Having problem monitoring brg. in far away American Air fans delivered by Osprey airflow is normal air. They give like 21, 28 gE from CMVP20 vib pen, ISO vibe is like 6-7mm/s max and 350mV AC RMS from 100mV/g accel. fixed mounted. Accel´s bottom out from overload so xducer resonance must get excited. No hi brg. temp, 35-40 DEG C, no hi noise. Anyone have had similar problems w. fans like that or others? I have currently no way to get frequency information. Any input of value. Olov

Olov,
I'm having trouble keeping up: ).25 IPS max? But 21-28 g's OA RMS from a vib pen; and, 3.5 g's RMS from a 100 mV fixed transducer......... Which is reliabile? I'd have a tendency to go with the fixed transducer and if 3.5 g's is OA over 10 kHz then I would think it is within tolerance. Am I off in left field here?

Well, if the freq. it generates hit the mounted resonance of the xducer it will give anything and that is what it does. I don´t think 28g is normal for a fan blowing air? If they put down the xducer on the floor it show 0 both vibe and brg. cond. If they hold it by hand it is lower and if they have a plastic material between it get normal. If the RMS voltmeter they used to measure raw signal from the probe have limited freq. range that is not uncommon, that may be much more. I tried to make them use a oscope to verify but it´s so fancy nobody can make it display anything useful. Anyone seen signals in the KHz range from a fan that can excite accel resonance? Anyone have a magic material that can give the classic accel problem solver, a mechanical filter that will kill the upper KHz range? Olov

Olov,

3.5 g-rms from a 100 mv/g accelerometer cannot be overloaded, since they typically respond up to +/- 50 g in their linear range. Based on the available data, I would say bearing is running dry and needs lube (oil or grease). Worth a try, isn't it? Bruel &Kjaer is the only company that I know of that sells a "mechanical" filter for accelerometer mounting.

Walt

I asked them to grease the first thing, they say it makes no difference and that bugs me. 350 mV is pending the unknown freq. range of the device used to measure with, if there are signals in KHz range that level may only be the part from 1xRPM at 17-20Hz or? It may also be the case that several bearings are bad but there is a problem convincing people on that from a distance. We are shipping a instrument so they can collect some more data. Yes I only seen B&K filters also. Olov

Olov,

The CVMP 40 measures gE (enveloped acceleration) over frequency range of 10kHz to 30 kHz. This is outside the range of common 100-mv/g accelerometers and probabaly not in frequency range of the instrument that read the 350-mv (3.5 gRMS). You are comparing apples to oranges. SKF information states that "velocity of 4 mm/sec RMS with gE of 10 or higher usually indicates a lubrication problem." Possible other causes (my list) are bearing seal rub, loose retainer, or split-race bearing that do not produce much low frequency vibration. I have measured this effect with ultrasound meter. The person using the hand-held pen may not have held it in firm contact with a clean surface on bearing housing, so impacts at the sensor tip were occuring.
As you know, if someone said they greased a bearing, it does not mean that it was properly greased. The grease may not have reached the rolling elements for various reasons.

Walt

Walt, We (or our customer rather) are installing 28 accels on 14 fans, a handful of fans are of this brand and making the problem. We are entering brg temp, ISO vib and L value brg condition as a HP filter of 600Hz velocity RMS, to a PLC. gE value is equally hi on those fans giving trouble with the mounted accels 5-6 of them 2 accels on each fan. Greasing was performed by the installing guy to troubleshoot so he should have tried properly but who knows, still is a possibility that they never grease their stuff and grease never get where it´s supposed to, funnier things happen. I have not been reading up on the CMVP 20 so it was good info that it does measure so hi in frequency it would point at the possibility that some signal or metal metal contact could excite the xducer resonance. It is not likely that the guy with the handheld toy would mess up 30% of the readings and that they would coincide with the funny fixed accel readings.
I have seen it before at a gbx at 2xGearmesh at 18KHz. In that case it made the accel overload so it did output something that was indicating 45mm/s in the ISO range. So the factory say that the brg. temp is low on all fans, they are not noisy, not likely that so many brg´s as that are that bad, at least one fan have recently got new brg´s so they ask why the measurements are so crazy. They never cared about the gE reading before... So there must be or could be something else beside brg problems generating a hi frequency signal at the grease zerk position. I just can´t figure out what it could be in a standard belt driven fan that transport air. I asked them to look for something metal that can touch the shaft but they can´t find anything. Anyone seen that kind of signal from that brand of fan, do they have some strange seal design? Any other ideas? Olov

High gE's can also indicate electrical arcing through bearings and stator shorts. I suggest you take an acceleration reading to 360kCPM with 3200 lines; look for haystacks with sidebands spaced at BPFO.

Improperly grounded fans can also generate static discharge but this will probably be random.

Other sources of high gE's- belt squeal, loose keys and hubs, loose bolts, loose motor stators.

It seems unlikely, but is there any chance these fans have ceramic bearings? I've seen a lot of cases where ceramic bearings in good condition generate enough high frequency noise to overload a 100 mV.g accelerometer.

You might try some 25 mV/g accelerometers to gain 12 dB of headroom....or high freqeuncy accelerometers to raise the natural frequency.

Jon
Spintelligent Labs

There is no numerical evidence presented that indicates to me that the 100 mv/g accelerometer is overloaded. Its overall level is 3.5 g rms. These things can take 50 g's!

Walt