Went on a emergency balancing Friday. Overhung fan belt drive 3010RPM. First rotor some time ago lost a part of the rotor back plate, big dent in cover. Next fan was wrong in inlet cone, current mounted Thursday had 42 mm/s vibration 1xRPM dominating, there was weights on it from factory. Balancing was moved in time as bearings had to be swapped, they passed exchange levels. One brg popped off easy the other not, it needed 10-15T to be removed and the new was not entering despite cooling with fireextinguisher so we took it down minimal with paper in a lathe, shaft was oval 5/100 mm or more and oversize at entering area and undersize at brg. fit. No spare shaft was at hand. So time got a bit pressed by operations. I got from 42mm/s to 8-9mm/s passing both side of center a couple of times when welding indicating to much play. Operation decided, that was good enough, they had to start operation. Fan came off easily and was oversize compared to the others but didn´t fit "easily". A typical Friday situation. Brg. values was anyway 10 times better than before. So when adding all stuff to run it, specially belt guards resulted in 18mm/s on the drive side brg and fan brg still 8mm/s. So my aim with this ranting is that the beltguard vibrated 95mm/s from start, how much of that will add to the level found at the brg it is closely attached to/ welded to? Obviously the belt guard have a perfect resonance problem, so it will be taken care of. It will be fun to see what part it have in the brg levels. All/most is 1xRPM, no obvious resonance part from the beltguard. Fan is also operated at 10-15% flow 95% of the time so that should be worked at, since it have a delta-Y starter I was thinking of leaving it in the low speed until full capacity is required. Anyone tried that? Olov

Guard attatched to bearing -- even welded on!!!

I would mount the guard to the floor as it sounds the structure is flemsly.

Ensure bearing to shaft alignent and bearing to bearings alignment then sheave alignment.

Run at low speed until high speed is need is common. Our Junker Furnace runs that way as an example. From its poor design I recently modified one fan. The modification has been so successful that I'm going to make the rest match.

I wouldn't venture a guess what would be the effect. I have seen some surprising cases of a small tail wagging a big dog. Here was one:
http://maintenanceforums.com/eve/forums/a/tpc/f/3751089...211084273#7211084273

As far as the attachement of the guard, most of our belt guards are attached to the fan housing, but not near the bearing (see attached). Personally I would prefer floor mounted, but I guess that costs more (more material).

Thank´s, in my case the similar struts holding the guard is much shorter and it´s welded very close to the bearing on the metal sheet that is the structure the bearing is bolted to, it´s tuned to about 50Hz and that´s pretty stiff for a belt guard. I also suggest mounting it on the machine base. Guard was loose and mount busted and we welded it back so I guess we tuned it back to vibrating massive.
We brazed it with a bit 2x4 against the floor to keep it from falling off. Suggestion also include running low speed unless really needing the flow, good to know that it´s used Sam, and getting a replacement shaft for the next time it need service is also on the list.
I added some 70 g at maybe 150mm radius out of the 5-600 mm fan radius so I find that a little more sensitive than normal, also bits falling of the fan rotor backplane is not that common so maybe there is a resonance somewhere possibly axial in the fan that made fatigue cracks or it was the turbulent flow from running it against nearly closed valve. Valve is a floor level up so there is a air column that may try to move around. Next visit may give some more answers. Olov

Guys,

Do you find running at the lower speed you get the rolling elements of the bearing skidding instead of rolling? A few times we had this probblem as the bearings were not loaded enough, what's your thoughts? have you seen or herd this in your data?

Hooch

quote:

Fan is also operated at 10-15% flow 95% of the time so that should be worked at, since it have a delta-Y starter I was thinking of leaving it in the low speed until full capacity is required. Anyone tried that?

The wye/delta wye starter applies lower voltage per winding in wye during start and then switches to delta which provides higher voltage per winding. Leaving it in wye (if that's what you're suggesting) would not be a good idea. With lower voltage per winding, the speed will not be significantly lower, but the current can be 1.7 times as high or more.

Currently the transient when starting is short so I can´t say anything on bearing load. It´s a belt drive so there is always a load. Ok I knew there was a catch with the idea, when starting there is a significant speed change, it goes smoother until the full speed is on. Normally the motor plate indicate what you get in speed with different connections but it was not to be seen on this as it was covered in some white pulp stuff. I will add the argument. Thank´s. Maybe just swapping the belt ratio would be better if they rarely need full flow, the brief info I got was that it was needed when running on oil and that is not to be used these days. Olov

Well when things act funny there are reasons. Shaft broke off at midnight today, it was replaced by lunch and by a small 400 km Sunday outing it was balanced mainly by extracting 60 g of previous weight and replacing it by 10 g. So I guess the shaft was cracked already 3 weeks ago and that´s why the balancing was tricky. Never knowingly balanced a thing that was cracked before. There was no noticeable 2xRPM, 3 weeks ago only 1xRPM. I held the shaft in my hands 3 weeks ago and could not see anything. When balancing with piping disconnected it run at 2800 RPM. I did a coast down waterfall and there is a very nice resonance at 2500RPM confirmed by knock test at fan center vert at 52Hz. Original it was run above 3100-3200RPM when the fan broke, dropping parts of the blade backing, suggesting axial fan resonance so where should you run such a beast? Maybe the major problem is that they only use 10-15% of the flow normally so the flow noise excite all resonances making it fall apart. Drive side bearing of the fan has a whacked bearing house where a big washer is holding down the remains, that side is vibrating more. After balancing free side bearing was ISO 2.5-2-8 mm/s driveside was ISO 7 mm/s and readings not so stable as normally is. So have anyone tested the difference of vibration levels between a whacked bearing house bolted down with washers and a good one? Olov