Most rotating or reciprocating equipment needs to have proper vibration isolation to prevent noise problems elsewhere within the building. Machinery vibration that gets into the building structure can travel great distances, and can create noise quite far from the original source.
One high-rise building had a mysterious noise occurring on the fifth floor, but nowhere else. After much investigation, the noise was tracked back to the cooling tower circulation pumps on the 21st level penthouse. The pipes were poorly isolated from the cooling tower pumps, and vibration traveled via the pipe walls. Then the seismic restraint cables supporting the pipe riser were unduly taut on the fifth floor. Pump vibration conducted through the piping and tight cables into the shaft wall, creating significant vibration-induced noise -- but only on the fifth floor. The solution required adding an effective neoprene bellows pipe isolator at the pump, which kept most of the vibration from getting into the piping in the first place. Loosening the restraint cables in the shaft at the fifth floor then prevented the transfer of the remaining pipe vibration into the wall. The noise disappeared.
When properly vibration isolated, it should be possible to easily rock the equipment, even for very large machines. Standing on the corner of the isolation base should be more than enough force to easily see movement. With small equipment, fingertip pressure should be used. If the equipment cannot be readily moved, or if the deflection under load is minimal, the isolators are too stiff to be effective.
Vibration isolators are described in terms of the "minimum static deflection". Static deflection is simply the amount that a spring compresses under the weight of the equipment. If a spring is 2" long when unloaded, and compresses to 1" when carrying the weight, we say that it has 1" of static deflection. The more static deflection, the "squishier" the spring. Static deflection is directly related to vibration isolation efficiency. 1" static deflection equates to about 88% isolation; 1/2" static deflection isolates roughly 50% of the vibration energy. Static deflection is also related to the lowest frequency of vibration. To isolate very low frequencies (less than 20 Hz) requires very large static deflections -- 4" or more. Beyond a certain point, springs can no longer do the job, and special air isolators are needed.
A neoprene "waffle" pad is
typically rated at 0.25" static deflection, and does not work for
frequencies below about 300 Hz. Such pads often get placed under
electrical transformers (vibrating primarily at 120 Hz and 240 Hz). These
pads provide very little actual isolation at these low frequencies.
Do not over-tighten the fasteners. A neoprene pad may provide adequate isolation when unloaded, but excess compression can crush the pad. When it is crushed, it becomes essentially rigid and no longer provides isolation. A good rule of thumb is no more than 25% compression of the pads under load or under fastener tension.
Putting the equipment on springs or rubber pads is only part of the picture. Piping, duct and conduit connections must also be adequately flexible, or else the vibrations will travel around the isolators. We refer to this as a "short-circuit".
Inertia bases are usually not an important part of the vibration isolation scheme. For continuous vibration from rotating equipment, the inertia base mostly minimizes transient movements of the equipment during start-up or operational cycling, which reduces stress on piping and duct connections. Inertia bases are not that important for airborne noise either. The sound level at the corner of the concrete base is virtually identical to the sound level a few inches away, which transmits through the thinner structural slab.
Braided metal "flex" connectors
are NOT effective as vibration isolators. They are simply too stiff,
especially when pressurized. The braided connector does help prevent
start-up motions from damaging the pipe or conduit, but it does not isolate
equipment vibration from the piping. On many occasions we have actually
measured higher vibration levels on the "isolated" side of the
braided metal flex connector than on the equipment side.