I never cease to be surprised by the wind tunnel results, and Cobb’s article today is no exception. The biggest surprise was that the lowest drag came from the front mounted Profile water bottle, presumably including the straw. The second lowest drag came from the downtube mounted water bottle. Cobb didn’t test combining these two together, but since they are on different lines presumable they would not interact and this might be the fastest combination.
On the other hand, the results out of the wind tunnel prove repeatedly that well reasoned assumptions are typically wrong.
Someone please explain why a high water bottle in the rear would be better than a low one. While you are at, explain to me why a water bottle on the downtube would speed things up. If it were that simple to speed up the frame by a change in design, why wouldn’t the manufacturers take advantage of that fact?
I do love to see this though. I am planning on using a Profile front water bottle and a downtube bottle at Lake Placid this month. Looks like that should save a minute or two. If I can figure how to mount 60 of them, I might be competitive on the bike this year after all.
Now I’m really p*ssed at John. Based upon his earlier recommendations I sold my front mounted bottle on ebay and designed my own rear mounted “never reach” style system. Just as I finally get it working properly he comes out and says that I should have saved myself the effort and kept the front mounted bottle all along. Jeeze.
whatever you go with, the biggest difference is about 3:26 over an ironman distance, and :39 over a 40k. is that enough to make it matter to a MOP AGer? absolutely not.
one test yields one result, while another will reveal something different. i find it seems that you can often keep repeating the same test until the data says what you want it to (or you can interpret it that way).
i agree with the advice, use whatever is most comfortable for you and will encourage you to drink and stay hydrated.
AJ,
I thinkt he rerason the water bottles mounted behind high are probably slower is that they stick up into the wind. Where as the lower ones are shielded from the wind by the body a bit.
On another note, I’ve been using a down tube bottle for years. The others, if behind are prone to launching which is a danger to others and are hard to reach. If mounted in the front mine always seemed crooked or they fell off somehow. The way it seemed to me no water made for a slow then slower run which ultimately costs more time.
The biggest differences between these tests and the earlier ones JC did were, from my reading:
-rider on bike
-fatter tubes on bike
This being the case, the results were different than when he tested for the bike alone, that being a skinny steel frame. How? From what I read…
The water bottle on the frame DT is actually a boon with a larger diameter frame member. The larger tube (relative to a skinny steel tube,) in front of the bottle, makes for a smoother transition, and the wind sees this as an increase in surface area/depth of the DT. The tube is effectively deeper, and may also be cowling the seat tube, minimizing the negative effect of the big hole in the middle of a diamond configuration frame.
With a rider on the bike, a lot changes, and the placement of a profile-style bottle on the front may well provide some sort of a fairing benefit. Perhaps it is helping with the dirty air in the knee/lower torso area? I would like to see what effect the bar/bottle height has on this number.
The rear mount bottle #'s are really interesting, and are counter to what JC has been saying for a while. These numbers are saying that higher is better. Very curious.
I have a sneaking suspicion that the riders back profile could have a profound effect on this #.
I would love to see these #'s with a larger sample size. This is VERY interesting stuff, but the rider on the bike is adding a pretty substantial variable to the testing, and I would liike to see how much these #'s change in a more varied sample/with other riders…
As usual, Fredly is “right on the money”. I use the NeverReach simply because it gets me the most fluid in the easiest manner, and I can refill it so quickly and easily. I actually use a wide mouthed water bottle on my seat tube for my spare and tools, not water. I don’t have a front-mounted water bottle system, simply because I don’t want any extra steering forces up there, I run a Hed 3 up front, and that’s enough to deal with in windy conditions without adding more surface area up front.
How does my setup compare to the latest from the windtunnel, and how does it compare to my “best” aero position? Who knows? I’m larger than the test rider, I doubt we are the same shape although we are almost the same in height, my knees may not be nearly as close to the top tube as the test rider…all of those things make a difference. But, probably not enough to sweat about for someone like me doing local triathlons, even if I’m FOP instead of MOP.
Comfort, function (getting the fluid you need), and bike handling are the important things that most of us need to fuss over as far as water bottles go. Your position on the bike will have a much greater effect on your drag than your water bottle(s) location. And unless you get in a wind tunnel, or do some well controlled tests, even body position is a guess…although it certainly seems that there are some rather obvious beneficial positioning guidelines to follow.
It’s interesting stuff, though. But not stuff that may be directly related to any particular individual’s drag numbers, except for the actual subject that did the test.
Re: “whatever you go with, the biggest difference is about 3:26 over an ironman distance, and :39 over a 40k. is that enough to make it matter to a MOP AGer? absolutely not.”
From time to time on this site, there is discussion that slower folks gain more from good aerodynamics than faster ones. If that is true, and John’s tests were all based on riding at 30 MPH, are some of us 23 mph guys going to save more than 39 seconds over 40K?
I’ve noticed that a front mounted water bottle tends to hide the brake/ shift cables that appear on the front-end of a syntace style base/aero bar setup. Maybe this cleaning up of the cable drag contributes to the bike with profile bottle being faster than a bike with no bottle. If you’ve already cleaned up these cables, perhaps the time savings dissappear…? Just a guess.
While the results are interesting, I am not sure they represent reality. What was tested was the drag resulting from carrying the bottles, not carrying AND using them. If one has to sit up to drink out of a downtube bottle 2-4 times in 40 km, drag increases while out of the aero position. Does this drag more than offset the minimal advantage over a front-mounted Profile bottle or a Never Reach wherein you do not have to sit up?
John’s tests were all based on riding at 30 MPH, are some of us 23 mph guys going to save more than 39 seconds over 40K?
Nope, you will save less time because the drag force increases with speed (not linear) so at a lower speed you will have less force slowing you down. I think the take home, which John mentions, is use what you like and will drink from the most.
“you will save less time because the drag force increases with speed” has always been my logic. But take a look at http://www.cervelo.com/qa/cervelo-calc.html, where Gerard says otherwise - I think.
One explanation I’ve read elsewhere is that drag amounts to a fixed wattage, such that a 200 watt cyclist who saves 25 watts in his setup gains 12.5% in power, whereas a 400 watt cyclist who saves that same 25 watts gains half that much. Again, I’ve always understood that drag force goes up geometrically with wind speed, so I can’t get my head around the explanation.
Could it be as simple as saying that the 30 mph cyclist will save less time than the 23 mph cyclist simply because there is less time to work with at the higher speeds?
For those that use the NeverReach (even the home made ones) or Camelback, which based on these tests would seem to be less of an advantage vs. bottles, wouldn’t the gains be from not having to come out of aero to drink. I agree that the differences are close to negligible vs. proper hydration. So if you are able to drink while in aero then you (the rider) will not add significantly to the drag that is now been captured in these tests. It would be interesting to see how the temporary increased drag attributed to reaching for and drinking from bottles compares to the steady state of the NeverReach or Camelback systems.
I missed something there. John mentions testing at 30mph, but the numbers are all around 1:07 over 40k which is closer to 22-23 mph. Maybe with no drag it is 30 mph and the resulting times are taking the drag into consideration.
they do the tests at 30 mph because that is the industry standard speed setting for the wind tunnel, or at least, the speed that is commonly used, therefore valid for comparison. They then extrapolate the drag coefficients that result from the test based on calculated times assuming a rider with a certain wattage output at that drag coefficient. Or something like that. I think he says he figured 200 watts for the time trial and 150 watts for the Ironman.
In any event, they probably have a set calculation that they just plug the drag coefficient into and it yields an approximate time split, given all other variables are fixed.
From this data I would put the biggest advantage of the NeverReach system to the fact that it holds enough water for all but the longest races (you don’t have to fill it all the way up on the shorter races if you want to save a little weight) such that one would never have to slow down at an aid station to get water. I am sure that is worth a minute or two over an IM distance. This was partially addressed in the PK Racing test but not here which could account for the difference findings.
JC did not address this aspect of the aerodynamic effect of having to reach for a bottle to drink or slowing and reaching to refill in his analysis. And what about those who have skinny tube bikes. Proof that studies designed to settle controversies frequently generate them.
I suspect another reason they test at 30 mph is it is high enough that any differences will be reliably discernable by the instruments they use. The relative numbers would be the same for 5 mph but the force differences would be milligrams. Once one has a number, then it is mathematically convertible to another speed as long as the airflow remains essentially the same - laminar vs turbulent.