Skip to main content

Dear Forum members,

Anybody using multiplane balancing spread sheet.( more than 2 planes)

Vibration : housing vibration / 1X shaft vibrations

First run - initial vibration

2nd run - trial mass on plane 1  and take response run. (remove wt after taking run)

3rd run - trial mass on plane 2  and take response run. (remove wt after taking run)

4th run - trial mass on plane 3  and take response run. (remove wt after taking run).

Then calculating IFC for all 3 runs with 1st run and calculating correction weights to suppress the vibrations.

Appreciate your responses.

Ravishanker

Original Post

Replies sorted oldest to newest

For a high speed balance in a shop, one covers multiple speeds.  This calculator stuff, may not have what you need.  If you plan to balance only at speed, this may result in a poor balance when taken to the field.

Talking about three plane balances, surely you must be speaking of flexible rotors.  These need to be balance through the speed up to running speed or a selected speed (which was selected based on good technical reasons).

Last edited by Registered Member
@Registered Member posted:

3 plane balancing calculator (vibromera.com)

sent by Walt strong.

At least what I see as far as instructions, there is a statement reading "Panel Run#0 Original vibration, Vo1 and F1 for Plane1 and Vo2 and F2 for Plane2".  I would expect that the variables "F1" and "F2" refer to the phase lag angles. 

The instructions don't seem clear on the trial weight location but I haven't really looked them over that carefully.

The last comment is interesting.  Is this a low speed balance.  Multiplane (above 2) software will have the problem 'overdetermined' mathematically.  This is trouble if this is what you are doing.

Even with a low speed 2-plane balance one can use distributed weights, but this is note really the same as using more than 2 (linearly) independent balance planes. 

The effect (as a vector of measurement points, possibly including different speeds) vector is the run with the trial weight minus the reference run (vectors all).  Call these Ei - effect of adding ith balance weight set.

The reference run is the run without the trial weight (weights can be removed, too - negative amplitudes or 180 degrees out). The trial weights are vectors with amplitudes and angles as are the vibrations at the measurement points.  Call these vectors Ti - for trial weight set i.

Use A for the matrix of influence coefficients, ICs.

Now if you use only one plane for each balance weight the calculations are simple; however, they are not as general, and it is difficult to keep track of balances if maintaining ICs for future balances (can be used for fleets of machines).

Then A*[Ti] = [Ei] + errors is a regression equation with unknown A, the ICs.  This is how I calculate multiplane ICs (mostly).  Regression minimizes the errors in some way - the usual way is to minimize the sum of the squares of the errors.

Note [Ti] and [Ei] are matrices formed from the effects and the trial vectors, e.g.

[Ti] = [T1 T2 ... Tn] for the n trial weights.

When applying multiplane trial weights this will solve for ICs.  Also, subsequent balances and trims can be added to refine the ICs.

Have I used this in practice.  Over a 100 times with great success.

Dear sir,

Thanks for your explanation, but i could not understand little bit,

request you to explain with one example vibration reading as per your explanation,

so that i can understand easily.

as per my post:

First run - initial vibration

2nd run - trial mass on plane 1  and take response run. (remove wt after taking run)

3rd run - trial mass on plane 2  and take response run. (remove wt after taking run)

4th run - trial mass on plane 3  and take response run. (remove wt after taking run).

Then calculating IFC for all 3 runs with 1st run and calculating correction weights to suppress the vibrations.

Ravishanker

@Registered Member posted:

Dear sir,

Thanks for your explanation, but i could not understand little bit,

request you to explain with one example vibration reading as per your explanation,

so that i can understand easily.

as per my post:

First run - initial vibration

2nd run - trial mass on plane 1  and take response run. (remove wt after taking run)

3rd run - trial mass on plane 2  and take response run. (remove wt after taking run)

4th run - trial mass on plane 3  and take response run. (remove wt after taking run).

Then calculating IFC for all 3 runs with 1st run and calculating correction weights to suppress the vibrations.

Ravishanker

If it the basic procedure of calculating the IC then you should perhaps review the presentation at ROTOR BALANCING & INFLUENCE COEFFICIENT ANALYSIS.  In particular the slides 19-23 demonstrate how to calculate the parameter.

High speed balancing is often more efficient when using distributed sets of balance weights rather than applying a single plane weight each time.

Also, if you make the vibration response better you should not be inclined to remove the  balance weight that is helping - just use a different reference run.  There is no need to remove weight if helpful, as is there no need to use single plane balance weights to calculate the influence coefficients.

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×