Anti-vibration units.

[Ronnie Sullivan]

I have a customer that has a huge centrifugal hydro dryer.
They put up to 10 ton of wet yarn into it and 'spin dry' it.

It looks a bit like this one but much much bigger.

I retro-automated it a long time ago, so it has 'been my baby' ever since.

The operatives (who are paid from the neck down) don't care if they put the load in unbalanced (even though they are told spread the product evenly) and when 10 ton gets a lick-on the whole thing shakes like (as the engineer put it so eloquently) like it's trying to climb out of its pit.

It came with an anti-vibration unit - a very imaginative ball bearing sat in a groove that is thrown out of the grove onto a micro switch when it vibrates.

The trouble with that is - setting it up is almost impossible. Too fine and it trips the switch on nearly every start up. Too coarse and it either doesnt work or hits the switch on every revolution with mild vibration
I tried programming it out of the start up but it's my opinion that it is not good enough.

It came to a head a while back when the main shaft sheared and it was out of action for 2 weeks while it was repaired. It was probably due to, in part excessive vibration over a long period.

Does anyone know of anti-vibration switches that are digital or analog that can be monitored throughout the running of the machine?

A normal run of this machine would be as follows. On start up it has a 30 second ramp up time. it starts to gently lurch at about 15 seconds (this is ok)
As it gets faster it settles down but can still vibrate about 1cm 1/2 inch off center (this is still acceptable, it has massive springs to take this up)
If the load is very unbalanced - this vibration can increase to 2 to 4 inch and that's when I want to stop it................hmmm any suggestions?

 

[AustralIan]

In addition to Anti-VB, Mount it on load cells and tell them to get the balance right before allowing them to run.

Depending on the spin speed, you can probably use the same load cells to measure vibration while running.

 

[JesperMP]

A ball in a groove, hmmm...
How about this.
The ball sits in a recess, if the vibrations are too severe, the ball will jump out of the recess.
Besides the recess, the bottom is shaped like a funnel, so that the ball automatically resets by itself (Pat. Pending. ! ).
A proximity sensor sits vertically above the ball. The sensor has to detect the presence of the ball to not trip.
Now, to keep the ball from jumping out in the initial startup, the prox moves down and rests on the ball, keeping it in place. When started up, the prox moves up and the sensor is "armed". The moving up and down of the prox can be by a pneumatic actuator.
Here is a sketch with the prox down (left) and the prox up (right).

 

[milldrone]

A normal run of this machine would be as follows. On start up it has a 30 second ramp up time. it starts to gently lurch at about 15 seconds (this is ok)
As it gets faster it settles down but can still vibrate about 1cm 1/2 inch off center (this is still acceptable, it has massive springs to take this up)
If the load is very unbalanced - this vibration can increase to 2 to 4 inch and that's when I want to stop it................hmmm any suggestions?

Use a couple of analog laser sensors at 90 degrees to the axis, to look for the off center?

 

[Tom Jenkins]

You are talking about a significant amount of displacement!

If the G-force is proportional, which it should be, you could frame mount some piezo type vibration switches to the frame and use them to stop the machine:
http://www.balmacinc.com/products/vibration-switches

Another possibility is to use a proximity switch picking up the drum and set to trip when the vibration amplitude exceeds the desired level.

 

[Pete.S.]

Put in a 4-20mA vibration sensor. Then you can set the limits in software and also make it speed dependent or dependent on the load (if you have that info).

 

[Bit_Bucket_07]

Put in a 4-20mA vibration sensor. Then you can set the limits in software and also make it speed dependent or dependent on the load (if you have that info).

I agree. I don't know a name brand off the top of my head, but I've used analog vibration sensors in numerous applications with excellent results.

 

[jstolaruk]

If the machine vibration sheared the main drive shaft, sounds like the mechanical design needs to be reviewed.

 

[RET]

I agree. I don't know a name brand off the top of my head, but I've used analog vibration sensors in numerous applications with excellent results.

We use Metrix quite a bit..model 162VTS,..4-20mA loop-powered devices in several ranges.

 

[Ronnie Sullivan]

Thank you for your answers. I knew I would get some good results here.

Still pondering the best way to go

 

[Toine]

In addition to Anti-VB, Mount it on load cells and tell them to get the balance right before allowing them to run.

Depending on the spin speed, you can probably use the same load cells to measure vibration while running.

I always treat load cells as somewhat fragile parts. Perhaps this is biased by my needs to read weights reliably, but from my point of view this sounds like an application where load cells would not last long.