Web Analytics

Monitoring Hardening Processes on Gear Components Using Eddy Current
From Industrial Heating.com - January 2008

by Bill Buschur
January 9, 2008

High production volumes and just-in-time manufacturing call for a reduction in destructive testing in order to maintain production flow. Eddy-current systems are an ideal way to monitor quality and still save time and money.

In the hardening shop of a large automotive factory, gear components for transmissions are hardened in continuous carburizing furnaces with very high production volumes.

Checking of the hardening depth was formerly made by microscopic examination of microhardness tests on cut and polished randomly selected samples. Under normal conditions, four pieces per shift of each of 11 parts (132 gears per day) were checked by this destructive method.

Problems with this situation were recognized as follows:

  • Under wrong process conditions or failure of the destructive test, the number of destructive test pieces was significantly increased to try to assure that no wrong-hardened parts were shipped.
  • Results from the destructive tests were not known for 30 minutes at the earliest.
  • Only pieces known with 100% certainty to be good were the destroyed ones.
  • The properties of all pieces in a batch could only be assumed.
  • High costs for the destructive testing.
  • High costs incurred solving problems associated with limitations of sampling and destructive testing.

The high cost of destructive testing and examination, negative quality factors and high costs incurred resolving hardening errors provided motivation to improve the effectiveness of the hardness-testing program by implementing a nondestructive-testing program supported by only limited destructive testing.

A nondestructive-testing system was implemented utilizing the eddy-current test method in accordance with the “Preventive Multi-Frequency Testing” (PMFT) protocol. Since installation of the first system, two additional systems have been installed, including a second for gears and one for input and output shafts.

Economic Considerations for the Nondestructive-Test System

Fig. 1. Chart comparing cost of full destructive testing versus
nondestructivetest supported by limited destructive testing

Cost justification was based upon reduced metallurgist labor, reduced number of destroyed parts and reduction in saw blades consumed (Fig. 1). Other significant factors that were only subjectively considered included saved consumables other than saw blades, costs associated with delayed knowledge of failed samples and cost of increased destructive testing of suspect batches.

Technical Basis of the Inspection System

Fig. 2. Monofrequency method of testing

The technical basis of the system is an ultra-fast, multiple-frequency eddy-current testing method developed by ibg NDT Systems Corporation called PMFT.

The method provides the following prerequisite solutions for the system:

  • Reliable testing due to the PMFT method
  • High sensitivity to detect parts with material properties just beyond allowable tolerances for hardness and case depth
  • Repeatable test results over time (measured values of calibration parts – voltage vector values – are valid over long time periods)
  • Easy test setup and use after only short training of factory personnel
  • Easy test to maintain
  • Fast testing – seconds per part
  • Documentation of inspected parts and of calibration parts
  • Easy to increase testing to 100% of parts when required

What is Preventive Multi-Frequency Testing?

Fig. 3. Preventive multifrequency method testing

Fig. 4. Diagram representing the need to satisfy all tolerance fields (parameters)

Modern eddy-current test instruments that operate according to the PMFT method operate in a completely different manner than previous eddy-current instruments (Figs. 2 & 3). Based on well-established experience that different defects cause different signals in eddy-current instruments, a large number of test frequencies are now used. Only a statistically significant number of OK parts are required to calibrate the test instrument.

Apart from use of a large enough number of frequencies, it is important that a broad range of frequencies be used. The ratio between lowest to highest should be 1:1000 or higher to guarantee reliable testing.

Use of modern electronic components permits test times for frequencies to be considerably reduced. From the standpoint of time, it does not matter whether one, two or eight frequencies are used. All required frequencies can now be used for testing. Thus, all the information contained in the material is available to be read by the eddy currents. So it is now possible to really test preventively to reliably find all the possible hardening errors in the material.

Another advantage is the multidimensional evaluation of modern test systems. A separate tolerance field is created for every test frequency (Fig. 4). The fields account for all the information available from the voltage vectors (amplitude and phase) produced by the transducer for every test frequency. Only when all tolerance fields are satisfied can one assume that the part is OK. If a part is not OK in only one tolerance field, it is classified as not good. Any change in the low, middle or upper frequency range is displayed clearly.

With modern test protocols like PMFT, it has become possible to use eddy-current testing on mass-produced heat-treated components to reliably detect all significant hardening errors caused by mistakes in the heat-treating process. IH

For more information : Bill Buschur - ibg NDT Systems Corporation, 20793 Farmington Rd., Farmington, MI 48336; tel: 248-478-9490; fax: 248-478-9491; e-mail: sales@ibgndt.com; web: www.ibgndt.com

Additional related information may be found by searching for these (and other) key words/terms via BNP Media SEARCH at www.industrialheating.com: eddy current, gear testing, nondestructive testing

Realization of the Systems

Fig. 5. Test system 1 for testing of gears

Comments on Test Systems I and II

  • Instrumentation and a PLC interface permit automatic testing of up to 16 different parts utilizing 16 test channels.
  • There are two systems that utilize 11 of the channels, one each for 11 different gears.
  • A third system utilizes four of the channels to test two induction-hardened shafts (input and output shafts) at four locations on each shaft (Fig. 6).
  • Each different part or location has its dedicated test channel.
  • 15-inch color touch screen with several graphic display options for interpretation
  • On a bench-type assembly, each of 11 channels are realized by a set of test and compensation coils. Each coil set is dedicated to one of the 11 gears.
  • The instrumentation, PLC, control elements and printer are all mounted to the benchtop assembly (Fig. 5).

Test procedure is very simple and is organized as follows:

Fig. 6. Locations for testing of the transmission gear assembly for test system 2
  • A button is associated with each channel and its set of test coils.
  • A gear to be tested is placed in its assigned test coil positioned precisely by a plastic sleeve.
  • The button is pressed once to activate the test channel and to call up its respective stored calibration for this gear. The calibration was previously established by taking and storing measurements on 15-25 correctly hardened gears.
  • The button is pressed a second time and the gear is tested for correct hardening. A green light is illuminated on an OK test result and a red light for a not OK test result. A test report is automatically printed, and the test result (an x,y voltage vector value or its graphical representation) is stored.
  • Parts that test not OK may be evaluated by destructive testing to accurately determine the cause for their failure. These are few in number, and costs associated with cutting them are minimal.
  • Parts that test only marginally not OK may be added into the existing calibration based upon the results of further evaluation on them by the met lab.
  • Test results, calibrations and instrument setups can be stored via external PC for backup.


Three installations are now realized at this plant. The test systems were recently utilized to significantly shorten the set-up time of new high-throughput vacuum-carburizing and hardening furnaces. These furnaces release new batches of parts every nine minutes, creating a potentially enormous and unmanageable destructive-testing workload. Use of the nondestructive-test systems during setup significantly reduced the related destructive-testing requirements to manageable levels. Additionally, the PMFT test results are now used in the plant's statistical process-control procedures.

The basis of the decision by this manufacturer to follow this concept was its sensitivity and reliability; its proof of stability and repeatability under long-term evaluation prior to implementation; the simplicity of use and maintenance; and the significant cost savings realized by its implementation.

Bill Buschur
ibg NDT Systems Corporation, Farmington, Michigan.