Within the automotive sector, new gear prototypes must be produced continuously, for which conventional production methods are not efficient. A possible alternative is selective laser hardening, integrated on a production centre that also performs other operations on the gears. The possibilities were investigated on three different types of gears.

Within the automotive industry, newly prototyped automotive drivetrain gears are continuously designed to satisfy the need to handle higher loads combined with weight and size reduction. The conventional process chain for the manufacturing of prototype gears is very time and energy consuming. Prototype gears are machined in a soft state followed by a hardening process and if still required, a finishing step. However, these classical production technologies proven to be efficient for large batch sizes, but not for prototyping. The development and validation of a more sustainable process chain for the efficient production of prototype gears became imperative.

The production of automotive drivetrain gears involves several steps that require a lot of clamping, adjustment and transport, which increases production time and costs.

A study conducted by Flanders Make, in cooperation with KU Leuven and Sirris, focused on the implementation and validation of a more efficient and sustainable process chain achieved by integrating a hardening set-up within a machining centre where roughing and finishing of automotive gears can be performed. Thus, machining and hardening of the gears can be done at one production centre.

In this research 3 different types of prototype automotive gears are manufactured, representing several geometrical features of a gear as well as materials.

Advantages of laser hardening

• By using an integrated laser hardening setup, uniform hardness profiles were achieved on the different features, while all measured quality parameters were within the required tolerances.
• Transportation logistics, re-clamping, re-aligning are no longer needed, and the lead time and energy consumption during gear production are significantly reduced.
• Due to the elimination of large energy consuming furnaces and transportation, laser hardening has a lower environmental impact compared to conventional hardening.
• In addition, the laser hardening processes is a self-quenching process which does not use water or oil, as the (cooler) material around the laser hardened processing zone is sufficient to support the martensitic transformation.
• Further, laser hardening, often done in a selective manner, only heats up the workpiece locally, avoiding large deformations. This results in a minimum number (often none) of hard finishing operations.

Three prototypes

For this research 3 different types of prototype automotive gears are manufactured, representing several geometrical features of a gear as well as materials.

The first industrial case (a) tested is a C45 straight cut transmission gear, shown in figure 1a, with a diameter of 98 mm and a thickness of 16 mm used in the drivetrain of sports cars. In this case, all the gear flanks need to be hardened to a depth of 150μm.

The second industrial case (b), a C45 bias gear (figure 1b), having the same chemical composition as the first case, is an auxiliary gear/spring combination, mounted onto a neighbouring gear. The function of the bias gear is to remove backlash from the automotive gear train. Since the function of this gear is not to transfer power but to counteract backlash, the thickness of this gear is only 4 mm. The contact areas indicated in orange sustain repetitive metal-to-metal contact and need to be trough hardened to prolong the lifetime of the gear. The spring area requires sufficient yield strength to avoid plastic deformation during the spring loading.

The third industrial case (c), a 42CrMo4-sprocket gear (diameter: 120 mm, thickness: 5 mm) is connected to the camshaft by a timing chain to synchronizes the rotation of the crankshaft with the engine cylinders intake and exhaust strokes. The sprocket-gear requires hardening of the teeth and of the rim on the backside of the gear to a depth of 200μm.

Both the gear of the first case and of the third case have been pre-machined on the same machining center as where the integrated laser hardening operation will be performed. Despite that the gear of the second case has not been made on the same machining center, the integrated laser hardening can still prove beneficial to harden this gear. Due to its limited thickness, conventional hardening techniques would deform the bias gear considerably resulting in the need for additional finishing operations. These finishing operations, involving, milling, profile grinding and flat grinding multiply the cost of these gears by a factor 10. Initial testing done, indicated that all the machined gears were out of tolerance and required finishing operations after hardening them in a furnace.

Test equipment

In this research, the laser hardening operation has been integrated into a DMG MORI NTX2000 Mill turn center. This is achieved using a laser head which can be picked up as a regular tool (using a Capto tool holder). The laser source used in the developed setup is a 500W (High Power Diode Laser, HPDL) with a wavelength of 940 nm. With the use of an optical fiber (core diameter: 600 μm), the laser light is transmitted from the laser source towards the machining center.

Most flanks of automotive gears are wider than the spot size used for laser hardening. Hardening the full width of the flank of a tooth requires multiple adjacent laser tracks, which may result in a non-uniform hardness. This can be avoided by a scanning motion of the laser spot achieved by deflecting the laser beam using a mirror. In the set-up of this research, the mirror is integrated inside the laser head. This scanning motion omits the use of multiple tracks to harden large surfaces and ensure a uniform hardness over an 18mm wide track to a depth of 0,3mm on C45 steel.

Laser hardening in action

During the laser hardening operation, especially of shaped surfaces such as gear flanks, the laser light has to stay in focus over the entire profile of the gear tooth. To achieve this, a laser programming software calculates the corresponding machine axes values. This software allows the user to select certain sections of the gear to be hardened. As an example, only the flanks of the gears can be selected for hardening, while other parts of the gear will be left untreated.

Due to the profile of a tooth flank, the angle of incidence of the laser light changes during laser hardening, resulting in a change of the absorbed laser power and hence the quality and hardening depth of the hardened layer. To compensate for this absorption change a feed-forward power regulation function is implemented.

The analysed samples are cut by a wire-EDM machine, embedded in a polymer, ground, polished and etched. All hardness tests performed on these samples are Vickers micro hardness tests with a 300g load. The hardness measurements on gears are performed on pitch radius of the gear. The maximum hardness measured, is the maximum hardness measured starting from 50μm below the surface of the sample. Hardness on the surface can be higher than the maximum measured hardness in this research. The hardening depths are measured on the cross-section of the hardened zone.

Results

Laser hardening on gear flanks

First trials on the straight cut C45-gear showed that it is impossible to heat treat the undercut region of the gear flanks. This has been solved by moving the laser head (Y-axis) out of center of the spindle axes. This combined with the required spindle rotation, a more uniform laser beam incidence is obtained over the whole tooth profile. The left position hardens the right gear flanks, while the right position hardens the left gear flanks. This offset strategy is also implemented into the programming software.

Initial testing proved that an offset from the center of 12mm (Y direction), with feed-forward power regulation combined with feed-backward control, reached the best results. The programming software estimated the variation of required applied laser power over the profile of the gear tooth between 420W and 490W. Results showed that a uniform hardened layer, with a maximum hardness of 645 HV at the surface and a hardness of 551HV at a depth of 115μm, was formed across the profile and flank profile of the gear tooth. Essential for uniform tooth hardening is to harden the tooth flanks from the top to the root circle (top-down strategy).

(a) Bottom-up strategy; (b) Top-down strategy; (c) High feed/High power density hardening; (d) Hardened 42CrMo4-sprocket

In the above figure (a) and (b) show the hardened zone of a bottom-up and a top-down hardening strategy using the above-mentioned parameters. Using a bottom-up strategy (i.e. laser moves from the root to the top circle), heat is accumulated in the top of the tooth resulting in an annealing of the opposing tooth flank that was hardened before (figure a). The hardness measured at the annealed side dropped to an average of 400HV. A small overlap region with reduced hardness is detected near the root of the tooth, resulting from the stop location of the last hardened flank.

(d) shows the hardened layer achieved on the sprocket gear using the above-mentioned parameters: the hardened layer had a maximum hardness of 669HV at the surface of the layer and a hardness of 570HV at a depth of 290μm. Both overall hardness and hardening depth meat the required specifications for this type of gear.

Additional experiments were performed to test the maximum reachable hardening depth without annealing the other side of the tooth, using the top-down strategy. Very high feed rates are required to prevent the tooth from reaching the annealing temperature. Following these experiments, a maximum hardening depth was reached using a feed rate of 500mm/min, a scanning frequency of 50Hz, a scanning width of 7mm and a power of 500W. Reducing the scanning width to 7mm was required to gain sufficient power per surface area to harden with these high feed rates. To implement this strategy to the full scanning width of 16mm, a higher laser power then 500W is required. With these parameters, a maximum hardness of 660HV was achieved at the surface of the hardened layer and a hardness of 526HV was reached at a depth of 352μm.

To harden the flanks of the sprocket wheel, initial experiments on flat bars of 42CrMo4 were performed. The results of these experiments showed that a hardening depth of 0.3mm could be reached with a scanning width of 5mm (= tooth width), a scanning frequency of 120Hz, a feed rate of 700mm/min. Solutions found for hardening gears in the C45-gear case were implemented. An offset position of the laser head of 8 mm (instead of 12 mm in case of the C45 gear), for a more uniform incident angle, in combination with a top-down hardening strategy was used. The feed rate used for this gear is significantly higher than the feed rate used in the C45 case. This is to prevent annealing of one side of the gear tooth due to the fact that the tooth is smaller at the top and thereby heat up faster compared to the C45 case. Due to the more uniform angle of incidence of the laser light, and thus a more consistent hardening process compared to the C45 case, a feed-back control system was sufficient to control the process.

Laser hardening of features with a constant angle of incidence

The rim feature (42CrMo4-sprocket gear, figure 1d) and the metal-to-metal contact feature (C45 bias gear, figure 1b) are hardened by focusing the laser beam on the surface features to be hardened but kept parallel to the gear axes. Hardening with a constant angle of incidence does not require additional control systems. A fixed power and feed rate strategy is used to harden these features. The thickness of the gear (and also the features) is small enough to have a through hardening. To harden the rim of the 42CrMo4-sprocket gear, the following process parameters are used: fixed power: 500W; fixed feed rate: 1000mm/min, scanning width: 5mm and scanning frequency: 120Hz.

As a result of non-uniform heat dissipation throughout the rim, different hardening depths occurred over the width of the rim. Heat concentration on the outer side of the rim allowed more material to reach the austenization temperature resulting in larger hardening depths (maximum hardness of 683HV at the surface and 555HV at a depth of 340μm). The other rim side (close to the gear disk) had a maximum hardness of 715HV at the surface and a hardness of 565HV at a depth of 270μm. The higher hardness achieved on the left side is due to the higher cooling rates achieved in this section of the rim.

To harden the metal-to-metal contact points of the C45-bias gear, flat samples with the same thickness, were tested. The experiments resulted in the following process parameters for optimal hardening: fixed spot size: 2.5 mm, laser power of 360W and feed rate of 50 mm/min. A low feed rate is required to allow the full depth of the gear to reach the austenization temperature. The hardened zone is the smallest at the bottom of the gear (opposite side of the laser beam). In this region, a maximum hardness of 635HV was measured at the contact point and a hardness off 450HV at a depth of 0,9mm. Both overall hardness and hardening depth meat the required specifications for both applications.

(a) Hardened 42CrMo4 rim (left side of the rim is close to the gear disk); (b) Cross section metal-to-metal contact of the C45 bias gear

Improving spring properties through laser hardening

To analyse the spring properties of the bias gear (figure 1b), beam-shaped samples with the same width and thickness as the spring are investigated. It is essential for the spring to remain within the elastic region during load. Using a 3-point bending, samples with both single spot and scanned laser hardening strategies were tested. Initial experiments proved that the yield strength of the beams increases in relation to the hardening depth.

None of the hardening methods using the 500W laser (implemented on the multi-axis machining platform) reached the required yield strength specified by the gear manufacturer. Therefore, additional tests were performed on a similar fixed spot integrated hardening setup, using a 1.1 kW diode laser. With this higher laser power, the required yield strength could be reached using a fixed spot (5mm) hardening strategy with a feed rate of 200 mm/min.

Dimensional measurements and finishing requirements

The straight cut C45-gears were measured along with the profile and trace, of 4 different teeth, of both left and right flanks. All measured gear quality parameters of laser hardened straight cut C45-gear hardened to a depth of 115μm stayed within the required specification, no significant deformation was detected for the gear. No additional finishing operations were required. Deformation outside the specified requirements was detected in both profile and trace variation of the straight cut C45-gear hardened to a depth of 350μm. Although this gear needs additional finishing operations, the time to market and costs compared to conventional hardening will be reduced. These finishing operations can be performed in the same machining center and can be programmed using conventional CAD/CAM software directly after the laser hardening operation. Transportation to and from a hardening facility, re-clamping, re-aligning are eliminated in the process chain.

No measurable deformation was found after laser hardening the 42CrMo4 sprocket gear The flatness, roundness and the gear functional dimensions showed no significant variations after laser hardening the bias-gear and no additional finishing was required.