[Amps] HV transformer issue

[David H Craig]

    Hello all,

    I am seeking input on a potential problem with my "almost new" B+ transformer.  It is a 240V input, output 4 taps from 2800-3000V rated 2 amp full load.  Transformer was tested in situ today, first stand-alone.  Has a slight hum but not significant, and as mounted no detectable vibration on the laminations.  AC output tested fine on all taps.  Second test was with full wave bridge feeding a 12 x 800MF (series) homebrew cap bank attached- expected DC Voltage measured fine on the cap bank.  3rd test was with 100k ohm 250w R across the cap bank from B+ to B-, ~100 volt drop.  
    Now the "problem":  The transformer is a typical E I Core (roughly 9"x7"x7") with primary taps on one side fed to binding strip & secondary on the other side to an EU style strip opposite the mains. Core is held together with 4 pieces of allthread, nuts on each side, around 1/4" in diameter.  Transformer is not potted.  With mains fed to the transformer & even with cap bank removed from the secondary side, the 4 allthread "bolts" get hot. Not a "little hot" but "blister your finger hot" with just a minute or 2 of mains voltage applied.  I understand this was new when I bought it- that said, I can smell some heating varnish, apparently from the bolts.  The laminations themselves stay completely cool, but granted it is so massive (84lbs) that it would take a good long while for it to heat up. I am not sure but the allthread appears to be standard steel.  I have never run into this strange effect before- this can't be normal, can it?  I am guessing this is hysteresis, magnetostriction and heavy eddy currents in the bolts- does this sound plausible?  What is going on here? 
    Because they are getting so hot after just a short time under power I am concerned about using the transformer in the amp (GS35b 50 MHz in mid-build).  It has been suggested that I pull one of the allthreads and check for hotspots on it, which I plan to try.  If they are steel would it pay to replace them with non-magnetic stainless bolts & get rid of the allthread completely?  Also, in its current configuration I cannot tell whether the bolts are actually a proper fit for the holes in the E I plates- it appears that it may be slightly undersized, and of course they are threaded the whole way, ans therefore may not be effectively transferring heat to the laminations. 
    Any input would be most appreciated.

    73 Dave N3DB