A converter is one of the main pieces of equipment used in making steel. The bearings in the converter trunnion are characterized by low speeds, heavy loads and large bearing dimensions and revolve slowly when molten steel is added or emptied. Their life expectancy is 20 years. Their failure reduces the output of the converter, incurring losses in the entire product line. Therefore, it is important for the bearings to operate smoothly.

Vibration and oil analysis are two techniques used to monitor the condition of rotary machines. But normal vibration analysis cannot be performed on converter trunnions because of the limitation of their sensors in collecting signals for the rotary frequencies of the bearings; the signals are too low.

Analytical and emission spectroscopy produce results regardless of the operating characteristics of the machine at the time of sample collection. Therefore, spectroscopy and ferrography are ideal tools for analysis of this type of machinery.

This article examines how oil analysis was used to monitor the bearing wear on the drive side of the No. 5 converter during and after the run-in period, using spectroscopy and ferrography. The run-in period was observed to be 19 months because of the relatively short accumulated working time of the bearing.

Bearing Parameters
Model 240/1320 CAME4C4, the drive-side bearing of the No. 5 converter, is the equipment used to conduct this experiment. The bearing’s outer and inner ring and rollers are constructed of alloy steel including iron, chromium and manganese. The holder is made of high-intensity brass. The outer diameter of the bearing is 1,800 mm and the inner clearance (windage) is 1.2 to 1.6 millimeters. It is lubricated with Cosmo 320 lubricating oil and the volume of the container is approximately 30 liters.

Elemental spectroscopy and ferrography are used to monitor the wear conditions of the bearings. Such techniques utilize the following pieces of equipment:

  • Baird multielement oil analyzer (MOA) (now Antares Analytical)
  • direct-reading (DR) ferrograph and rotary particle depositor (RPD) made by Predict DLI
  • dual ferrograph analyzer made by Standard Oil Company of Ohio (Sohio), USA
  • BX-60 optical microscope made by Olympus Company of Japan. The microscope studies the morphologies of the wear particles, using both white reflected light and green transmitted light.

Sampling Condition
Shortly after the No. 5 converter was commissioned in May 1998, an abnormal noise was identified on the bearing’s drive side. Oil analysis was used to monitor the wear on this bearing. Lubricating oil was changed approximately once per month, the length of time required to produce 400 “melts” or batches of molten steel. Samples were taken before the oil was changed.

A vacuum pump was used to obtain a representative sample. During the two-year monitoring period, lubricating oils from bearings in the other three converters (Nos. 1, 2 and 3) were sampled to compare their wear rates. These converters had been running for more than 10 years with no evidence of abnormal wear.

Spectrometric and DR Ferrograph Analysis
Iron exists in the bearing’s outer and inner shell and journal, while copper exists only in the bearing’s inlay. The location of the worn part and its wear severity can therefore be estimated using iron and copper concentrations. Spectrometric analysis and DR ferrographic analysis data are shown in Figures 1 and 2. The dashed line in Figures 1 and 2 represents the multinomial concentration of copper and whole particle concentration (WPC). The inlay is made of brass with a lower hardness than that of the chromium alloy steel roller. The surface roughness of the inlay is also lower than that of the roller as well, and as such, is prone to the most rapid rate of wear. Bearing wear can be tracked by measure of the steel and copper content in the lubricant.

Figure 1. Spectrometric Analytical Results


Figure 2. DR Ferrograph Analytical Results

As shown in Figure 1, the copper concentration increased sharply from May 27, 1998 to October 26, 1998, then it declined gradually, becoming stable in December 1999. This illustrated that bearing wear accelerated, then evened out, which is consistent with a wear-in pattern for newly commissioned equipment. This trend could be observed more clearly from the trend curve of the copper concentration. Furthermore, the iron concentration trended the same as copper. In Figure 2, the WPC, direct read large (DL) and direct read small (DS) trended similar to iron and copper in Figure 1.

Prior to December 1999, the iron and copper concentrations in Converter No. 5 were considerably higher than concentrations for Nos. 1 through 3 converters.

However, they fell to the same level after December 1999, suggesting that the No. 5 converter became more stable after December 1999. Direct read (DR) ferrographic analysis showed the same trend.

Wear Particle Analysis
The copper alloy appeared yellow or reddish brown, while most other free metals appeared silver white. From May 1998 to December 1999, large amounts of large ferrous particles and copper alloy particles were found in the lubricating oil. The copper alloy and steel particles diminished in size and number until December 1999. Most large particles obtained from the oil sample before December 1999 had a long, thin, straight-edged shape and a striated surface. Some showed characteristics of severe sliding wear (Figure 3), typical of wear particles produced during run-in. After December 1999, most of the particles had smooth surfaces and jagged edges, and the average size of copper alloy particles was approximately 20 microns (Figure 4). This suggested different wear stages.

Figure 3. Copper Alloy
Particle During Run-in
Figure 4. Copper Alloy
Particle After Run-in

Compared with the bearings in Nos. 1 through 3 converters, the number of large wear particles including iron and copper was less than those in the bearing of No. 5 converter before December 1999. Most copper particles measured approximately 20 microns, and the average iron particle in the oil samples of Nos. 1 through 3 converter was 30 microns, similar to those in No. 5 converter after December 1999.

Therefore, the conclusion can be drawn that bearing wear on the drive side of No. 5 converter accelerated first, then evened out, and December 1999 can be regarded as the end of the run-in period.

Why did the bearing in No. 5 converter experience such a long run-in? This could be explained by its working condition. Considering that the bearing revolved only a few cycles for every “melt” produced, then its accumulated working time was short, though the operating time of the converter was rather long. It was the relatively short working time of the bearing that led to a run-in lasting 19 months.


  1. The drive-side bearing on the No. 5 converter experienced a 19-month run-in period, then the wear rate became stable. This was attributed to the relatively short accumulated working time of the bearing.
  2. It was feasible to use atomic emissive spectrum and ferrography to monitor the working condition of the converter trunnion bearing. Its wear rate was the same as general machine wear.
  3. Oil analysis has its advantages over other techniques in the condition monitoring of low-speed and heavily loaded mechanical equipment.