When Anti-Vibration Mounts Transmit Instead of Isolate: A Case Study from a Vietnamese Gas Processing Plant
In early 2025, the maintenance team at a gas processing facility in Ba Ria-Vung Tau Province, Vietnam reported an unusual pattern of structural degradation. The concrete plinth supporting a 450 kW reciprocating compressor — commissioned less than two years earlier — had developed visible cracks radiating from all six mount anchor points. The compressor itself was operating within its mechanical specifications. Oil analysis showed no abnormal wear. Alignment checks, performed quarterly, showed a progressive drift that required correction every six months. The problem was not the machine. The problem was what sat between the machine and the foundation.
The facility's engineering team contacted ChuncoTech with a straightforward question: the mounts were rated for the compressor's static weight — why were they failing? The answer required a site visit, a bump test, and a recalculation of the one parameter the original specification had omitted: the mount's loaded natural frequency relative to the machine's excitation frequency.
What the Static Load Rating Did Not Predict
The compressor operated at 585 RPM, producing a dominant excitation frequency of 9.75 Hz. The original mounts were six cylindrical NR/SBR pads at Shore A 60, each rated for the compressor's distributed static weight. The selection was correct by the load-per-mount calculation. It was incorrect by the dynamic criterion that governs isolation performance.
A bump test conducted on-site recorded the actual vertical natural frequency of the mounted compressor at 7.1 Hz. At this natural frequency, the frequency ratio r — the excitation frequency divided by the natural frequency — was 1.37. The threshold for vibration isolation is r > 1.414 (√2). Below this value, the mount does not isolate — it amplifies. At r = 1.37, the transmissibility was approximately 1.8, meaning each piston stroke delivered roughly 1.8 times the machine's dynamic force directly into the concrete plinth.
The root cause of the discrepancy between the catalogue stiffness and the measured natural frequency was threefold. Under operational preload — the sustained weight of the compressor plus gas load on the pistons — the mounts experienced a mean compressive strain above the linear region of NR's stress-strain curve. The effective dynamic stiffness was approximately 1,250 N/mm per mount, not the catalogue value of 520 N/mm. The Mullins effect — progressive stiffening of filled rubber under repeated loading — compounded this over the first several hundred operating hours.
The cylindrical mount geometry provided stiffness in the vertical compression axis only. Under the compressor's combined vertical and horizontal loading from the reciprocating piston mass, the mounts developed a rocking mode — a rotational degree of freedom that the single-axis specification never accounted for. This introduced a second resonance path through the lateral plane.
Vietnam's climate added a third factor. The compressor building's ambient temperature averaged 35–38°C year-round, with the compressor casing radiating additional heat. At these sustained temperatures, NR compounds undergo thermo-oxidative crosslinking — a progressive hardening of the polymer network. Hardness measurements on the extracted mounts confirmed Shore A had increased from the specified 60 to approximately 71 after approximately 1,500 hours of thermal exposure.
The Engineering Response: Geometry, Compound, and Configuration
ChuncoTech's technical team proposed a three-variable solution: change the mount geometry to engage shear deformation, change the compound to resist thermal aging in Vietnam's climate, and change the mounting configuration to suppress the rocking mode.
Mount geometry. The replacement specification used bell-type anti-vibration mounts from the CBT-M series. Unlike a cylindrical pad that behaves as a single-axis compression spring, the concave bell profile distributes the supported load across both compression and shear planes. Rubber is inherently softer in shear — the shear modulus of typical NR compounds is approximately one-third of the compressive modulus at equivalent strain. By routing a portion of the vertical load through shear, the bell mount achieves a lower vertical natural frequency for the same static load capacity. After installation, the measured vertical natural frequency was 4.3 Hz, restoring the frequency ratio to r = 2.27 and transmissibility to approximately 0.24 — from 80% amplification to 76% isolation.
Compound selection. Vietnam's sustained high ambient temperature was not an incidental condition — it was the primary environmental driver of the original mounts' stiffness drift. The CBT-M mounts were specified in an EPDM-modified compound at Shore A 55 ± 3, with a compression set below 12% at 70°C/72h per DIN 53517. EPDM's saturated polymer backbone resists the oxidative crosslinking that hardened the original NR compound. The marginally lower hardness grade — 55 versus 60 Shore A — further reduced dynamic stiffness, while the bell geometry's larger effective load-bearing area maintained the required static load capacity.
Mounting configuration. The number of mounts was increased from six to eight, with the two additional mounts positioned at the compressor's center of percussion to suppress the rocking mode. Each mount was fitted with an integrated height-adjustable leveling plate — the original mounts had been shimmed with steel plate offcuts that introduced point-loading at the mount-plinth interface, a contributor to the concrete cracking.
| Parameter | Original Mounts (Cylindrical NR/SBR) | ChuncoTech CBT-M (Bell-Type EPDM) |
|---|---|---|
| Mount geometry | Cylindrical — compression only | Bell-type — compression + shear |
| Compound | NR/SBR, Shore A 60 | EPDM-modified, Shore A 55 |
| Measured natural frequency (vertical) | 7.1 Hz | 4.3 Hz |
| Frequency ratio r at 9.75 Hz | 1.37 (amplification zone) | 2.27 (isolation zone) |
| Vibration transmissibility | ~1.8 (80% amplification) | ~0.24 (76% isolation) |
| Compression set (70°C/72h) | 28–34% (extracted mounts) | < 12% (DIN 53517) |
| Number of mounts | 6 | 8 (anti-rocking layout) |
| Height compensation | Steel shims (point-loading) | Integrated leveling plates |
Implementation and Measured Outcome
The mount replacement was executed within a 36-hour planned maintenance window. The compressor was lifted on hydraulic jacks, the original mounts and steel shims removed, and the concrete plinth surface ground flat at each mounting point using a portable surface grinder. The CBT-M mounts were installed with leveling plates adjusted to a common reference plane using a laser level. Mounting bolt torque was set to 85 N·m ± 5 N·m, verified with a calibrated torque wrench and recorded against each mount position on the installation checklist.
Vibration measurements were taken at three milestones: commissioning, three months, and twelve months. The control room — located 40 meters from the compressor building — had registered 6.3 mm/s RMS before the replacement, exceeding the ISO 10816-3 Zone A limit. After commissioning, the reading was 2.8 mm/s RMS — a 56% reduction, within the acceptable zone. At three months, the reading was 2.9 mm/s RMS, within measurement repeatability. At twelve months, it was 3.0 mm/s RMS. The dynamic stiffness had remained stable across the monitoring period with no drift.
The concrete plinth showed no new crack formation over the twelve-month observation window. Existing cracks — epoxy-injected during the mount replacement — showed no propagation when re-inspected at the six-month and twelve-month marks.
A representative CBT-M mount was extracted at the twelve-month service interval for laboratory evaluation. Shore A hardness measured 56 — within the 55 ± 3 specification. Compression set after approximately 8,000 operating hours was 9% — within the sub-12% specification. The mount's dynamic stiffness, measured under representative preload, had shifted by less than 6% from the as-installed baseline.
Frequently Asked Questions
Three field indicators that do not require specialist equipment. Visual inspection: surface cracking on the rubber body — particularly radial cracks around the bolt holes or compression set visible as a permanent flattened profile — indicates the rubber has aged beyond its elastic recovery range. Vibration measurement: if the vibration level at the foundation or adjacent structure has increased measurably since commissioning — even if the machine itself is unchanged — the mounts have likely stiffened and shifted out of their isolation window. A before-and-after comparison using a handheld vibration meter at the same measurement point is sufficient to detect the trend. Alignment drift: rotating equipment that requires more frequent alignment correction than the plant's scheduled interval — without any change to the machine, coupling, or foundation — is often sitting on mounts that have lost thickness through compression set.
Rubber mounts provide multi-axis isolation through the material's inherent damping. Rubber mounts are the default choice for mid-to-high frequency isolation (above 20 Hz). Spring mounts provide lower natural frequencies — typically 2–5 Hz — than rubber mounts, making them the choice for low-frequency isolation (below 15 Hz excitation). Spring mounts have near-zero damping and require separate damping elements.
The specification sequence is: measure machine's operating RPM; calculate excitation frequency. Select a mount with a loaded natural frequency below half the lowest excitation frequency (target r > 2.0). Confirm the mount's load rating covers the actual distributed weight plus dynamic load allowance. Verify the compound is rated for ambient conditions (NR for <35°C; EPDM/CR for higher temperatures or chemical contact).
What This Case Means for Mount Specification
The Ba Ria-Vung Tau case is not unique to Vietnam. It is a predictable outcome whenever anti-vibration mounts are specified by static load rating alone, without accounting for the dynamic stiffness under actual operating preload, the compound's aging behavior at the site's ambient temperature, and the mount geometry's ability to constrain the machine's degrees of freedom.
A mount's static load capacity only confirms it will not collapse under the machine's weight. Whether it isolates or amplifies vibration is determined by the ratio of the excitation frequency to the loaded natural frequency.
The compound matters as much as the geometry. In climates like Vietnam's, sustained ambient temperatures above 35°C accelerate the hardening of standard NR compounds. Specifying EPDM or CR-based formulations prevents the progressive stiffness drift.