Background:

Noise has always been and will continue to be a subject of discussion between the gearbox and motor manufacturers with their respective customer base. By its very nature it remains a controversial and subjective topic. The term “it’s too loud or it’s making noise” is often used to relay complaints from the end user back to the manufacturer.

In this article, the goal is to provide some basic understanding of the dynamics involved and try to give some points of reference to help explain the mysteries of noise related issues within the power transmission industry.

Fundamentals of noise and how it’s measured:

The terms “Sound Power and Sound Pressure” are used to define the category of the measurement taken. The most commonly used is the “Sound Pressure or Sound Level” which determines a noise measurement taken at a known distance from the source. Typical distances are measured (in feet) away from the source (i.e. 1 foot, 3 feet or 5 five feet, as examples). A decibel meter is used to collect readings from the noise source. The decibel meter or (dB meter) can be a simple hand held device typically used in an open shop environment or an expensive sophisticated instrument set up in laboratory conditions.

Sound Power is measured exactly at the source using the decibel meter and without any noticeable distance involved. Both terms (power and pressure/level) will be measured on the dBA scale and reported as such. The decibel meter approach will only measure overall noise levels. In other words “how loud is the machine element” and expressed numerically, (as an example; 62 dBA). This numerical value will help take some of the subjectivity out by giving a value that all parties involved can relate to.

While we now have a numerical value to compare against and we can all agree with, that’s just the start. The dBA scale is a composite value without consideration for frequency and amplitude, it’ just an “unfiltered” noise level for volume only.

For general comparisons about noise related issues please refer to the following. All measurements are based on a distance of 6 feet (industry standard):
• 30dBA is a whisper
• 45 dBA is rustling of leaves, background music
• 52 dBA is typical desktop computer
• 60 dBA is normal conversation
• 68-73 dBA is typical Boston Gear WGSR
• 75 dBA is average radio, vacuum cleaner
• 80 dBA is busy office
• 82 dBA is inside coach section of typical passenger jet
• 85 dBA steady sound levels for a working shift of 8 hours of is the maximum generally permitted as per the 1983 OSHA Published Standards.
• 100 dBA tractor or power saw
• 120 dBA is chain saw, jackhammer or snowmobile
• 135 dBA is jet taking off, rock concert
• 140 dBA is threshold of pain, gunshot or siren

Above information is referenced in the Machinery’s Handbook-25th Edition page 1238 and internet research…

As mentioned previously, the above referenced are just noise levels or volume. What are missing are the characteristics of noise. These values are expressed in terms such as frequency and amplitude. It should be worth noting that the equipment required to record frequency and amplitude is typically expensive and tested under laboratory conditions. While I won’t bore you with the science behind frequency and amplitude, they do play an important role in mathematically describing the type of noise thus leading to very good problem solving techniques.

In short, the take away on frequency and amplitude could be described by the following example:

We have a customer who is complaining that his system is too noisy. The decibel reading measured is 75 dBA. In relative terms, this is a quiet system but the customer is still saying it too noisy. As an example only, by using specialized sophisticated equipment to determine the frequency and amplitude, we discover that the frequency could be in the 3000-4000 Hz (hertz) range. The human ear is most sensitive to the frequencies in this range. At low sound levels, the human ear is most sensitive to high pitched tones than low pitched tones. At very high sound levels the human ear is virtually equally sensitive for all tones. The volume of the noise is not loud by measurable terms it’s the frequency of the noise level which is the issue. Knowing the frequency and amplitude is the key to problem solving. This can direct the manufactures of equipment to focus on certain areas of noise generation. Once the frequency is determined (expressed in numerical values usually between 0-20000 Hz) then “root cause analysis” can direct us to certain elements.

Examples of gearbox noise generators could be gear tip relief, lead and profile error, pitch error, burrs, nicks, bearing noise, fan noise, motor noise, just to name a few.

Examples of motor noise could be windings, bearings, fans, motor misalignment, and loose bolts at the connection points.

Examples of systematic conveyor noise could be structural vibration, loose connection points, bearings, rollers and overlapping natural frequencies of different thickness of steel beams and plates.

All of these mentioned have specific frequency characteristics in the hertz range.

Practical Use of Decibel Meters:

Noise levels produced by a single machine in a relatively quiet environment (no background noise) can be recorded via the hand held decibel meter. This is a straight forward measurement and value.

The real question that comes into play is when you have multiple pieces of equipment running, how do you determine which is producing the loudest noise or if other background noise is present?

So if we don’t have the specialized equipment available to evaluate the noise condition, how will taking a decibel reading help you in your initial troubleshooting exercise?

By using the hand held decibel meter you can rule out surrounding noise from other pieces of equipment if you apply the following logarithm. Logarithmic scales are used to define normally large values into a smaller scale, thus preventing huge long numbers.
For every 3 dB increase in sound level, this doubles the perceived sound for humans.

Step 1:
At the machine of interest, measure the total noise (t) and record the value. Hint; simply hold the hand held decibel meter in the air at arms length in front of you and in the general area of concern.
This value will consider the total noise around the machine and any background noise.

Step 2:
Turn off the suspect machine and repeat the measurement procedure for background noise (b).

Step 3:
Plug in the values for (t) and (b) into the formula and solve for M (machine noise).

Example 1:
A speed reducer assembly complete with motor has been identified as noisy. It’s attached to a conveyor line but it is really hard to determine exactly where the noise is coming from. Is it the gearbox? Is it the motor? Is it the conveyor?

Using the 3 step method above will help determine where the noise could possibly come from. In this example,
Step 1 would record the total noise of the system a (t=75)
Step 2 would disconnect the chain sprocket from the conveyor and record the background noise (b=70)
Step 3 would be:

M = 10 log [10(75/10) – 10(70/10)] = 73.34 dBA

(Would be the noise of the gearbox & motor without conveyor background noise).

To take it one step further you could now measure the gearbox with and without the motor connected.
Step 1 would record the total noise of the gearbox and motor running without conveyor (t=71)
Step 2 would disconnect the motor from gearbox and run motor (b=73)
Step 3 would be:

M = 10 log [10(71/10) – 10(73/10)] = 68.67 dBA
(Would be the noise of the gearbox without motor background noise).

What we have just determined with this example, the motor is louder than the gearbox. In this case, we could say with confidence that we have localized the noise generator and have a decibel value associated with it.

Since both pieces of equipment (motor and gearbox) were close together in decibel readings, it would have been difficult to distinguish by the ear which unit is actually louder than the other.

This approach will not tell you what is causing the noise but it will help the engineers and equipment manufacturers in their evaluation to determine exactly is causing the noise within that piece of equipment and develop a corrective action

Feel free to comment or ask questions by emailing to engineering@bostongear.com