Topband: elevated radial question

[Grant Saviers]

I'm finalizing the layout for a tree hung T loaded vertical for TB, 
about 85' to the top, two x 45' mostly horizontal loading wires, and 
radials elevated 10' above (much?) better than average soil, at least 
this time of year when I have standing water among the trees. (Redmond, WA)

I'm taking W8JI's advice and going with the top loaded vertical rather 
than a delta loop, particularly after I determined I can squeeze seven 
fairly symmetrical 130' radials in (with a cooperative neighbor).  I 
plan a switched series capacitor feed for bandwidth, with the antenna 
resonant at 1815KHz or so.

Now two questions before stringing wire -

1. My new DXE 4 square receive array is outside the radial field, with 
the center of the square 82' from the radials perimeter.  Does it matter 
if the end of one radial is about 30' from a corner 4sq antenna, or 
should I pitch the radials to maximize the separation? Even as much as  
a 90 degree (or more) segment with no radials?  At 90 degree pitch the 
nearest radial ends would be about 80' from their nearest 4sq antenna.

2. Now the unusual circumstance - there is a 56' x 70' steel building 
entirely inside the radial circle, but at the perimeter. Steel roof, 
walls, and Ufer grounded to the perimeter footing.  My thinking, not 
sure I'm correct, is to NOT attach any radials to the building (12' to 
eves, 14' to peak), but nestle it between 2 radials with about 15' feet 
of minimum clearance.  OTOH, I can connect one or more radials to the 
steel, or run one or more insulated radials over the roof to a support 
off the perimeter end of the building. And then OTOH, the steel sure 
makes this part of the radial field pretty high conductivity.  This one 
is definitely not in the handbooks or in ON4UN or in the N6LF QEX 
articles. btw I have a SteppIR BigIR vertical going on the center of 
this roof as a secondary/backup HF antenna.

Inputs appreciated,

Grant KZ1W