The electric drive system serves as the core powertrain unit in new energy vehicles (NEVs), and its performance and reliability are paramount. Within this system, the precise control of axial clearances or interference fits in the transmission assembly directly impacts critical attributes such as gear mesh alignment, bearing life, noise, vibration, and harshness (NVH), and ultimately, the overall efficiency and durability of the drivetrain. The process of selecting the correct thickness shim for bearings is a fundamental step in achieving this precise control during assembly.
Historically, shim selection relied on manual trial-and-error methods or costly, offline coordinate measuring machine (CMM) inspections, leading to inefficiencies and inconsistent results in controlling final axial gear train float. The challenge is particularly pronounced for assemblies using tapered roller bearings, where the “working clearance” is influenced by the axial preload applied to the non-integral bearing cup and cone set. Therefore, modern assembly lines for high-volume electric drive system production have adopted in-line measurement techniques. These methods measure the installed height of bearing components and the corresponding depth of the housing, using these values to calculate and automatically recommend the required shim grade. The accuracy of this automated selection is vital. This paper investigates the factors influencing the precision of this bearing shim selection process, analyzes the key error contributors, and proposes methodologies for optimization to ensure the electric drive system operates within its designed parameters.
The current industry-standard process for shim selection in electric drive system assembly, as implemented for a typical 180kW system (referred to here as System S180), involves an automated “measure-calculate-recommend” cycle. The target assembly is often the differential carrier assembly, which uses tapered roller bearings pressed into an aluminum housing. The process requires selecting a shim that sits between the bearing cup and a shoulder in the rear housing. The fundamental measurement principle is outlined below.
The axial clearance (or the space to be filled by the shim) is determined by measuring two key dimensions and applying known compensations. An in-line station measures the distance from the face of the bearing cup (pressed onto the differential carrier) to the mating surface of the main housing. This is denoted as dimension \(X_1\). Concurrently, another station measures the depth from the same housing mating surface to the seat for the bearing cup in the rear housing cover. This is denoted as dimension \(X_2\).
The theoretical gap \(X_3\) before shim insertion is then calculated by accounting for the deformation under specified preload. The bearing assembly and the housing material exhibit slight elastic deformation when the final axial preload is applied. This deformation is characterized as a stiffness compensation value for the bearing set (\(K_1\)) and for the housing (\(K_2\)). The formula for the required shim thickness or the nominal gap is:
$$X_3 = X_2 – X_1 – K_1 – K_2$$
Where:
\(X_3\) = Calculated assembly gap / Required shim thickness
\(X_2\) = Measured housing depth
\(X_1\) = Measured bearing cup height
\(K_1\) = Bearing stack stiffness compensation under design preload
\(K_2\) = Housing stiffness compensation under design preload
The calculated value \(X_3\) is compared against a predefined binning table, which maps gap ranges to specific shim thickness grades. The system then automatically recommends the correct shim for assembly. The accuracy of the final shim selection hinges on every variable in this equation and every step in the preceding and subsequent processes.
Key Factors Influencing Shim Selection Precision
The precision of selecting the correct shim for the electric drive system bearing is not governed by the measurement system alone. It is a systemic outcome influenced by the state of components before measurement, the stability and accuracy of the measurement process itself, and the integrity of data handling and matching post-measurement. A failure in any of these stages can lead to an incorrect shim being installed, compromising the performance of the electric drive system.
| Process Phase | Key Influence Factors | Potential Impact on Shim Selection |
|---|---|---|
| Pre-Measurement (Part Preparation) | Bearing press-fit quality onto shaft. Bearing cup press-fit quality into housing. Dimensional conformity of components. |
Incorrect installed height (\(X_1\)). Non-uniform cup face, tilted axis. Introduces error in baseline dimensions. |
| Measurement Process | Measurement System Analysis (MSA) capability. Sensor calibration and repeatability. Clamping force and measurement preload consistency. Data point fitting algorithm. |
Direct error in \(X_1\) and \(X_2\) values. Unstable readings, high variation. Alters the simulated preload state, affecting \(K_1, K_2\). Inaccurate plane derivation from points. |
| Post-Measurement (Data & Selection) | Correct data binding (part to measurement). Accuracy of compensation values \(K_1, K_2\). Precision of gap-to-shim binning logic. |
Wrong data used for calculation. Systematic offset in \(X_3\). Misclassification of required shim grade. |
Pre-Measurement: Study of Bearing Press-Fit Conditions
The condition of the sub-assemblies before they reach the measuring station is the first critical link in the chain. For the electric drive system under study, this involves two key press-fit operations: the tapered roller bearings onto the differential carrier shaft, and the bearing cups into the main aluminum housing.
1. Tapered Bearing Press-Fit onto Shaft: This operation must ensure the bearing inner race is seated squarely against the shaft shoulder with the correct interference fit. An excessive press force can cause plastic deformation of components, while insufficient force may lead to creep under load. More critically, if the bearing is pressed on tilted or does not seat fully around the entire circumference, the effective axial distance between the two bearing cups (the “spacer length”) changes. This directly alters the \(X_1\) measurement. Monitoring the press force vs. displacement curve is essential. A nominal curve shows a steady rise until the bearing contacts the shoulder, followed by a sharp increase in force. Deviations indicate potential issues like misalignment or contamination. For the S180 system, the press force is monitored to be within a window (e.g., final force >18kN but within a safe limit to avoid damage), and a 100% check for shoulder contact gap is performed.
2. Bearing Cup Press-Fit into Housing: The cup must be pressed squarely and completely into its bore until it bottoms out against the housing shoulder. A tilted cup creates a non-uniform running surface for the bearing rollers, leading to premature wear and anomalous axial movement, which undermines the stability the shim is meant to control. The press curve for this operation is monitored for a similar “hard-stop” signature indicating full contact. Consistent and verified press-fit processes ensure that the datum face of the cup, from which \(X_1\) is measured, is in its correct and repeatable position. Our analysis confirmed that a well-controlled press process eliminates this as a major source of variation for the subsequent shim selection in the electric drive system.
Measurement Process: In-Depth Analysis of System Stability
The heart of precision shim selection lies in the capability of the in-line measuring stations. These stations, one for bearing cup height (\(X_1\)) and one for housing depth (\(X_2\)), must provide highly repeatable and accurate data. Their performance is validated through a comprehensive Measurement System Analysis (MSA).
First, the clamping force (which holds the assembly) and the measurement preload force (which removes internal bearing play) are calibrated and verified for stability. These forces must be consistent to ensure the mechanical state during measurement matches the assumptions in the stiffness compensations \(K_1\) and \(K_2\). A sample verification is shown below:
| Measurement Cycle | Preload Force (N) | Clamping Force (N) |
|---|---|---|
| 1 | 502 | 5324 |
| 2 | 501 | 5330 |
| 3 | 502 | 5325 |
| 4 | 499 | 5327 |
| 5 | 499 | 5326 |
| Range | 3 | 6 |
| Specification | 500 ± 20 | >5000 |
The core MSA for the measuring equipment involves the following steps:
1. Equipment Zeroing and Master Part Calibration: Certified master parts (small and large value) with known dimensions traceable to national standards (e.g., CNAS) are used. The sensor readings are zeroed against the “true value” (Nominal) of the small master. The process capability of measuring this master repeatedly (Type 1 study) is evaluated. The key index, often the Capability Index (\(C_g\) or \(C_{gk}\)), must exceed 1.67, indicating excellent measurement capability relative to tolerance.
$$C_g = \frac{0.2 \cdot T}{6 \cdot s}$$ where \(T\) is the tolerance and \(s\) is the standard deviation of measurement variation.
2. Verification Across Measurement Range: The large-value master is then measured to confirm accuracy across the expected working range of the electric drive system components. A repeatability study on this master further validates system stability.
3. Correlation with CMM Benchmarking: The ultimate test is correlating the in-line gauge results with a trusted benchmark. A sample of production parts (differential carriers with cups and bare housings) is measured using a high-precision CMM. For bearing height (\(X_1\)), a special fixture that applies the same design preload is used with the CMM to replicate the in-line gauge’s mechanical state. The CMM-measured value is treated as the reference “true value.” The in-line gauge results are then compared against these values. A Gage R&R (Repeatability & Reproducibility) study is conducted on multiple parts measured multiple times by multiple operators (if applicable) on the in-line system. Key metrics are evaluated:
- %EV (Equipment Variation): Should be low.
- %AV (Appraiser Variation): Ideally negligible for automated systems.
- %GRR (Total Gage R&R): The primary metric. According to AIAG standards, %GRR < 10% is considered excellent, < 30% may be acceptable depending on application. For critical electric drive system shim selection, targeting < 10% is necessary.
- Number of Distinct Categories (ndc): Should be >= 5.
4. Optimization of Measurement Methodology: A significant finding during MSA for tapered bearing cup measurement is the effect of cup face runout or minor tilts. A static measurement at three fixed points can be susceptible to errors if a point lands on a high or low spot. An optimized method involves a dynamic measurement: the differential carrier is rotated at a constant speed, and sensor data is sampled over a stable period (e.g., the last 5 seconds of rotation). Software filters outliers and fits a plane to hundreds of data points, yielding a highly stable and representative average cup face height (\(X_1\)). This method significantly improves repeatability and reduces measurement noise, directly enhancing shim selection confidence for the electric drive system.
| MSA Metric | Result (Example) | Acceptance Criteria | Interpretation |
|---|---|---|---|
| %GRR (Total Variation) | 8.5% | < 10% (Excellent) | Measurement system variation is a small fraction of the process variation, suitable for precise shim selection. |
| Number of Distinct Categories (ndc) | 12 | >= 5 | The measurement system can adequately distinguish between different part variations. |
| Correlation with CMM (R²) | > 0.98 | Close to 1.0 | In-line gauge results show excellent linear agreement with the CMM reference values. |
Post-Measurement: Data Integrity and Shim Matching Logic
With accurate and stable measurement data for \(X_1\) and \(X_2\), the final step is flawless data processing and shim grade assignment. This phase involves several critical, yet often overlooked, steps that are vital for the electric drive system assembly.
1. Data Binding and Traceability: It is imperative that the measured \(X_1\) from a specific differential carrier assembly is correctly paired with the measured \(X_2\) from the specific housing it will be assembled into. Barcode or RFID tracking throughout the line ensures this one-to-one data matching. A mix-up at this stage guarantees an incorrect shim.
2. Accurate Application of Compensations: The stiffness compensation values \(K_1\) and \(K_2\) are not arbitrary; they are derived from engineering design calculations and physical tests (e.g., load-deflection tests on the bearing and housing). Using incorrect or outdated compensation values will introduce a systematic bias into every \(X_3\) calculation. The formula must be consistently and correctly applied:
$$X_{3\text{ (calculated)}} = X_{2\text{ (measured)}} – X_{1\text{ (measured)}} – K_{1\text{ (design)}} – K_{2\text{ (design)}}$$
3. Shim Binning Logic: The calculated \(X_3\) value represents the theoretical gap. The available shims come in discrete thickness grades (e.g., 1.20 mm, 1.25 mm, 1.30 mm…). The binning table defines which \(X_3\) range corresponds to which shim grade to achieve the final, desired assembly preload or clearance. This logic must be meticulously aligned with the design specification. For instance, if the target is a slight preload, the shim chosen will be slightly thicker than \(X_3\). The rule must be unambiguous and correctly programmed into the selection system.
4. Verification of Selection: As a final check, the shim selection process can be verified by manually calculating the expected final gap/preload using the “true” CMM-measured dimensions and the actual shim thickness. This predicted outcome should match the design intent. A sample verification is shown below, confirming that the automated system’s recommendation aligns with a manual calculation based on correlated data.
| Assembly Sample | Gauge-Calc Gap \(X_3\) (mm) | Manual Calc. from CMM Data (mm) | System Recommended Shim | Manual Shim Selection | Match? |
|---|---|---|---|---|---|
| 1 | 1.2495 | 1.2495 | 1.40 mm | 1.40 mm | Yes |
| 2 | 1.2101 | 1.2101 | 1.35 mm | 1.35 mm | Yes |
| 3 | 1.1973 | 1.1973 | 1.35 mm | 1.35 mm | Yes |
| 4 | 1.2135 | 1.2135 | 1.35 mm | 1.35 mm | Yes |
| 5 | 1.1543 | 1.1543 | 1.30 mm | 1.30 mm | Yes |
Conclusion
Ensuring the precision of bearing shim selection is a holistic endeavor critical to the performance of the new energy vehicle electric drive system. It requires a closed-loop, systemic approach that controls quality at every stage:
1. Foundation in Part Preparation: Robust and monitored press-fit processes for bearings and cups are essential to establish correct and repeatable datum geometries for measurement.
2. Cornerstone of Measurement: A rigorous and ongoing Measurement System Analysis (MSA) program is non-negotiable. This includes force calibration, master part correlation, and dynamic measurement techniques to mitigate errors from component runout. The measurement system for the electric drive system must demonstrate excellent %GRR (<10%) and strong correlation with reference methods.
3. Integrity of Data Processing: Flawless part-to-data tracking, accurate application of design-based compensation values (\(K_1, K_2\)), and precise shim binning logic are the final determinants. They translate stable measurement data into a correct action (shim selection).
By methodically addressing each of these influence factors—validating pre-measurement part quality, optimizing and certifying the measurement system’s stability, and ensuring data handling integrity—the assembly process can achieve highly precise and reliable bearing shim selection. This precision directly translates into optimal axial positioning of the gear train, minimizing unwanted axial float, reducing NVH, improving bearing life, and maximizing the efficiency and durability of the electric drive system. This research provides a validated framework for quality control in electric drive system assembly, with methodologies applicable to other precision clearance-control operations within the powertrain and beyond. As the industry evolves towards higher speeds and power densities, the principles of systematic error control in shim selection will remain fundamentally important.