Seed Mechanisms revisited

[rwest]

There is a new 'seedmech' branch for us to implement the changes to seed mechanism behaviour that we discussed in the meeting today.
You can follow its progress at master...seedmech

So far, seed mechanisms are now read in to the edge not the core, which was the first major change.
Still to be done:

• including the seed mechanisms in the chemkin files even though they're on the edge.
• checking to see if I broke restart
• changing the pressure-dependent behaviour: use FAME to fill in gaps in P-Dep networks, but the seed mechanism to over-ride specified rates.
• ... add more below!

Please feel free to contribute :-)

git fetch official
git checkout -b seedmech official/seedmech

[rwest]

Some more progress.

It now includes seed mechanisms in the chemkin files (species, thermo, reactions, and transport data).

If a reaction is in a seed mechanism, be it with pressure dependence or
otherwise, then it is removed from any PDep networks before they are
sent to the ODE solver, or printed in the chemkin files.

[rwest]

Thus far it seems to be working.

For the record, this is what we're aiming at, based on the meeting we had:

Name	Treatment	Pdep?	React with core	Prunable?
Seed Mechanism species specified in condition file	Included in core at start	k(T,P) as specified	Cross-react at start	No
Seed Mechanism (species not in condition file)	Placed on edge at start AND added to chemkin files.	k(T,P) as specified	React only when included in core	No
Reaction Library	Placed on edge.	k(T,P) or k_inf chosen by user in condition file	React only when included in core	Yes
Kinetics Library	Used if RMG predicts reaction exists via family template.	k_inf	React only when included in core	Yes

I think we need some work to achieve the desired PDep column.

[rwest]

I'm going to create a pull request to continue this "issue" as I think it's more appropriate. See http://help.github.com/pull-requests/ for details.

[rwest]

I'm going to create a pull request to continue this "issue" as I think it's more appropriate. See http://help.github.com/pull-requests/ for details.

[rwest]

No benefit to leaving this 'issue' open as it's superseded by pull request #153.