peterws

all our signposts say stuff like 1 in 4, or 1 in 6. 1 in 6 is hard work, 1 in 4 is suicidal if your going downhill . . . I have had some scary mopments on those, on bikes and in cars.when the handbrake doesn't hold.going up, and you meet someone coming the other way!

Bald Paul

I use Garmin Connect whenever I'm going on a ride where the route is published beforehand, usually in RWGPS. Just export the RWGPS file as .TCX and then import it into Connect. The climbs and rest stops will show up on the Garmin map, and when you send the route to your unit, they will show up as you ride. ClimbPro will tell you the gradient of the climb and how much longer you'll have to suffer.

PeteHski

This also works if you just import a route directly to your Garmin head unit. It doesn't need to go through Garmin Connect. I really like ClimbPro. Knowing what is coming up ahead and how far you have left to climb is far more useful than knowing the gradient that you are currently on.

ofajen

Around here, really steep stuff is short as we don’t have big elevations. Just pointing out that single speed forces high pedal force on the hills and around 15% the pedal force will lift me out of the saddle. That’s a lotta watts and something I only do for five to ten seconds.

I ride with platform pedals and shoes I can comfortably run or walk in. There are certain routes that have hills that I just get off and walk. One in particular is very tall for this area, very steep and loose gravel. I’ve climbed it in 40x32 but traction is a really challenge. Kind of a relief with SS to just get off and walk.

Otto

John E

How about Fargo St. near downtown Los Angeles? 32%

https://losangelesexplorersguild.com...%20the%201970s.

They have an annual bicycle climb event. I would not even think about riding down that thing.

Update: the city hosed us over by changing this to one-way downhill, as they did to my favorite climb, Tuna Canyon in Malibu, destroying that one, as well.

rm -rf

Grade profiles!
Ha, these look extremely easy when drawn. Using the profiles vs real life:
5%: Will I even notice this? vs "a real hill. breathing harder, set a maintainable pace here."
10%: Easy. vs "no more lower gears. How long is this section?"
20%: Might have to use some effort on this. Not too difficult. vs "Using my 1 minute max power. Hope I reach the crest by then!

rm -rf

Finding grades
ridewithgps.com can show average grades on a selected portion of a climb. It can show max grade and grade at a point in the route, but those can be inaccurate, due to using known data points from the mapping to estimate a point in between those. Usually reasonably close, though. The average is quite accurate most of the time.

Like this example. I dragged a portion of the red elevation chart to show it's stats in the box, and moved the cursor to find the local grade there (which can sometimes be anywhere from 8% to 12-13% in real life.)
8.9% isn't that hard for me at 169 feet tall. If it was 400 feet, I'd have more trouble maintaining my pace up the hill.



~~~~~~~~~~~~~~
Accurate spot grades with a phone app
A couple of years ago, I was interested in a couple of the "steepest climb around" hills. These are very hard to measure accurately with a GPS device, since the bike is going slow, there's often tree cover to interfere with GPS, and any small data error will throw off the calculations significantly.

I have a bubble level app on my phone. It's accurate to well within 1% grade, tested using my accurate wood shop 4 foot level. I have the phone sitting on it and the level shimmed so the bubble is exactly centered.
Next, I set the bike on a known level floor. The top tube slopes to the back, so I zeroed the app on an exactly level tabletop,then set it's bottom edge on the top tube near the stem.

It read "14.3 percent" with my bike level. So I can hand tilt the phone to 14.3%, the press the Zero setting. now it reads "zero %" correctly when sitting on the bike, and I just stop the bike, feet on the ground, phone on the top tube, and read off the actual grade at that point in the road.

(That steep climb was measured when I rode it downhill and stopped to measure. 21% part way down, 22% nearby. Ha, it's not easy to come to a complete stop on a hill that steep!)

rsbob 

You are right. I confused altimeter with gradient. The Avocet did not display gradient, only total gain or loss.

andrewclaus

After a short but steep climb in the Tucson area, a visiting friend asked what the gradient was. I had no idea, so I guessed 10%. He disagreed, saying it felt more like 13%. When we got home, I checked the state highway website, and they actually listed the gradients of the steepest roads in the state, probably for truckers, RVers, etc. It listed this particular road (Gates Pass) at 13%! I was impressed with my friend's guess.

PeteHski

If you do a lot of climbing and are aware of what various slope gradients feel like then it is quite easy to guess based on your effort and gear ratio. I think I could guess within a couple of percent, although it gets harder to guess above 20% where everything just feels super hard!

Polaris OBark

Mine lags by about 60 sec. Enough to be annoying.

PeteHski

Yeah mine lags too, but not by that much! Seems more like 5-10 seconds max. Maybe I just climb slow 😆

terrymorse 

To be fair, determining grade from the data is not easy. The GPS data are noisy, and even if it were perfect, the best you can get is the grade behind you.

I wrote a tiny web app to show elevation and grade, so I could tell the missus how hard our hike was. Noisy data makes for noisy grade estimates.

https://myelevation.terrymorse.com

RChung

It's not that hard if you have an accurate power meter and speed sensor. Much better than GPS.

PeteHski

Yeah I’m not impressed with GPS live speed on my Garmin. But not bothered enough to run an additional speed sensor.

PeteHski

That’s a good point about it showing the grade behind you. I guess that explains the apparent lag. Other than that the readings I get seem to reflect known grades on climbs. But ClimbPro is what I find actually useful in showing what grade profile is coming up ahead of me.

terrymorse 

Hmm. So you'd use the power vs speed ratio to determine grade? What about wind?

I knew a guy who was working on a device using 2 accelerometers and a speed sensor to display grade. I don't think he ever got it to work.

RChung

I came up with virtual elevation (and virtual slope) 20 years ago back when I was trying to figure out a way to assess the quality of power meter data. Since then, people have mostly used it to estimate CdA and Crr but I can pretty reliably spot small changes in gradient. Here's a plot of the calculated gradient along the Bay Trail in Berkeley alongside I-80 from University Ave south one mile. This was for an out-and-back on a "flat" trail, and there was a tiny tail wind on the out (headwind on the return); I attribute the slight asymmetry in the two halves of the profile to that. Note the scale. As an aside, I hadn't noticed that first dip at about 150m from the start/end. That's about a 25cm "local minimum." I was skeptical, so I went out there the next day, rolled out 150m from the start, and saw that there was dried-up silt at that point. That's where water was collecting.



I can also differentiate turns from the straights on velodromes. A friend has been using RTK GPS to validate the virtual slope calculations. So far so good.

big john

Measuring down to the meter? I'm thinking for a big climbing ride there would be a lot of movement in the line but it would look like noise in the overall profile?

RChung

Depends on the scale you're interested in. I was showing a flat ride so you could get a sense of the precision. Here are two other plots for hillclimbs.

The first is from the 2002 and 2003 New Year's Day San Bruno Hillclimbs (I told you I was doing this 20 years ago). The rider was the same, but 2003 was much windier. San Bruno is pretty open, there's no tree cover so it's pretty susceptible to wind. Nonetheless, you can see how the profile derived from power and speed match.




The second plot is from a climb of Mt. Palomar, along with elevation from the barometer of an Avocet 50. You can see that this works much better on the climb than on the descent, where the rider was using his brakes.

big john

Very interesting stuff!

Iride01 

Grade is much like Calories. I never understand why people expect the device to show them accurately, and especially instantly. Or even consistently between any two rides.

I-Like-To-Bike

Possibly for the same reason some people (probably the same ones) seem obsessed concerned with collecting and recording all kinds of detailed route data and/or personal performance "stats" for every ride, as well as the various devices they can buy and use to better collect and record that presumably all important data.

RChung

Yeah.

In a big picture kind of way, they want grade and elevation change so they can see how hard a ride is, was, or will be; or to communicate to someone else about a ride's difficulty. But if you have a power meter you don't really need grade or elevation change: power tells you much more about how hard it is, or was (not so much will be). Power for speed is a direct measurement of how hard it is, and those get updated every second with pretty high precision and accuracy, so "translating" power and speed into grade is an easier problem than trying to translate approximate grade or elevation change into difficulty.

terrymorse 

Well, power tells you how much power you were putting out. How “hard” it was to generate that power seems grade dependent. 4 W/kg at 6% feels a lot easier than 4 W/kg at 15%. To me.

I just finished some climbs with 15% sections, and my quads feel it. Aerobically, not so bad.