Design of Experiments to OPTIMIZE design solutions for a Power reduction Gearbox

Project Information

OPTIMIZE

Grant agreement ID: 641542

Closed project

Start date 1 October 2014

End date 30 June 2016

Funded under

• Specific Programme "Cooperation": Joint Technology Initiatives

Total cost € 599 709,00

EU contribution € 449 781,75

Coordinated by

DESARROLLOS MECANICOS DE PRECISION SL
Spain

Final Report Summary - OPTIMIZE (Design of Experiments to OPTIMIZE design solutions for a Power reduction Gearbox)

Executive Summary:
Both Airbus and Boeing have estimated around new 30.000 aircrafts over the next 20 years, whose worth reaches 3,6 b$. Demand comes mainly from fast-growing markets in Asia, Middle East and Latin America but also the replacement of thousands of aircraft in more mature markets of Europe and North America.
In addition 22,000 units of business jets are expected by 2033 . This growth leads to a promotion of Business Regional Aviation in order to keep resources efficiency since 90% of travellers within Europe are able to complete their journey, door-to-door within 4 hours, statement declared by the European Commission.
In the next fifteen years commercial aircraft and business jets incorporated more than 140,000 engines with more than 575,000 million euros, representing more than 25 % of the volume estimated for the aeronautical sector market until 2030.
On one side future engines are based on new configurations to make them more efficient and more environmentally-friendly. Strangely enough, these engines, contra-rotating open rotor ones (Rolls-Royce, Snecma) and Geared Turbofan ones (MTU, Pratt & Whitney), have incorporated an epicycloidal reducer as a key system. It also seems that Business Regional Aviation and jets will be based on turboprops whose key system is a planetary.
So these mechanical systems are set to play an even more important role in the future. In fact, engine manufacturers are already counting on companies like DMP to align them with this trend. On the other side a new global scenario is forcing component manufacturers like DMP to get more knowledge in order to give more added value. Very soon it could be difficult for European SMEs to compete underprice strategies with developing countries manufacturers such as Brazil, Russia or China since offset contract model is imposed. Some OEMs are not willing to lose control of core technological culture and they have set suppliers development plans giving them a new and more important role. This implies that these SMEs must assume an adaptation for developing new capabilities such as designing and testing.
The OPTIMIZE project´s goal is the realisation of a dedicated DOE (Design of Experiments) to evaluate the effect of tolerance variation in the gear efficiency. Users of cylindrical gears in aeronautics demand continuous improvement such as increased power density, lightweight, compact size, reduced manufacturing costs, reduced noise, increased reliability and reduced operating costs.
The project has brought to light some relevant breakthrough:
- The influence of mainly one design parameter (alpha) and one manufacturing tolerance (Fp) on the Gearbox performance.
- A DoE strategy in combination with adapted standard test rigs is feasible for these types of investigations.
- LTCA is coherent with experimental results.
- It is demonstrated that changes in microgeometry by using LTCA benefit gears performance.

With regard to the exploitation the Protection Right situation could have been an issue for DMP however this knowledge cannot be protected and it will remain within the company.
The potential customers are extremely demanding and some certifications will be asked: DOA (Design Organisation Approval), and POA (Production Organisation Approval). The aeronautics sector represents a huge opportunity. However a mayor threat is the fact that the market could not identify DMP as a reliable gear supplier as it still is a little unknown. Therefore specific dissemination action has been planned.
As business opportunities grows a business plan, including post-project financing plan, must be elaborated. Aeronautics is based on high-tech equipment acquisitions and the size of DMP is a weakness.
After the OPTIMIZE project TEKNIKER has broadened its gear testing capacity, in particular in the field of Tribology. This fact has positioned TEKNIKER as an experienced partner for solving tribological problems where wear, contact fatigue and efficiency studies could be carried out giving a stronger position in comparison to other competitor laboratories.

Project Context and Objectives:
This project aims to analyse the effect of deviation of Total cumulative pitch deviation (Fp), Total profile deviation (Falpha) and Total helix deviation (Fbeta), more related with manufacturing, and Pressure angle (alpha), this one related to design, on Load capacity (contact pressure, root bending stress), Teeth wear (pitting, scoring/scuffing), Noise (transmission error) and Power losses.
With the help of Load Tooth Contact Analysis tools (LTCA) it is also possible to predict the transmission error graph and the grids of the bearing pattern of the gear set, considering loads, or/and a shaft angle.
Currently, there is a lack of available validation studies for these prediction tools. The combination of specific numerical and experimental study within this project will allow gaining knowledge regarding the accuracy of these tools.

The main objectives are:

- To define the experimental test campaign based on a Design of Experiments (DOE).
- To design and manufacture the Test Articles.
- To carry out the Test campaign and analyse the data results.

The main functional characteristics of the current available FZG test rig to be adapted in the project were defined as starting point
Then it was also clearly defined and agreed the definitive testing campaign, which allowed the establishment of the manufacturing of the test articles (gear pairs). It was agreed to perform two different groups of tests:
• GROUP 1: tests are carried out using “Non-Optimized” geometry gears, meaning gears with induced microgeometrical errors generated during the manufacturing process. In this group a full factorial 24 would be performed. The objective of this group of tests would be to investigate the effect of 4 factors (alpha, Fp, Falpha, Fbeta) in different performance variables or effects (wear, load capacity, transmission error, power losses).
• GROUP 2: tests are carried out using “Optimized” gears, that is, gears including microgeometrical modifications based in LTCA models, whose purpose is to improve the gear performance in the same four effects explored in GROUP 1. In this case, a full factorial 22 would be performed.
The definitive procedure for monitoring the four effects to be studied was clearly defined. The standard FZG test rig had to be modified to be able to control some of the effects. For doing that, the test rig was implemented with force sensors, accelerometers and encoders. It was also developed software to manage all the signals from the different sensors. The definitive testing conditions (time, speed, torque, reference oil, etc.) were also agreed.
All the different pairs of gears with the defined geometries and errors have been manufactured. The expertise in manufacturing has been very critical for this project because the gears needed to have very narrow differences in terms of geometry and tolerances in order to get significant results after testing.
The manufacturing stage has been divided in 3 stages:
• Manufacturing Drawing preparation
• Raw material procurement
• Test articles machining, heat treatment and metrology
Previously the manufacturing process has been analysed. In every step a specific drawing, specific instructions and type of materials were indicated in order to obtain the same manufacturing characteristics. The last step, i.e. the final grinding of the gears, was different per each gear to satisfy the DOE objectives.
Raw material quality should not disturb the experiment result therefore it has been bought in specific batches certified under AMS6265 standard and ordered to a supplier certified by DMP within aeronautic sector.
Each gear follows a defined route wearing a unique traceability number. The heat treatment and its test of hardness have been carried out in batches by a certified provider whose results were confirmed by IK4-Tekniker. The final metrology report was confirmed by the Topic Manager.
All of the gears of both GROUP 1 and GROUP 2 were manufactured. At the end of 2015 almost all of the gears of GROUP 1 had been manufactured and eventually during the 2nd period all of the gears were manufactured, measured and tested.
The DOE1 was satisfactory finished, so the results of 16 experiments were available to investigate the effect of different macro and micro geometrical features on gears. Information concerning wear resistance (scuffing, pitting), transmission errors and power losses were recorded. A total of 63 different experimental parameters were reported for each test. Results of DOE1 were analysed using Minitab software, specialised software for statistical analysis of data.
Parallel to these tests the behavior of the gears was simulated by using LTCA software. The simulation considered the same factors that the experimental tests.
With the empirical results and LTCA a new batch of experiments was defined considering new microgeometrical requirements (GROUP 2). This new design has also been simulated by LTCA.
The tests of DOE2 on gears with modified geometry were also started but an unexpected gear failure occurred and the test rig was affected. An investigation was launched and the results showed that a tooth broke due to a fatigue failure.

Project Results:
The definition of a complex experimental test (in which different test articles geometries, many input variables and few effects from different nature must have been harmonised) has been a success itself. This configuration shows the potential for future research in gears.
Another important result is the fact that according to analytical simulation (LTCA) alpha has been found as the most important parameter, so some of the negative effects can be really optimised during the design.
The effects simulated correspond to contact pressure, i.e. Von Misses Stress at pinion and wheel in contact position and fillet and peak to peak loaded transmission errors. The stresses vary depending on the contact gear location so in order to compare the effect of the different manufacturing errors only the maximum level has been considered. The effect of the errors has been normalised.
The conclusion is that alpha is the most important parameter in several cases, so some of the effects can be optimised during the design. The other conclusion is that in the case that a low transmission error is needed the Fp is the most important parameter to monitor during manufacturing. However this is the most challenging parameter to achieve on the shop floor.
These results have been confirmed by the experimental tests. Four factors (pressure angle (alpha), Total profile deviation (Falpha), Total helix deviation (Fbeta) and Total cumulative pitch deviation (Fp)) were combined to study their influence in 63 different subeffects related to teeth wear resistance, transmission errors, power losses and load capacity during gear operation at specific testing conditions. Regressions analysis showed that 13 of those effects had a fitting with R2(adj.)≥80%, and just those ones were considered as acceptable predictions

Parameter alpha was involved in 10 of the 13 considered interactions, being 6 times the main factor. Parameter Fp was involved in 6 of the 13 found considered interactions, being 5 times the main factor.
In general, alpha and Fp seemed to be the most influencing factors. The main contribution of alpha was related to vibrations whereas the main contribution of Fp was related to transmission errors.
Then, a new LTCA was carried out for the OPTIMIZED gears. It shows that the stress is reduced for the new design for 20º. This is a very important point in order to obtain a longer life for the gear. However in the case of 22.5º the stress at the fillet and contact area is similar to the previous microgeometry. This result is because as we have confirmed the effect of alpha is one of the most important factor, so with 22.5º the optimisation of microgeometry has no sense and it is better to modify the macrogeometry design.
The transmission error is also improved with this new microgeometry in the case of 20º, however for 22.5º is quite similar

In the case of contact pressure the results are similar, for 20º the contact pressure is reduced with this new microgeometry and for 22.5º the improvement is minimal. The contact pressure is very important in order to reduce the wear of the teeth.

So the conclusion is that the new microgeometry design improves significantly the contact stress capacity of those gears whose alpha is 20º. In the case of 22.5º the effect of alpha in the parameters is more important so the effect of the microgeometry improves the behaviour minimally.
DOE2, defined to investigate the effect of modification in the microgeometry of gears, was not completed because of an unexpected fatigue failure that affected test bench. The results are not presented in the reports, but the two pending tests will be executed in a short period of time using internal resources of IK4-TEKNIKER and DMP to complete the technical outcome of the project.
The last relevant achievement is the knowledge gained by the project team regarding the manufacturing process and machine function. The test articles have been manufactured within a very narrow range of tolerances. Conceptually manufacturing tolerance is that error one can admit considering functionality. This is something unpredictable and consequently the generation of two different and relevant errors in a tight range (in order to get relevant results after testing) is challenging.

Potential Impact:
The OPTIMIZE Project is a response to the Aeronautic market evolution since it is moving to geared engines in order to reduce fuel consumption compared to turbofan engines. Therefore, one of its critical systems, the Power transmission Gear Boxes (PGB), needs a new design approach taking into account the manufacturing tolerances.

The expected final results are listed below:

1. Characterization of gear behaviour for aeronautic
2. A new testing protocol to investigate multi-effects in relation to dynamic gear behaviour
3. A new complementary tool to investigate factors and effects in gear design: gear test rig + DOE
4. New gear design and profile optimisation method by using LTCA
5. Pre-Qualification as aerospace gear designer and supplier

These results will impact on:

1. Strengthening the Aeronautics European competiveness through efficiency, manufacturing cost, and reliability, important drivers for advanced gearboxes.
2. Increase of innovation capacity of the partners. From one side DMP will keep improving its LTCA and gear design knowledge leading to a new status within the supply chain. On the other side, IK4-TEKNIKER will improve its knowledge and capabilities in gear testing at laboratory scale.
3. European SMEs growth, as both partners are SMEs.
4. Environment; noise reduction. The lower the transmission error is the lower the vibration, and consequently the noise. This effect can be achieved by optimising the contact pattern between teeth.
5. Society; Local Universities and Technical Schools agreements to teach this project field.

Considering the result achieved, it can be said that this project will impact on the design method of a gearbox as alpha and Fp has an important influence on efficiency, reliability and noise. Also the results will impact on DMP manufacturing processes as a new knowledge has been gained. This will reduce manufacturing costs and improved gearbox components quality, main drivers for growth.
Finally IK4-TEKNIKER has gained a crucial knowledge for applying its testing background in aeronautics, diversifying its activity to other sectors.

List of Websites:
http://www.optimizecleanskyproject.com/

DMP

Person: Jose Amores
Email address: [e-mail]
Mobile: [phone]

Ik4-TEKNIKER

Person: Beatriz Fernández
Email address: [e-mail]
Mobile: [phone]

Related documents

• final1-report-pictures.pdf