I just mounted a 100mv/g accelerometer on the bottom of a vertical pump 25' below. The acceleration and Spike energy (HFD) levels are really high. 5 g's and 15 gse. Now I recall in the past I saw this before on some units with stud mounted accls. It makes sense since the mounting gives the assembly a higher resonance roll off as compared to a magnet. What types of amplitudes and differences have you seen in your stud mounts vs. magnet collected levels?

jbalko,

What frequencies are you seeing increased in the g's reading? A pump with flow turbulence should certainly excite the spike, and probably tickle the resonance too.
On some gearboxes, I have had to drop down to a 50 mv/g transducer because of the resonance excitation problem.
On a centac compressor, we saw differences of
4X - 5X higher at some higher frequencies, but they are trending OK. There is a world of difference in those little high freq. transducers between magnet and stud mount.I'm not even sure why they provide magnets with them as they are about useless for any high freq data.

Dave

The accel is a CTC AC102-1a It is 2" by .8" in size. Here are the plots. Its mostly raised floor and the 600K plot does not even encompass all the energy. The velocity overall was ok about .12 in/s but there were frequent pulses of energy.

Vtpumpplot.doc (60 Kb, 41 downloads)

Make sure you haven't preloaded the accelerometer by torquing it down on an uneven surface. This distorts the crystal and acts just like a huge acceleration force. Remember, if you are going to permanently mount an accel, you should spot face the mounting surface first, and make sure the stud mounting hole is perpendicular to it.

Would what you are saying mean that if I checked this while the unit was down, vibration may not be zero and may have some false energy? Would testing this while down prove anything?

Ron is talking about base strain. If the base of the accelerometer is distorted by an uneven mounting surface, a non-perpendicular mounting hole, or the stud bottoming out in the mounting hole, a force with be transmitted to the piezo crystal. The real problem occurs when there are large displacements, which result in changes forces on the cystal and false signals.

Fortunately, base strain isn't much of a factor with shear accelerometer. The design isolates the crytal from the base. It is a problem with compression accelerometers. The acceleromters currently on the market are mostly shear type, but check the spec sheet to be sure. Compression accelerometers also have a large response to thermal transients.

The CTC AC102-1A is a shear mode accelerometer so base strain shouldn't be a problem.

Jon
Spintelligent Labs

Along these lines, a customer had someone install "stud mounted" accels on some large motors for ease of monitoring. They used the little "spot face" tool from CTC to drill and tap the holes for the stud mounting. Only about 1 of 8 was actually square, flat, with well-formed threads. I got some mounting pads with studs on both sides, cleaned up the existing stud holes with a tap, and then screwed and glued the "double stud" bases using Loctite Depend adhesive. I figure what little high frequency response I lose with the adhesive, will more than offset what was being lost by out-of-square mounting, lack of full contact, etc.

I have the spot face tool as well, but I think I will only be using it to prep mounting locations for the "double stud" pads....

Good morning all
Rusty
Could you please provide the source where you buy those double stud mounts.
They sound like a very worth while investment, very meaty mount by the sounds of them. I also agree, what you lose at the high end by using the depend is probably minimal. The fact that this will or should be a very robust mount has
tweaked my interest and I would like to use these type of mounts on a few of the applications we have. I know that the mounting pads are very good but the ones you speak of sound like the cadillac mount.

I usually recommend using a glue-down base to our customers for permanent accelerometers for two reasons:

1. It's far easier than milling, drilling and tapping, so it's likely to get done rather than put off for another day.

2. It's more likely to be done correctly than a screw-down installation.

If the mounting bases are installed properly, there's not much impact on frequency response. A proper installation means removing paint down to bare metal and keeping the glue line as thin as possible with a rigid adhesive.

Loctite Depend is a reliable adhesive...just be sure it's not past its shelf life. For short-term testing, "Superglue" is excellent, but it won't last over the long term.

A secret to remember...superglue needs moisture to cure. If you're trying to install an accelerometer on a wet pump position, use superglue and just wipe the spot dry and slap the pad down.

Jon
Spintelligent Labs

I like the idea of using glue but in my application the mounting is under water. I feel that glue would be too risky.

Jbalco,

Why did you use the CTC AC102-1A accelerometer for underwater use? What type of cable connector and how did you seal it? I recently did a underwater pump installation with CTC accelerometers with side exit and integral cable. I did not measure any significant high frequency vibrations. Are you measuring Cavitation vibrations on your pump installation?

Walt

The acc has an integral cable. CTC assured it would work under water. It could be cavitation?
I believe I have seen this in the past and it would make sence to have higher levels since the roll off I am used to (magnet) occurrs near 140,000 cpm and stud mounted the rolloff should be closer to 600,000 cpm.

The "double stud" pads are from CTC, part no. MH138-1A. Click here to see them.

I agree with Jon that for most applications, the double stud pad is overkill. The regular glue on pad is more than adequate for most installations. But when you absoulutely, positively want to be sure your accel stays in place and is properly mounted, the double-stud pad is the way to go.

I did a test just this week, where I mounted the single-stud pad to a clean steel surface using only Loctite Depend adhesive. After 24 hours I mounted an old accelerometer and using an inch-pound torque wrench, the pad of course didn't budge when I maxed out the torque at 125 in-lbs. Given that the maximum torque I've seen for accels (1/4" hole) is 60 in-lbs, the mount is more than adequate. Next I used a wrench to tighten the accel further, and the threads in the accel started to strip out. Then I used a large wrench on the flats of the mounting pad, and I'm not sure how much torque it took, but I'm sure it was well over 100 ft-lbs to break the pad loose.

As Jon says, take the mounting surface down to bare metal (I use an 18Volt Dewalt cordless grinder with a "stringer bead" knotted wire wheel), clean it well (I used contact cleaner, and then denatured alcohol), and use Loctite Depend with the activator, and it will never come loose. You may knock it off, but it will never fall off.

For walk-around mounting pads, the magnetic stainless steel flat pads that Jon sells have worked extremely well for me. There are in the middle of this picture.

Rusty's double studs would have been perfect for an application I had a few years back. The customer wanted assurances that the pads could not fall off. I couldn't provide that 100% assurance, but I did explain that the only time I'd had pads come loose were when the accelerometer was being tightened. No good enough.

The customer? NASA

The application? Bridge cranes to lift the Humble telescope lens when there was a plan to bring it back from space for repairs.

Having a pad drop and break the lens would be one way to become famous....or should I say infamous?!?

Jon
Spintelligent Labs

Another place I've used them is above a calendar stack where you have two polished steel rolls running in close proximity (0.040")... a chunk of metal there would do serious damage. We even drilled tie wire holes through the edge of the base and secured it with stainless tie wire.

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