Croda Int Plc is a developer and manufacturer of an innovative phyllosilicate-based surface treatment additive technology for gears and bearings. The particles with a platelet shape use lubricants as a carrier and build through their adsorption a protective phyllosilicate-based coating on the surface. The modified surface has a significantly lower surface roughness, which ensures a better load distribution and lower local pressure. Additionally, due to the special layered material structure, the particles can be sheared in the tribological contact, which leads to a significant reduction in friction. More detailed mechanism description can be found in our scientific publication, Tribological properties of a phyllosilicate based microparticle oil additive Wear 426-427 (2019) 835-844 . All in all, when applying the products, treated systems can run better with reduced friction, wear, surface roughness, and temperature. These effects lead to higher efficiency and longer lifetime.
1 Challenges in gearbox reliability in wind turbines
While wind power is enjoying significant growth, it is also important to understand its limitations. Multiple studies have confirmed that wind turbines suffer from reliability issues: The EU’s RELIAWIND study aimed to examine current reliability of large wind turbines. The study found electrical systems accounted for the highest failure rate, but gearbox failures accounted for the highest amount of downtime (14 days) . Another study, by the National Renewable Energy laboratory (NREL), also found gearbox failures contributed the greatest amount of downtime of any single wind-turbine component, while the majority of wind-turbine gearbox failures (76.2%) are caused by bearings. Gears were the second leading cause of failures (17.3%) [2,3,4]. A later review by  also suggested reliability of gearboxes has not improved over time and drivetrain technology has not yet fully matured. Figure 2 displays the annual failure rate and downtime per failure by component.
For offshore wind turbines, reliability is even more critical, with marine operations costing significantly more than onshore. Efforts have been made to remove the gearbox from offshore wind-turbine designs completely, with Siemens Gamesa announcing in 2017 that all offshore wind turbines will be direct drive in the future . The market leader, Vestas, however, has committed to offshore wind turbines with gearboxes, with one source suggesting up to a 10 percent increase in material costs for a direct drive vs. wind turbine with gearbox . A review by  revealed gearbox failures for offshore wind turbines happens at about three times the onshore rate.
Maintenance of wind turbines is essential to prevent and mitigate failures. However, ongoing operation and maintenance (O&M) is costly, representing about 25 percent of the total cost of the wind turbine over its lifetime . Unscheduled maintenance is estimated to contribute between 30 and 60 percent of the cost of O&M . These high costs cause some wind-turbine owners to skip maintenance: Insurer G-Cube cited the top cause of a claim as poor maintenance, at 24.5 percent of total claim costs, with claims involving gearbox failure costing on average $380,000 to rectify .
As we have described above, wind-turbine maintenance and component failure is expensive, with bearings and gearboxes contributing significantly to downtime costs. Here we present a proven technology that is a microparticle-based Phyllosilicate-Additive for lubricants to reverse existing damage and protect the system for the future, thus improving reliability in wind-turbine gears and bearings.
Phyllosilicate-Additive technology helps to significantly reduce or even prevent the damage, whereby an application is recommended for both new and already damaged systems. The technology is an innovative lubricant additive with a protective and repairing effect, which mainly consists of phyllosilicates in the form of micro and nanoparticles. When using the term repair, repaired, or repairing, the effect can be described as creating a smoother surface by filling in micro-valleys and micro-pits, which creates a more evenly distributed oil film, which in turn reduces asperities from contacting. The particles use lubricant as a carrier to reach the rubbing metal surfaces and to coat damaged areas by adsorption. The new, modified surface is optimized and protected from a tribological point of view, so that surface roughness, friction, wear, and temperature in the system are reduced. This leads to a significant improvement in efficiency and lifespan.
The wear can be calculated in theory by V = k⋅N⋅x/σ (Wear volume = topography coefficient k ⋅ Load ⋅ sliding distance / material hardness). We are not able to change the parameters such as load or material hardness in already running systems, but it is possible to modify the surface to ensure the optimal load distribution and to minimize wear.
2 Gear and bearing surface modification as a solution
2.1 Scientific tests
For the further improvement and development of the technology, especially for the first fill applications, Croda has been working closely with several research institutes, universities, and OEMs for many years, which can perform tribological tests with high accuracy and reliability in the lab and in the field. The tribological performance was proven with different test bench configurations such as 2-discs, pin-on-disc, FE8, 4-ball, MTM, SRV, and other model tests. Standard concentrations of Phyllosilicate-Additive of 0.2% is used in treating gear oils, and 2 percent is used in treating greases.
2.1.1 Efficiency improvement (SRV FZG test)
FZG efficiency test is a well-known and established test method for the performance analysis of the friction modifiers. To simplify the study and to keep costs low, we started with an FZG simulation on the SRV test bench with the test load values provided by Optimol. The following parameters were used: temperature, 98°C; frequency, 50 Hz; and stroke length, 4 mm. The measurement was performed in a common wind turbine gear oil Castrol Optigear Sythetic X320 (gray curve) and in the same oil modified with Phyllosilicate-Additive (blue curve).
The load was increased step by step, and the friction behavior was observed. The experiment shows Phyllosilicate-Additive can reduce the friction significantly, almost at all loads, and the maximal friction reduction was 38 percent. The wear during the test by measuring material loss of tested components (cylinder and plate) was reduced by 12 percent.
2.1.2 Friction reduction & surface protection in gears
The friction process in gears can be simulated by the 2-disc assembly bench to reproduce the contact pressure and the sliding velocity of gears at one particular point along the tooth profile. Therefore, numerous tests with different gear oils in combination with Phyllosilicate-Additive were tested in this configuration at the University of Mannheim, Germany.
Figure 4 shows a friction measurement in Amsoil PTN 320 oil with Phyllosilicate-Additive added after 20 hours. The measurement was performed at 60°C with a load of 1 GPa and rotation speeds of 424 RPM/339 RPM (20% slip).
A friction reduction of 46 percent was achieved by adding the Phyllosilicate-Additive. The surface analysis of the contact surfaces from the 2-disc tests, Figure 5, also shows a protective and repairing effect of the Phyllosilicate-Additive, so that the surface roughness is significantly reduced, and the surface topography is more homogeneous, which leads to reduction of local load, tribological stress, and temperature.
Multiple gear oils were tested: Castrol Optigear Synthetic X320, Mobilgear SHC XMP 320, Klübersynth HEM 4-320N, Fuchs Unisyn CLP 320, Amsoil PTN 320, Shell Omala S4 GX 320, or Klüberbio EG 2-150. Testing was in accordance with the above test parameters outlined in Figure 4 and were analyzed with and without Phyllosilicate-Additive. Results show that the Phyllosilicate-Additive significantly reduces friction, wear, and surface roughness in all the oils. See Table 1.
2.1.3 FE8 test – surface protection
A further test for the understanding of the Phyllosilicate-Additive effect on the surface topography is FE8, which was performed with and without Phyllosilicate-Additive in the Castrol X320 oil with the following parameters: speed, 7.5 rpm; test duration, 80 h; temperature, 80°C; and load, 80 kN.
The surface treated with Phyllosilicate-Additive looks more homogeneously without sharp wear traces, which leads to a better load distribution. Treated bearing measured mass had 17 percent lower wear based on material losses.
2.1.4 Friction reduction in bearings
The previous tests show the tribological behavior of the Phyllosilicate-Additive technology in the oil lubricated systems. The following MTM test shows the friction performance in a common grease with and without Phyllosilicate-Additive at different rolling/sliding ratios. The following test parameters were used: load, 70 N; temperature, 23°C; velocity, 700 mm/s; and time, 172 s.
The diagram shows that the friction coefficient can be significantly reduced for all bearing relevant (0 to 40 percent) rolling/sliding combinations by up to a maximum 38 percent.
2.1.5 Surface protection – Improved weld load in greases
A measurement of the weld load in grease-lubricated systems is part of the standard test methods for the grease qualification and can be performed on a four-ball-test bench. For this test, the standard DIN 51350 part 2 (welding force) was used: rotating speed 1,450 RPM, 1 min for every load step.
The experiment was performed with the pure grease and with the same grease modified with Phyllosilicate-Additive and shows a significant weld load improvement by 400 N (15%).
2.1.6 Surface protection – Standstill damage prevention
Standstill damages are a well-known problem, especially in wind turbines, but they occur generally in all systems where vibrations appear during standstill. Although common lubricants help reduce the damage, they can’t prevent it, or prevent it to a minimum, because of the very poor lubrication due to lack of movement in the system. For this reason, it’s useful and necessary to optimize classical lubricants by adding Phyllosilicate-Additive with its silicone-based nano- and micro-particles, which adsorb on the steel surface and stay on it, even if the lubrication is deficient.
We have scientifically analyzed the behavior of Phyllosilicate-Additive particles in standing systems through the so-called False-Brinelling-Test in cooperation with the University of Mannheim, Germany (25 Hz, 750 N load, room temperature). The experiment results show a reproducible significant wear reduction of up to 76 percent and a clearly visible reduction of corrosion. Furthermore, the Phyllosilicate-Additive technology can not only reduce the damages for the future but also partially repair already damaged surfaces.
3 Examples of application – Surface protection
For a meaningful evaluation of the application, it is necessary to analyze and to compare the steel surfaces in the system before and after the treatment. Replica-set technology makes it possible to perform the surface analysis of gears or bearings by making imprints of the representative areas of the surface with high precision and reliability. The subsequent analysis of the imprints with light microscopy provides valuable tribological information about the surface condition.
3.1 Gear tooth flank surface analysis
Figure 11 displays the tooth flank surface of a high-speed shaft in a wind turbine with a power of 1.5 MW. The first imprint (a) of the high-speed shaft was taken before the application of Phyllosilicate-Additive. It shows running traces at the upper area and notable micro-pitting at the middle and lower area. The second imprint (b) was taken three months after the product application. The same area of the tooth flank is clearly visible. The striking difference is that the micro-pitting is repaired, and there are no running traces. There is a significant improvement in the surface condition through the repairing and protective effect of the Phyllosilicate-Additive. This surface modification ensures more homogenous load distribution in the system and, in this way, protects the surfaces. The optical analysis of the worn areas shows the Phyllosilicate-Additive application was successful and achieved its aim: repairing and protection of surface.
3.2 Spherical roller bearing surface analysis
In this example, we show a similar analysis of the main bearing in a common wind turbine (1.5 MW). To analyze the wear development over a period of one year, we made several imprints of the representative areas of the surface with high precision and reliability. For a quantitative evaluation of the application, we measured the roughness within the track in the middle of the imprints. This provides valuable tribological information about the surface condition.
The roughness calculated before the wind turbine was treated with Phyllosilicate-Additive was Ra = 0.556 μm. Five months after the application, the roughness was reduced by 28 percent to Ra = 0.403 μm. At the third measurement 12 months after application, we noticed a reduction of roughness within the track by 60 percent to Ra = 0.225 μm.
The surface analysis clearly shows the structure and roughness of the surfaces are significantly improved. That means less stress on the mechanical components and a substantial increase in the lifetime. Besides the significant improvement of the surface condition, a reduction in vibration was also observed.
4 Influence on the system lifetime
In this section, we would like to present the lifetime calculation of a grease-lubricated bearing (1.5 MW) with and without the Phyllosilicate-Additive application. CRODA contracted Sentient Science to perform a computational life prediction (CLP) of its Phyllosilicate-Additive surface treatment at 2-percent concentration compared to a “do-nothing” scenario using its DigitalClone for Engineering (DC-E) software computational testing package. The goal was to quantify the computational life test result benefits based on before and after treatment of Phyllosilicate-Additive.
4.1 Surface roughness and material characterization
Measured surface roughness distributions, along with material properties such as grain size and residual stress profiles, are used as inputs to the CLP process (see Figure 13). These inputs allow the model to represent surface asperity interaction as well as the material’s response to these stresses.
4.2 Generate damage equivalent loads for use as input by system and component models
Damage equivalent load calculations are performed according to DNV GL IV-1, leveraging the actual measured wind speed for a chosen turbine (generally three years minimum data) and running the wind through an aeroelastic model specific to the turbine technology. The resultant loads are used as input to the CLP reliability simulation process.
4.3 MTM lubricant testing
The MTM or Mini Traction Machine testing allows one to investigate friction and traction using a variety of materials, geometrical contacts, oil temperatures, and contact pressures. The resulting curves are used as input into the CLP reliability simulation process. These measurements indicate the Rewitec additive improves (decreases) the level of friction between the raceway and rolling elements over a range of contact conditions typically experienced in main-shaft bearings.
4.4 Creation of system and sub-component models
System models and sub-component models are generated so the areas of highest load and the specific contact stresses can be obtained on the sub-component of interest. The resultant contact stress is used as an input to the CLP reliability simulation process.
4.5 CLP (Computational Life Prediction) simulation
The CLP reliability simulation considers the surface profile, geometry, material properties, loading, and lubricant properties, among other considerations, using actual historical operating data of the wind turbine. The software produces a reliability curve of failure vs. time.
4.6 Mixed-EHL solution for life prediction
To take the influence of asperities into account for determination of probabilistic fatigue life, Sentient mixed EHL (elastohydrodynamic) solver uses simulated surface roughness profiles in an explicit- deterministic calculation of surface tractions. Surface traction refers to the pressures and surface shearing stresses transmitted between two surfaces through a lubricant and/or direct metal-to-metal contact.
The simulation shows that with Phyllosilicate-Additive, the probability for a bearing failure significantly decreases, which leads to a lifetime extension of up to 17.3 years (Figure 14).
The tests confirmed that using Phyllosilicate-Additive provides improvements in friction reduction on metals in both grease and oil applications. Additionally, Phyllosilicate-Additive also reduces most adhesion wear to include pitting, most abrasion wear, and some tribo-chem reaction wear in accordance with DIN 50320. Besides the classical friction and wear reduction, Phyllosilicate-Additive technology can repair partly damaged surfaces, which increases the lifetime enormously. All in all, using a Phyllosilicate-Additive leads to higher efficiency, higher reliability, and a longer lifetime.
- Final Publishable Summary of Results of Project ReliaWind, ReliaWind Project Nr 212966.
- Wind Turbine Drivetrain Condition Monitoring During GRC Phase 1 and Phase 2 Testing S. Sheng, H. Link, W. LaCava, J. van Dam, B. McNiff, P. Veers, and J. Keller
- National Renewable Energy Laboratory S. Butterfield and F. Oyague Boulder Wind Power.
- Wind turbine reliability data review and impacts on levelized cost of energy Cuong Dao | Behzad Kazemtabrizi | Christopher Crabtree.
- Wind turbine reliability: A comprehensive review towards effective condition monitoring development Estefania Artigaoa, Sergio Martín-Martíneza,b, Andrés Honrubia-Escribanoa,b, Emilio Gómez-Lázaroa,b,*.
- Wind turbine reliability data review and impacts on levelized cost of energy Cuong Dao | Behzad Kazemtabrizi | Christopher Crabtree.
- Chizhik et al., Wear 426-427 (2019) 835-844.
Printed with permission of the copyright holder, the American Gear Manufacturers Association, 1001 N. Fairfax Street, Suite 500, Alexandria, Virginia 22314. Statements presented in this paper are those of the authors and may not represent the position or opinion of the American Gear Manufacturers Association. (AGMA) This paper was presented November 2021 at the AGMA Fall Technical Meeting. 21FTM10