Impeller Machining: Five-Axis CNC Process

Impellers are widely used in aircraft engines, gas turbines, and fluid machinery. The spatial variations in their blade surfaces are complex, making machining particularly challenging.

With the development of high-end equipment manufacturing, impeller machining now demands higher standards for surface accuracy and surface quality.

Five-axis CNC machining can achieve complex surface profiling through coordinated multi-axis motion;

However, in actual machining, the machining results are still prone to being affected by path changes, improper tool axis adjustments, and incorrect parameter settings.

Therefore, it is necessary to review and improve machining processes in light of the structural characteristics of impeller surfaces, thereby providing a reference for enhancing the quality of impeller surface machining.

Machining Characteristics and Precision Requirements of Impeller Surfaces

• Structural Characteristics of Complex Surfaces

Impeller blades are typically three-dimensional, twisted, thin-walled structures.

The pressure and suction surfaces vary continuously along the blade height, with tight transitions between the leading edge, trailing edge, and root fillet, resulting in narrow flow channels.

The curvature of the blade surface varies unevenly in both the radial and circumferential directions, and local areas may also exhibit torsion and deflection, making it difficult to approximate these geometric features with simple surfaces.

The blades themselves have relatively thin walls and are prone to elastic deformation under the combined effects of cutting forces and heat; even slight vibrations can be amplified into profile errors.

Furthermore, the connections between the hub and the blades are typically small-radius fillets, which restrict the space available for tool entry and exit.

The combination of these factors makes impeller surfaces one of the more challenging part types in both CNC programming and on-site machining.

• Suitability for Five-Axis Machining

When machining impellers, five-axis machining centers can use two rotary axes to adjust the tool axis orientation, ensuring that the side edge of the tool always maintains an optimal angle of incidence close to the blade surface.

This allows for continuous machining of most blade areas within a single setup, reducing the accumulation of errors caused by station changes;

The tool axis orientation can be adjusted according to changes in the surface’s orientation, which helps avoid areas prone to interference—such as the leading and trailing edges—and reduces the concentration of local cutting forces.

For areas with deep flow channels and rapid changes in curvature, appropriately adjusting the tool axis inclination improves chip evacuation and cooling conditions.

Overall, five-axis machining is better suited for the forming of complex surfaces like impellers in terms of accessibility, contour conformity, and reduction of clamping errors.

• Key Aspects of Surface Accuracy Control

Surface accuracy control for impellers focuses on the profile, blade thickness, leading and trailing edge shapes, and root transitions.

Since blade curvature varies continuously along the blade height, the tool contact position and actual cutting depth change during machining, making local overcutting or undercutting likely;

Thin-walled blades have low rigidity; under the influence of cutting forces and heat, they may experience deflection and springback, leading to deviations in blade thickness and height.

The leading and trailing edges are small in size and exhibit concentrated changes in curvature, placing higher demands on toolpath continuity and cutting stability; even slight vibrations can easily result in tool-to-surface marks or contour deviations.

The space around the blade root fillet and the hub connection area is narrow, limiting the tool’s range of motion; this is a high-risk area for interference and residual errors.

Therefore, machining should focus on maintaining uniform machining allowances across all areas, smooth toolpaths, and controlled deformation of thin-walled blades.

Operators should also maintain consistent surface roughness and dimensional accuracy between blades, providing a solid basis for subsequent process parameter optimization and toolpath adjustments.

Process Optimization

• Toolpath Planning

Impeller blades have long profiles and rapid changes in curvature; if toolpaths are improperly planned, steps and irregular patterns are likely to form at tool transition points and areas of abrupt curvature changes.

During actual programming, the impeller is typically divided into three regions: the main working surface of the blade, the inlet and outlet transition zones, and the root fillet area.

Roughing, semi-finishing, and finishing toolpaths are set up separately for each region.

In the central region of the blade, where curvature changes gradually, isoparametric or constant residual height side milling paths are prioritized.

This ensures that the cutting direction aligns with the blade’s development plane, guaranteeing continuous surface texture.

In areas such as the leading edge, trailing edge, and flow channel bends, the step size must be reduced, and “edge-hugging” machining in the direction of the tool path should be employed to minimize contour errors caused by path turns.

Typical path schemes are shown in Table 1; further refinements can be made based on machine tool performance and tool rigidity.

Table 1. Toolpath Arrangement for Different Impeller Machining Regions

• Tool Spindle Orientation Adjustment

The tool spindle orientation determines the contact state between the tool and the surface and is one of the key factors in controlling the accuracy of the impeller profile.

Generally, a reference tilt angle is selected to ensure that the tool cuts primarily with its side edge, followed by local fine-tuning based on different areas.

On the one hand, appropriately increasing the tool spindle’s rake angle can reduce the extent to which the tool tip participates in the cutting process, thereby minimizing tool indentations;

On the one hand, in areas with deep flow channels and sharp leading edges, the tool must be steered using a lateral tilt angle to ensure that the tool envelope matches the target surface while avoiding interference.

In actual programming, engineers can set the “maximum orientation change angle” and “minimum allowable tilt angle” in CAM software.

These limits help reduce servo tracking errors caused by excessive orientation changes.

For the junction between the blade root and the hub, the orientation change is performed in stages by adding transition tool positions, thereby reducing localized “kinked” trajectory patterns and ensuring a smooth toolpath.

• Coordination of Cutting Parameters

During the machining of impeller surfaces, cutting parameters must be reasonably set based on three key practical requirements: machine tool load capacity, blade vibration suppression, and surface scratching control.

Generally, a reasonable cutting speed is first determined based on the material and tool diameter, and then the spindle speed and feed rate are calculated accordingly.

A commonly used relationship is

Where:

• v_c is the cutting speed;
• D is the tool diameter;
• n is the spindle speed.

A higher v_c can be used for aluminum alloy impellers, while it should be appropriately reduced for nickel-based superalloys.

The corresponding spindle speed is calculated using Equation (1) to ensure the tool operates within the recommended speed range.

› Feed Rate Calculation

Once the spindle speed is determined, the feed rate v_f is calculated based on the feed per tooth f_z using the following formula:

Where:

• v_f is the feed rate; f_z is the feed per tooth;
• z is the number of teeth on the cutting tool.

› Parameter Adjustment by Machining Stage

During the roughing stage, f_z and the radial depth of cut should be appropriately increased to improve material removal rate;

For semi-finishing, the depth of cut and feed per tooth should be reduced to ensure a more uniform distribution of the remaining material.;

During the finishing stage, a higher spindle speed should be maintained while reducing f_z and the axial depth of cut.

This, together with the earlier-determined step size, reduces cutting forces and minimizes the heat-affected zone.

For thin-walled blades, a “multiple light cuts” approach can be adopted during finishing; for example, dividing a planned single cut depth of 0.5 mm into 2 to 3 passes, and reducing the feed rate by 10% to 20% in critical areas.

Although these parameters are somewhat conservative, they significantly reduce the impact of deflection and springback on surface accuracy, which helps maintain stable surface quality during mass production.

• Machining Control of Critical Areas

The leading edge, trailing edge, root fillet, and narrow flow channels of the impeller are areas where errors tend to accumulate and therefore require separate control.

For the leading and trailing edges, it is advisable to use independent finishing toolpaths and small-diameter ball-nose or form cutters, with feed rates reduced by 30%–40% compared to the blade center section and step sizes controlled between 0.05 and 0.10 mm.

If necessary, add sweeping cuts along the leading edge direction to minimize tool marks.

The blade root fillet has low stiffness and is prone to stress concentration; a layered, light-cutting approach can be used, dividing the total cutting depth into 2 to 3 passes, and incorporating soft-in and soft-out sections at the start and end of the fillet to reduce springback.

For narrow flow channels, schedule an intermediate inspection after rough machining; adjust the finishing allowance and feed pattern based on the cross-sectional dimensions and fillet radius, and perform make-up cuts if necessary.

Tools, parameters, and inspection requirements for critical areas should be listed separately on a process card and strictly followed during batch production.

Process Optimization and Validation

• Experimental Platform and Impeller Test Specimens

To verify the applicability of the aforementioned process scheme, a five-axis vertical machining center equipped with a swivel head and rotary table was selected as the experimental platform.

The machine features a maximum spindle speed of 18,000 r·min⁻¹, along with a high-pressure cooling system and an in-process tool measurement device.

The CNC system supports five-axis interpolation and spatial tool axis control, and features post-processing interfaces for common impeller machining, facilitating the execution of toolpaths according to predetermined paths and orientations.

The fixture employs a combination of a specialized expansion sleeve and face positioning to secure the monolithic impeller blank on the rotary table via its center hole, while auxiliary supports are installed to minimize deflection during machining.

The impeller test specimen features a common monolithic structure with a diameter of approximately 120 mm, 12 blades, and a blade height of 30 mm.

The blades have a three-dimensional twisted surface, and the transition between the hub and the blades features a small fillet.

The material is 7075-T6 aluminum alloy, and the blank is rough-turned from a forging to form the outer shape with an allowance for machining.

Based on the geometric characteristics of the test specimen, areas such as the mid-section curved surface of the blades, the leading and trailing edges, the root fillets, and the narrow flow channels are designated as key observation points for subsequent accuracy and surface quality assessments.

• Optimizing the Process Implementation

Based on the 3D model and process reference, a set of reference toolpaths was first generated according to the workshop’s existing practices: rough machining used conventional contour-following toolpaths, while the finishing of the blades primarily employed simple side milling with constant parameters; the tool axis orientation and cutting parameters were set based on previous similar parts.

This approach is designated as the “original process set.” Subsequently, following the path planning, tool axis orientation adjustments, cutting parameter coordination, and key area control strategies proposed in this paper, adjusted toolpaths for the “adjusted process set” were generated on the same computer-aided manufacturing (CAM) platform:

For the blade’s central section, side milling with a constant residual height was employed;

independent finishing programs were created for the leading and trailing edges and the blade root fillet; local re-machining paths were added to narrow flow channels;

and upper and lower limits, as well as maximum rates of change, were set for the tool axis tilt and side tilt angles to ensure smoother attitude transitions.

Regarding cutting parameters, both groups maintained consistency during the roughing stage to minimize interference from variables;

During the semi-finishing and finishing stages, the spindle speed and feed rate are re-matched based on calculated cutting speeds and feed per tooth, and the finishing depth of cut is divided into multiple shallow cuts.

The machining sequence involves first completing the impeller using the original process group, then machining the impeller using the adjusted process group under the same fixture and environmental conditions, while simultaneously recording spindle load, vibration alarms, and the time per piece for each process step to provide a basis for subsequent result comparisons.

• Comparative Analysis of Machining Results

After machining both sets of impellers, a coordinate measuring machine (CMM) was used to measure the blade leading edges, root radii, and mid-section profiles, extracting the maximum contour error and root mean square (RMS) error;

Surface roughness Ra was measured at the center of the pressure face and in the root transition zone;

The per-piece processing time for the finishing stage was obtained from machine tool records. The comparison results of the impeller test specimens are shown in Table 2.

Table 2. Comparison of Machining Results for the Impeller Test Pieces

› Profile Accuracy and Surface Quality

As shown in Table 2, the maximum contour error at the blade leading edge and blade root fillet positions in the process-adjusted group was significantly lower than that in the original process group.

The root mean square contour error also decreased from 12.2 μm to 7.6 μm, indicating better overall conformity.

In terms of surface roughness, the Ra values for both the pressure face and the root transition zone improved, and issues with localized tool marks and irregular patterns were mitigated.

› Blade Deformation and Machining Efficiency

The reduced variability in blade height and thickness indicates that multiple light cuts and speed reduction at critical locations help control deformation of thin-walled blades.

Although finishing time increased slightly—from 42.3 min to 45.1 min—the machining accuracy of the curved surfaces and surface quality were significantly improved, making the process fully engineering-acceptable under existing production cycle constraints.

Conclusion

This paper focuses on the five-axis CNC machining of impeller surfaces.

Based on structural characteristics and precision requirements, it outlines key process aspects such as toolpath planning, tool axis orientation, cutting parameters, and machining of critical areas, and validates these through comparative testing on test specimens.

The results indicate that this process scheme can reduce contour errors and surface roughness while maintaining essentially controllable machining time, thereby improving consistency among the blades.

It provides valuable reference for developing on-site machining processes for similar impeller components.