So I’m riding my Kalibur the other day, and the shifting is a little off as its a brand new bike, and the cables have probably stretched a little. So I reach down while on the fly, give the barrel adjuster a little twist, and I’m good to go. This got me thinking…is there any real advantage to having internal cable routing? It does look a little more squared away, but in terms of function, I’m thinking that external routing is better. With those skinny cables, I can’t believe that they really are that much of a factor aerodynamically, either, but I could be wrong. Anyway, just wondering, as it seems so many high end bikes have internal routing…
So I reach down while on the fly, give the barrel adjuster a little twist, and I’m good to go. This got me thinking…is there any real advantage to having internal cable routing?
I would think that to those who work on their own bike, no advantage at all - I’d trade mechanical access and adjustability for those precious 3sec/40k any day.
But for those who don’t know their way around a barrel adjuster? It’s worth huge advantages!
I cannot think of any significant advantages to internal cables.
Aerodynamics: cable housing can be a very significant source of drag. Bare cables are better, but then you have to consider the impact of the additional ‘tiddly bits’ (as English framebuilders used to refer to the various braze-on fittings of steel frames, e.g., cable stops).
With those skinny cables, I can’t believe that they really are that much of a factor aerodynamically, either, but I could be wrong.
Chet Kyle gives the drag of cable housing as 18 g/in (yes, he used those mixed units).
true, but how much cable housing is removed from the equation with internal cables? Virtually none, and for some designs there is actually more cable housing than for non-internally routed designs (thinking of my old P2 that used cable housing around the BB for both derailleu cables)
I don’t know from any windtunnel data, but as an example, the trussing on a old style biplane produces more drag than the combined wings do during low angle of attack flight (alpha <= 5deg, if I remember correctly). This was a homework problem during one of my classes on aerodynamics. Just something to consider. Round objects, such as cables, are a bluff body, which can be a significant source of drag, even if they are very thin.
true, but how much cable housing is removed from the equation with internal cables? Virtually none
Now that entirely depends on the design, now doesn’t it? Hooker bikes, for example, routed the rear brake and both shift cables all the way through the aerobars (or through their road stem) and into the frame behind the headset, while the front brake was actuated by a bare cable. The result was that absolutely no cable housing was visible when the bike was viewed from the front.
I think, just guessing though, that the aerodynamic effect of running internal cables would be considerably less than the effect of swapping an 18 spoke front wheel for a 16 spoke front wheel, (keeping the rim shape, hub shape, and tires constant)
.
I don’t disagree with you there. Most currently available frames do not go to the same extremes as Hooker. The closest would be Trek, with the cables entering the frame behind the steerer.
I think, just guessing though, that the aerodynamic effect of running internal cables would be considerably less than the effect of swapping an 18 spoke front wheel for a 16 spoke front wheel, (keeping the rim shape, hub shape, and tires constant)
You do realize, don’t you, that there are actual data on this, and that they don’t support your guess?
Sure - the love affair you will have with your bike after to do the cable routing yourself on your damn Cervelo Seriously, I’ve done my SoloistC myself, and while tough, it is the little details (super thin headtube, etc. - compare the headtube on your SLC w/ the huge Kalibur one becuase of the bottom race), internal cable stops that Cervelo has, etc. that do add up to an aerodynamic difference. If we don’t care about that, then no real reason reason to run a Zipp 404 over a Ksyrium then. As for cable “stretch” - use aramide ones
While looking at Guru’s new Cron’Alu at the lbs, and comparing it to mine that’s a couple years old I noticed the new internal routing. I was there to buy a cable cutting tool, and the guy at the shop told me if I had the internal routing I’d probably end up having to pull the bb to change the cables. I have larger aero problems than cables, and am happy to have a bike that’s a little easier to maintain.
Mate, you’ve found it spot on. The mechs run internally are nothing but a bother to me. I reckon they are more bother than better. I say you’re right- external is better.
Badger, I’ve replaced the cables on my Chron’alu, and yes you need to pull the BB. Pretty tough, but if you have some time it’s kinda fun. Much harder than replacing cables on an externally-routed bike.
With those skinny cables, I can’t believe that they really are that much of a factor aerodynamically, either, but I could be wrong.
Chet Kyle gives the drag of cable housing as 18 g/in (yes, he used those mixed units).
I realize I’m responding to a 5-year old post, but this has been quoted several times on ST and has never added up for me.
Cable housing has an OD of less than 0.200". Assuming a Cd of ~1 for a cylinder perpendicular to airflow with Reynolds number of 3000, a 1-inch long cylinder has a CdA of 0.000130 M^2. Using your ROT, this should be worth only 130mg of drag, not 18g. Is that Kyle figure off by a factor of 10?
This leads into a related question regarding housing-less center-pull brakes. Does eliminating, say, 4 inches of housing in a brake setup (a la Hooker) have a measurable CdA reduction? (Using the .000130 M^2/in number, the CdA reduction is only .0005 M^2 – worth maybe 0.5W. And, that ignores the fact that a sheathed cable is not entirely out in the wind, but is partially shielding the headtube.)
With those skinny cables, I can’t believe that they really are that much of a factor aerodynamically, either, but I could be wrong.
Chet Kyle gives the drag of cable housing as 18 g/in (yes, he used those mixed units).
Cable housing has an OD of less than 0.200". Assuming a Cd of ~1 for a cylinder perpendicular to airflow with Reynolds number of 3000, a 1-inch long cylinder has a CdA of 0.000130 M^2. Using your ROT, this should be worth only 130mg of drag, not 18g. Is that Kyle figure off by a factor of 10?
F = CdA * 1/2 * rho * v^2
= ( 1.27e-04 m^2) * ( 0.5 * 1.2 kg/m^3 * 12 m/s * 12 m/s ) @ 12 m/s
= 0.0109 N
Not sure why one would want to translate this to grams, but multiplying by velocity will give the power lost:
P = 0.0109 N * 12 m/s
= 0.132 W
This leads into a related question regarding housing-less center-pull brakes. Does eliminating, say, 4 inches of housing in a brake setup (a la Hooker) have a measurable CdA reduction? (Using the .000130 M^2/in number, the CdA reduction is only .0005 M^2 – worth maybe 0.5W. And, that ignores the fact that a sheathed cable is not entirely out in the wind, but is partially shielding the headtube.)
With those skinny cables, I can’t believe that they really are that much of a factor aerodynamically, either, but I could be wrong.
Chet Kyle gives the drag of cable housing as 18 g/in (yes, he used those mixed units).
Cable housing has an OD of less than 0.200". Assuming a Cd of ~1 for a cylinder perpendicular to airflow with Reynolds number of 3000, a 1-inch long cylinder has a CdA of 0.000130 M^2. Using your ROT, this should be worth only 130mg of drag, not 18g. Is that Kyle figure off by a factor of 10?
F = CdA * 1/2 * rho * v^2
= ( 1.27e-04 m^2) * ( 0.5 * 1.2 kg/m^3 * 12 m/s * 12 m/s ) @ 12 m/s
= 0.0109 N
Not sure why one would want to translate this to grams, but multiplying by velocity will give the power lost:
P = 0.0109 N * 12 m/s
= 0.132 W
This leads into a related question regarding housing-less center-pull brakes. Does eliminating, say, 4 inches of housing in a brake setup (a la Hooker) have a measurable CdA reduction? (Using the .000130 M^2/in number, the CdA reduction is only .0005 M^2 – worth maybe 0.5W. And, that ignores the fact that a sheathed cable is not entirely out in the wind, but is partially shielding the headtube.)
Interesting question.
Thanks, Andy. The Kyle citation was in g/in, that’s why I used grams. It looks like our numbers pretty much match. I wonder if Kyle’s original figure was off by a factor of 10 and should have been 1.8g/in…
Seriously, I’ve done my SoloistC myself, and while tough, it is the little details (super thin headtube, etc. - compare the headtube on your SLC w/ the huge Kalibur one becuase of the bottom race), internal cable stops that Cervelo has, etc. that do add up to an aerodynamic difference. If we don’t care about that, then no real reason reason to run a Zipp 404 over a Ksyrium then. As for cable “stretch” - use aramide ones