FEATURE ADVANCED MATERIALS & PROCESSES | NOVEMBER/DECEMBER 2025 44 T o provide stronger, more durable, and more precisely engineered components, heat treatment remains an indispensable step in manufacturing. One of the more recent transformative processes in this domain is low-pressure carburization (LPC). Unlike traditional gas carburization, LPC takes place in a near-vacuum environment where carbon is delivered to the steel surface in a series of carefully timed “boost” and “diffuse” steps, shown in Fig. 1. This high-efficiency approach avoids intergranular oxidation and also offers manufacturers tighter control over case depth, carbon profiles, and cycle time. However, the high reactivity, rapid saturation, and steep gradients also make it challenging to design recipes. Small deviations in timing or pressure can lead to excess carbide formation, or inconsistent hardness. For gear makers, aerospace suppliers, bearing manufacturers, and other high-precision industries, the ability to model and predict LPC behavior before committing to furnace time is now critical for reducing costly mistakes. This article first outlines the fundamentals of LPC, then explores a case study using a software tool (DANTE VCarb) to design and optimize an LPC recipe for gear flank and root geometries. It concludes with how modeling can accelerate process development and reduce costs showing why LPC is fast, reliable, and increasingly indispensable. In LPC, a carbon-rich gas, often acetylene, is pulsed into a sealed furnace at low pressure. Under these condiADVANCING LOW-PRESSURE CARBURIZING THROUGH SIMULATION This case study demonstrates how using software to design and optimize a low-pressure carburization recipe saves time and material and increases confidence in final part performance. Jason Meyer* DANTE Solutions Inc., Cleveland tions, the steel surface quickly reaches a carbon saturation limit, after which carbides can form. Unlike gas carburization, where carbon potential is governed by a continuous atmosphere, LPC saturates the surface rapidly and requires alternating boost and diffuse steps to achieve the desired profile (shown in Fig. 1). Boost steps introduce carbon to enrich the steel surface, and diffuse steps pause the carbon supply, allowing carbon to diffuse deeper into the part and dissolve any incipient carbides. Balancing the length and number of these steps is essential. Boost steps that are too long or diffuse steps that are too short can leave undissolved carbides, undermining fatigue strength and mechanical properties. Conversely, boost cycles that are too short can leave a shallow, under- carburized case and lead to longer processing times. Because LPC is nonlinear and sensitive to material, geometry, and furnace characteristics, it is difficult to design by feel. Software tools like VCarb aim to remove the guesswork by simulating the physics of LPC, which is based on Fick’s laws of mass diffusion. They allow engineers to specify alloy composition, geometry, desired case depth, and furnace conditions to generate a complete boost/ diffuse schedule along with predicting carbon, hardness, and carbide profiles. CASE STUDY: DESIGNING AN LPC RECIPE FOR A SPUR GEAR GEOMETRY This case study focuses on a spur gear made of AISI 5120, a low carbon alloy steel used for a wide range of applications in the automotive industry due to its versatility. In LPC, convex (outer) surfaces and concave (inner) surfaces affect the carbon uptake differently due to local diffusion dynamics. Concave surfaces like the inner radius of the root fillet will often have a shallower case depth and lower carbon than convex surfaces like the flank of the tooth. The selected gear geometry is a spur gear with an outer flank radius of about 6.2 mm and an inner radius of about 0.76 mm. The goal is an effective case depth of 1 mm with 0.35 wt% carbon (~50 HRC) at that depth and a final surface carbon of ~0.8 wt%, producing about 50-60 HRC after quenching. *Member of ASM International Fig. 1 — Schematic representation of idealized LPC boost/diffuse steps. 10
RkJQdWJsaXNoZXIy MTYyMzk3NQ==