Topband: Perimeter Wires, Fat Radials...thoughts?

[PaulKB8N at aol.com]

I've had very good success with a 24' end-loaded vertical dipole on  40M.  
I'm wondering how principles that I've used there might apply to 160  and 
80M.
 
First of all, end hats can be quite small and still be very effective,  
especially if a perimeter wire is installed around the spokes.  In fact, it  is 
generally accepted that a perimeter wire can reduce the radius of a top  
hat by 30 to 40% compared to one without.  Those on my vertical dipole are  
only 4' square.
 
Can/should this principle be applied to short elevated radials? Will a  
perimeter wire help or hurt performance?   I have used 16 30-35'  radials on 
the ground with fair success in the past.  If I go elevated, I  can go 30' in 
at least a few directions with no complaints from the  XYL.  Should I place 
a perimeter wire that connects the ends of my  radials and therefore make 
the radial field appear electrically larger?
 
Alternatively, could I install "fat" radials that could be  deployed 
radially from the vertical and form a "fat" pie-shaped  radial consisting of two 
30' foot pieces connected at the distant end along  the fence line with a 
10-15' connecting wire.  Would that be a better  alternative to several 30' 
radials by themselves?
 
Yet another option would be to deploy T shaped radials. Would a few of  
these (longer electrical length) be better than several shorter radials?
 
My goal is to have an effective system with as little visual distraction as 
 possible.  
 
I'm not aware of any research done on these questions, if you have any  
references, please let me know.  Conjecture, based on experience, is  certainly 
encouraged!
 
73,  Paul, K5AF