I have noticed that CTC offers "Motor fin Mount probe/pads"for mounting in heatsinks. I am wondering weather this is a good product, I am sure it is better than a pad or washer spanning a heatsinks.Any comments out there, does the stem need to be completely bottomed out in the valley?, will it work well if a loose fit must be taken up with a large amount of epoxy? How much time does it take to install one, I assume that the paint must be chipped off.


Any comments would be appreciated.

Thanks
Al

Alan,

The fin mounts are a good solution where there aren't other alternatives. Sometimes there are wider "ribs" where the motor is bolted together which can be a good location.

The fin mount definitely must bottom-out in the valley, with metal-metal contact. There are a variety of sizes of the fin mount to get the best fit. Removing the paint is best, but a thin layer of factory-applied paint shouldn't have much effect.

I'm not sure what I'd recommend for an adhesive. I'm a fan of Loctite Depend for most applications but it won't fill the large gaps in an installation like this. The adhesive must be extremely rigid when cured. A two-part epoxy is the best bet.

Finally, the fin mounts are only effective for single-axis measurements. This is not an acceptable way to mount triaxial accelerometers.

CTC has an application note on installing fin-mounts:

https://www.ctconline.com/uploadimages/%7B9A985CFA-16E9...1652%7D_MNX10003.pdf

Jon
Spintelligent Labs

Alan,

I've had good luck on the fin mounts with the Wilcoxon two part epoxy. As John stated, bottom out the mount in the valley. I usually end up machining some part of the rascals to fit properly. I have four different sizes (2 different lengths with 2 different thicknesses of each).

If your not usuing a transducer with the BBBBig magnet (65# pull), then the fin mounts are a good way to go.

Dave

Here is an alternative source for cost effective Motor Fin Mounts:
http://www.reliabilitydirect.com/accelerometeraccessories/cmcp205.htm

http://www.reliabilitydirect.com/accelerometeraccessories/cmcp205kit.htm

I always bridge across 2 fins with a pad wide enough that the sensor magnet sets directly over a fin. I really don't see why you would lose any response when taking the data on a pad epoxied to the top of a fin compared to epoxying narrow bar with a flat top to the motor housing between the fins.

Why bother with special fins pads at all?

Attaching a dual rail magnet to the top of a SINGLE fin directly so that the fin crosses both rails squarely and holding it down by hand won't reduce reponse or distort the signal much.

In general, most energy of a bearing fault signal is already significantly attenuated anyway in the interface between the end bell and motor body. If the proposed above method attunates it further, the overall attenuation won't be much affected by it. Just an opinion...

Jim,

How can you arrange an accel to be seated directly over the top of a fin by having two fins bridged with a pad?

This message has been edited. Last edited by: David_G,

David,

I'm currently going through a process of attaching magnetic pads to non-magnetic machines. Previously I was holding the accelerometer in place by hand. I've noticed a significant improvement in response in my spike energy readings.

Gentlemen

Thanks, maybe I should have mentioned that I would like to use peakvue (which is new to me) on many of these points..will that work just spanning the heatsinks with a pad, washer or polled magnet? Quite a few of the points are on 1800 and 3600 rpm motors, only a handful are 1200 or 900 rpm, I would assume most of the HP filters are 2000, or 5000hz.

Regards,
Alan

David,
As I said, we make our own pads. They are just rectangles and if you overlap each fin by a little you can position the sensor more directly over one fin or the other. I really don't know if this makes that much difference or not but that is the way I prefer to do it.

Alan,
For PeakVue I prefer not to use washers, they aren't usually very flat and a flat magnet can rock back and forth. It depends on what you are trying to detect with the PeakVue reading. Any decent attachment, as long as it is pretty clean and don't have a lot of paint will give good indications of bearing defects. If you are going to detect lubrication faults accurately you will need a good, clean flat surface and a flat magnet. I have done a lot of testing with lube problems here due to problems with some of our lubrication practices and can state for sure there is a difference in sensor mounting. If lube faults aren't a big concern just remember to be consistent if you want the trend values of the max peak waveform to be useful.

Per Jim's comments, I've never had good results using washers. The glue will form a high spot in the center of the washer, keeping the magnet from making good contact. I also find them a pain in the butt to install and usually end up getting glue everywhere but where I want it.

Spintelligent Labs offers economy magnet targets. These are almost identical in diameter and thickness to a Susan B. Anthony Dollar and made from magentic stainess. They cost $60 per bag of 100. Using them provides the best mounting surface for the accelerometer and the data gets taken from the same location every time.

Jon
Spintelligent Labs

Jon,

1. In what sizes do Lab's magnet targets come?

2. What is the minimum radius curve they can be glued onto?

3. If a target is used, will accel with a FLAT magnet be significantly advantageous over that with a DUAL RAIL as far as bearing defect or lube problem detection are concerned? (Jim Crowe, have you experimented with different magnets?)

4. Is it sufficient to use just one magnet target per bearing from a bearing fault detection point of view and measure vibration in other directions with no targets?

Thanks.

Our economy pads are 1.06" (26.9mm) in diameter x 0.081" (2.08mm) thick. This is nearly the exact size as a Suzan B. anthony sliver dollar.

These work pretty well down to a radius of maybe 1.5 inches using an adhesive that can fill gaps.

I believe that on a flat, machined surface, a flat magnet will out-preform a rail magnet but I don't have any hard data to back this up.

I recommend using mounting pads at every mounting location. This is both to ensure the best possible data is acquired, and that the measurement is taken in a consistent location each time. If there is a pad for each measurement, there is no doubt about the locations and less chance of forgetting to make a measurement at each location.

If you have a data collector that allows listening to the data, it's easy to prove how much difference the mounting surface can make. With the accelerometer on a rough surface, listen to the data as you move the accelerometer slightly and remove and replace it. You'll be amazed how much difference you can hear when you do this. and if you can hear the difference, you'll be able to see it in the spectra too!

Jon
Spintelligent Labs

Jon,

Thanks for the info. I do realize this is not a commercial site but I have no other way of comminication on the subject of ordering the targets from the Labs. How can I order them?

Thanks

David,

Sorry I didn't think to include contact information! You can e-mail me:

jon@spintelligentlabs.com

Our phone number is 206-855-9284, but today is one of those crazy days when we may be out most of the day.

Jon
http://www.spintelligentlabs.com