Induction Heating: Everything You Wanted to Know, But Were Afraid to Ask (February 2018)
February 12th, 2018
Question of the Month: We are planning to use induction to heat composite metallic materials prior to diffusion bonding. Is there any concern with respect to excessive noise? Have you found that certain frequencies are noisier than others? If so, do you have a solution for containing the noise level?
Answer: In the vast majority of induction heating applications, industrial noise does not reach an appreciable level. Therefore, there is no reason to be concerned about excessive noise, although there are a few exceptions. Lower frequencies typically result in higher coil current thus increasing electromagnetic forces and coil turn vibration. So it is reasonable to expect that a system applying line frequency (60 Hz) would be associated with a greater magnitude of industrial noise compared to 3 kHz or 30 kHz. In addition, systems with a greater amount of kW most likely would also be associated with more noise. Therefore, assuming all other factors are identical, 500 kW can produce greater noise than 100 kW.
When discussing audible noise, we must bear in mind the two main factors that impact how humans are affected by industrial noise—magnitude and unpleasantness/discomfort level. For example, low-frequency audible noise (e.g., 50 60 Hz) could have a higher magnitude, but it might not be as unpleasant to the human ear compared to a lower magnitude noise at an elevated frequency (e.g., 1 kHz).
Following are the four main sources of noise generation during induction heating:
- Noise generated by the power supply. Numerous power supplies are available on the market. For induction heating, power and frequency combinations that use single-module inverters vary from line frequency to several hundred kHz and power levels exceeding 1000 kW even for a frequency in the 800 kHz range. Design features of modern power sources based on semiconductor technology result in nearly silent operation. Therefore, noise stemming from the power supply is not usually a concern.
- Noise generated from vibration of copper coil turns. The two main approaches to building induction coils can be categorized as either open-wound or refractory-encased [Ref. 1]. The open-wound method provides more convenient repair in the event of failure, but coil turns must be secured using studs and proper fixtures to eliminate or minimize vibration and noise. An encased coil using a castable refractory (for example, special grades of cement) offers durability and longer life, eliminating or dramatically reducing the vibration of coil copper turns. Properly designed induction coils do not exhibit problematic noise levels.
- Noise generated by workpiece vibration or resonant sound waves (amplifying effect). If a workpiece consists of some loose parts, then they may vibrate and produce noise. To assess this possibility, the specific workpiece geometry would need to be reviewed. Pressure can be applied in certain diffusion bonding applications to minimize the possibility of individual component vibrations. When induction heating hollow workpieces (e.g., induction heating of relatively thin-walled pipes or tubes), certain frequencies in combination with sufficiently high power densities could emit resonant sound waves of an appreciable magnitude, exceeding the audible limit. In cases like this, audible noise can also be a dominant factor that greatly affects the selection of frequency. Each tube has its own structural resonant frequency (SRF), which depends on the diameter, wall thickness, material, length, and other factors. When the heating frequency of the induction coil is sufficiently close to the tube’s own resonant frequency, then high amplitude vibration and excessive audible noise may occur. In other words, as a radio receiver transforms electromagnetic waves into an audible sound, somewhat similar mechanisms can occur that cause a tube or pipe to act as an amplifier when dealing with certain frequencies. A decision as to whether noise could be reduced by choosing a different (higher or lower) frequency depends on a combination of the structural self-resonant frequency (SRF) of a particular tubular workpiece and the applied frequency of the induction heater. Therefore, in cases like this, it is beneficial to measure the workpiece SRF by simply hitting it with a hammer and measuring the resonant audible noise with some kind of audible receiver with the ability to detect the frequency and amplitude of the measured signal. As a result, more intelligent decisions can be made in determining whether a certain frequency would improve the noise level or not compared to a frequency that produces unacceptable noise. For example, if the SRF of a certain workpiece is 300 Hz, then the further away the selected frequency is, the lower the noise that will be produced. In this case, 6 kHz would produce a noticeably lower noise compared to 1 kHz. In contrast, if the SRF is 5 kHz, then the use of 6 kHz will make it worse compared to 500 Hz. Many times, a protective or reducing atmosphere is used while heating metallic materials. Therefore, if your induction system will be located inside a gas-tight chamber/enclosure, this will substantially reduce the noise level.
- Noise generated by vibration of power cables, buses, and fixtures. There is a very small probability of using the wrong design of fixtures, power cables, and bus bars. However, any concerns may be addressed by an induction heating expert.
More discussion on these points can be found in Ref.1.
Dr. Valery Rudnev, FASM
Director, Science & Technology
- V. Rudnev, D. Loveless, and R. Cook, Handbook of Induction Heating, 2nd Edition, CRC Press, 2017.
Beginning in July 2016, the Professor Induction column started a new article series called “Induction Heating: Everything You Wanted to Know, But Were Afraid to Ask.” The most commonly asked questions related to different aspects of induction heating and heat treating will be reviewed and explained. All are welcome to send questions to Dr. Rudnev at email@example.com. Selected questions will be answered in this column without identifying the writer unless specific permission is granted.