FEATURE ADVANCED MATERIALS & PROCESSES | NOVEMBER/DECEMBER 2025 45 Furnace temperature was set to 925°C, with a partial pressure of 1.5 mbar acetylene atmosphere during the boost cycles. For process design, parameters like furnace control resolution (time steps), temperature, gas pressure, and the target case/carbon values are used as inputs. A monitor point just below the surface is also defined to control the upper and lower carbon bounds reached at that point. VCarb then automatically calculates boost step time until the monitor point reaches the desired value and then switches to a diffusion step until it drops to the lower bound. This cycle repeats until the target depth and final surface carbon are achieved. After the parameters are entered, VCarb calculates the full LPC schedule. For the example gear flank and input parameters, the resulting process required about 28 boost/diffuse steps (14 pairs) totaling just over 6 hours of furnace time: approximately 1000 seconds of boosting and 21,000 seconds of diffusing, yielding an R-value (diffuse/boost ratio) of about 21. A plot of predicted carbon concentration shows the targets were met by the analysis. Around 0.35 wt% carbon was achieved at the 1 mm case depth, which approached 0.8 wt% at the surface, corresponding to roughly 62 HRC if fully martensitic, Fig. 2. EXTENDING THE RECIPE TO GEAR ROOT FILLET With the flank recipe developed, attention turned to the root of the gear tooth, which has a concave inner radius of about 0.76 mm. Because diffusion behavior differs on concave surfaces, applying the flank recipe directly will yield a different case depth or surface carbon. Using VCarb’s prediction module, the previously developed recipe file is imported and executed. All material and process parameters remain the same except for the inner radius geometry input. The model then predicts how the root will carburize under the same boost/ diffuse steps. Figure 3 shows the results of the analysis with a surface carbon in the root fillet predicted to be just under 0.7 wt% and a case depth of about 0.75 mm for 0.35 wt% carbon. FURTHER CONSIDERATIONS If results show excessive carbon at the surface or deeper than desired case depth, engineers can adjust the partial pressures to modulate carbon uptake, temperature to accelerate or slow diffusion, or delete or add boost/ diffuse steps to achieve the desired cycle. Each step time, temperature, partial pressure, and gas flow are adjustable, allowing engineers to tweak parameters to optimize the recipe before any actual furnace time. For example, if cycle time is a concern, raising the carburizing temperature to 950°C with the same recipe increases carbon diffusion. In this case, it was predicted to cut cycle time to about 4.5 hours. By iterating these changes virtually, engineers can converge on an optimized schedule for both flank and root while balancing processing time and reducing the risk of forming harmful carbides. Another valuable modeling feature is the ability to predict carbide precipitation and dissolution during each step of the process. For AISI 5120, simulations showed negligible carbide formation under the case study conditions, but higher-alloy steels like 9310 displayed carbides up to 39 µm deep with the developed recipe. This case study illustrates how modeling transforms LPC process design from trial-and-error into a predictable, data-driven process. Traditional recipe development might require multiple furnace cycles to establish a set of boost/diffuse pairs. With a modeling tool, cycle times can be estimated, different materials or geometries can be compared, and carbide formation and hardness profile can be predicted. The result is a shorter path to production-ready recipes, improved quality, and lower cost. 11 Fig. 2 — Plot of carbon profile vs. depth from the spur gear flank surface. Fig. 3 — Plot of carbon profile vs. depth from the spur gear root fillet surface.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==