I approach the numerical simulation of forged gear blanks as a manufacturing problem that links geometry, material flow, thermal history, tooling contact, and final mechanical performance. Gears are among the most important transmission elements in lifting equipment, marine machinery, rolling mills, mining equipment, and many other engineering systems. In my work, I treat precision forging not merely as a shaping operation but as a route for producing forged gear blanks with favorable grain flow, high material utilization, short process chains, and improved load-bearing capacity. I use an axisymmetric finite-element model in MARC.SuperForm to examine how two candidate billets behave during forging, and I compare the resulting metal flow, temperature distribution, die filling, and deformation uniformity. The central question I ask is simple: which billet geometry produces forged gear blanks with the most uniform deformation and the most continuous flow lines while remaining compatible with the die cavity and the press schedule?
My starting point is the conventional process chain for gears, which typically includes billet preparation, heat treatment, blank machining, tooth shaping, end machining, surface heat treatment, datum correction, and finishing. Precision forging of forged gear blanks offers an alternative logic. Instead of removing material from a solid blank, I form the blank close to the required geometry, improve the continuity of the metal flow lines, and reduce the amount of subsequent machining. In my assessment, material utilization above 70 percent, gear strength improvement above 20 percent, and productivity improvement of roughly 40 percent are realistic targets when the forging process is designed around the correct billet and die sequence. These targets guide my comparison of forged gear blanks rather than serving as decorative claims.
I define the forged gear blank as a disk-like axisymmetric part with a central hub region, a web region, and an outer rim region. The material is C45 steel, corresponding to a medium-carbon steel with good forgeability and adequate hardenability for many gear applications. Because the part is axisymmetric, I set the symmetry axis along the X direction and use a two-dimensional axisymmetric formulation. This choice reduces the computational cost while preserving the essential deformation and thermal fields of the forged gear blanks. I model the die as a rigid tool and the workpiece as a deformable body. The moving die starts at an initial position of 200,0 and terminates at 31.5,0, so the total prescribed displacement is 168.5 mm. I set the initial workpiece temperature to 1250 degrees Celsius, the friction coefficient to 0.2, and the remeshing option to OVERLAY QUAD with a maximum element edge length of 0.3 mm. The remeshing body is the workpiece, the deformation force option is HYDRAULIC PRESS STAGE, and the termination criterion is based on complete node contact with the workpiece.
Simulation framework. I use MARC.SuperForm because it is a volume-forming finite-element environment that combines robust remeshing with visualization of temperature, strain, stress, and flow patterns. In my workflow, the software modules are not isolated steps. I move between geometry preparation, material definition, process-parameter input, finite-element solution, and post-processing until the predicted forged gear blanks satisfy the intended process window. The first module allows me to draw the billet and die geometry with sufficient accuracy. The second module defines the material response, including flow stress as a function of strain, strain rate, and temperature. The third module specifies the initial conditions of the billet and die, the motion of the moving die, the friction law, and the press characteristics. The fourth module performs the finite-element calculation. The fifth module converts the results into color maps and line plots so that I can judge whether the forged gear blanks are likely to fill the cavity without laps, folds, or excessive temperature gradients.
| Simulation module | Role in my study | Output I examine for forged gear blanks |
|---|---|---|
| Geometry preparation | Defines axisymmetric billet and die contours | Initial contact area, cavity volume, symmetry |
| Material definition | Assigns C45 flow behavior and thermal properties | Flow stress, strain hardening, temperature sensitivity |
| Deformation parameters | Sets die motion, friction, temperature, press stage | Stroke history, contact sequence, load response |
| Finite-element analysis | Solves equilibrium, incompressibility, and heat transfer | Effective strain, stress, temperature, damage indicators |
| Visualization | Maps results onto the deformed mesh | Flow-line continuity, fill quality, thermal gradients |
In my model, the equilibrium of the deformable workpiece is expressed through the principle of virtual work. For a volume V with prescribed traction on the boundary, I write
$$
\int_V \sigma : \delta \varepsilon \, dV
–
\int_{S_t} t \cdot \delta u \, dS
–
\int_V \rho b \cdot \delta u \, dV
= 0
$$
where sigma is the Cauchy stress tensor, delta epsilon is the virtual strain-rate tensor, t is the surface traction, b is the body force per unit mass, and rho is the density. For the large deformation of forged gear blanks, I use an updated Lagrangian description and remesh whenever the local element distortion exceeds the acceptable limit.
I describe plastic flow using the Levy-Mises relation and the associated flow rule. The deviatoric stress is
$$
s_{ij} = \sigma_{ij} – p\delta_{ij}
$$
and the plastic strain increment is
$$
d\varepsilon_{ij}^{p} = d\lambda \, s_{ij}
$$
where d lambda is a positive scalar determined by the consistency condition. The effective strain, which I use to compare the deformation severity in different regions of forged gear blanks, is
$$
\bar{\varepsilon}
=
\sqrt{
\frac{2}{3}
\left(
\varepsilon_{11}^{2}
+
\varepsilon_{22}^{2}
+
\varepsilon_{33}^{2}
+
2\varepsilon_{12}^{2}
+
2\varepsilon_{23}^{2}
+
2\varepsilon_{31}^{2}
\right)
}
$$
The effective strain rate follows the same definition with strain-rate components. I use these quantities to identify whether the center of the forged gear blanks deforms early, whether the rim fills before the hub, and whether any region remains effectively undeformed until the final stroke.
For the C45 steel workpiece, I represent the flow stress as a function of strain, strain rate, and temperature. A general form that I use for interpretation is
$$
\sigma_f
=
K \bar{\varepsilon}^{n}
\dot{\bar{\varepsilon}}^{m}
\exp\left(\frac{Q}{RT}\right)
$$
where K is a material constant, n is the strain-hardening exponent, m is the strain-rate sensitivity, Q is the activation energy, R is the universal gas constant, and T is the absolute temperature. In the high-temperature regime relevant to forged gear blanks, I also consider the Zener-Hollomon parameter
$$
Z
=
\dot{\bar{\varepsilon}}
\exp\left(\frac{Q}{RT}\right)
=
A\left[\sinh\left(\alpha\sigma_f\right)\right]^{n}
$$
This relationship reminds me that the resistance of forged gear blanks to deformation is not a fixed number. It changes as the billet cools at the die contact, as friction heating raises the local temperature, and as the strain rate increases near the final closing stroke.
Friction is critical because it controls the contact shear stress and the temperature rise near the die surface. I use a constant shear friction model,
$$
\tau_f
=
m k
=
m\frac{\sigma_f}{\sqrt{3}}
$$
where m is the friction factor and k is the shear yield stress. In the present study, the friction coefficient input is 0.2. I interpret this value as a moderate friction condition for hot forging of forged gear blanks with lubricated dies. The friction stress affects the contact layer, the local temperature, and the load required by the hydraulic press stage.
The thermal field in forged gear blanks is governed by conduction, plastic dissipation, and friction heating. I write the energy equation as
$$
\rho c_p
\frac{\partial T}{\partial t}
=
\nabla \cdot \left(k_T \nabla T\right)
+
\dot{q}_{plast}
+
\dot{q}_{fric}
$$
where rho is density, c_p is specific heat, k_T is thermal conductivity, and the source terms represent plastic work and friction work. The plastic dissipation term is
$$
\dot{q}_{plast}
=
\eta
\sigma_f
\dot{\bar{\varepsilon}}
$$
where eta is the fraction of plastic work converted into heat. The friction heating term is
$$
\dot{q}_{fric}
=
\tau_f
\left\|v_t\right\|
$$
where v_t is the relative sliding velocity at the contact interface. I use these relationships to explain why the contact zones of forged gear blanks often show the highest temperature and why the temperature gradient can become asymmetric when the initial billet diameter exceeds the die entrance diameter.
Billet geometry and volume control. The die entrance diameter is 85 mm. I compare two round billets: one with diameter 85 mm and length 130 mm, and one with diameter 100 mm and length 94 mm. The first billet matches the die entrance diameter, while the second billet exceeds it. This geometric difference is not minor. It changes the first contact location, the direction of metal flow, the amount of barrelling, and the time at which the center of the forged gear blanks begins to deform.
For a cylindrical billet, the initial volume is
$$
V_0
=
\frac{\pi d_0^2}{4} h_0
$$
I calculate the two candidate volumes and compare them in the table below. The volumes are nearly identical, so the comparison isolates the effect of billet aspect ratio and initial diameter rather than the effect of a large volume difference.
| Candidate | Diameter d0 (mm) | Height h0 (mm) | Volume V0 (mm3) | Volume difference from candidate 1 | Geometric relation to die entrance |
|---|---|---|---|---|---|
| Candidate 1 | 85 | 130 | 737,685 | 0.00 percent | Diameter equals die entrance |
| Candidate 2 | 100 | 94 | 738,274 | +0.08 percent | Diameter exceeds die entrance |
The volume difference is only about 0.08 percent, which is small enough that I can attribute the different simulation behavior to the billet geometry and contact sequence. In my view, this is an important point for forged gear blanks: two billets with almost the same mass can produce very different flow-line patterns if their aspect ratios and initial contact conditions differ.
Modeling assumptions and mesh strategy. I use a two-dimensional axisymmetric model with the X-axis as the symmetry axis. The radial displacement on the symmetry axis is zero,
$$
u_r(0,z)=0
$$
and the radial heat flux on the symmetry axis is also zero,
$$
\left.
\frac{\partial T}{\partial r}
\right|_{r=0}
=
0
$$
These conditions reduce the model size while preserving the physics of the forged gear blanks. I divide the workpiece into quadrilateral elements. The remeshing rule uses OVERLAY QUAD with a maximum element edge length of 0.3 mm. I monitor the minimum Jacobian and remesh when
$$
\min J < 0.3
\quad \text{or}
\quad h_e > 0.3 \, \text{mm}
$$
where J is the element Jacobian and h_e is the element edge length. The objective is to avoid negative-volume elements and to maintain a mesh that can resolve the contact layer and the shear bands in forged gear blanks. Remeshing is applied to the workpiece, not to the rigid die, because the die geometry remains fixed in the simulation.
| Modeling item | Setting I use | Reason for the choice |
|---|---|---|
| Model type | Axisymmetric, X-axis symmetry | Matches the disk-like forged gear blanks and reduces computation |
| Workpiece material | C45 steel | Medium-carbon forging steel suitable for gear blanks |
| Initial workpiece temperature | 1250 degrees Celsius | Represents the start of hot forging |
| Friction coefficient | 0.2 | Moderate hot-forging friction condition |
| Moving die initial position | 200,0 | Defines the open-die starting state |
| Moving die final position | 31.5,0 | Defines the final closing stroke |
| Total stroke | 168.5 mm | Difference between initial and final positions |
| Remeshing type | OVERLAY QUAD | Maintains a structured quadrilateral mesh during large deformation |
| Maximum element edge | 0.3 mm | Resolves local flow and contact gradients |
| Press option | HYDRAULIC PRESS STAGE | Represents the forming equipment response |
| Termination criterion | 100 percent node contact with workpiece | Ends the simulation when the die cavity is closed on the workpiece |
I also define the contact condition in the usual complementarity form. The normal gap must be nonnegative, the normal traction must be compressive, and their product must vanish:
$$
g_n \ge 0,
\quad
t_n \le 0,
\quad
t_n g_n = 0
$$
The friction traction opposes the relative sliding velocity,
$$
t_t
=
–
m k
\frac{v_t}{\left\|v_t\right\|}
$$
These conditions are important for forged gear blanks because the contact area grows rapidly during the final stroke. A small change in the initial billet diameter can change the contact area history and, therefore, the load and temperature history.
Candidate 1: billet with diameter 85 mm and length 130 mm. I begin the simulation with the billet aligned with the die entrance. The moving die advances along the negative X direction. In the first stage, the two ends of the billet contact the fixed die and the moving die. Because the billet diameter matches the die entrance, the initial contact is distributed around the full circumference rather than concentrated on an overhanging shoulder. As the stroke increases, the metal flows radially into the available cavity space, and the center of the billet begins to deform. I observe that the contact points and the central region experience the most intense deformation at an early stage. This early center deformation is desirable for forged gear blanks because it helps close internal porosity and produces a more continuous flow-line pattern through the hub and web.
As the moving die continues to advance, the deformation spreads from the contact zones toward the mid-plane. The metal sliding against the die cavity generates friction heating, and the temperature in the contact layer rises. The temperature field remains broadly symmetric about the mid-plane because the upper and lower contact conditions are similar. I note that the flow lines remain continuous through the web region, and the material that will form the rim is drawn outward in a controlled manner rather than being folded over an initial overhang.
Near the later stage, the metal that has flowed into the upper cavity reaches the die top, and the filling front moves toward the gear-root region. The final stage is dominated by the closing of the rim and the establishment of full contact. In my simulation, the temperature distribution at the end of the stroke is left-right symmetric to a good approximation, and the deformation is more uniform than in the second candidate. I attribute this to the matched initial diameter, which allows the billet to enter the die cavity without an abrupt shoulder contact.
| Stage in candidate 1 | Step index | Step fraction | What I observe in the forged gear blanks | Process implication |
|---|---|---|---|---|
| Initial positioning | 1 | 0.000 | Billet ends contact both dies; axisymmetric alignment is preserved | Stable starting contact |
| Early compression | 34 | 0.363 | Contact zones and center deform; metal moves into cavity space | Good center consolidation |
| Intermediate flow | 54 | 0.582 | Radial flow increases; friction heating appears in contact layers | Controlled flow-line development |
| Advanced filling | 64 | 0.692 | Deformation range expands; center remains active | Uniform strain distribution |
| Rim filling | 84 | 0.912 | Metal reaches die top and begins filling gear-root region | Progressive cavity fill |
| Final closure | 92 | 1.000 | Temperature field is approximately symmetric; deformation is relatively uniform | Favorable final forged gear blanks |
I use the normalized stroke and step fraction to compare the two candidates. The step fraction is
$$
f_s
=
\frac{N-1}{N_{total}-1}
$$
and the normalized die position is
$$
s_n
=
\frac{x_0-x}{x_0-x_f}
$$
where x_0 is 200, x_f is 31.5, and x is the current die position. These normalized quantities let me compare events that occur at different total step counts. Candidate 1 requires 92 steps, while candidate 2 requires 107 steps. The larger step count in candidate 2 is a first indication that the forming sequence is more complex and that the solver needs more increments to resolve the contact and remeshing history.
Candidate 2: billet with diameter 100 mm and length 94 mm. I now replace the billet with the larger-diameter, shorter candidate. The initial volume is almost the same, but the initial diameter exceeds the die entrance diameter of 85 mm. This changes the first contact. The portion of the billet that overhangs the die entrance contacts the die shoulder first, and the contact is concentrated on the overhanging region. The center of the billet does not deform at the same early stage as in candidate 1. Instead, the outer shoulder region carries the initial deformation, and the metal flows toward the upper cavity while the central region remains relatively rigid.
As the stroke continues, the overhanging metal is forced inward and upward. The contact area between the billet and the die shoulder becomes a high-friction zone. The temperature in this zone rises because of friction and plastic dissipation, and the temperature gradient becomes steeper. I observe that the deformation is less penetrating at the center of the forged gear blanks. The flow lines near the shoulder are more distorted, and there is a greater risk of a lap or a fold if the inward flow meets the outward flow in an unfavorable sequence.
In the later stage, the central region finally begins to deform, but by then the outer region has already experienced a different strain path. The final temperature field is not as symmetric as in candidate 1. The contact layer near the moving die shows a higher temperature, and the gradient between the contact surface and the interior is larger. This is consistent with the longer step count of 107 and with the more localized deformation history.
| Stage in candidate 2 | Step index | Step fraction | What I observe in the forged gear blanks | Process implication |
|---|---|---|---|---|
| Initial positioning | 1 | 0.000 | Billet overhangs the die entrance; shoulder contact is imminent | Nonuniform initial contact |
| Shoulder contact | 25 | 0.226 | Overhanging metal contacts die; center remains largely undeformed | Delayed center deformation |
| Localized flow | 50 | 0.462 | Outer region flows upward; contact friction rises | Steep local temperature gradient |
| Constrained filling | 80 | 0.745 | Flow is constrained by the shoulder; distortion increases | Higher risk of flow-line discontinuity |
| Late center deformation | 93 | 0.868 | Center begins to participate after outer region has already deformed | Nonuniform strain history |
| Final closure | 107 | 1.000 | Temperature field is less symmetric; contact layer is hotter | Less favorable forged gear blanks |
Comparison of deformation uniformity. I compare the two candidates using several indicators: the time at which the center begins to deform, the symmetry of the temperature field, the continuity of the flow lines, the maximum effective strain gradient, the predicted load trend, and the fill sequence. The first candidate produces an earlier center deformation, a more symmetric temperature field, and a more continuous flow-line pattern. The second candidate produces a localized shoulder deformation, a delayed center response, and a steeper temperature gradient. For forged gear blanks, these differences are meaningful because they affect grain flow, internal soundness, and the probability of forging defects.
I define a simple uniformity index based on the standard deviation of effective strain over the cross-section:
$$
U_{\varepsilon}
=
1
–
\frac{
\sqrt{
\frac{1}{N}
\sum_{i=1}^{N}
\left(
\bar{\varepsilon}_i
–
\bar{\varepsilon}_{avg}
\right)^2
}
}{
\bar{\varepsilon}_{avg}
}
$$
where N is the number of sample points and epsilon_avg is the average effective strain. A value closer to 1 indicates a more uniform distribution. In my comparative assessment, candidate 1 gives a higher uniformity index than candidate 2. I also define a thermal symmetry index,
$$
U_T
=
1
–
\frac{
\sqrt{
\frac{1}{N}
\sum_{i=1}^{N}
\left(
T_i
–
T_{mirror,i}
\right)^2
}
}{
T_{avg}
}
$$
where T_mirror is the temperature at the mirror point about the mid-plane. Candidate 1 has a higher thermal symmetry index because the upper and lower contact histories are more balanced. Candidate 2 has a lower thermal symmetry index because the shoulder contact and the overhang create a more asymmetric flow and friction history.
| Comparison indicator | Candidate 1: 85 mm by 130 mm | Candidate 2: 100 mm by 94 mm | My interpretation for forged gear blanks |
|---|---|---|---|
| Initial contact | Full-circumference end contact | Shoulder contact from overhang | Candidate 1 gives a more stable start |
| Center deformation onset | Early | Delayed | Candidate 1 improves center consolidation |
| Deformation penetration | Broad and progressive | Localized near shoulder | Candidate 1 favors uniform strain |
| Flow-line continuity | Continuous through web and rim | More distorted near shoulder | Candidate 1 is preferable for forged gear blanks |
| Temperature symmetry | Approximately symmetric | Less symmetric | Candidate 1 reduces thermal gradients |
| Maximum contact temperature | Moderate | Higher near moving die | Candidate 2 may increase local softening |
| Step count | 92 | 107 | Candidate 1 has a simpler numerical history |
| Risk of lap or fold | Lower | Higher | Candidate 1 is more robust |
| Overall preference | Preferred | Not preferred | I select the 85 mm by 130 mm billet |
Flow-line evolution. Flow lines in forged gear blanks are not cosmetic. They indicate how the material has been reoriented by the forging sequence. When the flow lines follow the intended contour of the hub, web, and rim, the forged gear blank has a favorable grain-flow pattern. When the flow lines are abruptly turned, folded, or interrupted, the risk of fatigue initiation and stress concentration increases. In candidate 1, the flow lines remain smooth through the web because the billet enters the die cavity without a shoulder-induced reversal. In candidate 2, the overhanging material creates a local reversal near the shoulder, and the flow lines become more distorted.
I use the flow-line continuity as a qualitative criterion and supplement it with the strain-gradient criterion. The effective strain gradient is
$$
\left\|\nabla \bar{\varepsilon}\right\|
=
\sqrt{
\left(
\frac{\partial \bar{\varepsilon}}{\partial r}
\right)^2
+
\left(
\frac{\partial \bar{\varepsilon}}{\partial z}
\right)^2
}
$$
A large gradient means that adjacent regions of the forged gear blanks have experienced very different amounts of deformation. This can lead to residual stress, hardness variation, and an increased risk of cracking during cooling. Candidate 1 has a lower average strain gradient in the web region, while candidate 2 has a higher gradient near the shoulder. For this reason, I regard candidate 1 as the better route for forged gear blanks.
Temperature field and friction heating. The thermal history of forged gear blanks is coupled to deformation. Plastic work and friction work raise the local temperature, while conduction transfers heat into the die and into cooler regions of the workpiece. The competition between heat generation and conduction determines the final temperature distribution. In candidate 1, the contact heating is distributed over the end faces and the cavity walls in a relatively balanced way. In candidate 2, the shoulder contact produces a local hot zone, and the temperature gradient near the moving die is steeper.
I estimate the local temperature rise from friction using
$$
\Delta T_{fric}
\approx
\frac{
\tau_f \left\|v_t\right\| \Delta t
}{
\rho c_p \Delta V
}
$$
where Delta t is the time interval and Delta V is the local volume. This equation is not a substitute for the finite-element solution, but it helps me interpret why the contact layer of forged gear blanks heats up quickly when the sliding velocity and friction stress are high. In candidate 2, the shoulder region experiences a longer sliding distance under high contact pressure, so the local temperature rise is more pronounced.
I also examine the temperature gradient normal to the die surface,
$$
G_T
=
\left\|\nabla T\right\|
$$
A high gradient can produce a hard surface layer and a softer interior, which may be desirable in some cases but can also increase residual stress. For the forged gear blanks in this study, I prefer a moderate and symmetric gradient. Candidate 1 provides this more effectively than candidate 2.
| Thermal indicator | Candidate 1 | Candidate 2 | Reason |
|---|---|---|---|
| Initial temperature | 1250 degrees Celsius | 1250 degrees Celsius | Same starting condition |
| Contact heating pattern | Distributed over end faces and cavity walls | Concentrated near shoulder and moving die | Different initial contact geometry |
| Mid-plane symmetry | Approximately symmetric | Less symmetric | Candidate 2 has overhang-induced flow |
| Maximum local temperature | Moderate | Higher near contact layer | Longer sliding and localized friction in candidate 2 |
| Temperature gradient | Lower | Higher | More uniform deformation in candidate 1 |
| Implication for forged gear blanks | More uniform final properties | Greater property variation | Candidate 1 is preferred |
Stress, load, and press response. The load required to forge the gear blanks is the integral of the contact pressure over the die-workpiece interface,
$$
F
=
\int_{A_c}
p
\, dA
$$
where A_c is the contact area and p is the contact pressure. The load is not only a function of the material flow stress; it also depends on the contact area, the friction stress, and the constraint imposed by the die. In candidate 2, the overhanging shoulder increases the contact pressure in a localized region and requires more steps to resolve the flow. I observe a steeper load increase in the later stage of candidate 2, which is consistent with the more constrained flow and the larger contact area near the shoulder.
The mean pressure can be approximated by
$$
p_{avg}
=
\frac{F}{A_c}
$$
and the normalized load can be written as
$$
F_n
=
\frac{F}{F_{ref}}
$$
where F_ref is a reference load at the final stroke of candidate 1. I use these normalized quantities to compare the two candidates without relying on a single absolute load value, which would depend on the exact flow-stress data and press characteristics. In my assessment, candidate 1 has a smoother load history and a more gradual load increase. Candidate 2 has a more concentrated load response because the shoulder contact restricts the early flow.
The plastic work dissipated in the forged gear blanks is
$$
W_p
=
\int_{0}^{t}
\int_{V}
\sigma_f
\dot{\bar{\varepsilon}}
\, dV \, dt
$$
and the specific forging energy is
$$
e_f
=
\frac{W_p}{m}
$$
where m is the mass of the forged gear blank. A lower specific forging energy is not automatically better, because some deformation work is necessary to consolidate the material and fill the cavity. However, an excessively concentrated energy input near the shoulder can produce local overheating and an unfavorable strain path. I prefer the candidate that distributes the deformation work more evenly through the forged gear blanks.
| Mechanical indicator | Candidate 1 | Candidate 2 | Interpretation |
|---|---|---|---|
| Contact area growth | Smooth and progressive | Rapid near shoulder | Candidate 1 is more stable |
| Load increase | Gradual | Steeper in later stage | Candidate 2 is more constrained |
| Peak contact pressure | Lower and more distributed | Higher and more localized | Candidate 1 favors die life |
| Plastic work distribution | Broad | Concentrated near shoulder | Candidate 1 favors uniform properties |
| Specific forging energy | Moderate | Higher local concentration | Candidate 1 is preferable |
Die filling and defect risk. The fill sequence of forged gear blanks determines whether the final part has complete corners, a sound web, and a well-formed rim. In candidate 1, the metal reaches the die top and then fills the gear-root region in a progressive manner. The final contact is established without a sudden closure of an isolated cavity. In candidate 2, the shoulder contact creates a flow barrier, and the metal must turn inward and upward before filling the rim. This increases the risk of a lap or a fold. A lap can form when a free surface folds onto itself and is then forged into the body. A fold can create a stress concentration that survives subsequent heat treatment and machining.
I use a simple negative-Jacobian check as a numerical defect indicator,
$$
N_J
=
\#\left\{e \mid J_e \le 0\right\}
$$
where J_e is the Jacobian of element e. A successful simulation should have N_J equal to zero. I also monitor the remeshing frequency. A high remeshing frequency indicates severe element distortion and often signals a difficult flow pattern. Candidate 1 requires fewer remeshing events than candidate 2 in my assessment. This supports the conclusion that candidate 1 produces more stable deformation in the forged gear blanks.
Another indicator is the fill ratio,
$$
R_f
=
\frac{V_{filled}}{V_{cavity}}
$$
where V_filled is the volume of the cavity occupied by the workpiece and V_cavity is the total cavity volume. Both candidates eventually reach complete fill, but candidate 1 reaches a more uniform fill history. Candidate 2 reaches full fill after a more localized flow sequence. For production of forged gear blanks, a uniform fill history is preferable because it reduces the chance of entrapped oxides, laps, and cold shuts.
| Defect-related indicator | Candidate 1 | Candidate 2 | My judgment |
|---|---|---|---|
| Negative Jacobian count | Zero after remeshing | Zero after remeshing | Both are numerically stable with remeshing |
| Remeshing frequency | Lower | Higher | Candidate 1 has a simpler flow history |
| Fill sequence | Progressive from web to rim | Constrained by shoulder | Candidate 1 is more robust |
| Lap risk | Lower | Higher | Candidate 1 is preferred |
| Fold risk | Lower | Higher | Candidate 1 is preferred |
| Final cavity fill | Complete | Complete | Both fill, but candidate 1 fills more uniformly |
Material utilization and process economy. I evaluate material utilization because precision forging of forged gear blanks is often justified by savings in material and machining. The material utilization ratio is
$$
\eta_m
=
\frac{V_{forged}}{V_{billet}}
\times 100\%
$$
where V_forged is the volume of the final forged gear blank and V_billet is the billet volume. If flash is present, I write
$$
V_{flash}
=
V_{billet}
–
V_{forged}
–
V_{scale}
$$
where V_scale is the volume lost to scale. A higher material utilization ratio means that more of the billet becomes the forged gear blank and less becomes flash, scale, or machining chips. The source process target of more than 70 percent material utilization is consistent with a well-controlled precision forging route. I do not treat this number as automatic. I treat it as a target that depends on billet selection, die design, flash allowance, and the repeatability of the press stroke.
The machining allowance also matters. If the forged gear blank is too close to the final shape, the machining allowance can be small, but the die cost and the risk of incomplete fill increase. If the forged gear blank is too conservative, the machining allowance is large and the material saving is reduced. I therefore examine the forged contour and the expected machining stock together. Candidate 1 gives a more uniform deformation field, which in my view makes it easier to control the machining allowance around the hub and rim.
| Economic and material indicator | Candidate 1 | Candidate 2 | Effect on forged gear blanks |
|---|---|---|---|
| Billet volume | 737,685 mm3 | 738,274 mm3 | Nearly identical |
| Material utilization potential | High | High but more sensitive to flash and scale | Candidate 1 is more robust |
| Machining allowance control | More uniform | Less uniform | Candidate 1 favors near-net shape |
| Die fill risk | Lower | Higher near shoulder | Candidate 1 reduces scrap risk |
| Process repeatability | Better | More sensitive to billet alignment | Candidate 1 is preferable |
Sensitivity to process variation. A forging process is only useful if it tolerates normal variation. I consider small changes in billet diameter, billet length, initial temperature, friction, and die alignment. The first candidate is less sensitive to small diameter variation because its nominal diameter already matches the die entrance. A small increase in diameter creates only a small overhang, while a small decrease creates a small clearance that can be accommodated by the initial compression. The second candidate is more sensitive because its nominal diameter already exceeds the die entrance by 15 mm. A further increase in diameter increases the overhang and the shoulder contact force. A decrease in diameter reduces the overhang but can change the contact sequence abruptly.
I express the diameter sensitivity as
$$
S_d
=
\frac{\partial F}{\partial d_0}
$$
and the temperature sensitivity as
$$
S_T
=
\frac{\partial \sigma_f}{\partial T}
$$
In my assessment, candidate 2 has a higher diameter sensitivity because the shoulder contact is strongly affected by the initial overhang. Candidate 1 has a lower diameter sensitivity because the billet enters the die cavity in a more axisymmetric manner. This is important for forged gear blanks produced in repeated production runs, where billet tolerances and die wear are unavoidable.
| Variation source | Effect on candidate 1 | Effect on candidate 2 | Robustness implication |
|---|---|---|---|
| Billet diameter tolerance | Moderate | High | Candidate 1 is more robust |
| Billet length tolerance | Moderate | Moderate | Both require volume control |
| Initial temperature scatter | Moderate | Moderate to high | Candidate 2 has a more localized thermal response |
| Friction variation | Moderate | High near shoulder | Candidate 1 is less sensitive |
| Die alignment error | Low to moderate | Moderate to high | Candidate 1 preserves symmetry better |
Interpretation of the simulation results. I do not treat the numerical simulation as a replacement for physical trials. I treat it as a way to reduce the number of trials and to identify the most promising billet geometry for forged gear blanks. The simulation shows that candidate 1 and candidate 2 have almost the same volume, but they do not have the same forming history. Candidate 1 produces an earlier center deformation, a more continuous flow-line pattern, a more symmetric temperature field, and a more gradual load response. Candidate 2 produces a shoulder-dominated contact, a delayed center deformation, a steeper thermal gradient, and a higher risk of flow-line distortion. These results lead me to select the 85 mm by 130 mm billet for the forged gear blanks.
I summarize the decision using a weighted score. The score is
$$
S
=
w_1 U_{\varepsilon}
+
w_2 U_T
+
w_3 C_f
+
w_4 R_d
+
w_5 M_u
$$
where U_epsilon is the strain uniformity index, U_T is the thermal symmetry index, C_f is the flow-line continuity score, R_d is the defect-risk score, and M_u is the material-utilization score. The weights w_i sum to one. In my qualitative assessment, candidate 1 receives a higher score because it performs better on strain uniformity, thermal symmetry, flow-line continuity, and defect risk. Candidate 2 is not disqualified, but it requires tighter process control and a more careful die-shoulder design.
| Weighted criterion | Weight | Candidate 1 score | Candidate 2 score | Reason |
|---|---|---|---|---|
| Strain uniformity | 0.25 | High | Medium | Candidate 1 deforms the center earlier |
| Thermal symmetry | 0.20 | High | Medium | Candidate 1 has a more balanced contact history |
| Flow-line continuity | 0.25 | High | Medium to low | Candidate 2 distorts flow near the shoulder |
| Defect risk | 0.20 | Low risk | Medium to high risk | Candidate 2 has a greater lap and fold tendency |
| Material utilization | 0.10 | High | High | Both have nearly identical volume |
Design implications for forged gear blanks. The most important design implication I draw is that billet diameter should be matched to the die entrance whenever the process allows it. A matched diameter promotes axisymmetric contact, early center deformation, and a smooth flow-line pattern. An oversized diameter may seem attractive because it reduces billet length and can increase buckling resistance, but it introduces shoulder contact and a more complex flow history. For forged gear blanks, the flow history is part of the product quality. It influences grain flow, hardness distribution, residual stress, and fatigue performance.
I also conclude that the initial temperature of 1250 degrees Celsius is not simply a scalar input. It sets the starting flow stress, but the local temperature evolves because of plastic work and friction. A process that looks acceptable at the starting temperature can develop an unfavorable hot zone near the die shoulder. I therefore evaluate the final temperature field, not only the initial temperature. The same logic applies to friction. A friction coefficient of 0.2 is a reasonable input for the simulation, but the local effect of friction depends on the sliding velocity and contact pressure. The shoulder contact in candidate 2 amplifies this effect.
Another implication is that the remeshing strategy is part of the process design. The maximum element edge of 0.3 mm and the OVERLAY QUAD option allow the simulation to follow large deformation in the forged gear blanks. Without remeshing, the mesh would become distorted and the predicted flow lines would be unreliable. I monitor the remeshing frequency because it is an indicator of flow complexity. A process that requires frequent remeshing in a small region is a process that deserves a closer look at the die geometry and billet shape.
Practical process window. Based on my simulation, I define a practical process window for the forged gear blanks. The billet diameter should be close to the die entrance diameter. The billet length should be chosen to provide the required volume without excessive flash. The initial temperature should be high enough to keep the flow stress within the press capacity but not so high that scale growth and grain coarsening become excessive. The friction condition should be controlled by lubrication and die surface condition. The press stroke should be monitored so that the final contact is reached without an excessive overload.
| Process variable | Preferred condition for forged gear blanks | Reason |
|---|---|---|
| Billet diameter | Close to die entrance diameter | Promotes axisymmetric contact and early center deformation |
| Billet length | Set by volume balance and flash allowance | Controls fill and material utilization |
| Initial temperature | Sufficient for flow without excessive scale | Balances press load and microstructure |
| Friction | Moderate and repeatable | Controls contact heating and flow-line distortion |
| Die alignment | High precision | Preserves symmetry of forged gear blanks |
| Remeshing control | Fine mesh near contact and shear zones | Maintains numerical accuracy during large deformation |
Final selection and conclusion. I compare the two candidate billets for forged gear blanks and select the 85 mm by 130 mm billet. The simulation with this billet requires 92 steps and produces a deformation pattern in which the center participates early, the metal flows progressively into the cavity, the temperature field remains approximately symmetric, and the flow lines remain continuous. The 100 mm by 94 mm billet requires 107 steps and produces a shoulder-dominated contact, a delayed center deformation, a steeper temperature gradient, and a greater risk of flow-line distortion. Although both billets have nearly the same volume, the matched-diameter billet is clearly more favorable for the forged gear blanks.
In my final assessment, the numerical simulation of forged gear blanks is valuable because it reveals the interaction between billet geometry, die contact, friction, and thermal history before a physical trial is performed. I can see where the metal flows, where it heats, where it strains, and where it may fail to fill. This information allows me to choose the billet, refine the die, and define a more reliable process window. For the present case, the 85 mm by 130 mm billet is the better choice for producing forged gear blanks with uniform deformation, continuous flow lines, and favorable mechanical performance.
| Final decision item | Selected condition | Basis |
|---|---|---|
| Billet for forged gear blanks | 85 mm diameter by 130 mm length | More uniform deformation and better flow-line continuity |
| Material | C45 steel | Appropriate forgeability and gear-blank strength |
| Initial temperature | 1250 degrees Celsius | Hot-forging condition used in the simulation |
| Friction coefficient | 0.2 | Moderate hot-forging friction |
| Moving die stroke | 200,0 to 31.5,0 | Defined closing schedule for the forged gear blanks |
| Remeshing limit | 0.3 mm maximum element edge | Accurate tracking of local flow and contact |
| Preferred outcome | Candidate 1 | Symmetric temperature, early center deformation, lower defect risk |
The broader lesson I take from this study is that forged gear blanks should be designed with the same attention to material flow as to final dimensions. A gear blank that looks correct on a drawing can still produce an unfavorable flow pattern if the billet does not enter the die in a symmetric and progressive manner. By using an axisymmetric finite-element model, a matched-diameter billet, and a controlled remeshing strategy, I can improve the uniformity of forged gear blanks and support the development of a more reliable precision-forging process.