hirschmann / pid-autotune Goto Github PK

Are the units returned from the cool function correct?

If thermal conductivity is in W / [ m.K] , surface area in m^2, delta T in K, duration in T (seconds) then cooling power has units

W / [m.K] . m^2 . K / T = W.m / T which is kW m / T after dividing by 1000

Then, with heat capacity in kJ / [kg.K], Joules in W.T and mass in Kg, the subsequent get_deltaT function returns units:

[kW.m / T] . T / [kJ / [kg.K] . Kg] = [kW.m ] / [kJ / [ K] ] = [kW.m ] / [kW.T / [ K] ] = K.m / T

Which is not K!! Unless I'm mistaken I think the duration in this function should in fact be the width or depth of the material, which is consistent Q being a heat flux in W / m^2 and will ensure the cool function returns temperature change in units of temperature.

According to this, the conductivity of steel should be 25 (stainless) or 50 for some other steel: https://www.sciencedirect.com/topics/engineering/thermal-conductivity-coefficient

Hi Hirschmann,
I could find no other way to contact you directly so ... Therefore i Created this issue.

I followed the guide for install

1. Install git and python3
2. git clone https://github.com/hirschmann/pid-autotune.git
3. pip install matplotlib

Then : python sim.py --atune --volume 50 --power 4
and first python
and then : sim.py --atune --volume 50 --power 4
(getting errors)

Do I need to do it some other way ? or can you direct me to more help?

I run your code.

However, it seems that the provided several parameter pairs from the autotune do not work ....
I mean those parameters make convergence eventually, however they are not good parameters...
and I think the brew type works only well as a good parameter tuning.

I assume that brew is assigned by yourself, right?

I have one more question, there are divisor variables in each method, and I don't think they do not quite follow the table in this website: https://en.wikipedia.org/wiki/Ziegler%E2%80%93Nichols_method

How did you come up those divisors?