Fe IV emission lines are observed in stellar sources, such as hot stars and
white dwarfs, (Penston
etal. 1983 and Holberg etal. 1994; Vennes etal.\
1995). Becker & Butler (1995) have modeled non-LTE
line formation for Fe IV and synthesized the predominantly UV spectra from
hot subdwarfs. According to photoionization models
Fe IV is expected to be a dominant ionization stage of iron in nebulae
(e.g. Baldwin etal. 1991 and Rubin etal. 1991).
However, the Fe IV spectra from H II regions remained elusive until the
recent HST observations of the
Orion nebula by Rubin etal. (1997), who reported the first
detection of Fe IV in the UV [Fe IV] lines at 2836.56 Å arising from
the transition
, and the blend of
due to 
. From the [Fe IV] lines and the predicted
fluxes from photoionization models, they derive an iron
depletion in Orion, relative to solar, of up to factors of 70 - 200.
This appears to be excessive; Rubin etal. (1997) suggest the
need for improved modeling and a reexamination of the atomic data,
especially the electron impact excitation collision strengths used in
the analysis.
The only earlier published work on electron impact excitation
collision strengths
of Fe IV is by Berrington & Pelan (1995) in Paper XII of
this series on the Vanadium-like ions. They included the five lowest sextet
and quartet
LS terms 
in a non-relativistic
(NR) calculation, and used algebraic recoupling to obtain
excitation rate coefficients for transitions
between the 12 fine structure levels of the first four terms.
In our earlier study on Fe III (Zhang & Pradhan 1995a) using the NR and Breit-Pauli (BP) close-coupling R-matrix methods, we concluded that the relativistic effects are small for the forbidden transitions between the low-lying levels, and that the resonances and the coupling effects arising from a large coupled-channel wavefunction expansion dominate the collision strengths. Therefore, as described in Zhang (1996, Paper XVIII), we carried out an 83CC NR calculations for Fe III, neglecting the relativistic intermediate coupling effects. Fine-structure collision strengths were then obtained from the LS coupling data by algebraic recoupling using the code STGFJ.
Similarly, for Fe IV we expect the relativistic effects to be small,
especially for the forbidden transitions from the
ground level to the low-lying even parity levels since
there is no fine-structure splitting for the ground term 
.
Before carrying out a large-scale calculation, we made test calculations
with the BP R-matrix method (Berrington etal. 1995)
using a 16-level target expansion (corresponding to the above five terms)
and obtained results similar to those in Paper XII by Berrington
& Pelan (1996).
These BP calculations, albeit with a small target
expansion, showed that the
relativistic effects are indeed not important for Fe IV, and also provided
the basis for the present calculations, which are much more extensive than those
reported earlier in the IP series.
Therefore, we have carried out the calculation for Fe IV with
a 49-term target expansion (49 CC) using the NR R-matrix method in
the close-coupling approximation. Again, the algebraic recoupling
method was used to obtain results for a large number of
fine-structure transitions.
The maxwellian-averaged rate coefficients or effective collision
strengths are calculated and tabulated over the temperature
range in which Fe IV is most abundant in astrophysical sources.
A brief description of the computations and the results
are given in the following sections.
The present work is part of an international collaboration known as
the IRON Project (Hummer etal. 1993, referred to as Paper I)
to obtain accurate electron-impact excitation
rates for fine-structure transitions in atomic ions.
A full list of the papers in this Atomic Data from the IRON Project
series published to-date is given in the references. A complete list
of papers including those in press can be found at
http://www.am.qub.uk/projects/iron/papers/, where abstracts are also given
for each paper. Information on other works by the present authors and their
collaborators, including photoionization and recombination of ions of
iron and other elements, can be found
at http://www-astronomy.mps.ohio-state.edu/
pradhan/.