 7lbs lighter than the Ritchey. The tale of the tape:



Ritchey effort, 8/10/19, 25.32 miles, 1h14m19s

speed: 20.4 avg, 27.5 max
cadence: 88 avg, 105 max
HR: 148 avg, 168 max
power: 234 avg, 680 max

Cervelo effort, 8/25/17, 25.32 miles, 1h13m24s

speed: 20.7 avg, 25.1 max
cadence: 89 avg, 106 max
HR: 153 avg, 176 max
power: 217 avg, 677 max


More aero, lighter bike, narrower tires, in roughly identical conditions... comes up to 54 seconds faster. A difference of 0.3mph. In the KOM effort, I spent 7:52 in Z4 and 5:13 in Z5. In the slower one, 11:56 in Z4, and just 33 seconds in Z5.
Mind you, I didn't set out on the day to make this a test, it was coincidental. I was just surprised to see how close the two efforts were in all performance categories-- I guess blind testing helps.
So 7% more power to go 1.2% slower. That casts off any doubts I may have had about the rolling resistance and/or aero penalty of the big wide Pirellis. Nothing like a real hill, so a weight penalty apparently doesn't come into play.



I do know that from previous efforts, getting a single lap under 10 minutes requires an average output right around 250W.
...of course, it's like The Game-- now that it's in my head, any subsequent 40k effort will be tainted. These two will stand side by side forever.

asgelle

So the lighter, more aero bike was faster on less power. I'm not sure I get your point. Since you did multiple laps, why not use Golden Cheetah to estimate CdA and Crr for the two bikes and see where the difference is coming from.

canklecat

I wasn't seeing much difference between my 25 lb steel road bike ('89 Ironman) and sub-20 lbs carbon bike ('93 Trek 5900). The carbon bike *felt* lighter on climbs, but we don't have any long climbs. In actual practice my ride speeds/times on familiar routes I ride often are about the same.

But I suspected there was some difference due to the poor fit of the carbon bike. The early '90s Trek 5900 had a lovely, elegant, lightweight titanium stem that's ridiculously long, around 140mm. The bike was technically my size, but due to age, joint problems and injuries (neck, back and shoulder), it felt wrong.

I've tweaked the Ironman quite a bit to fit well so it feels comfortable. I figured the Trek might respond as well. So I switched to a shorter 90mm stem, compact drops with shorter reach, and planned to try again this weekend. Didn't work out, though. Felt rotten and sluggish, bailed out at 35 miles of a planned 50-60 mile ride, with only my usual average time. The bike felt better with the new stem and bar. I just didn't have the energy to make good use of it. So I'll try again next weekend.

It's difficult to judge my performance from Strava unless I crop the beginning and end of a ride. I need 30 minutes just to warm up and another 15-30 to cool down, finishing very slowly around 10-12 mph. I don't usually crop my rides but I do select the middle 20-30 miles to compare with previous efforts, via private segments that are interesting only to me. My average speed has increased during my maximum effort segments of 10 to 20 or more miles. But I still don't see significant differences between the two bikes, despite the 5 lb weight difference.

Gearing is different too. I've been experimenting with the old Shimano Biopace on the Trek and I'm not convinced there's any advantage. There's no disadvantage either. But it feels odd, kinda grindy. I'm a spinner and I suspect non-circular chainrings, or at least the Biopace, may be better suited to a lower cadence. So the two bikes aren't really comparable, with only weight as the main difference.

Seattle Forrest

I bought my Cervelo because it handles exactly the way I want a bike to.

But 1.2% fastest for 7% less effort is pretty significant in my book. And that's with almost no climbing which is the bike's strong point.

Speedskater

Sounds like 'self fulfilling expectations' to me.

DrIsotope

If I was setting out to confirm a bias, I would hope that the lighter, more aero, more expensive bike would be more than 1.2% faster.

If I was setting out to confirm a bias, I wouldn't wait almost two years between efforts.

But okay.

DrIsotope

This is back-to-back days, The "heavier" one on the left, the other one on the right. Course was about 98.5% the same, I took some opportunistic rights when available on the lighter bike, so it ended up being
that third of a mile longer. Some of the speed comes from weight, some from a more aero position, some from more gearing above 30mph, and I'm sure some from luck at lights.

Again, I would have imagined the variance to be much larger. Even with a slightly higher intensity (normalized power within 2%) it's something like 12 seconds per mile.
I mean, I guess that is a lot when viewed over time (over 20 minutes in an imperial century) but I think on some level we all imagine a change in orders of magnitude. But alas,
it's just a little bit faster at the same effort, and a fair bit faster with a little more juice put to it. I sweat a lot more the second day, when it was cooler. Go figure.

Seattle Forrest

Road bikes are basically pretty similar. The weight differential between two bikes isn't huge once they have a rider, and it only matters on hills. If you really want an order of magnitude more speed, you need a motor.

PS - thanks for sharing your data. You ride a lot more than the average bear and have a treasure trove of info. It's one thing to read an industry white paper, but real world experience is a lot more relevant to 99% of us.

DrIsotope

Elevation profile from Tuesday, route ridden "forward."


Route profile from this morning, route ridden "reverse."


So there is a clear advantage in my case to having that first 10 miles with next to no climbing, because it gives me time to properly warm up.
I had some advantages on the reverse day, it was significantly cooler, and despite making a lot of left turns (which are rights in normal mode)
I only had a few stops, they were just long-- 30 seconds or more, usually. With the bulk of the elevation gain coming in the first 8.5 miles,
I had to spend a bit more gas on the gentler downhill, trying to make up time that I lost by going into that slog of a climb essentially cold.

The elevated HR on the first ride is I think mostly due to the elevated temps-- it was around 95º when I got home, compared to ~75º today--
because the NP and intensity are spot on. I think I could have pushed a bit on the middle climb and made up at least half of the time,
but I was trying to do an RPE thing and see how well I could copy the previous effort's intensity. A couple made lights could have been
all the difference in the total time.

Wilfred Laurier

This agrees with my observations. I haven't had a 'road' bike for a few years, but have done many road miles, including fast group rides, on my touring bike with 700X35mm tires. When riding solo there is a tiny difference in speed, but I have no problem keeping up with a group.

Mid last year I changed my rear tire from a sweet and supple Panaracer Pasela to a burlier Specialized tire with 'blackbelt' flat protection and feel like that made more difference than going from road bike to touring bike.

DrIsotope

Related to another thread concerning riding via RPE-- yeah, it doesn't work. Of these two rides, the one on the right felt significantly easier.
But it was basically the same effort, separated by just 69 seconds. I honestly don't see where the differential in relative effort is coming from.
I guess it could be that I did a more steady, measured output on the higher relative effort, while on the other I reserved on the flats and
hammered on the punchy climbs.

DrIsotope

Ignore the calories, now that we know that Wahoo ignores power data altogether. Nearly identical route (a little more turns here and there to avoid stopping in the one on the left) but
two otherwise very similar efforts three days apart. The added stress/power from the second effort is all from one section about a mile and a half long.

The CF bike spends more time north of 25mph than the steel bike-- it just descends much faster-- so the average speed jumps more than the power differential would indicate.