Advanced Carburizing Technology: An Industrial Perspective
Oerlikon Fairfield has been a technology leader and product innovator in gear and drive design for almost 100 years. Its people, knowledge and resources help to provide unique solutions for a wide range of application needs.
With manufacturing operations in the U.S., India and China, Oerlikon Fairfield has the capability to produce up to AGMA class-14 custom gears with spur, helical or bevel forms from 20 mm to 2 meters in diameter. The company also designs and builds custom drives for mobile equipment and stationary industrial machinery with torque outputs from 800 Nm to well over 4,000,000 Nm. Other products designed to provide integrated solutions for mechanical, hydraulic or electrically driven systems require Torque Hub® planetary drives, drop boxes, right-angle drives, transfer cases, specialty transmissions, differentials and differential carrier assemblies, housings and custom drive assemblies.
Need for Gas Carburizing
To meet both customer product performance requirements and service-life expectations, parts require a hard, wear-resistant surface; a soft, ductile core; and the ability to withstand not only the high Hertzian stresses present along the active flank of gear teeth but also significant bending moments in the root.
Gas carburizing is an ideal technology and a cost-effective solution to these challenges. As such, Oerlikon Fairfield operates a large heat-treat shop incorporating numerous batch and continuous furnaces running endothermic gas and operating 24/7.
Increased Productivity Investigation
Oerlikon Fairfield decided to investigate a process technology being offered by Heavy Carbon Co., LLC (Pittsford, Mich.) that manufactures an Endocarb system “Endocarb,” which claims to dramatically reduce cycle time. This is accomplished by alternately increasing and then decreasing the carbon potential in the furnace, thereby producing parts more quickly. Endocarb differs from a conventional gas-carburizing system in several important ways:
- Endocarb is the source of process gas and control. It is mounted directly on the furnace, simplifying gas piping and avoiding transmission issues. The close proximity also minimizes the risk of leaks and decreases the amount of upkeep necessary to maintain the system. The system can be easily set up so that the furnace runs at reduced gas flowrates.
- At 925°C (1700°F), the process can be set to run at a carbon potential of 1.5%, which is significantly higher than conventional values often in the range of 1.05-1.10% but certainly no higher than 1.2%.
- The increased carbon potential decreases the run time for the load. The carburizing time to achieve a 2.3-mm effective case was decreased by nearly 18%.
While conventional carburizing maintains a constant carbon potential during the boost phase, Endocarb uses “carbon cycling” (i.e., the process starts at a much higher potential and is then rapidly lowered well below the initial setpoint). This cycling continues throughout the run. In addition, at certain points, air is pumped into the furnace and a “controlled lean-out” performed. As a result, atmosphere is rejuvenated and the furnace itself remains clean and soot-free. A typical Heavy Carbon cycle can be seen versus a conventional gas carburizing cycle in Figure 1.
With a lower cycle time at the same process temperature, Endocarb would have a relatively quick payback when compared to conventional carburizing based on reduced cycle time and improved furnace cleanliness. In order to assess potential cost savings with equal or better quality, Oerlikon Fairfield decided to conduct an investigation to determine whether the time savings claimed could be achieved without sacrificing metallurgical quality or the performance characteristics of gears.
Two identical gear sets were chosen for the trials along with five different material test bars (8620, 4320, 8822, 4820 and 4817) and eight different segments of various part numbers with varying diametrical pitches. The gear used was a 230-kg bull gear (Fig. 2).
Metallurgical results were compared to both internal and customer specifications and included checks of surface and core hardness (flank and root), effective case depth (tip, flank, root), case and core microstructure, carbide morphology, surface carbon content, retained-austenite percentages, NMTP percentages, decarburization, and IGO/IGA.
For testing purposes, the austenitizing and carburizing temperatures were held constant for both trials. The parts were furnace-cooled after carburizing from both processes and then reheated and quenched in the same furnace to eliminate any difference related to the quenching process.
Carbon content was analyzed for each of the samples sent to the commercial heat treater whose furnace was equipped with an Endocarb unit installed on it, namely Euclid Heat Treating (Euclid, Ohio), as well as those run at Oerlikon Fairfield. For clarification, the cycle run at Oerlikon Fairfield will be referred to as “Conventional” and the cycle run at Euclid Heat Treating will be referred to as “Heavy Carbon.” The carbon percentage at various depths was taken, and charts of identical samples sent to each testing facility were compared. Two of the results are reported in Fig. 3.
Figure 3 illustrates that there is no significant difference in carbon weight percentage at varying depths on similar samples. These results suggest that there is no carbon composition difference in the products produced by both the conventional and Endocarb carburization methods. This suggests that the surface carbon, case depth at 0.40% carbon and other carbon-potential characteristics used to determine the mechanical properties of the produced part are virtually identical between the two sample sets.
Effective case depth (50 HRC) results as well as surface-hardness measurements of the parts created by both carburization techniques were also compared (Tables 1 and 2 respectively). These show the measured effective case-depth (ECD) difference and percent difference between the two measurements. Samples from both techniques were tested by Oerlikon Fairfield, Euclid Heat Treating and independently by The HERRING GROUP, Inc. (Elmhurst, Ill.).
The data indicates that, once again, there are few differences between all four measurements conducted for each tooth of the bull gear. The main experimental difference (over 10%, or 3 HRC) comes from the ECD measurement of bull-gear tooth #2 (or Sample B as referred to by Oerlikon Fairfield). The third-party measurements, however, were very similar for the ECD measurements of this tooth in both instances. Overall, there is only a slight difference between measurements, indicating again that there is only a slight variance of the essential characteristics within the parts produced by the two carburization techniques.
Although the balance of all measurements (not reported here) supported this conclusion, microstructural analysis was conducted to characterize the microstructure (Figs. 4-6) before and after etching.
The microstructure is consistent between samples (the color difference in Figure 6 is due to an etching effect and differences in lighting). These microstructures confirm that there is virtually no difference between the parts produced via both systems.
In order to justify a change in the carburizing process, a business case must be built using an SEQCDM (safety, environmental, quality, cost, deliverables and morale) model.
- Safety - The project is expected to meet safety goals by making the gas quality easier to monitor while adhering to the same quality standards as the previous method. More study is needed to ensure safety compliance, though Euclid Heat Treating has been utilizing this method for years.
- Environmental - The direct injection of endothermic gas would reduce the potential leakage coming from the piping that currently requires continual maintenance and monitoring. Endocarb also requires less gas use than the conventional carburizing method, hence the overall air quality throughout the facility would improve.
- Quality - The parts produced by Endocarb would maintain the same quality standards as those produced through traditional gas carburization.
- Cost - Endocarb would decrease the typical run time for gas carburization of the specified part analyzed here by 18%. As a result, payback will be relatively short given the ability to carburize more parts in less time. The maintenance savings are also significant.
- Deliverables - The product produced with Endocarb meets or exceeds customer expectations in less time than more-traditional gas carburizing. Case-hardness and effective case depth are virtually equivalent between the two methods.
- Morale - Endocarb is directly attached to the furnace, allowing for an easier path with less complications for the technicians to deal with. Morale would improve over time due to the simplification of the process.
Thus, from both an economic and quality standpoint, Endocarb satisfies the SEQCDM model by producing the same overall quality of parts more quickly and efficiently.
The Endocarb system was determined to have achieved the same results with the part studies in these trials as those produced from conventional endothermic gas carburizing. The principal conclusions reached were:
- Carbon potential - The higher and lower carbon-potential setpoint throughout the process produced less soot accumulation within the furnace, resulting in less “housekeeping” for the pyro technicians. Sooting from conventional processing is known to cause significant equipment problems and process variation.
- Temperature - The temperature fluctuations necessary to control the varying carbon setpoints were not shown to adversely affect the quality of the heat-treated products.
- Time - The cycle time for gas carburizing was found to decrease by approximately 18%, allowing higher productivity and associated cost savings.
- Microstructure, case depth, surface hardness, retained austenite - The analyses of both the conventional and Heavy Carbon treated parts produced similar microstructures. No significant differences in mechanical properties are expected from identical parts produced by both processes.
Given that Endocarb appears to produce identical quality in less time at lower cost, the heat-treatment department will operate more efficiently, shorten product lead times and still meet or exceed customer expectation.
The authors would like to thank the following individuals and companies for their contributions to this study: Mr. George Barbour (Heavy Carbon Co., LLC), Mr. Jon Vanas (Euclid Heat Treating), Mr. Marc Abney, (retired, Oerlikon Fairfield), Mr. Ryan Wilmes (American Axle and Manufacturing), and the Oerlikon Fairfield MetLab Technicians (Oerlikon Fairfield).
For more information: Contact Aaron Flesher, Chief Metallurgist, Oerlikon Fairfield Manufacturing 2400 Sagamore Pkwy S, Lafayette, IN 47903; tel: 765-722-4326; e-mail: Aaron.Flesher@oerlikon.com; web: www.oerlikon.com/fairfield
- Heavy Carbon Company (www.heavycarbon.com/overview.html)
- Atmosphere Engineering Company (www.atmoseng.com/documents/technical/GEN-ENDO-BASICS.pdf)
- Euclid Heat Treating (www.euclidheattreating.com/index.html)
- Independent third-party analysis, The HERRING GROUP, Inc. (www.heat-treat-doctor.com)
- Herring, Daniel H., “Method for Accelerating Atmosphere Carburizing,” Industrial Heating, December 2017