Dear Member ,
It is mounted on soft foundation . Has the movement of the soft foundation been compared to the movement of the fan bearings' balancing positions (amplitudes and phase)? Is the "soft foundation" springs or rubber pads?
Just a week before , i had Balanced Two Overhung Blowers Running at 3000 r.p.m and resonating at the speed (Vibration 100 mm/sec in Horizontal direction and 20 mm/sec in vertical direction  
and not responded for initial balancing attempt and it had directional resonance .i had change the stiffness of the structure by Welding stiffners in 'X' shape on pedestal in Three directions .It is responded well and vibration decreased to 5 mm/sec in Horizontal direction and 3 mm/sec in Vertical direction .Those Blowers are running smoothly till now . This must have been a "hard foundation" fan since strengthening the "structure" changed the "response" and allowed for a better balance, HUH? 
Is there any such issue for this I.D Fan.This fan also not responded for initial Balancing ,after that we found cracks on discharge duct and damaged Nuts and Bolts at Suction duct flange .After the repairs ,it is responded and the response is not to the satisfactory .And the trail weight location is the same angle as first balancing attempt .But the response is little ,Fan had responded after Correction weights attached .Was the "damage" to the duct work confirmed to have been caused by the unbalance movement of the actual fan pedestal or from the movement of the "soft foundation" and the fan pedestal combined? 
Please verify once again the balancing report.So that you could understand better I will consider all the feed back ,when i get opportunity to Balance the Fan .After reading your balance report carefully, as suggested, I notice some things, especially the "reference run data", that raises a question or two in my thinking.These questions are:
1. I am not sure if the "OBH" position is the bearing closest to the fan or closest to the sheave, because Some analyst use "OB" as next to the sheave. I, personally consider "OB" as next to the fan.
2. I see, in the reference data, where the phase between the OBH and IBH points are almost the same, (70 and 73). The amplitudes are also fairly close, one to the other, (54 and 42). I asked myself, Why would I not have tried a single plane balance first rather than a complicated 2 plane balance procedure? Just my thinking.
3. After running the numbers from the balance report through CSI's balance program, I came up with corrections weight of 102 @ 215 degrees and 67 @ 70 degrees. This would have been with the removal of the original trial weights.These numbers are based on the assumption that the tach light and transducers are sitting in the same plane and facing the same direction.
4. I also read in the report where, for the trim run data, the weight was changed(?) from 80 and 18 to 70 and 20 at almost the same positions of the 1st correction weights, but not sure if the trim weights of 70 and 20 were added to the 1st correction weights or replaced the 1st correction weights.
With an amplitude of 54 mm/sec Peak, and that seemingly converts to near 30 mils Peak to Peak, and unless some of this movement is being absorbed or shared or caused by the soft foundation, this is extremely high, IMO.
Is there some way you might could temporarily stiffen the "soft foundation" long enough to see just how out of balance the fan really is?
Only my opinions and I could be totally wrong.
Thanks and Have a Great Day and a Happy New Year,
Ralph
