Machine Foundation Design for Vibrations

[J2CUB612]

How are rotating unbalanced forces calculated by equipment manufacturers? A 10 megawatt generator for example.

We frequesntly design foundations for large rotating equipment. Generators, pumps, compressors etc. A recent example was a smaller 5MW combustion turbine generator set.

We use the standard techniques found in text books such as "Vibrations of Soils and Foundation". We frequently estimate the dynamic unbalanced force for the equipment (say the generator) using this formula

Unbalanced Force=Rotor Weight*Rotor Speed(rpm)/6000.

Or by using the rotor shaft eccentricity (provided by the manufacturer) and mass to come up with unbalanced force using this formula F=m(e)(4*Pi^2)*(f^2). m=mass, e=eccetricity of the shaft, f=frequency of rotation.

Sometimes we get manufacturers data giving us the unbalanced force directly which is usually just one number such as, "1248 pounds at 1800RPM" for a generator lets say. This information is usually always a lower value than the equation above estimates.

Recently a generator manufacturer gave us forces much higher than the formula would estimate and gave different forces for vertical and horizontal directions. They indicated that the forces were calculated using bearing stiffness(k) which was different in each direction but that they assumed a constant displacement(x) in each direction. The result of calculating F=kx gave high horizontal forces and even higher vertical forces.

Can someone familiar with machine design discuss how unbalanced forces (for use by the structural engineer) are usually determined by manufacturers?

It seems that the basic unbalanced rotor force based on shaft eccentricity is always the same in each direction and that would represent the basic vibration load we are interested in structurally.

Thanks

 

[GregLocock]

"Unbalanced Force=Rotor Weight*Rotor Speed(rpm)/6000 "

I can see no reasonable explanation for that equation as a general purpose description. It isn't dimensionally consistent.

Cheers

Greg Locock


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

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

 

[J2CUB612]

Your correct Greg. At face value there are no consistent units. The factor 1/6000 is not a dimensionless value. It contains all of the missing units to result in an unbalanced force with units in pounds. Rotor weight must be used in pounds.

It's really a simplified version of the second equation with an assumed value for e and unit conversions from frequency in hertz to RPM and mass to weight in lbs. I quoted it because it's a common foundation design text book formula that many in my field will recognize.

The error in this equation is usually the overly conservative assumption of e.

That's where I'm trying to learn more from some machine design specialists.

How do they determine the value of unbalanced force given their technical knowledge of the equipment.

 

[Tmoose]

For rotating rotors it's from worst case expected unbalance.

For big reciprocating equipment it includes the inertial effects that the configuration does not counteract.

http://www.authorstream.com/Present...-6-Lecture-lect-Entertainment-ppt-powerpoint/

 

[TPL]

You can use ISO1940 to get an idea of what typical levels of acceptable residual unbalance should be (with units of gram/mm)- you can use this to generate resulting unbalnce forces. Use engineering judgement to accomodate worst case factors, say a factor of 10. For design of foundation, use half of force applied in phase to each end (bearing location) for translational modes and out of phase for couples.

 

[baw1]

ACI (American Concrete Institute) has a publication that gives different approacheds.
ACI 351.3 R-04 "Foundations for Dynamic Equipment."

 

[J2CUB612]

Thanks baw1. You're correct. We use this document as well as several other design guides and text for designing the foundation.

I'm looking for a technical understanding of how the data we use as input, the unbalanced force, is actually determined by manufacturers when they supply it.

I can better detect bad data from suppliers if I understand more about what they are doing.