Excellent choice of software g - though I get a deflection for the
timber of 9mm - it's a load-sharing system.

I am but a novice. I simply modelled a single beam with 1kN per metre
distributed over the length. This is based on 2 joists per meter and 2kN
loading/m2.

As your key question, I think it's a matter of timber being a much more
friendly material to work with, especially in the confines of a loft. How
do you hold up the ends of these joists:

I sort of presumed they sat on top of the walls with the aid of whatever
bearing structures are normally used. The main "fabrication" would be how to
incorporate the stairs. However, only the top landing is problematic here. A
bit of ASCII art:

-----4m---- ----4m-----
-1m-
########################### ^
# # | # |
#------------# +----# 1m
# # | # |
#------------#-------+----# \/
# # #
#------------#------------#
# # #
#------------#------------#
# # #
#------------#------------#


## = structural wall
-- = steel joist
| = trimmer joist

As you can see, only one part is problematic. The 4m joists (or centre
supported 8m) lay on brickwork at the ends. Where the stairs come up, it
will be sufficient to just stop at the centre wall. However, a landing needs
to be incorporated into the end where the missing joist section is. The
landing is bounded by one untrimmed joist on one side, and structural
brickwork on two sides. I presume the solution will be to use a trimmer
joist or suchlike. I wouldn't be surprised if a timber structure would be
appropriate here. What do you think?

BTW, the roof needs redoing at the same time anyway, so the slates will be
off when the joists are craned into position. The old roofing system will be
replaced by an unventilated Kingspan system.

Christian.

P.S. You seem to know a lot about loft conversions, what are my chances of
the foundations needing underpinning before all this can be done? House
built 1909.

Christian.