1. Introduction
The high-pressure pump is the centerpiece of the common rail system for diesel cars and is its most important component. It is driven by the engine crankshaft via a toothed belt connected to gear wheels or a chain at a 1:1 gear ratio, and its main role is to compress the fuel to pressures in excess of 2000 bar. Fuel under pressure is delivered from the pump to the engine combustion chamber through the high-pressure pipes, the high-pressure rail, and the injectors [
1].
Over time, there have been various constructive solutions for high-pressure pumps, depending on the motion transmission mechanism. There are cam–follower and roller–shoe transmission mechanisms. The roller–shoe mechanism is notable for its effective performance and economical production costs. Therefore, it has become a favored solution that captures the attention of all automotive manufacturers. The roller, identified as a rolling component within this mechanism [
2], has a cylindrical shape, with its profile being an important factor in wear reduction. This is crucial because, upon contact with the work surface, maximum pressure is applied to the ends. Manufactured from a wear-resistant material characterized by high hardness, it exhibits excellent resistance in high temperature scenarios dictated by the operational requirements. The subsequent component in the mechanism, namely the shoe, is composed of a robust material that is perfectly compatible with the roller. Designed to withstand shocks and crushing, the shoe also facilitates the optimal adhesion of the lubricating fluid. Its durability and hardness are enhanced through the application of a Diamond-Like Carbon (DLC) coating [
3]. The camshaft serves a pivotal role in transmitting the movement of the roller–shoe mechanism, maintaining constant contact with the roller. This is a critical component due to its continuous interaction with the roller. Given the stringent performance requirements of the high-pressure pump, its components experience increasingly heavy loads, resulting in severe wear. Over time, various strategies have been developed to improve their efficiency, including modifications in size, materials, or surface characteristics [
2]. These modifications aimed to mitigate wear on the high-pressure pump.
The roller–shoe motion transmission mechanism is the central subassembly of the common rail high-pressure pump. Its wear is important because it can lead to the total failure of the pump. For this reason, it is necessary to find solutions to increase the lifetime of its components. By examining the operation of the motion transmission mechanism, along with its construction and overloading, various approaches are explored. These include the study of lubrication, coatings, and surface profiles.
An essential factor influencing the wear of the mechanism is the quality of the lubrication. This is important to minimize energy loss, heat generation, the wear of the mechanical components, as well as smooth machine operation. Production and productivity are highly dependent on the lubrication system [
4]. The lubrication process in high-pressure pumps involves the use of fuel, specifically diesel in our context. The efficacy of diesel fuel in preventing or mitigating wear on the contacting components is diminished, primarily due to its low viscosity. Contact surfaces within high-pressure pumps are safeguarded against wear through the transition from hydrodynamic lubrication to boundary lubrication regimes. In the hydrodynamic lubrication regime, a fluid film acts as a barrier, preventing direct contact between the surfaces [
5]. It represents an ideal case of roller–shoe assembly operation. In this way, the lubricant thickness does not allow direct contact between the two components, which leads to the long service life of the components. If a high-pressure pump could constantly run in this regime, a coating for the components would not be necessary. However, this case is not often encountered in roller–shoe contact. The boundary lubrication regime is often reached when there are high loads on the roller, or even more so, when it is operating at low speeds [
6]. In this way, the two components come into direct contact. Initially the contact is between the outer diameter surface of the roller and the Diamond-Like Carbon (DLC) layer that is applied to the shoe. Due to high loads, the contact surface extends to the intermediate layer (or the connecting layer) and finally reaches the base material. These aggravating cases lead to scuffing.
Due to the specific working conditions of high-pressure pumps (low speed, high pressure or high fluid temperature), which cannot be modified, it is necessary to investigate new solutions for wear reduction. As a result, our paper is focused on roller profile improvement and the impact of this change on the coating layer. For this purpose, we identified the most used profiles in the literature and assigned them to a cylindrical roller, adapting them according to the requirements of use. To be able to analyze the effect of their use, we performed a simulation in a finite element analysis program and validated the best result through experimental endurance tests. Finally, we performed an analysis of the shoe coating and the material.
3. Roller Profile Theoretical Improvement
The durability of cylindrical roller assemblies is closely connected to the loads that occur within these mechanisms. In both bearings and roller–shoe assemblies, due to the higher pressures on the edges of the roller rather than the central area, even at moderate loads, they can have premature component wear. In order to avoid plastic deformations, it is necessary to use a modified roller profile that is adapted according to the demands, which leads to a uniform pressure distribution [
25].
A straight profile is the simplest solution to achieve the contact surface of the roller with the shoe or the shaft. The advantages of this solution are represented by the reduced costs for part design and the manufacturing process. Over time, various tests have been carried out that highlight the fact that a roller with such a profile has high loads on the ends [
26]. Also, the performance requirements of high-pressure pumps (Delphi, Iasi, Romania) have increased, necessitating new, more efficient solutions. The performance of a high-pressure pump is closely related to the fuel pressure it can deliver. As this pressure increases, the wear on its components will increase due to overloading. Thus, the operating parameters will be changed and, implicitly, technical improvement solutions will need to be sought to reduce wear. Using two basic mathematical functions, we built the following graphic representation (
Figure 4).
High-stress distributions on the straight profile roller ends can be reduced by modifying the profile [
16]. Thus, a new profile solution for the roller–shoe assembly is an elliptical profile. This can be obtained based on mathematical functions (Equations (1) and (2)) with the help of which we can create an elliptically-shaped profile generator. The role of the constant k is to globally determine the shape of the elliptic curve. A graphical representation of the profile based on Equations (1) and (2) is shown in
Figure 5.
Another roller profile, as a solution for improved contact and reducing wear, is a logarithmic one. As we can see in
Figure 6, the logarithmic profile of a cylindrical roller refers to its cross-sectional shape, which is obtained by applying Equations (3) and (4). In this instance, the roller’s diameter undergoes a specific variation along its axis, ensuring that the cross-section of the roller adheres to a logarithmic curve.
Equations (3) and (4) are developed with the help of elementary functions composed according to the requirements imposed (generation of a logarithmic profile curve). The curvature constants α and β determine globally (starting from central area) and locally (at the roller ends) the shape of the logarithmic curve. These can be arbitrarily chosen so that the resulting profile can minimize contact pressure distributions on overloaded areas.
To observe the impact of each profile in minimizing stress, we performed a finite element analysis. Thus, we can study the evolution of contact pressures, equivalent stress, specific deformations, and contact forces under real operating conditions.
4. Simulation of the Transmission Mechanism Operation
A numerical simulation is carried out on the entire transmission mechanism, composed of shoe, roller, and shaft cam. The shaft cam is vertically positioned, thus the piston is at the top dead center, compressing the fluid in the compression chamber.
4.1. Component Design in Catia
The program used for the design of the element’s parts is Catia (Catia V5 r19) (
Figure 7).
Designing the straight profile roller is a simple process, which does not imply a special mathematical function in the design program. Thus, using the basic commands and taking into account the dimensional characteristics, we were able to obtain a 3D variant for the straight profile roller (
Figure 8).
Using the Equations (1) (
Figure 9) and (2) and the values from
Table 1 is obtained the elliptical profile (
Figure 10) of the cylindrical roller in the Catia V5 program. The values from the table are the measured dimensions from a real roller–shoe transmission.
The logarithmic profile is another improved version compared to the previous ones. Referring to the complexity of the roller manufacturing process, creating a logarithmic profile is a high-precision operation that is susceptible to processing deviations.
By entering Equation (3) and the values from
Table 2 in the Catia program (
Figure 11) we obtain the logarithmic profile generator (
Figure 12).
4.2. FEM Analysis Technical Details
The purpose of this simulation is to observe the evolution of the pressure distribution at the roller–cam contact when the roller has various profile shapes (straight, elliptical, and logarithmic profile). The simulation is statically performed using the ANSYS program. To obtain a realistic result, we take into account the forces acting in the vertical direction (Fr) and the friction coefficients from
Table 3. These friction coefficients were obtained using the Castro relation [
27] and also considering the constructive and usual running conditions of the mechanism [
28]. The roller–shoe and roller–cam friction forces are calculated directly using the ANSYS program, at the required pressures, and it is only necessary to enter the friction coefficients according to the type of materials.
The sum of the forces acting in the vertical direction in the four phases of operation (phase 1—cranking, phase 2—idle, phase 3—maximum torque, phase 4—maximum power) of the high-pressure pump will be taken into account (
Table 4).
In order to achieve an optimal balance of convergence and validity of the results, with the aim of obtaining applicable results in a similar manner to all the simulations in this study, we finally chose hexahedral elements (with six faces). These elements are known for providing high precision results.
Following the discretization tests, with different values of the mesh, we found a fine configuration to be optimal, as a coarse discretization mesh provided imprecise distributions of pressures, equivalent stresses, and specific strains. A fine discretization mesh also offers the advantage of eliminating the pressure and stress concentration values, with optimal element size values leading to the convergence of finite element analysis results, whilst also involving higher processing time.
For the simulations, we considered Augmented Lagrange frictional contacts (using the friction coefficients established in
Table 3). Alternative contact formulations such as Normal Lagrange, MPC (Multiple Point Constraints), or Pure Penalty would have provided similar but inconsistent results across the duplication of simulations, discretization grid, and solver parameters (Analysis Settings).
We opted to exclusively examine the interaction between the roller and the cam, as the inclusion of the shoe amplifies the non-linearity of the assembly. This causes convergence issues. Consequently, our simulations concentrated only on the components that best illustrate the impact of the roller profile. In order to avoid cases of non-convergence of the analyses, we opted for a simple application of the contour conditions. This consists in embedding on the side faces of the cam and pressing the roller with the force values (
Table 4). Additionally, using a translational joint, we ensured that the roller has a displacement in the vertical direction, thus restricting the model even better and ensuring full success when running the simulations (
Figure 13).
4.3. ANSYS Simulation Results
We studied, in simulation, the three roller profiles.
The straight roller profile develops contact forces and pressures, equivalent stress, and moderate specific strains, but with peaks at the ends of the roller contact area. This is visible in both the surface and the section distributions (
Figure 14).
The simulation results of the contact between the cam and the roller with an elliptical profile show increased values of contact pressures, stresses, and strains (
Figure 15). The load concentrations on the central area, together with the high values of the analyzed parameters, lead to high wear for the roller and cam.
The simulation of the roller–cam contact with a logarithmic profile reveals a noticeable improvement when high forces are applied to the mechanism. A better dispersion of the contact pressures and the equivalent stresses are visible. Also, the contact force vectors along the roller show a stress-free distribution at the ends (
Figure 16), which are shown by the ANSYS program for each individual node, the legend presenting the value of the maximum contact force in that area.
4.4. Simulation Comparative Analysis
Figure 17 provides an overview of the simulation results.
4.4.1. Quantitative Comparisons between Straight Profile Roller and Logarithmic Profile Roller
The simulation outcomes at a force of 878.42 N reveal an approximately 22% variance in the maximum contact pressure between the straight profile (47.75 MPa) and the logarithmic profile (61.66 MPa), with the latter exhibiting a disadvantageous result. This pattern persists in terms of specific strains and contact forces, showing differences of approximately 15% and 25%, respectively. It is important to mention that this force corresponds to a short operating phase of the vehicle, with little impact on the wear of the motion transmission components.
When applying the same force (878.42 N), the equivalent stress shows a percentage difference of 9% in favor of the roller with the logarithmic profile (32.56 MPa compared to 29.6 MPa).
Comparatively analyzing the results when applying the other forces, an obvious improvement is visible for all parameters when we use a logarithmic profile.
4.4.2. Quantitative Comparisons between the Roller with an Elliptical Profile and the Roller with a Logarithmic Profile
Compared to the maximum contact pressure of 168.38 MPa provided by the straight profile at a force of 878.42 N, the logarithmic profile develops a maximum of 61.66 MPa, showing a decrease of about 63%. This trend is maintained for all parameters, with the percentage difference remaining close to 60%.
The same favorable percentage difference is maintained in the case of the other applied forces.
5. Experimental Validation
The high-pressure pumps wear can be highlighted based on the performance of experimental tests. The evaluation must be carried out by simulating real operating conditions, which is why the testing process becomes a complex and important stage. The roller–shoe mechanism operates at high forces for most of its working time. After the simulation at high forces, the logarithmic profile shows lower values of the contact parameters (contact pressure, equivalent stress, equivalent elastic strain, and contact force) compared to the straight one. All these parameters are involved in the production of wear and friction [
29]. Considering the favorable results of using the logarithmic profile in the simulation and the limited financial resources, we chose to perform a physical test on a single pump, whose transmission mechanism contains such a roller profile. Thus, from a batch of components from the manufacturer, we chose a roller with an identical profile to the one in the simulation and mounted it on a high-pressure pump. We then performed an endurance test on the resulting assembly.
5.1. Testing Methodology
High-pressure pump tests are carried out on special test machines (Bosch EPS 815, Istanbul, Turkey) (
Figure 18). They are equipped with measurement sensors for pressures, flows, temperatures or speeds, parameters that can be checked in real time. The test algorithm is adapted so that we can obtain a wide range of values for these parameters of interest. Based on this test, a detailed analysis of the contact surface of the roller–shoe is pursued. For this reason, the chosen test is a static one, through which we can maintain constant values of the above-mentioned parameters for a period of approximately 20 h. The aggravating test conditions are achieved by reducing the fluid film between roller and shoe. Thus, we imposed a reduced speed (approx. 400 rpm) for the pump shaft, a high pressure for the fluid exiting the pump (approx. 2000 bar) and, at the same time, a high temperature for this fluid (approx. 120 degrees). The testing fluid has diesel-like properties.
5.2. Testing Data Analysis
The analysis of the data obtained from the test bench is carried out by carefully monitoring the evolution of the important parameters of the high-pressure pump. The monitored parameters are shaft speed, pressure, and temperature. The experimental data are taken from the test bench, after which graphs are made. They indicate the evolution of the parameters throughout the test (
Figure 19).
The results do not indicate deviations from the values initially imposed on the parameters. The speed and pressure keep setpoints constant during the test. The temperature on the pump return has a slight increase due to the friction of the pump’s internal components. Two peaks are visible on the graphs (at the beginning and at the end). The first is due to the initialization period of the pump (in which the parameter values were imposed). The one at the end is caused by the maintenance period (oil and filter change) of the test bench.
5.3. Drive Train Component Visual Analysis
The high-pressure pump (Delphi, Iasi, Romania) is disassembled in order to observe the condition of the transmission mechanism elements (
Figure 20). We perform this visual inspection to confirm the results of the test bench data analysis.
There are no friction marks on the ends of the roller, the surface condition being the same along its entire length. The inner surface of the shoe is similar to the surface of a new component. No areas of coating removal or circular contact streaks are observed. This indicates a very low level of wear due to uniform contact with the roller. The shaft cam shows no signs of wear or overheating. The contact area of the cam with the roller is easily visible, without overheated areas on the edges. Considering the experimental test and also the simulation results, the logarithmic roller profile is beneficial for the wear behavior of the transmission mechanism. Finite element analysis is a basic tool for the design of any product, and the comparison between the straight and the logarithmic profiles showed the superiority of the latter in terms of wear reduction. Therefore, experimental testing is performed only on the assembly that shows the best results, without the need to waste resources.
7. Conclusions
The components inside a pump are lubricated with diesel fuel. Its viscosity is low compared to other lubricants but it is impossible to modify the lubricant, considering its use in the common rail system. The lubrication of a roller–shoe assembly can migrate in various regimes, over different periods of time, depending on the operating mode of the vehicle. Considering that diesel fuel is the only fluid that can circulate in a pump, we focus on the roller profile and its influence on the shoe contact surface. Improving the roller profile can be easily implemented in the production process. Thus, initially we theoretically analyzed different types of profiles that are already used on cylindrical rollers. In order to observe their impact on the roller of the transmission mechanism, we performed a simulation of its operation in the ANSYS program. The simulation results show an obvious improvement of the pressure distribution when we use a logarithmic profile calculated for the cylindrical rollers compared to the straight and elliptical profiles. This result was validated by an experimental test, which indicated a guaranteed applicability. The coating analysis is an additional analysis carried out to observe the roller profile improvement impact on this. By analyzing the clean surface of the shoe, free of defects, we can consider that the DLC is an anti-friction solution that can still be used on such parts. This is also confirmed by the sufficient adhesion, the result of which falls within the acceptable area. These analyses were performed on the actual mechanism and are not found in the literature. The study is based on the interaction between theory and automotive exploitation conditions, where high-pressure pumps are used. The developed solution was designed so that it can be applied in the industrial environment, respecting the trend of improving performance indicators (time and cost).