In-depth discussion about "slope mode" - flattenting power sensitivity around a desired cadence

[John Wurts]

In summary: I'm looking for a more in-depth discussion and derivation of what happens during slope mode, specifically how a set "slope" maps to a resulting power when operating in "manual slope mode".

More context:
I'm interested in "flattening" my Wahoo Kickr's required power versus cadence in the viscinity of a desired cadence and set power. The specific application is I'm trying to do a maximal effort interval, such as 20 minute effort, at a target cadence around 95 rpm. I can't complete this in erg mode as I can't stop to adjust intensity, and in manual slope mode I find the resulting power at 93 or 97 rpm to be too large a difference compared to 95 rpm, hence the desire to flatten it. Unfortunately, I've found shifting results in too large of a change in rpm to achieve my desired effort. As a sanity check, shifting from the 20 tooth to 19 tooth ring results in a 5% reduction in cadence for a constant bike speed, and hence constant power, causing this cadence issue.

My approach was to use slope mode and unrealistically change my weight, CdA, Crr, gradient, and other parameters to achieve my target wattage at a selected velocity, which for a non-shifting ride results in a selected cadence. However, the closest I've been able to identify Golden Cheetah's ability to relate this model is in the Options->Data Fields->Processing->Estimate Power Values. However, https://groups.google.com/g/golden-cheetah-users/c/vAhgiHnY5LI indicatres this functionality is focused on estimating power from an outdoor ride, not using a road gradient to derive velocity. I found a "Virtual Speed" discussion (https://github.com/GoldenCheetah/GoldenCheetah/issues/229), but I can't find other documentation for how the "Virtual bike for simulated rides in Train mode" is calculated.

In reading the source code, the closest I could come is in https://github.com/GoldenCheetah/GoldenCheetah/blob/cabe078453cbe8e136b8adf8d4187a5be878671b/src/ANT/ANT.cpp#L383 , which suggests slope mode simply sends a gradient to the ANT+ FE-C. Line 401's todo suggests there is something in the works where GC will calculate the power load for a given gradient, Crr, wind, etc. This might be the virtual power disucssion prior, as a closed-loop feedback for adjusting target power is not implemented (to my knowledge).

As an added complication, Wahoo Kickr looks to have 4 operating modes (https://support.wahoofitness.com/hc/en-us/articles/204281764-What-Are-the-Different-Modes-On-My-Smart-Trainer-), which I'm not sure how GC interacts with. Best I can tell GC can operate in erg mode, publishing a target power, or operate in a "levels" mode where GC publishes some form of resistance (https://github.com/GoldenCheetah/GoldenCheetah/blob/cabe078453cbe8e136b8adf8d4187a5be878671b/src/ANT/ANTMessage.cpp#L999). It is unclear to me how a negative slope in GC translates to resistance curve for the Kickr.

One issue that has become apparent to me is some ambiguity regarding slope, gradient, and resistance. The wahoo documentation suggests the "set resistance" scales the "level" for the kickr's resistance (https://support.wahoofitness.com/hc/en-us/articles/204281734-What-Are-the-Power-Curves-for-Level-Mode-), but I don't know if the ANT message coming from GC is structured to do this. Another ambiguity is the term "slope" used by GC train mode may not have the same meaning as "gradient" used virtual speed or "slope" used by wahoo.

So, is it possible to effectively flatten the power sensitivity around a desired 95 rpm cadence?

[Ale Martinez]

El viernes, 16 de septiembre de 2022 a la(s) 05:14:36 UTC-3, John Wurts escribió:

In summary: I'm looking for a more in-depth discussion and derivation of what happens during slope mode, specifically how a set "slope" maps to a resulting power when operating in "manual slope mode".

Mapping from slope to power and “wheel” speed is up the trainer exclusively, GC just sends the target slope according what you set in manual mode or what’s specified in the workout. For some trainers, s.t. Wahoo Kickr weight, and other parameters are sent to help the trainer a better mapping.