Dave : Bike/Wheelsize Testing

Testing, testing, and more testing

Dave — Tue, 21 Aug 2007 10:19:00 GMT

In winter 05/06 I was on a quest to answer the question: which bike is more efficient, a Trek Fuel or a Salsa Dos Niner? It just so happens that one uses 26" wheels while the other uses 29" wheels...

These have proved to be rather popular posts, most of them getting over 10,000 hits and still counting. To this day, it is the only published attempt at quantifying the difference in wheel formats using power. I could be flip and say it doesn't really matter anyway, we're just riding bikes for fun, right? Well, based on the popularity (infamy?) of these posts it clearly matters to some folks out there. Probably you ;)

My stable has grown - it now includes a FS 29er (again), lots of 26ers, and soon a 55er (26+29=Siren Fifty-Five). And if that isn't confusing enough, there's an outside chance I'll be testing a new entry to the wheel size mix, the 650B/27.5" by Kirk Pacenti.

So guess where my interests lie now? Yep, learning the strengths of each of these formats. It's a technical nightmare as James at CN has learned. Realistically, the only power meter I trust on an MTB is the Power Tap. The major (MAJOR) hitch is they don't have a disc compatible hub yet. I've been leaning on them for a few years but I'm only once voice. If you are interested in the same hit with an email. The Lev is rear disc only...so any testing on that bike is limited to terrain where a rear brake is optional.

Fun things to look forward to, more charts and geekdom.

So with all the downtime in the last 6 weeks, aside from moving and a few other things...I've got my music situation dialed in. Some folks get cable TV, I get a monthly music service with Rhapsody. Their top plan is called Rhapsody to Go but it requires specific MP3 players to utilize it - it always seemed overkill to me. Well my trusty little Creative Nano finally croaked and forced my hand...so I got a Rhapsody compatible player. All I have to say is I'm glad that Nano bit the dust. In the past once I found the music I liked online, I had to buy it somewhere. It wasn't so much the money but the time that was a PITA and kept those efforts to a minimum. Rhapsody to go - so far - has changed all that. I can now upload anything from Rhapsody's catalog to the MP3 with a 30 day license - this even includes custom stations. Pretty sweet deal considering I work at a desk and use Rhapsody most of the time already...the new MP3 just upped my music addiction.

Ain't technology somethin'? I guess this means I'm no retro-grouch. Gimme gears, suspension, music, and power baby. That and a touch of crack makes for a lot of flow.

Back from the dead: 29er studies

Dave — Thu, 08 Jun 2006 11:46:00 GMT

With the Dos stripped of components and hanging in my garage, let's just say 29er studies aren't at the forefront of my interests right now. I found the answers to my questions. However, Cycling News is hitting the issue hot and heavy .

It's nice to see an organization with some industry ties and a bigger budget is taking the time to look at the issue. They do have some interesting analysis techniques proposed, such as using GPS and MotionBased to analyze specific sections of a route, comparing between the two bikes.

There are a couple of deal killers in the study, though. The first one hit me in the head so hard I nearly passed out. They aren't using power! Without measuring rider output, how can efficiency possibly be determined? It can't. You could determine which bike was faster over a short course by repeated time trials, sure. But that doesn't apply to the endurance racer, the segment of the cycling market that is gung ho on 29ers. Efficiency is king to the endurance racer. We have a limited supply of energy, and the more distance we get from that limited supply, the better. Gotta have power.

An example: Robert Chung of the wattage list (I think he is a statistician by trade, but not sure) assisted with the data analysis of my files from the Old Pueblo 24 hour event. He looked closely the files comparing a couple of laps with similar times but different power, one on the 26, the other on the 29. What he found was that for most of the time, the power between the two bikes was very similar. There was 1 section, though, that required a lot more power on the 29 when compared to the 26. For those familiar with the course, this was the Corral trail. This section is slightly downhill, twisty but very fast. Time to cover that section wasn't much different, but power was much different. That sort of observation would have been lost without the use of power.

The second issue is that they've gone to great lengths to normalize the cockpits between the 2 bikes - and I'm assuming this implies there is only 1 test rider. It's been my observation that larger riders tend to prefer 29ers more than smaller riders. By larger, I'm talking about weight more than height. There is probably a reason for that...results from one person's runs will likely be different from anothers, especially if they differ in size.

In short, CN has tightly controlled for equipment variables but without objective energy measurements, their results are just going to stir the pot rather than provide any sort of definitive conclusion. They could change all that with an SRM or Ergomo...are you listening???

Don't settle for 80% CN, this ain't the pareto principle.

KTR, moonbeams, and dead rabbits

Dave — Mon, 15 May 2006 12:56:00 GMT

Pre-race quote: "Mike, I want to thank you now for organizing this race 'cause I ain't going to be thankful at the finish"

- anonymous rider to race organizer Mike Curiak 15 minutes before the midnight start of KTR.

Well said amigo. That was a tough race. I'm hard pressed to think of a tougher race I've done. But at the start, midnight Friday at the Slickrock trailhead ouside of Moab, Utah - it was all grins, smiles and anticipation. What an exciting time and place! As the minutes ticked off, Mike would announce the number of racers...45...50..52! 56!!! No way, 56 enduro nutcases showed up under a full moon in the desert to embark on what would be a grueling day's ride where only slightly more than a third of the field would finish.

By the numbers, I had a pretty good race. I finished 2nd to Jon Brown (who ripped it on a singlespeed), coming in with a time 20+ minutes under the previous course record. The numbers lie. I got my a$$ handed to me out there, and I'm not embarrased to admit it. As the first unsupported epic race in my experience, I had a lot to learn. Still do for that matter...

It's hard to convey the scene at the start. Slickrock trailhead was crowded, there were so many people and vehicles there. We were a bit behind, and when the trailhead first came into view it looked like a small active city with all the moving lights. Everyone was excited, you know, that sort of excitement that finally gets release after months of anticipation...old friends were reunited as well. It was a magical time, a magical place.

So there we are at the edge of the parking lot, MC in front of us about to turn us loose...and he does so with the most subdued "go." It was nearly a whisper, and I laughed outloud at the contrast of the actual start to this epic with the anticipation of the crowd. No warm up meant I was going to take it easy for awhile. That was my plan anyway, to start relatively easy to moderate up the initial climb, then push it a bit more on the second climb to the end of the pavement.

I had problems immediately. The first one was technical - the sweet lights I've been cooking up? Well, I decided to add some power to my bar setup by adding 2 x 3W luxeons to the 5W already there...all run at 700 mA. It is an absolutely amazing setup. All together it draws only about 10 W and is much brighter than a light in motion HID according to a side-by-side comparison. My mount, however, wasn't sturdy enough for the extra weight (it was now 150 g) and on the very first bit of washboard on Sandflats road the light came undone, found the front wheel and broke lenses...and wiring. 5 minutes into the race and I'd lost 2/3 of my lighting. Sweet. Luckily, I still had my helmet mounted setup...

The next problem came quickly as well - and it was painful. I used a Wingnut Hyper 2.5 for the event and added a 100oz camelback bladder to the main comartment. With 200 oz fluids in the pack and clothes bungied on the back of it, it squeezed the pack so that something with the shape and texture of a football pressed firmly into my back, right at the kidney zone. OMG that was some intense pain, I thought for sure I was bound for a DNF. Some repositioning helped, but it would haunt me the rest of the day.

The first 1.5 hours were pretty tough, I must say. Very unusual for me...but anyway, things got better when we hit the pavement. After railing the descent (I'd just ridden it recently so remembered the hairpins) I found myself with the leaders, although I didn't know it at the time. I thought Jon was still up the road...but he was in this group. IIRC, all but one of these riders were on singlespeeds. I was amazed at their ability to remain out of the saddle for long extended climbs. Jon Bailey was there and crushing it. 40 min of climbing later, Jon (Brown) and I crest the top of the climb together and pause for a bit to take in the view of Fisher (or is it Castle?) valley to our left and the LaSals glowing to our right under the full moon. Extraordinary..."that's pretty clean" as a racer friend once muttered under similar circumstances in the Soul Ride...This sport is simply amazing in terms of the currency used to reward hard efforts.

Jon was crushing the climbs on his singlespeed, and since it was a no-drafting event we didn't even bother to stay near each other most of the time. He'd fly up the steeper sections, I'd diesel back up on the mellower sections...but once at the top of the LaSals, with SS gearing he spun out immediately and I started to move ahead. This was about 2:50 am when we started descending, and for the next 4.5 hours I continued on by myself. There were areas where I could see lights behind me and get an idea of the gap. After the descent off Beaver Mesa there's a short steep climb...where there were 4-5 hardy souls waiting for riders and cheering at 4am! How cool is that? Anyway, from the top of the short climb, I could see a light and estimated the gap at 12 minutes or so. The Beaver Mesa descent was kind to me. I was on the Dos Niner, and the big wheels were just the ticket. Deep sand, rocky ledges, it was no trouble on the Dos. The Dos found redemption in this race... My helmet LEDs were working flawlessly, and being the first one through meant there was no airborne dust that is typically an eye irritant in lap courses. I was in the flow, feeling great now and loving life. What can be better than racing off the front in the middle of the night through epic terrain?

Soon I came to Rose Garden hill, aka Magpie hill. Magpie hill redeemed itself as well as the Magpies were absent and replaced with happy morning chirpy bird sounds. The short hike up the hill was great for my back and while at the top I took a little break for some solid food, an almond butter and honey tortilla. I could see that the gap was still growing and was at least 15 minutes now.

Thank goodness for the pre-ride Lynda and I did a couple of weeks ago, becuase without it the next 2 hours would have killed me. Instead, I knew what was coming and just took it nice and relaxed, kept hard efforts to a minimum, and tried to be smooth. Very technical, ledgy, rocky, and quite a few bits where I got off the bike. It felt good though, and when finally reaching top of the world where MC's truck was, it was about 5:30 am with glimmers of sunrise in a pink sky. Magical again! I hit Dewey about 5:55, and there were some folks cheering there too. Now for the Yellow Jacket section, the sandy, climby, semi-tech bit of 9 miles. The sun came up during this section to reveal the Lasals in the distance, and a day that would become hot. Here's a pic from the pre-ride:

Of course, water is the big issue in this event. How to stay on top of hydration in the desert??? My plan was to stop and filter at McGraw Draw out of the Colorado River. I had tested this when the river was muddier, and knew it work fine. So where the trail becomes singletrack next to the river, I sat on a ledgy rock next to the river to fill up. This was the turning point of the day...

Pumping...nothing. Huh? Well, this is something I haven't encountered before being a water filtering neophyte and all. All the pumping in the world wasn't getting any water through the filter. Nada. I wasn't dry and had plenty of fluids to get to Westwater, but didn't want to take the 15 minute penalty of the detour. So I fiddled with that #$%^%^$$ filter until it worked. 30 minutes later everything was packed up, bladders full and I was ready to continue. But my lead (which was 28-29 minutes) had been erased. Jon came by just as I was re-packing. Also, unknown to me during the race is that Jason Stubbe came by when I was filtering. Also unknown to me is that I passed him when he was watering the flora.

Wind and SS gearing, these were my 2 thoughts. I know Jon is tough as nails and I was glad he was on a SS. The forecast had called for east winds in the am (headwinds) switching to the west (tailwinds) after it got warm. If that wind switch happened, I figured I'd be able to fare better with the higher speeds with full gearing than Jon with SS gearing. I didn't know what his gearing was, but it had to be low enough to tackle the LaSals. But alas, the wind remained a steady headwind for the entire race, keeping speeds low. When Lynda and I pre-rode the second half, it took us 1.5 hours less than it took me race day. We had rippin tailwinds that day. We rode together for awhile, but he had a bit more juice on short climbs and started to pull away. In the meantime, riding into the "blow dryer" was beginning to take a heavy toll on me. The comfort of cool night riding was replaced with hot desert...and I made the mistake of putting drink mix in much of my filtered fluids, so plain water was limited. The gut wouldn't tolerate much of anything but plain water...and the not so slow drain of fluids from my body was turning this into a survival event. My goals were changing rapidly...no longer chasing Jon, I simply wanted to finish before 2pm.

Finally...FINALLY...came the descent to Salt Creek. I had been looking forward to this cause all I could think of was taking a dip and dumping some of this heat. I was surprised to see Jon still only 5 min ahead of me. He must be hurting too, and rightfully so since he started with 170 oz fluids and didn't stop for more along the way. Anyway, into Salt Creek I went - literally - and forgot about the camera hanging around my neck. But I didn't care...the creek was cool and offered a moment of relief.

The relief was short lived. The next 15ish miles were brutal. Very steep tech singletrack and hikeabike, and at one point, my upper body was so weak it took a long time to get past a few boulders over which the bike had to be carried. As blown as I was, I still passed at least 20 rec riders out on the trails to the finish...I remember those days ;) Finally, the finish was mine in 13:55. Mike was waiting with the coldest, sweetest water ever to pass my lips and for that I will be forever grateful.

Here is the man, Mike Curiak, taking a shot of my bike pre-race:

Post Mortem:

Lynda will tell her own story in time, but that desert rat had a strikingly different race. She had bionic competition which pushed her hard and she rose to the occasion with a time that will surely stand for a LONG time. Huge congrats to her for a beautiful effort. She gets more confident and savvy with every endurance event and I am privelidged to be witness to her growth.
Both Lynda and I were sick as dogs the night of the race. Can you say heat stroke/electrolyte loss?
Sadly, this will likely be the last running of the KTR. It seems that many competitiors didn't follow the rules clearly laid out by MC. MC isn't paid for this - he does this to return something to the sport that has provided so much to him - and this is a disrespect he can't tolerate.
Huge kudos go out to Jon Brown, who nailed the course in 13:26. The previous course record was 14:19 (on a geary!) set on a day when conditions were tailwinds. This year's race conditions were slow - headwinds from start to finish. How he did that on 170 oz baffles me, I drank twice that!
Many, many thanks to Mike Curiak for having the patience to put up with all the silly pre-race questions, getting the organiztion dialed in, and being so easy going on race day. It was a pleasure to meet him and I hope the eventually takes on a positive sheen for him. Ah, and thanks for that water from heaven...

Unintentionally stirring the pot

Dave — Sat, 11 Mar 2006 15:29:00 GMT

It's a rainy morning here in southern Arizona, the first in 140+ days so no complaints...

So. My quest to determine which bike is faster for me, the Dos 29er or the Trek Fuel, kicked up a bit of a ruckus in places. The Cyclingnews piece in particular got folks talking, er, yelling? In case you missed it, it's this one:

http://www.cyclingnews.com/mtb.php?id=news/2006/mar06/mar03mtbnews

Then yesterday g-ted posted a link to bikemag, which features an article on Old Pueblo and the bike test.

http://bikemag.com/features/columns/030806_mike/. Mike Ferrentino doesn't exactly have a gentle writing style...

It has really baffled me why so many folks get so worked up over my little quest. How many folks are riding 29ers strictly for racing & speed? Very few. It's as though readers see my test and think it's a slam against their belief system. Some bring up points that would be valid if I was doing a blanket across the board study...but I am not a research facility!

Entertaining in a "gee this doesn't make me so proud to be a human" sort of manner. Here's a random collection of the links I found to MTBR threads concerned with the testing, either directly or indirectly. From initial reactions, to "where'd the love go", to "I'm outta here", to "wait a minute, we ride these for fun!".

Phew! Someone finally got it in the end.

http://forums.mtbr.com/showthread.php?t=170033

http://forums.mtbr.com/showthread.php?t=170202

http://forums.mtbr.com/showthread.php?t=170114

http://forums.mtbr.com/showthread.php?t=170650

This one is about the 96er. Anything 26" gets attacked here.

http://forums.mtbr.com/showthread.php?t=170773

http://forums.mtbr.com/showthread.php?t=170061

http://forums.mtbr.com/showthread.php?t=171388

http://forums.mtbr.com/showthread.php?t=171890

http://forums.mtbr.com/showthread.php?t=170790

http://forums.mtbr.com/showthread.php?t=170491

http://forums.mtbr.com/showthread.php?t=170411

http://forums.mtbr.com/showthread.php?t=166486

http://forums.mtbr.com/showthread.php?t=155598

On the endurance racing MTBR board:

http://forums.mtbr.com/showthread.php?t=141317

Funny thing is, I love riding my 29er. 26er too, for that matter.

Every thorn has it's rose

Dave — Wed, 22 Feb 2006 01:25:00 GMT

Drum roll please...as you look at this plot, keep the following in mind:

This compares a Trek Fuel 26" wheeled bike and a Salsa Dos Niner 29" wheeled bike
Bikes were similarly equiped - near identical weight
Identical tread pattern, Specialized Fast Traks all the way around
I changed the crankset on the Dos to accomodate a 30T middle ring to more closely match the Fuel's gearing

The data is the result of daylight laps at the 24 Hours in the Old Pueblo race. When I say race, I mean it was full on...see the OP race post if interested. The upshot of this was that I did everything possible to go as fast as possible, within the limits of the event. No prior subjectivity could have skewed the results. The displayed lap times are the time on the course. Pit times are not included. I threw out the night laps for publication because of the eyesight issue - the 9th and 10th laps are extreme outliers due to extenuating circumstances.

OK...giving this a minute to sink in...each point represents the x, y pair of average power and lap time. In other words, the dots in blue show the relationship between power and lap time when riding the Fuel. More power = faster laps. Make sense? The black lines show a linear regression (done by MS Excel) to illustrate a trend, or predicted values. The two pink squares show what happened on the Dos Niner.

I added the red rectangles to show the "what if" scenarios. Consider the case where average power over a lap is 188 watts. Based on this data, a lap time of 67 minutes would be predicted for the Fuel, while the predicted time for the Dos is just under 69 minutes. Looking at it from another angle, for a lap time of just under 69 minutes, I'd have to average 175 watts on the Fuel but 186 watts on the Dos.

So there you have it. It's cut and dried in my mind. The 29ers in my testing have lost in rolly, twisty singletrack, and also on climbs. They might fare better on downhills, and almost certainly in sand and maybe even rocky stuff...but right now I'm losing interest fast.

Anyone interested in a tricked out Dos Niner with a Power Tap hub?

26 vs 29 singletrack shootout: tie

Dave — Sat, 11 Feb 2006 01:23:00 GMT

Well now, (AHEM!), looks like I jumped the gun yesterday in revealing results. That's what I get for seat of the pants analysis.

The testing of 29er vs 26ers continues on. Quick recap: the first round of tests looked at how climbing on the Dos 29er compares to climbing on the Fuel 26er. Those results showed a small but repeatable advantage to the Fuel for climbing dirt roads. That wasn't a big surprise to me, however, many took it as a direct attack on their lifestyle ;) We're just talking about a slightly larger wheel size, right?

This round of tests was done on a 3.2 mile singletrack loop in McDowell mountain park outside of Fountain Hills, AZ. If you've done the Mesa NORBA nats, you've ridden this trail. Swoopy, twisty, fast in spots, steep rollers, bad braking bumps in spots, a few very tight turns. It starts with a ~ 3 min big ring climb, finishes with a more rolling but trending downwards section of whoops and twisties. I call the latter section the "descent" in this analysis.

I did the rides on 2 different days separated by a 5 day period. The first ride tested the 29er, the second ride tested the Fuel. The results for each test are broken down by overall time, average power, and normalized power. Here's how it turned out:

Other relevant data not reported in the above sheet: energy expenditure was identical across bikes and runs at 185 kJ & 184 kJ for the Dos, 185 kJ and 182 kJ for the Fuel.

At first I thought based on average powers that the Fuel was significantly faster on the descent portion - which is completely non-intuitive. The Dos felt like it held speed better in turns, so I expected the Dos to be faster on the descent. If we just look at average power, that conclusion would stand. However, normalized power comes to the rescue. Normalized power was just enough higher in the Fuel tests to indicate I had a bit more snap in the legs on that day. In rolling terrain, it is very important where you put the power down, and how much at a time. On the Fuel's day, I just had a bit more to give on the steep ups, and rested more on the descents. The averages were the same, but Pnorm tells all...another factor that may have attributed to faster times on the Fuel was that I got 3 runs in on the Dos 5 days prior, so I was more familiar with the trail. The Fuel did feel faster in the tightest terrain, oh wait, that's subjective, nevermind.

Here's a chart showing how the distribution of power changed between the two rounds of tests. There's a considerable shift to the right for the Fuel's test, which simply means I was feeling better that day.

As for the climb - look how tight the data is between the first run on the Fuel, and the second run on the Dos. Looks identical to me.

Round 2 conclusion: it's a tie.

So what's next?

One test will be similar to this one, except at a much more relaxed pace. These runs were at a XC like pace, far above that of an endurance race. What happens if power spikes are much lower & most of the riding is in the saddle? Will the bigger wheels conserve energy somehow in the twisty/rolly stuff under those conditions?

How about some pure downhill runs? There aren't any around here I can think of, that'll probably have to wait until late-spring back in Colorado. The great thing about a bunch of downhill runs means I get to do a bunch of climbing :)

Signing out for now, fire away, but keep it civil.

First results of 26 vs 29 inch wheels

Dave — Mon, 02 Jan 2006 20:04:00 GMT

Note (added 2/3/06): comments to this post have been disabled due to malicious spammers. You know who you are. Contact me through the "contact me" link if have something you'd like to add. If there is enough interest, I'd consider opening a public forum for the issue.

Note (added 2/12/06): this test looks at comparisons for climbing relatively smooth but variable pitched long dirt road climbs. For an analysis of rolling, twisty singletrack, check out http://teamhealthfx.com/blogs/dave_harris/archive/2006/02/10/510.aspx.

Thanks!

-------------------------------------------

Which one of these are faster? This is the question of the hour. Every 29er fan would have you believe 26" wheels are "kiddie" wheels and also that 29" wheels are faster. Time to cut through all the subjectivity and put some numbers on the problem. The short answer: for this comparison, the 26" wheels outperformed the 29" wheels, but it's not an overwhelming difference. Then again, the margins between winning and losing are often quite small...

I've got a Salsa Dos Niner and a Trek Fuel 110, both setup with power tap hubs. In the past week I did the same ride twice (4 Peaks road from hwy 87 to hwy 88 and back), a 60 mile dirt road/jeep trail route with approximately 11,000 feet total climbing. Two sections of the ride are compared: the first climb, which is partly rolling with some steep climbs and one long climb, rising about 4,000' in the process. The second section climbs from hwy 88 back to the high point; this climb is quite steep, climbing ~4,000 in 8 miles. There are a few short descents.

Bike setups: The Dos was setup with Specialized Fast Trak tires, tubeless ala Stans. The Fuel was setup with a tubeless (plus Stans latex) Panaracer fire XC pro rear and a WTB Weirwolf 2.3 front, tubeless ala Stans. The Dos is .6 lbs lighter than the Fuel, as measured by a Tanita scale (.2 lb increments). I started each ride with the exact same amount of water & food so as to eliminate rider weight changes. I assumed my weight to be the same on both days.

First off I just went out and rode the routes. I made no attempt to ride even paces for any sections of the ride, and in fact, the latter ride (on the Fuel) was done at a more spirited pace. I just planned on sorting out the data post-ride.

Next, I had to come up with correction factors for the power meters. Power taps are strain guage devices, and as such, have a small but significant precision variability. I performed a stomp test for each power meter (including my road bike PT), here's how they turned out:

Based on the stomp test, the 29" PT measures .86% high, while the 26" PT measures 1.75% high. This leads to correction factors of .9915 for the 29" and .9828 for the 26".

Now there's nothing left but to look at the data for each climb:

The road surface is sandy in spots, rocky in others, and fairly loose decomposed granite throughout.

Conclusion: in this test, the 26" wheels outperform the big wheels. The first "climb" has a lot of rolling terrain, I was interested to see how this would turn out. In the end, though, a consistent advantage exists for the 26" wheel, whether rolling or straight up climbing. On the steep climb ("climb 2"), one would expect a linear decrease in time for a linear power increase. IOW, a 10% power increase should decrease time by 10%. This is because the major resistance to overcome is that of gravity - a constant. Yet the Fuel saw greater speed increases than linear with power increases. I could feel this on the ride - any change in speed necessitates wheel accelerations, and the big wheels are simply slower to accelerate. The smaller bike feels much more "lively" or "responsive".

Another interesting detail is how closely the kJ tracked for each climb. Identical for the big climb! This is somewhat surprising given that the Trek was .6 lbs heavier.

Clearly, a test on a dirt road is only valid for similar conditions...so I'll do more tests off-road. But I'm becoming biased now. The acceleration issue of the big wheels is tough to overcome...I suspect they will be best suited for rocky/tech stuff.

One more data point: at Moab, I did a lap each on the Fuel, Dos, and the 292. I never did a lap on the 292 that was within a minute of the Fuel's lap time...

More to come!