Opened 5 years ago

Closed 5 years ago

#250 closed defect - convergence (fixed)

charge exchange raises H O ionization of high Z gas

Reported by: Gary J. Ferland Owned by: nobody
Priority: major Milestone:
Component: ionization convergence Version: trunk
Keywords: O H charge exchange Cc:

Description

four sims are attached. Only H and O are included, and it is a collisionally ionized gas at log T 4.1

There is a mysterious increase in the ionization of both species when CX is included in the high Z case. I cannot think of a physical reason for this.

There are four sims. All have a high density, log n=10. at solar O/H the gas is mostly neutral and CX does not change the solution greatly.

When O/H is increased by 30x the solution becomes more highly ionized, by a large factor, when CX is included. The ionization is similar to the low O/H case when CX is not included.

CX is simultaneously raising the ionization of both H and O by a large factor. Here is a table:

default 0.894 0.106 0.912 0.088
default no cx 0.931 0.0692 0.928 0.072
Z30 no cx 0.931 0.0693 0.928 0.0719
Z30 0.403 0.597 0.455 0.545

Attachments (2)

d10.tar.gz (339 bytes) - added by Gary J. Ferland 5 years ago.
four sims used to create table in ticket
eden.pdf (5.7 KB) - added by wangye0206 5 years ago.
electron density's change via hden and metallicity

Download all attachments as: .zip

Change History (5)

Changed 5 years ago by Gary J. Ferland

Attachment: d10.tar.gz added

four sims used to create table in ticket

comment:1 Changed 5 years ago by Gary J. Ferland

output generated with trunk r6922 on os x gcc 4.2.1.

But result was first noted by Wang Ye on dlx using icc / mpi at r6920 on WangYe? branch

Changed 5 years ago by wangye0206

Attachment: eden.pdf added

electron density's change via hden and metallicity

comment:2 Changed 5 years ago by wangye0206

Figure 'eden.pdf' may help to locate where the bug is.

comment:3 Changed 5 years ago by rjrw

Resolution: fixed
Status: newclosed

The problem was due to overpopulation of high levels by non-state-specific CX allowing collisional ionization, not by CX imbalance.

Fixed by r6948:6956

Note: See TracTickets for help on using tickets.