The effect of hills on average speed?

[BSRU]

"Mildly hilly" accurately sums up my commute, generally between 10 to 15 m per km.
Although if I exclude the start and finish in the town which is flat it would push the out of town part to nearer "hilly".

 

[GrasB]

Opps, just realised the last band is 20-30 not 20-25

 

[Nigelnaturist]

so its hilly then lol

which makes the sportive i did onthe weekend hilly too......and very nearly in the very hilly category, no wonder i was tired at the end

But as GrasB points out over a longish route, I did 42.2miles yesterday and got 62.67ft/mile or 67.88Km with 11.62m/Km and that included a Cat3. it was the ride to it and home that brought it down.

 

[Globalti]

Oh, so it's the climb over the entire ride then?

 

[ColinJ]

Oh, so it's the climb over the entire ride then?

Yes!

The local Season of Mists audax is what I would call extremely hilly ...

... about 2,500 m of climbing in its 100 km, which fits in pretty nicely with GrasB's ratings.

 

[jowwy]

But as GrasB points out over a longish route, I did 42.2miles yesterday and got 62.67ft/mile or 67.88Km with 11.62m/Km and that included a Cat3. it was the ride to it and home that brought it down.

it was 4950ft of climbing over 49.1 miles which equates to 18.99m/km

 

[Nigelnaturist]

it was 4950ft of climbing over 49.1 miles which equates to 18.99m/km

Be about right then as mine yesterday was 61.36ft/mile (11.62m?Km) and 2588ft over 42.2 miles.

 

[PpPete]

To get an idea of your ride time on a hill climb you need to do is get a good idea of your nominal altitude gain (VAM) an hour, your flat road still air cruising speed (V), then also know your total distance (D) & total elevation gain on the ride (EG).
D / V + EG / VAM
For arguments sake say you have 500m/h VAM & 25km/h V. You're doing a 40km ride with 200m climbing
D / V = 40/25 = 1.6h
EG / VAM = 200/500 = 0.4h
1.6 + 0.4 = 2h
So 20km/h (or 12.5mph).

This will always over estimate your time if you're doing a ride that include down hill sections as it is extremely hard to estimate your down hill speed. Put it this way, descend like a demon & be sprinting out of the corners to get up to Vmax ASAP on the descent (5% will get you to 25mph fairly easily on the hoods a road bike) will have a huge impact on your average speed rather than gently feathering the brakes keeping the speeds under 25mph.

I've always reckoned that the effect of downhills on the average speed comes close to balancing out stretches where one is held up at junctions, deceleration/acceleration time being quite significant.
I've been tinkering with spreadsheets to try and estimate timings over long audax rides but the whole thing goes a bit a pear-shaped when you start working with VAM numbers.
Looking at my stats from the recent BCM 600, my VAM (from Strava) on the Saturday was averaging around 480, on the Sunday around 425 (pretty sh!t, I know) but working things backwards on my spreadsheet (which does GrasB's calculation for me) so that it corresponds with the times I actually took on each leg.... my V on the Saturday would have to be a highly improbable 48km/h and on the Sunday, an only slightly more plausible 36km/h.
I know there some cracking descents, the one on the A470 after Cross Foxes down to Dinas Mawddwy for example, provides a couple of km at speeds above 60km/h but the effect on the overall average is pretty negligible:
My average speed from the top of that particular descent to the control at Aberhafesp (50 km) : 22km/h
from the bottom of the descent to the control: average 21 km/h

 

[redcard]

I've always reckoned that the effect of downhills on the average speed comes close to balancing out stretches where one is held up at junctions, deceleration/acceleration time being quite significant.
I've been tinkering with spreadsheets to try and estimate timings over long audax rides but the whole thing goes a bit a pear-shaped when you start working with VAM numbers.
Looking at my stats from the recent BCM 600, my VAM (from Strava) on the Saturday was averaging around 480, on the Sunday around 425 (pretty sh!t, I know) but working things backwards on my spreadsheet (which does GrasB's calculation for me) so that it corresponds with the times I actually took on each leg.... my V on the Saturday would have to be a highly improbable 48km/h and on the Sunday, an only slightly more plausible 36km/h.
I know there some cracking descents, the one on the A470 after Cross Foxes down to Dinas Mawddwy for example, provides a couple of km at speeds above 60km/h but the effect on the overall average is pretty negligible:
My average speed from the top of that particular descent to the control at Aberhafesp (50 km) : 22km/h
from the bottom of the descent to the control: average 21 km/h

See post 3

 

[PpPete]

See post 3

so you saying that on a flat course I couldn't ride at pro peleton speeds ?

 

[GrasB]

I've always reckoned that the effect of downhills on the average speed comes close to balancing out stretches where one is held up at junctions, deceleration/acceleration time being quite significant.
I've been tinkering with spreadsheets to try and estimate timings over long audax rides but the whole thing goes a bit a pear-shaped when you start working with VAM numbers.
Looking at my stats from the recent BCM 600, my VAM (from Strava) on the Saturday was averaging around 480, on the Sunday around 425 (pretty sh!t, I know) but working things backwards on my spreadsheet (which does GrasB's calculation for me) so that it corresponds with the times I actually took on each leg.... my V on the Saturday would have to be a highly improbable 48km/h and on the Sunday, an only slightly more plausible 36km/h.
I know there some cracking descents, the one on the A470 after Cross Foxes down to Dinas Mawddwy for example, provides a couple of km at speeds above 60km/h but the effect on the overall average is pretty negligible:
My average speed from the top of that particular descent to the control at Aberhafesp (50 km) : 22km/h
from the bottom of the descent to the control: average 21 km/h

I'm probably using a slightly odd definition of VAM. In that I think of VAM the fastest I could climb up a vertical face on my bike in a vacuum. If that is assumed it works.

 

[PpPete]

I'm probably using a slightly odd definition of VAM. In that I think of VAM the fastest I could climb up a vertical face on my bike in a vacuum. If that is assumed it works.

That's the way I thought of it too..... and so was hoping to use VAM data (with a bit of arithmetic) in a similar way to Naismith's formula.
But looking more closely at the Strava VAMs, they seem to divide the ascent mtrs by the total time (which of course includes both horizontal and vertical components).
Some sort of fudgeration factor would appear to be needed. Based on a very limited dataset I think dividing by something like 1.67 will convert Strava VAMs to what you and I think of as "true" VAMs.

 

[GrasB]

VAM normally is taken to be climbing ability up a hill of around 20-25%. Your VAM increases as the road speed decreases (or as the hill gets steeper, but it's not the steepness of the hill that's key, it's the reduction of aerodynamic drag).