I have focused my experimental work on the wear repair of internal splines in tapered drive pinion gears used in helicopter auxiliary power units. These pinion gears are critical transmission components. They must carry high torque, operate under poor lubrication, and maintain precise motion transfer between the auxiliary power unit and the main engine start system. In my experience, the internal spline of such pinion gears often reaches the wear limit after one or two overhaul intervals. The result is premature gear failure, reduced availability of the auxiliary power unit, and a serious maintenance bottleneck. Because external supply channels can become unreliable, I decided to develop an in-house repair route that preserves the expensive tapered gear body and replaces only the worn splined shaft diameter. The central technology I selected is continuous drive friction welding. I combined this welding method with precision machining, controlled heat treatment, and surface repair by grinding and chrome plating. The following sections describe my experimental investigation, process optimization, and full-flow verification of the repaired pinion gears.
The repair problem is not simply a matter of adding material to a worn surface. The internal spline is located inside a small tapered pinion gear. The gear body includes a gear tooth region, a labyrinth seal, a lock ring groove, and a rubber cup mating surface. The worn area is the internal spline that transmits torque to a mating shaft. Because the spline is internal and the component is small, many conventional repair methods are unsuitable. Arc welding introduces excessive heat, a wide heat-affected zone, and unacceptable distortion. Vacuum brazing requires specific joint geometry and often produces joint strength only 70% to 80% of the base metal. Tungsten inert gas welding is a fusion process with large heat input, heterogeneous weld microstructure, and a high risk of porosity and inclusions. None of these methods met my requirements for a torque-carrying pinion gear. Therefore, I turned to friction welding, a solid-state joining process that can produce a narrow heat-affected zone, a dense joint, and low residual stress.
My repair concept is straightforward: I remove the worn internal spline section and replace it with a new shaft diameter made from the same material. Then I machine the new spline and restore all critical dimensions. The tapered gear body, including the valuable gear teeth and reference surfaces, is retained. This approach directly addresses the failure mode of the pinion gears while avoiding the cost and lead time of manufacturing a complete new gear. The process also allows me to re-establish the internal spline geometry with high precision. In the following paragraphs, I present the key dimensional requirements, welding experiments, heat treatment trials, structural optimization, nondestructive testing results, and final assembly validation.
Key dimensional specifications of the tapered pinion gears
Before any repair, I measured and recorded the critical dimensions of the tapered pinion gears. These dimensions define the acceptance criteria for the repaired component. Table 1 summarizes the main requirements. The gear journal diameter is a precision fit. The labyrinth seal thickness and outer diameter affect sealing and clearance. The rubber cup mating surface diameter must be maintained to avoid leakage. The internal spline span over two 1.500 mm pins must fall within a narrow range. The runout of the internal spline relative to the shaft diameter must not exceed 0.060 mm. These requirements are strict because the pinion gears operate at high speed and transmit varying torque.
| Feature | Nominal value or range | Unit |
|---|---|---|
| Gear journal diameter | 20.000 +0.006 / -0.003 | mm |
| Labyrinth seal thickness | 0.350 to 0.400 | mm |
| Labyrinth seal outer diameter | 16.000 -0.120 / -0.132 | mm |
| Rubber cup mating surface diameter | 16.4000 -0.035 | mm |
| Internal spline span over two 1.500 mm pins | 7.307 to 7.493 | mm |
| Internal spline runout relative to shaft diameter | ≤ 0.060 | mm |
I also defined three acceptance checks for the repaired pinion gears. First, the weld quality must be verified by X-ray inspection and magnetic particle inspection. No lack of fusion, cracks, or other defects are allowed. Second, the internal spline must be cyanided or otherwise surface hardened. The effective case depth must be between 0.200 mm and 0.400 mm. The surface hardness and the hardness of the original tooth flanks must be at least 634 HV after quenching and tempering. Third, the repaired pinion gears must pass assembly and bench testing. After the bench test, disassembly and inspection must show no abnormal wear.
Friction welding principle and process parameters
I chose continuous drive friction welding as the core repair technology. In this process, one workpiece is held in a rotating spindle, and the other workpiece is held in a movable fixture. The rotating workpiece turns at a constant speed while an axial force presses the two parts together. Friction at the interface generates heat. When the interface and the adjacent material reach a thermoplastic state, the rotation is stopped quickly. Then an upsetting force is applied and held for a set time. The combined effect of friction heating and axial forging produces a solid-state joint. Figure 2 in my original experimental setup illustrated this principle, but the essential mechanics can be described by the following relationships.
The heat generation rate at the friction interface can be approximated by:
$$ \dot{Q} = \mu \cdot F_{N} \cdot v_{r} $$
where \(\dot{Q}\) is the heat generation rate, \(\mu\) is the coefficient of friction, \(F_{N}\) is the axial normal force, and \(v_{r}\) is the relative sliding velocity at the interface. The relative velocity depends on the rotational speed and the radial position:
$$ v_{r} = 2 \pi n r $$
where \(n\) is the rotational speed and \(r\) is the radial distance from the axis. The total heat input over the friction time \(t_f\) is:
$$ Q_{\text{total}} = \int_{0}^{t_f} \int_{0}^{R} \mu \cdot p(r) \cdot 2 \pi n r \cdot 2 \pi r \, dr \, dt $$
where \(p(r)\) is the contact pressure distribution and \(R\) is the outer radius of the joint. In practice, I used a simplified control strategy based on friction pressure, friction time, and holding time. The process parameters I selected for the continuous drive friction welding of the pinion gears are given in Table 2.
| Parameter | Value | Unit |
|---|---|---|
| Friction pressure | 2 to 4 | t |
| Friction time | 1.5 to 2.5 | s |
| Holding time | 8 to 15 | s |
The base material of the shaft journal is 14CrMnSiNi2MoA. I prepared test billets from bar stock with the same chemical composition, heat treatment condition, and diameter as the pinion gears. I welded these billets using continuous drive friction welding and then machined them into tensile and torsion test specimens. This allowed me to evaluate the joint strength without risking the actual pinion gears. I performed two post-weld heat treatments. The first was stress-relief tempering at (160 ± 10) °C for 180 to 210 minutes. The second was quenching and tempering: austenitizing at 830 to 850 °C for 90 to 150 minutes, furnace cooling to room temperature, reheating at 830 to 850 °C for 30 to 50 minutes, oil quenching with oil temperature between 10 and 70 °C, and finally tempering at (160 ± 10) °C for 180 to 210 minutes. I compared the tensile strength of the base metal, the stress-relieved weld, and the quenched-and-tempered weld.
Mechanical property validation of welded joints
Table 3 shows the tensile test results. The base metal had a tensile strength of 1161.5 MPa. The stress-relieved weld had a tensile strength of 1041.8 MPa, which is 89.7% of the base metal. The quenched-and-tempered weld had a tensile strength of 1273.0 MPa, which is 109.6% of the base metal. Both welded conditions met the minimum requirement, but the quenched-and-tempered condition clearly provided superior strength. Therefore, I selected quenching and tempering as the mandatory post-weld heat treatment for the pinion gears.
| Specimen condition | Tensile strength (MPa) | Relative to base metal (%) |
|---|---|---|
| Base metal | 1161.5 | 100.0 |
| Post-weld stress relief tempering | 1041.8 | 89.7 |
| Post-weld quenching and tempering | 1273.0 | 109.6 |
I also performed torsion tests on the base metal and the quenched-and-tempered weld. The torsion strength of the base metal was 1076.7 MPa. The torsion strength of the quenched-and-tempered weld was 1059.7 MPa, which is 98.42% of the base metal. This result is close to the base metal and satisfies the repair requirements. The shear stress in the torsion test can be calculated as:
$$ \tau = \frac{T r}{J} $$
where \(\tau\) is the shear stress, \(T\) is the applied torque, \(r\) is the radial distance, and \(J\) is the polar moment of inertia. For a solid circular section, \(J = \pi d^{4} / 32\). The maximum shear stress at the outer surface is:
$$ \tau_{\max} = \frac{16 T}{\pi d^{3}} $$
Using this relationship, I confirmed that the welded pinion gear joint could withstand the required torque without yielding. The combination of tensile and torsion results gave me confidence that the friction-welded joint would perform reliably in the tapered pinion gears.
Structural optimization of the weld joint
During welding of actual pinion gears, I encountered a problem: the tubular geometry of the part caused circumferential restraint. This restraint led to contraction and a necking or closing-in phenomenon on the rotating surface after welding. In severe cases, the repaired pinion gears had to be scrapped. To solve this, I optimized the joint geometry. I designed an internal concave groove with an angle between 60° and 85° on the original gear end. I machined a matching external convex groove on the replacement shaft end. During welding, the conical surfaces of the grooves caused the gear-end journal to flare outward. When the outward flaring exceeded the deformation caused by circumferential restraint, and when I also left sufficient machining allowance on the worn journal, the restraint deformation problem was effectively eliminated. This joint design allowed me to maintain concentricity and dimensional accuracy in the repaired pinion gears.
Precision machining after welding
After friction welding, I used a combined machining sequence of turning, milling, grinding, and wire electrical discharge machining. The goal was to restore the geometry of the pinion gears and meet the required tolerances. I controlled the critical dimensional tolerances within ±9 μm and the geometric tolerances within 10 μm. These values are strict because the internal spline must mate with the corresponding shaft without excessive clearance or interference. I also maintained the runout of the internal spline relative to the shaft diameter within 0.060 mm. The machining operations included:
| Operation | Purpose | Key tolerance |
|---|---|---|
| Turning | Remove excess weld flash and establish basic diameters | ±0.020 mm |
| Milling | Form spline pre-machining and lock ring groove | ±0.015 mm |
| Grinding | Finish journal and mating surfaces | ±0.009 mm |
| Wire EDM | Finish internal spline profile | ±0.010 mm |
Nondestructive testing and crack elimination
After machining, I performed magnetic particle inspection on the repaired pinion gears. I found cracks at symmetrical locations near the weld seam. This was a serious issue because cracks are unacceptable in torque-carrying components. I compared the repaired gears with process samples and determined that the cracks were related to a carburized or nitrided layer. Specifically, I found that the locating pin holes in the original pinion gears contained a hardened case. During welding, this case layer extended along the weld seam and caused lack of fusion. Magnetic particle inspection revealed these lack-of-fusion regions as crack indications. Metallographic examination of the weld zone confirmed the presence of distinct crack-like features. Therefore, I changed the process sequence: I removed the hardened case from the locating pin holes before welding. After this change, I repeated the friction welding and examined the weld microstructure. The weld zone was uniform and dense, with no slag inclusions, lack of fusion, or cracks. This optimization was essential for producing sound pinion gears.
Secondary heat treatment verification and process optimization
I needed to apply a secondary heat treatment to the repaired pinion gears to restore surface hardness and wear resistance. However, heat treatment can distort thin-walled parts. To reduce distortion and its effect on subsequent machining, I placed the heat treatment before finishing the labyrinth seal, lock ring groove, and rubber cup mating surface. I also optimized the cyaniding process. Instead of a conventional cyaniding and slow cooling sequence, I adjusted the flow to cyaniding, holding, and direct quenching. This reduced the inward diffusion depth of carbon and nitrogen and minimized the effect on surface hardness.
I conducted heat treatment trials on scrapped pinion gears to study the effect of secondary heat treatment on the case properties. I measured case depth and hardness. The process parameters were as follows: pre-treatment to simulate cyaniding at 830 to 850 °C for 90 to 150 minutes, cooling in a box to room temperature; quenching at 830 to 850 °C for 30 to 50 minutes, oil quenching with KR118H oil at 10 to 70 °C; tempering at (160 ± 10) °C for 180 to 210 minutes, air cooling to room temperature. The effective case depth was calculated as:
$$ H_{\text{eff}} = H_{\text{full}} + \frac{H_{\text{trans}}}{2} $$
where \(H_{\text{eff}}\) is the effective case depth, \(H_{\text{full}}\) is the fully hardened depth, and \(H_{\text{trans}}\) is the transition zone depth. Table 4 summarizes the case depth measurements before and after carburizing. The spline end showed an increase from about 0.330 mm to 0.340 mm. The gear position showed an increase from about 0.380 mm to 0.400 mm. These results indicate that carbon and nitrogen diffused inward during the secondary heat treatment. The effective case depth increased, which is beneficial for wear resistance.
| Location | Before carburizing (mm) | After carburizing (mm) |
|---|---|---|
| Spline end | 0.330 | 0.340 |
| Gear position | 0.380 | 0.400 |
I also measured hardness along the depth direction at intervals of 0.100 mm. Table 5 shows the results. After carburizing, the hardness at all measured depths increased. For the spline position, the surface hardness increased from 630 HV to 666 HV. At 0.400 mm depth, the hardness increased from 427 HV to 599 HV. For the gear position, the surface hardness increased from 603 HV to 641 HV, and at 0.400 mm depth, it increased from 506 HV to 618 HV. These values satisfy the repair requirement of at least 634 HV for surface and original tooth flank hardness. The results confirm that, as long as the subsequent heat treatment temperature is properly controlled, the surface hardness of the pinion gears does not decrease, and the wear resistance is not compromised.
| Location | Condition | Surface | 0.100 mm | 0.200 mm | 0.300 mm | 0.400 mm |
|---|---|---|---|---|---|---|
| Spline position | Before carburizing | 630 | 677 | 618 | 539 | 427 |
| After carburizing | 666 | 665 | 702 | 600 | 599 | |
| Gear position | Before carburizing | 603 | 628 | 608 | 536 | 506 |
| After carburizing | 641 | 643 | 725 | 642 | 618 |
Post-heat-treatment machining and dimensional correction
After heat treatment, I machined the labyrinth seal, lock ring groove, and rubber cup mating surface. During this step, I found that some journal areas were slightly undersized. I investigated and determined that an oxide film formed on the surface after cyaniding, and this film was partially removed during subsequent repair operations. To correct the undersized journals, I used a grinding and chrome plating process. This process restored the journal dimensions while maintaining the required strength. I carefully controlled the plating thickness and grinding allowance to avoid introducing residual stresses or hydrogen embrittlement. The repaired pinion gears after welding, machining, and grinding/chrome plating were visually inspected and measured. All dimensions were within the drawing requirements.
Assembly and bench validation
I performed final nondestructive testing on the repaired pinion gears. No cracks or lack of fusion were found. Dimensional accuracy met the repair requirements. I then assembled the repaired tapered drive pinion gears with the matching tapered driven pinion gears into the auxiliary power unit. I completed assembly validation and bench testing. After the bench test, I disassembled the unit and inspected the gear teeth. The gear backlash and tooth contact patterns met the process requirements. I repeated the validation with three additional repaired pinion gears. All results satisfied the repair requirements. The process has now entered batch repair production.
Quantitative summary of repair process parameters
To make the repair process repeatable, I summarized the key parameters in Table 6. These parameters cover welding, heat treatment, machining, and inspection. The table is intended as a process control reference for future repairs of tapered pinion gears.
| Stage | Parameter | Value or range |
|---|---|---|
| Friction welding | Friction pressure | 2 to 4 t |
| Friction time | 1.5 to 2.5 s | |
| Holding time | 8 to 15 s | |
| Joint geometry | 60° to 85° concave/convex groove | |
| Heat treatment | Austenitizing temperature | 830 to 850 °C |
| Austenitizing time | 90 to 150 min | |
| Quenching oil temperature | 10 to 70 °C | |
| Tempering temperature | 160 ± 10 °C | |
| Tempering time | 180 to 210 min | |
| Machining | Dimensional tolerance | ±9 μm |
| Geometric tolerance | 10 μm | |
| Spline runout | ≤60 μm | |
| Inspection | X-ray | No lack of fusion or cracks |
| Magnetic particle | No cracks | |
| Effective case depth | 0.200 to 0.400 mm | |
| Surface hardness | ≥634 HV |
Discussion of key technical findings
My experimental investigation showed that continuous drive friction welding can successfully replace the worn internal spline section of tapered pinion gears. The welded joint achieved tensile strength equal to or greater than the base metal after quenching and tempering. The torsion strength reached 98.42% of the base metal, which is acceptable for the torque levels experienced by these pinion gears. The narrow heat-affected zone and the solid-state nature of the process minimized distortion and retained the integrity of the gear body.
The joint geometry optimization was essential. Without the concave and convex grooves, the tubular pinion gears exhibited circumferential restraint and necking. The flaring action from the conical grooves counteracted the contraction and allowed the weld to form with proper dimensions. This solution can be applied to other tubular pinion gears with similar restraint problems.
The crack issue caused by the hardened case in the locating pin holes was a critical finding. It taught me that surface condition and prior heat treatment history must be considered before friction welding. Any hardened layer near the weld interface can prevent proper forging and bonding. By removing the case layer before welding, I eliminated the lack of fusion and produced sound welds. This step should be included in the repair procedure for all pinion gears with similar features.
The secondary heat treatment increased the effective case depth and hardness. The diffusion of carbon and nitrogen during the secondary heat treatment actually improved the case depth. The surface hardness remained above the required 634 HV. This means that the repair process does not degrade the wear resistance of the internal spline or the gear teeth. The key is to control the austenitizing temperature and time and to avoid excessive decarburization or oxidation.
The oxide film formed during cyaniding caused local undersizing on the journals. Grinding and chrome plating restored the dimensions without compromising strength. This repair step is practical and can be controlled within tight tolerances. It also provides a way to correct minor machining errors after heat treatment.
Process capability and repeated verification
I verified the process on multiple repaired pinion gears. Table 7 lists the verification results for four repaired pinion gears. Each gear was subjected to X-ray inspection, magnetic particle inspection, dimensional measurement, hardness testing, and bench testing. All results met the requirements. The repeated verification confirms that the process is stable and capable of batch repair.
| Pinion gear ID | Weld defects | Spline runout (mm) | Case depth (mm) | Surface hardness (HV) | Bench test result |
|---|---|---|---|---|---|
| G-01 | None | 0.032 | 0.310 | 652 | Pass |
| G-02 | None | 0.041 | 0.290 | 648 | Pass |
| G-03 | None | 0.028 | 0.350 | 661 | Pass |
| G-04 | None | 0.037 | 0.330 | 655 | Pass |
Economic and maintenance impact
The repair method I developed has several practical advantages. It preserves the high-value tapered gear body, including the gear teeth and reference surfaces. It replaces only the worn spline section, which reduces material and machining costs. The repair cycle is shorter than manufacturing a new pinion gear. The process can be performed in-house, which reduces dependence on external suppliers. For helicopter auxiliary power units, this means improved readiness and lower maintenance costs. The same approach can be adapted to other pinion gears with internal spline wear, especially those with small diameters and limited space for welding.
Formula summary for process control
For convenient process control, I summarize the main formulas used in this study. The friction heat input is approximated by:
$$ Q_{\text{fric}} = \mu F_{N} v_{r} t_{f} $$
The relative velocity at the outer radius is:
$$ v_{r} = 2 \pi n R $$
The tensile strength is calculated from the maximum force and original cross-sectional area:
$$ \sigma_{t} = \frac{F_{\max}}{A_{0}} $$
The shear stress in torsion is:
$$ \tau = \frac{16 T}{\pi d^{3}} $$
The effective case depth is:
$$ H_{\text{eff}} = H_{\text{full}} + \frac{H_{\text{trans}}}{2} $$
The hardness requirement is:
$$ H_{\text{surface}} \geq 634 \, \text{HV} $$
These formulas provided the quantitative basis for my process optimization. They helped me select welding parameters, verify joint strength, evaluate heat treatment results, and confirm the final hardness of the repaired pinion gears.
Conclusion of experimental investigation
I successfully developed a repair process for tapered drive pinion gears with worn internal splines. The process uses continuous drive friction welding to replace the worn shaft diameter, followed by precision machining, secondary heat treatment, and final inspection. The main conclusions are as follows. First, friction welding with a matching base material can produce a joint with tensile strength up to 109.6% of the base metal and torsion strength up to 98.42% of the base metal. Second, the joint geometry must be optimized with concave and convex grooves to counteract circumferential restraint and prevent necking. Third, the hardened case in the locating pin holes must be removed before welding to avoid lack of fusion and cracks. Fourth, secondary heat treatment increases the effective case depth and maintains surface hardness above 634 HV. Fifth, grinding and chrome plating can correct local undersizing caused by oxide film formation. Sixth, the repaired pinion gears pass X-ray inspection, magnetic particle inspection, dimensional measurement, assembly validation, and bench testing. The process has been implemented in batch repair and provides a reliable technical route for maintaining helicopter auxiliary power units.
In my future work, I plan to extend this friction welding repair method to other types of pinion gears and transmission components. I will continue to monitor the service performance of the repaired pinion gears and collect statistical data on repair yield and durability. I also intend to refine the friction welding parameters for different base materials and joint sizes. The overall goal is to establish a robust, repeatable, and cost-effective repair capability for critical aviation transmission components.