Interesting side bar on that bar

A couple of weeks ago we aero tested a lady that had those bars. She seemed relatively flat and we angled her up about 7 degrees.

She ended up considerably slower. Re-run. Again
So we decided to bring her back down to get back to our A->B->A and the second A was even a bit faster than the original....hmmmmm......

Post ride I checked the accelerometer in the device and sure enough the "2nd" A was a smidge below the first.

So in the end she was maybe 1.5deg, 8.5deg, 0deg and the numbers were better, flatter.

Now we know that tilt is very individual, but it was quite pronounced. Like in 0.015 in her CDA (by memory)

I rarely see such a degradation going from 0ish 8ish degrees.

It was with that bar on a speed concept. I wish I would have spent more time on it.

I can post some of the charts if you're interested.

It's so complicated, it's not just the forearms, but the upper arms triangle too (elbow spacing) and head as well. There is the extensions to be considered and how they interact with one another, the forearms, the pads and the stem/riser(s)).

For two cylinders of equal diameter research suggests there is no interference drag at spacing greater than 4 diameters. It is a curve obviously, but drag peaks around 2 diameters and goes down from there.

I don't know how this would apply to cylinders of different diameters (peak drag at d/D x 2D where d is the small cylinder diameter and D the large one? Just a guess/hypothesis) or other shapes and at different attack angles.

This is best evaluated in testing, but for cyclist I think we operate in the 1 to 3 diameter range a lot so closing it all up can make sense. You create more pressure drag that way, but it can be worth it.

At what speed was the flat, angled, flat tested?

20ish mph, good headwind/tailwind
55-59yo female AGer (I think)

I was under the impression for slower speed flat is almost always fastest

I have never heard that. Do you have a source ?

On the other hand, I have seen many cases around the 36ish kph where tilt (slight) is faster.

Edit : I just checked, she was 5:20IM pace, around 34ish. km/h

I was reading through the arm tilt thread the other day and it was pretty definitive that you MUST test individually to know what is fastest. Some folks had no difference tilt to no tilt. Some folks had the same at flat and 20° with 15° fastest than either.

I think there's too many other variables at play, sometimes tilting allows you to move your shoulders into better or worse positions. Sometimes it creates weird air pockets between the body and arms. Sometimes it disrupts air before it hits your helmet.

I think we'll see folks like dd and the ero guy(Jim?) come into threads and give somewhat solid advice and this works for 75% + of folks advice but then it really just comes down to testing on an individual level.

I've tried something similar I think to what you're suggesting, have not tested it yet though. Arms are already very narrow (fore arms touching eachother all the way from wrist to elbow) so I hope this "shielding" will help a bit.

You won't have a hard time convincing me on the need for testing :-)

Absolutely. Without testing you are throwing the dice. Which if what Fredly says is true, ie he didn't even test that tape job, blows my mind.

You are absolutely right on tilt being individual. But usually, which such small tilt, it's rarely significant. In this case it was so reproducible and A->B->A it was surprising. As I said I wish we would have spent more time on it. She was the 105th or so test.......

A few years ago I was doing some work for a cycling team. At a training camp in Valencia we had a gentleman named Yago come to test the whole team. He is a bit of a legend in track aero testing, working for many of the world tour teams. He tested every rider on the team and all the riders had very small numbers between 0 and 10. A few tested fastest at 0, most around 10. I remember him telling me the sweet spot, ie where most riders tested fastest was around 12. But of that whole test, the best to worst range was still quite small.

My curiosity and I have nothing to prove/disprove it, is if that bar behaves differently, based on my super size data set of 1.

I'd be shocked if he didn't test it, he tested his socks so he clearly has access to a tunnel.

I just ordered a 20° uptilt for my bars going up from 10° which currently feels fine but I feel even better when I raise them more so hoping I don't see too big an aero penalty as you say is *usually* the case. :)

Thanks for the idea. I don't think I could do around the forearms since I'm more using an open style cockpit and have a BTA to worry about.

Though is that just duct tape folded in on itself?

Haha more or less yes. I have the Revolver Ergo Mono Mantis cups which are a bit higher and closed off at the back so I could easily tape an "outside" layer on there. And then I just taped on the inside of that tape so that the sticky part was covered. Also in the hopes that it doesn't come off the first time some water drips in between there.

Ditlev had something similar going on on his Felt IA before he got the Scott sponsorship.

Dang, there's a lot of hearsay and misinfo on this thread (comment not directed at you by the way).

The mistake I see over and over again when people "test" or "fit" tilt is that's all they do...tilt the aerobars. There are a minimum of three adjustments that must be made when tilting the forearms. When done correctly, it's more comfortable and faster for the majority of athletes no matter the speed. You can achieve similar numbers in a flat forearm "superman" type position, but that's not nearly as comfortable nor sustainable. How much faster is athlete dependent. Also, I pay no attention to any yaw numbers above 10 degrees and, really, 7 degrees. I've gathered enough data from numerous locations in significant wind conditions to know the majority of riding time is spent at 7 degrees or below. Varies with speed, of course, but most of the people interested in aero testing are fast enough to be 5 degrees or under for the majority of their saddle time. You'll see puffs of wind above that, but nothing sustained unless you're riding in very unusual conditions and, even then, only until the road turns away from that wind angle a bit. This is not just me observing these results, the top fit and aero specialists find the same thing, which is why most of the top cyclists tilt their forearms. It's not a coincidence folks.

I find multi-sport athletes are generally not very good at aero testing mostly due to monetary restraints, and I think many of you would be surprised how many pros just follow along without testing at all.

Lastly, when I see a result that doesn't make sense, I retest - especially when the initial data indicates a larger difference than I'd expect. You don't take the numbers at face value, you try to figure out why. The best way to get your answer? Retest (and observe the athlete riding in that position if possible). I've said it before, and I'll keep saying it. If you "test" something one time, you haven't tested it, you've only looked at it. Can you repeat the numbers? Did you? If not, I would not consider it a valid test. The more data you gather, the more accurate the result. This is the problem with going into a wind tunnel, velodrome, or outdoor session trying to test as many things as you can as quickly as you can. That's just bad procedure, but it's the mistake I see over and over again. I made those mistakes early on a decade ago, but learned over time as we all should.

---

Jim Manton / ERO Sports

I'd be curious to hear what specifically is misinformation in your opinion.




One very competent tester/fitter once told me, what he does with the hands is more to do with what he can do with the head and shoulders. But he does get a feel for what individual change does before attempting multiple permutations combinations of changes. So to oversimplify, he will do up down, tilt no tilt, in/out, shrug/turtle and then find the ultimate combo of the above. I'm good at measuring how much salt, how much pepper, how much garlic......I am in awe of the chef that knows how to take those measures and make the perfect dish. I wish I could do that.


Sub 5 degree is great, but I don't see it as the norm or a condition for testing. Sure you can test at hight speeds, but that isn't the speed many athletes test at. Let's not forget that 22.5mph is a 5hIM. Testing only at hight speeds/low yaw seriously limits the time you can test. Getting people to bail on testing because it's too windy is very limiting. I often see very reasonable results at 12 and even 15deg. Then we go back to if you have stats on the confidence of tests you can make informed decisions even at higher yaws. Limiting to low yaw testing disqualifies many people from testing and many test opportunities IMO. Wildly varying test conditions is another story.


Yes and no. There are ways to extract data from the data and get confidence without having to retest ad nauseum. Standard deviations are one way. Virtual elevations to diagnose anomalies is another. There are several. A good product will provide good diagnosis tools

You misunderstand. I test in all conditions, but what I and most who have the ability to measure yaw have found is that the majority of time spent riding is in low apparent wind direction when combining the cyclist's speed and ambient wind direction. What I was referring to, and probably explained poorly, was that I don't make position or equipment decisions based on yaw angles above 7 degrees because I know that over 50% of your time will be spent at or below that yaw angle. There are exceptions to everything, but in general I've found it's a mistake to consider higher yaw angles because few athletes experience sustained time at angles higher than 7 degrees (again, at least the athletes who come to us for aero testing which tend to be faster than average).

If a test results in an odd result, yes I could go in and analyze the data to pick out what I want, but that is not good scientific method in my opinion. You're inviting error looking for data you deem correct when the better option is to gather more data. It's not a problem to identify the potential source of the problem, but then I prefer to retest. And not just the option which produced the odd number, but the option we're comparing it to as well. Because who's to say that initial test is not the outlier? Again, if you can't repeat the numbers within a reasonable margin of error, you don't have a good test. Error could be caused by something simple like a brake rubbing, a faulty power meter, or it could be the position being attempted isn't sustainable over time. We see this in our testing. For instance, and athlete might be attempting to maintain a low head position, but as the test goes on over several minutes you see the drag number rise as their head begins to rise. The position simply isn't sustainable (though perhaps could be with more training). I was working on head position with a former World Champion time trialist in the last year or so who was convinced his head position was excellent throughout his TT in a grand tour. Then I showed him the video of his race and he realized his head was moving all the place: up, down, left, right...there was very little consistency. It's easy to keep your head in a specific position in a wind tunnel, or even on a track when that's all you have to focus on, but test for 15 minutes or so on the road and the truth is often revealed.

We always test at race pace because to do otherwise is inviting false, underinflated, or overinflated numbers. Air flows differently over the athelte at different speeds. Testing below race pace will not produce accurate results. At least not to our standards.

---

Jim Manton / ERO Sports

Testing does need to happen at race pace indeed. When looking at the literature it shows there is a significant difference in pressure drag profiles between 9 and 13 m/s, and extending below and above this, following a power law for a given position/kit. So with around a 15% difference in drag profile happening across that range it only makes sense to test and optimize for race pace.

The issue with only testing though is you see what's getting you that result, but do you understand why it is happening? Or at least can it help you refine your understanding of what's happening? To me this is critical. Start with a theory of how everything fits and interacts together and then empirically refine it.

This is such a captivating topic. I'm thinking about putting something together to share general tenets of bicycle aerodynamics and detailed analogies and examples. These boil down to filling in the voids, minding the gaps and tripping what you can't fix.

Understanding the why is sometimes very obvious, and sometimes not, but it's the best part of the process. This thread is a great example. Why does tilt lower drag? People are talking about how it may or may not affect air flow around the athlete. There's no definitive answer. I tend to lean toward the simplest explanation. Tilt, done properly, narrows the shoulders and pre-crane's the head making it more natural to keep the head lower. Boom, simple. And it may be that simple, but it may not. As with everything, it's athlete dependent, but it sure is fun finding consistencies and developing theories. Theories which, of course, break down anytime you're dealing with feet/lower legs. :-)

---

Jim Manton / ERO Sports

There's no denying it affects the airflow around the athlete. Minimizing A is very important, and it's easily checked. Take a picture from afar against a background and do a pixel count. But I guarantee you there are changes from tilt alone.

Going back to laidlow's mod, I'll give another analogy. What's the fastest way to reduce a car's Cd? You put a piece of cardboard over the radiator grille. Why? Because of the interference drag mess behind that grille on a typical car. You can design a car nicely so that air flows through the rad and exits in the low pressure zone over the hood, that's optimal, but if you can't or won't make it clean, you'll need to block it to improve Cd (forget you need to keep the engine cool fo a second). The same logic can be applied to bike cockpits in my opinion.

Regarding the leg, the principles still hold. For example, why is a lower stance width more aero? It's the same projected frontal area technically. The answer is with the interference drag as well here. There is quite a bit of interaction through the stroke with the seat tube, the stays and part of the top tube. This interference happens over the width of your stance. You can think of it as the effective frontal area in this case if you will. Because of interference, the gaps you see in the frontal area don't effectively count.

The leg is very interesting as there is not as much you can do with it compared to the rest of the body, it has to go along with the crank. So, you trip what you can't fix to reduce Cd.

It's super interesting in F1 the way they combine the need for cooling airflow with aerodynamics. A big hole on either side and a completely sealed, very large, tilted radiator panel that forces the entirety of the airflow to go through it Then dumping it out on an annulus around the exhaust pipe right in the low pressure zone under the rear wing. (There was some trickery a few years back with having the exhaust blow into the diffuser to help lower the pressure under the car, but that's probably beyond the scope here)

I aero tested this today.

Not super sophisticated, cost a few watts.

Your sensor is severely interfering with what you're trying to measure. Any way you can mount it near a brake? Keep taping further below and then add a "behind the arms" bottle like Sam. As I've said, when you close up the front you get rid of interference drag at the cost of pressure drag. Getting rid of interference drag does not necessarily reduce total net drag on its own. Now pressure drag is simple to deal with, just fill the low pressure zone in the back, with anything. Sam had a bottle there, but you can add stuff to make it better; gels, a taped up bag with your tools/flat kit, anything goes. Bento box is good too and while you're a it stuff a bottle down your jersey placed longitudinally with the flow.

I did this for hoots on a training run. Strap it on, ride 5km out& back 6 times. Very consistent numbers.

As for placing it on the brake, placing these devices offset to one side or another leads to poor measurement in yaw conditions.

As for the interference, the sensor has been testing multiple times in the A2 tunnel to not interfere with the overall system drag, in multiple positions in the front of the bike. There are multiple mounting options, I rarely do anything but the basics for a "for hoots run".

There are pros and cons of several positions as people that have tried placing it on the basebar and a pole two feet out have discovered.

If I had the slightest hope this would yield positive results, maybe I'd try.

That picture to me that isn’t the way Sam has it

Not that it would make any difference. I don’t know

I have a similar set up to Sam. Have had it for many years now. It starts with one piece of tape from underneath. And then add a few more

Yours looks like it is from the top?

I put a couple gels and my Wahoo on it. They stay because the sticky side is up

Above that I have a BTA

I am having flashbacks to when I was 7 and a guy told me I was taping my hockey stick wrong going from front to back rather than back to front :-)

Mine was literally done in 2 minutes. Roll of duct tape, top to bottom, then a piece along the center. I didn't even stop the Garmin while taping.

Maybe there is a good explanation why it's slower, but it's very consistently slower but a small amount and then back to baseline when I removed it. The consistency between the runs was exceptional.

Rumor is we may get a last day of warm weather tomorrow. Maybe I'll be a little more diligent about taping

Lol Re taping the hockey stick wrong

Made me laugh

I just slapped it on a few years back to carry stuff. I figured it probably made it a little slower but not enough to worry about

I use packing tape. Start underneath at the bottom. One piece up to the shifter on the end. Then I put one on each side of it. I don’t wrap it

Takes maybe 30 seconds??

Sam’s is the only other one I have seen like that