Understanding, quantifying and correcting noise sources

Air or structure borne?

There are ways to separate air and structure borne noises transmission paths. For example, a functioning rotating machine can produce air or structure borne noise. Airborne noises can come from a ventilation duct or an opening in a dividing wall (electrical plug, for example). Structure borne noises are transmitted to the structure that supports the equipment, which then vibrates and excites the air volume in the receiving room through the dividing wall.

When noise appears to be coming from several areas, the transmission paths must be tested individually. This can be performed using auxiliary noise sources.

To quantify airborne noise, speakers can be used in the space housing the machine in need of soundproofing. The speakers will emit an airborne noise and eliminate the structural component. The difference between the sound in the room housing the noisy equipment and the receiving room provides the level of airborne soundproofing provided by a structure. A hammer will offer the same solution, but from a structural point of view.

These insulation performances can then be applied to the noise in the room housing the machine. In other words, the overall noise (air and structure-borne) generated by the machine minus the measured airborne isolation performance measured will determine the structural component of the sound caused by the equipment. This exercise identifies the main source of noise (air or structural) and helps select the right solution to reduce noise in specific areas, while quantifying the maximum impact this solution will have.

Identifying airborne noise generation mechanism

Certain noise sources have a clear transmission path, but it is sometimes difficult to determine their exact origin. The obvious solution is to measure the noise using a sound level meter and to identify, using a scanning procedure, the approximate origin of a noise (nearfield measurement). However, this technique has a maximum precision and is not always efficient in a reverberant environment. Additionally, some equipment can be difficult to access, which makes nearfield analysis impossible. However, there are solutions to remotely identify problematic parts on a piece of equipment.

Sound intensity probes measure not only noise amplitude, but direction (vector). Noise vectoring can occur when two or more microphones are linked on an sound level meter. Recent technological improvements have made it possible to increase the number of microphones used simultaneously, thus increasing the vectoring capacity of sound level meters technique on distant planes (acoustic holography). Furthermore, the combination of these holographic instruments along with sensors attached directly to the studied structure allows the use of consistency factors to reduce the false effects of reverberation and centre the instrumentation on a specific piece of equipment.

This technique helps identify, for example, acoustic leaks in a wall divider and thus quantifies the relative contribution of individual leaks with respect to the overall noise that is heard. It is therefore possible to determine the maximum potential impact of a specific correction, which means it is possible to determine whether it is worth correcting defects in a structure or whether to think of other reduction methods to achieve the desired sound objective. This can save significant costs in comparison to a trial and error process.

Finding the main structure-borne transmission path

Understanding a vibroacoustic phenomenon can be particularly complex when it involves a structure or building, because of several probable transmission paths. That’s why it is best to first isolate the equipment from the structure wherever possible. However, when this is not sufficient or technically feasible, the solution is to study transmission paths using frequency response functions (FRF).

Using a hammer and accelerometers enables to study a structure’s vibratory isolation performance piece by piece. Measuring FRFs makes it possible to observe whether the structure naturally amplifies vibrations at specific frequencies (natural frequencies). It is indispensable to perform a step-by-step process, from the noisy equipment to the receiving room, to isolate the role each component plays in transmitting vibrations. Analyzing input and output frequency content variations of components in transmission paths helps determine the individual contribution of a structure’s components and their natural frequencies.

If the step-by-step process is sufficiently refined (many measurement points per component), a visual reproduction of component deformations and of the overall structure (modal analysis) can be created by synchronizing the amplitude and the phase of measured accelerations. Using finite element models makes it easier to visualize this process; however, it takes longer and is more costly than an FRF measuring process.

Superimposing the forced frequencies produced by noisy equipment over the structure’s natural frequencies helps identify components that play a primary role in an area’s sound contribution. The noisy component’s rigidity or damping can then be modified to estimate future behaviour of a structure after modifying the part generating the resonance.

Conclusions

It can be challenging to identify and understand vibroacoustic phenomena. A trial and error process can be costly, inefficient and might not produce the desired effects. That’s why it is best to use the analysis process to discern the type of transmission (air or structure-borne) and quantify the relative importance of the various transmission paths. Ultimately, this analysis helps to understand how a vibroacoustic system behaves, minimizes the time required to correct the issue and minimizes the investment necessary to attain the desired noise goals.