[PATCH 2/2] gas model: replace Redlich-Kwong with least-square quintic

[Lubomir I. Ivanov]

On 3 March 2016 at 03:36, Linus Torvalds <torvalds at linux-foundation.org> wrote:
>
> On Mar 2, 2016 5:17 PM, "Lubomir I. Ivanov" <neolit123 at gmail.com> wrote:
>>
>> attached is the comparison of the wiki and baue values for air.
>> seem pretty close given the possibility of error when measuring.
>
> Hope close does the polynomial we use end up being?

the biquad and quint ones for AIR pretty much goes through the 1-500
data set from wiki.

>
>> >  (c) the linear 21/79% mix of the O2/N2 (that we don't actually use,
>> > since I decided to keep air special, partly because I trust that table
>> > more, but partly because we just have a special fractional
>> > representation of air anyway)
>>
>> i can try that tomorrow, should the linear mix be compared to the AIR
>> values and if so to which of the tables (baue vs wiki)?
>
> It works be interesting to see the results of our polynomial mixing compared
> with both the baue plot and the Wikipedia air plot.
>

ok, understood.

> It would be especially interesting perhaps top see how it acts at higher
> pressures than the baue tables use - the baue tables end at 4000 psi, but
> Europeans might actually use higher pressures than that. It may not be
> *common*, but 300 bar is not overly strange, and it works be sad if our
> tables end up giving completely insane results just because we had some that
> never went high enough..
>
> The Wikipedia air table goes up to 500 bar which is a lot better range,
> really.
>

is there a way to obtain some of the upper values up to 500 bars for
all gases (we already have the ones for AIR for wikipedia)?
if not, we may have to hardcode reasonable estimations (but arbitrary,
obviously) and re-calculate all coefficients so that all polynomials
look reasonable.

preferably these values should be given to me.

lubomir
--