Queries on reducing shaft vibration 1

[vanuta]

I am having a problem on shaft vibration.

I have 3 shafts (york slip) that is running between a PTO and a pump.The three shafts are connected by universal joints. The three shafts goes through 2 sets of flange bearings, which are bolted to a steel wall. On one of the walls, there is excessive vibration and i suspect it might be due to the large universal joint angle.Is there a way to reduce this vibration without changing the universal joint angle?

1) would a heavier duty bearing help in reducing the vibration?
2) I do not think the wall is very straight as the ends of the wall are welded to some beams. While welding, there might be warping of the wall due to the extreme heat that might have caused uneven surfaces where the bearings are mounted. Could this be a problem?
3)Are there vibration damping ball bearings in the market?

I have attached a rough sketch of my configuration

 

[vanuta]

Will a double carden joint help? forgot to add in the previous thread

 

[tbuelna]

vanuta,

Make sure your U-joints are phased properly on each end of the shafts.

http://upload.wikimedia.org/wikipedia/commons/thumb/f/fd/UJoint1.png/800px-UJoint1.png

Hope that helps.
Terry

 

[vanuta]

Could you please explain how the graph should be intepreted and what does shaft angle mean? is it the vertical angle with respect to the wall?

 

[Tmoose]

Could you indicate the locations of the universal joints?
What are the operating angles of the u-joints?

What is the frequency and amplitude of the "excessive" vibration?
I'd make a map of the vibration (amplitude, frequency, phase if I can get it) all around the bearing(s) on the wall with excessive vibration.

Good u-joint phasing reduces the 2X speed variation, but the "secondary couple", dependent on joint angle and torque transmitted, and ocurring at 2X shaft rotation, remains full strength. My expectation is the vibration is at 2X rotating speed, and the wall is describing an "S" with a line of low vibration "into the paper" in your sketch.

I'd also make a map of runouts/motions using a dial indicator and rotating the train by hand. There is a pretty good chance that shaft runout, component eccentricity, or even just a "stiff" u-joint could be a big part of the problem.

 

[GregLocock]

A double cardan will eliminate most of the torque or speed variation. They are robust but heavy.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[vanuta]

Tmoose,
The universal joints are in each ends of the shaft.
The amplitute peak to peak vibration is about 1.2mm in the axial direction. the attachment is side view..i have posted a new thread with new finidngs in a new thread.... you had mentioned "Good u-joint phasing reduces the 2X speed variation, but the "secondary couple", dependent on joint angle and torque transmitted, and ocurring at 2X shaft rotation, remains full strength". I do not really understand what you mean..could you please elaborate on this? Thanks

Kunal will extra bearing support help. I have posted new findings in a new thread..please do have a look.

 

[Tmoose]

http://www.sdp-si.com/D757/Images/fig2_joints.jpg

If uniform torque enters the input shaft, the intermediate shaft's rpm varies up and down slightly twice each revolution - "non constant velocity." At the same time each joint tries to "straighten out" twice per rev due to "secondary couple." The support bearings must withstand this couple. The severity of the couple is proportional to shaft torque, and the joint's operating angle.

Orienting a second u-joint correctly will return the output shaft's rpm to "constant velocity." But there is still a secondary couple exerted on the second joint;s support bearings.

If the "wall" the first 2 bearings are attached to is steel sheet, it's bending stiffness may be too low to resist the secondary couple. The "axial" vibration might look more like the middle mode shape shown here.

http://www.lusas.com/case/bridge/images/newark_modes3_7_8_300.gif

More thorough vibration measurements with amplitude, frequency, and phase over an area would help make that clear.