Another question concerning balance machine Ref Thread 354-35253

[micjk]

Thanks to all for response to my last thread. Maybe you can assist in another concern. What we have done is fabricated supports for our spindles and rolls to simulate actual operating setups. I am not sure whether the Reps. that sold this equipment thoroughly understood what and under what conditions we were trying to balance. The machine that one of our former engineers purchased is a CEMB dual pedastal belt drive configured machine. What has happened is that in order to balance some of our rollers we have actually had fabricated (also by CEMB) a support platform that connects the 2 pedastals together. Then we had a machine shop fabricate a "support block", which weighs @ 80lbs, to support the entire spindle assembly inc. bearings. This support actually simulates the support system used on the equipment. The spindle and rolls do not actually contact either measurment pedastal. My concern is, with the 2 pedastals being connected together and the spindle, roll ect. not riding on either measurement pedastal are we not getting mixed vibration readings transmitted from one pedastal to the other, and are we not affecting the systems response by connecting the pedastals and then adding 80 lbs of weight to the other. I know this sounds sort of vague. I would be glad to e-mail anyone pictures of this setup to see if anyone agrees with me that the application we are trying to use this individual machine for is impossible with the exception of using it for nothing more than an expensive drive unit for my field balancing analyzer.
Regards,
MICJK

 

[GregLocock]

In theory you should be OK.

So much for theory.

In practice you should endeavour to get your accelerometers mounted to the bearing caps, to minimize cross coupling. You should also try and get your correction plane as close as you can to the bearing.

Does that make sense?



Cheers

Greg Locock

 

[micjk]

Thanks Greg,
The balance machine has "permanently" mounted accels in the pedastals, I use the term "permanently" loosely because they could be moved but that would "open up another can of worms". As far as where "my" accel's for my field analyzer ( Csi Ultraspec 8000 w/fastbal II program) are mounted, I have them placed Horiz. and vertically on this 80 lb support block directly in line with bearing locations. The correction planes are existing. I have no problem with balancing using my analyzer but am trying to prove that with the design "mod's" we have done to the balance machine itself will not give us reliable results using the machines balance program.

 

[jclough]

Could you e-mail me your pictures? Sounds like this is right up my alley


Jon

 

[jclough]

Sorry,

bal_pro@hotmail.com

 

[Ralph2]

I don't think that your setup will ever give reasonable results. The pedestals "must" be able to sense indepentantly and having them connected in any manner just will not work. That is one of the reasons that a HB machine has a crowned roller, to allow as much free movement at the other end as possible. If the roll was flat this would bind the other end so to speak. Not very much but a balance machine is extremely sensitive. I find that on the occasion of balancing something in its own bearing housing (mounted on a angle plate) the felt seal strips can influence the results on the other plane...... leading to a lot more runs then what should be the case and weights added the full circle.
Hope it helps..
Ralph V

 

[micjk]

Thanks Ralph,
That is pretty much what I thought. Now I am wondering why the heck Vibrasys ( one of CEMB's distributors) fabricated and sent us the piece we have connecting the pedastals together. Besides the fact that it was a heck of a commision for someone.
Regards
Roy G

 

[Ralph2]

Have no idea, I would seriously question just how much they know about the theory of balance machines.
Ralph

 

[jclough]

THis sort of set up is often used when balancing rotors in thier own housings or pieces when the parts C.G. is outside of the bearing surfaces. This system should work pretty well if you need to correct for static unbalance only.

If you need to do a two plane balance the connection between the two work supports will give you difficulties with plane separation.

Special work supports are available (at least with Hofmann equipment)which have internal round springs. Hofmann calls them Uniport work supports. These work supports allow rotation about the vertical axis when the standard flat springs are released. Though they don't allow for ideal plane separation, they work pretty well. Does your machine have the Uniport supports?

 

[GregLocock]

The connection between the measurement pillars will confuse things, but in principle two plane balancing should still work. For instance, we two plane balance driveshafts in cars, and could mount the accelerometers either to the diff and the trans, or we could mount them to the body, near each end of the driveshaft. The balance solution would still be the same, but the influence vectors would not be.

We tested this out when we were developing the production balancing system - obviously it is a lot easier to speedily attach accelerometers to the body than the aluminium bits.

However, the cross coupling is alot stronger, and the signals are generally weaker, so in the end we put the accelerometer on the diff, and single plane balancing was good enough once we sorted a few problems out at the other end.

Cheers

Greg Locock