top of page

AHU Vibration Isolation: Check Spring Deflection at the Lowest Operating Speed

Oct 22
5 min read
Illustrative enclosed machine base supported by compression-spring vibration isolators.

Illustrative engineering scene. The worked calculation below uses explicit example inputs.


A spring isolator can look satisfactory on a full-speed selection and become much less effective when the AHU fan slows down. The spring does not become stiffer at turndown. The forcing frequency moves closer to the natural frequency of the supported assembly, changing the fraction of exciting force transmitted to the structure.


Review the complete permitted operating-speed range before accepting a statement such as “90% isolation.” That percentage needs a frequency, a load and a mechanical model. The worked comparison below shows why 25 mm of actual loaded deflection may exceed an illustrative target at 900 rpm but miss it at 450 rpm.


Start with the supported assembly, not the spring name


Identify whether the isolators support a fan skid inside the AHU, the complete AHU, or a separate base. Use the operating mass on that isolated assembly, including components actually carried by it. A total equipment weight divided equally among four mounts is valid only when the centre of gravity and mounting geometry justify equal reactions.


Static deflection means the spring compression caused by the supported load, measured relative to its unloaded condition. It is not the overall spring height or the gap below the frame. Greenheck's isolation catalog distinguishes nominal from minimum deflection and makes the supported load part of selection. A nominal “50 mm” spring family does not establish 50 mm compression under every load.



Use a transparent screening model


For a linear spring supporting a rigid mass on a rigid foundation: fn = (1 / 2π) × √(g / δ). Here fn is natural frequency in Hz, g = 9.80665 m/s² and δ is actual loaded static deflection in metres. For once-per-revolution excitation, fd = rpm / 60 and frequency ratio r = fd / fn.


In the ideal undamped model, force transmissibility T = 1 / |1 − r²|. For T below one, the corresponding force-isolation percentage is 100 × (1 − T). T is transmitted force amplitude divided by exciting force amplitude.


ASHRAE's vibration fundamentals describe this single-degree-of-freedom model and its limits. Isolation begins only above a frequency ratio of approximately √2; near r = 1, resonance produces amplification. Real damping limits the ideal mathematical singularity, but does not make operating near resonance acceptable by default.



Keep this calculation separate from an airborne sound calculation. Ten percent transmitted force does not mean ten percent of the fan's sound power reaches an occupied room. The source force, structural response, connections and acoustic radiation still matter. This model also does not predict a vibration-velocity acceptance value in millimetres per second.


Grouped bars compare transmitted-force amplitude at 900 and 450 rpm. For 25 mm deflection the values are 4.620% and 21.454%; for 50 mm they are 2.258% and 9.688%.

An illustrative 10% force-transmission target is exceeded by the 25 mm mount at 450 rpm in this ideal undamped rigid-support model.


Original comparison at 900 and 450 rpm


  • Illustrative fan-skid operating range: 450–900 rpm.

  • Dominant excitation considered: once per fan revolution.

  • Case A: 25 mm actual loaded spring deflection.

  • Case B: 50 mm actual loaded spring deflection.

  • Illustrative review target: no more than 10% force transmission at the lowest continuous operating speed.

  • Assume linear springs, negligible damping and an effectively rigid supporting structure for this comparison only.


Case A natural frequency = √(9.80665 / 0.025) / (2π) = 3.152 Hz. At 900 rpm, fd = 15.0 Hz, r = 4.759 and T = 0.04620: approximately 95.38% force isolation. At 450 rpm, fd = 7.5 Hz, r = 2.379 and T = 0.21454: approximately 78.55% force isolation.


The 25 mm case exceeds the 90% illustrative isolation target at high speed but does not meet it at low speed. That result concerns the transmission ratio. It does not prove that the absolute transmitted force or room vibration is larger at 450 rpm, because the exciting force itself may change with speed.


Case B natural frequency = √(9.80665 / 0.050) / (2π) = 2.229 Hz. At 900 rpm, T = 0.02258, or 97.74% force isolation. At 450 rpm, T = 0.09688, or 90.31% force isolation.


Increasing actual deflection improves the low-speed transmission ratio in this model. The 50 mm result clears the selected target only narrowly, so it should not be treated as a generous performance margin after real support flexibility, load differences and manufacturing tolerances are considered.


Solve backward from the lowest speed


The same equations can reveal the minimum theoretical deflection before asking the supplier for a selection. For the isolation region, T no greater than 0.10 requires r² no less than 11. At 450 rpm, the forcing frequency is 7.5 Hz, so the permitted natural frequency is at most 7.5/√11 = 2.261 Hz.


Required ideal deflection δ = 9.80665 × 11 / (2π × 7.5)² = 0.048577 m = 48.58 mm. This is a calculated minimum actual deflection for the stipulated model and target, not a catalog designation or a complete isolator specification.


As an independent check, insert 48.58 mm back into the frequency and transmissibility equations. The result should return approximately 10% transmission at 450 rpm. Also note that doubling deflection reduces natural frequency by √2, rather than by two. These checks catch common metre-to-millimetre and square-root errors.


Check the load schedule and the supporting structure


Suppose a spring carries 250 kg and compresses 25 mm. Its stiffness is approximately 98.07 kN/m. If that same spring carries only 150 kg, its compression is 15 mm, not 25 mm. This simple load check explains why spring selection must follow actual reactions. An unevenly loaded multi-mount assembly requires a coupled review; separate mount calculations alone do not determine its rocking modes.


A flexible roof or floor adds another dynamic system. ASHRAE's applications guidance treats supporting-structure deflection and resonances as part of isolation selection. Where those effects matter, give the structural and vibration specialists the mounting locations, operating masses and speed envelope. A rigid-foundation screening result is not transferable without review.



Turn the calculation into commissioning evidence


  • Obtain the selected mount model, spring stiffness or load-deflection data, allowable load and expected installed compression at each support.

  • Confirm the operating-speed range and relevant excitation frequencies with the equipment supplier.

  • Check loaded deflections and clearances against the approved installation instructions.

  • Inspect for unintended rigid connections across the isolation boundary and unreleased shipping restraints.

  • Have qualified specialists evaluate startup, shutdown and any observed resonant response before changing operating limits.


The calculated natural frequencies correspond to about 189 rpm for the 25 mm case and 134 rpm for the 50 mm case. Startup may pass through those frequencies even when continuous operation begins at 450 rpm. The undamped steady-state formula does not predict the transient response during that crossing.


Get the complete AHU Selection & Fault Diagnosis Handbook for a coordinated review of fans, sound and vibration, installation and commissioning. Use its review structure alongside manufacturer isolation data and project-specific structural advice, rather than as a substitute for either.




Continue the calculation review




Sources and calculation basis


Primary references checked on 22 October 2026. The worked examples and diagrams are original illustrative calculations, not manufacturer selections or project approvals. Apply the stated limitations and the requirements governing the actual project.





Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page