1 Thermal elastohydrodynamic simulation of a slider bearing
1 Thermal elastohydrodynamic simulation of a slider bearing in a heavy duty diesel engine transmission Andrew Spencer Dynamics & Acoustics Engine Development SCANIA 2015 -10 -07 14: 10 Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
2 Background – V 8 engine gear transmission Camshafts Intermediate Gear Crankshaft Investigation: Can we replace the Intermediate Gear roller bearing with a slider bearing? Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
3 Why? Motivation for a change from roller bearing to slider bearing: 1. 2. 3. Noise reduction – lower transmission of meshing noise into the engine block Cost reduction Friction reduction – if the roller bearing has seals then total friction can be lower with a slider bearing Bearing friction measurement Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
4 Multi-Body Dynamic model development Right Camshaft Multi-Body Dynamic model of the gear train developed in AVL EXCITE Power Unit Left Camshaft Hub • Crankshaft is driven at a constant speed • Dynamic braking torque is applied to the left and right camshafts, and also to the Fuel Pump, Air Compressor and Power Steering Servo which are all driven through the Intermediate Gear (not shown in the illustration) • Intermediate Gear and Hub are modelled as flexible bodies using finite elements Intermediate Gear Crankshaft • Radial and Axial bearings between the hub and Intermediate Gear modelled with Elastohydrodynamic bearings • Different engine operating conditions are simulated Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
5 EXCITE Power Unit model development Elastohydrodynamic joints Rigid bodies with brake torque applied Flexible bodies Crankshaft is driven at a constant speed Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 Gear joints transmit torque and radial/axial forces between bodies 2015 -10 -07
6 Condensated bodies Condensation is performed in Nastran. The DOF’s that we want to keep (because we want to connect a joint to them, or observe their motion in our simulations) are specified, and then Nastran is run to reduce, or condense, the stiffness matrix down to just our specified DOF. This can hugely reduce the DOF in our model. Hub Flexible body Gear Flexible body From 16212 to 1612 DOF’s From 18882 to 596 DOF’s Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
7 Tribological joints The time dependent Reynolds equation with cavitation is solved for the radial and axial bearings. Radial and Axial bearing pressure profile between hub and gear For a given separation, the pressure in the lubricant film is calculated. This pressure is then applied to the flexible bodies and the deformation calculated (EHD). A full mixed lubrication model is implemented, if the separation becomes very small then the surfaces will come into contact (asperity contact) and the contact pressure is derived from a precalculated asperity stiffness curve. Flow factors are implemented in the Reynolds equation. Example of roughness used to calculate asperity stiffness Lubricant Supply Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
8 Simulation of thermal effects The Multi-Body Dynamic model presented so far is iso-thermal Why might we want to include thermal effects? Increase in friction Increase in the temperature of the components Reduction in lubricant film thickness Reduction in lubricant viscosity Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
9 Inclusion of thermal effects Run MBD Model. Results: Frictional heating & oil flow Apply friction heating and oil flow (cooling) from MBD to FEM thermal model FEM: Step 1, Heat Transfer FEM: Step 2, Thermal Expansion Apply new temperatures and clearances to MBD model Evaluate Results Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
10 Step 1 – Heat Transfer Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
11 Step 2 – Thermal Expansion Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
12 EXCITE & ABAQUS iterations for temperature Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
13 Results 1. Under certain load conditions the gear is forced backwards due to the axial loads applied through the helical gear Bearing Ax. Front Ax. Rear Avg. Oil Flow (l/min) 0. 286 0. 025 Heat flux to solid (W/m²) 4921 11251 2. The oil flow rearwards out of the radial bearing is very low (25 ml per minute). This is the limit for how much oil can lubricate and cool the rear axial bearing. 3. At the same time the rear axial bearing has the highest heat flux into the bearing, leading to the highest temperatures. 4. The rear portion of the hub also has higher temperatures as there is less surrounding material for the heat to be conducted away through. Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
14 Comparison with test data A thermocouple was used to measure the temperature on the back-side of the hub – At the highly loaded condition simulated a spike in temperature is observed during the engine test Engine Speed Oil Temperature Torque transfer through Intermediate gear Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
15 Recommendations from simulation results 1. Most likely cause of high temperatures in the rear axial bearing is too little oil supplied from the radial bearing 2. Solution would be to place axial, or spiral, grooves in the radial bearing Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
16 Design change test results Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
17 Conclusions and Future Work • The use of Multi-body Dynamic simulation with thermal effects and EHD bearing models led to a fundamental understanding of the tribological behaviour of the system, not possible to gain from testing alone • The model was predictive of the elevated temperatures observed during engine testing • Future work will entail expanding the semi-2 D heat transfer and thermal expansion FEM model to be fully 3 D so that local hotspots around the circumference of the bearing can be calculated Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
18 Info class Public Dynamics & Acoustics/Andrew Spencer/Tribodays 2015 -10 -07
- Slides: 18