FEM (MSC.Nastran SOL600) and Multibody (MSC.Adams flexible contact) solutions: an application example in helicopter rotor analysis

Transcription

1 FEM (MSC.Nastran SOL6) and Multibody (MSC.Adams flexible contact) solutions: an application example in helicopter rotor analysis Daniele Catelani MSC. Software - EMEA Aerospace Consultant Francesca Bianchi AgustaWestland Structural Analyst Rotor Department Stefano Orzi Politecnico di Milano

2 Introduction Contact analysis into model with flexible bodies Available technologies MSC.Nastran SOL6 (Marc embedded technology) MSC.Adams Flex2Flex contact feature Comparison between SOL6 and Adams Comparison and validation with Experimental Tests Example: Helicopter Tail Rotor control chain Workflow FEM model for experimental results correlation and validation Adams flex model correlation with SOL6 DOE for parameters tuning and validation Model Improvements: full kinematic and flexible model Dynamic, transient analysis implementation DOE for design also Pitch control Beam Actuator shaft Bushing Ring 2

3 Tail rotor: problem description Helicopter Tail Rotor provides Thrust to compensate Main Rotor Torque Control chain allows the pilot to change the blade pitch and - consequently to control the Thrust Control chain: assembly of moving parts mechanichally joined together, involving CONTACTS btw parts Control chain Stiffness has to be known for dynamic assessments Stiffness needs to be evaluated for different control configurations Evaluation is performed by test and by analysis on simplified models It is evident the importance of a simulation tool well correlated wrt experimental results to be used for preliminary design because allows quick changes of control configuration (i.e. pilot control input)

4 More than words.

5 Experimental tests Actuator shaft Scope Identify tail rotor control chain stiffness Evaluate coupling effect btw actuator shaft and bushing Identify CONTACT characteristics Equipment Actuators to apply the loads (statically and dynamically) Displacement transducers to measure the displacements at proper locations (i.e. Along the shaft and on the pitch control beam) Strain gauges to study the load paths A proper rig allowing the control chain installation with original parts Bushing Pitch control Beam Ring Tests Collective and cyclic loads Results Test Data Post-processing allows the calculation of the control chain stiffness for subsequent dynamic assessments: bending moment/load curves shows the effect of contact btw bushing, ring and shaft Momento flettente [UMM] Momento flettente - Prova sperimentale BB1prl BB2prl Momento flettente [UMM] Momento flettente - Prova sperimentale BB1prl BB2prl

6 SOL6 analysis Analysis SOL6 uses already existing Nastran bdf file (FEM data) All bodies modelled with solid elements (HEXA) Local Mesh Refinement for contact and stress evaluation Different materials are considered Rigid element for loads and constraints Applied loading conditions to reproduce physical tests Five contact bodies defined ( touched and glue types) Deformable deformable contact Parameters/Solution tuning Contact detection Tolerance parameters Bias distance Contact body/surface definition: Master/Slave The contacts mainly affecting the stiffness results are: actuator shaft bushing installed in the power shaft actuator shaft ring installed in the power shaft

7 SOL6 analysis Static analysis performed on the most complex sub-part of the assembly (pitch beam), properly clamped, for verifying correlation with experimental stiffness Static analysis to simulate the full control chain stiffness tests: 2 configurations of the assembly are considered Max pitch Min pitch Note 1: No Analysis with time dependent (performed in the lab test) loads Note 2: No Analysis with time-dependent change of the assembly configuration and rotating shaft (performed with Adams) Sperimentale [UMR] 1.28 Carico 1.27 positivo.954 Media 1.3 Delta % Carico.985 negativo.976 Media.98 Delta %.24 Spostamento [UMS] Analitica [UMR] Collettivo - LVDT5/8 - direzione Y - Complessivo LVDT5 -.2 LVDT6 LVDT7 -.4 LVDT Spostamento [UMS] Ciclico - LVDT5 LVDT6 LVDT7 LVDT8 LVDT5 LVDT6 LVDT7 LVDT Note 3: CPU time to complete one SOl6 run on high performance multi-processors hardware : ~2 hours

8 Correlation with Experimental tests A few correlation OUTPUT chosen: Bending bridges reading the flexural moments along the actuator shaft Displacement transducers to compute the chain stiffness VERY GOOD final correlation wrt flexural moments and displacements after some tuning: Changing the glue contact with the pitch control beam on top of the actuator shaft: rigid bars have been used Resizing the gap between the actuator shaft and the bushing Momento flettente [UMM] Momento flettente - Prova analitica BB1prl BB2prl Numerical Before GAP changing Momento flettente [UMM] Momento flettente - Prova sperimentale Physical BB1prl BB2prl After GAP changing VERY SMALL final differences btw Test and SOL6: about 1% on the pitch control beam displacements Momento flettente [UMM] Momento flettente - Prova analitica BB1prl BB2prl Numerical Momento flettente [UMM] Momento flettente - Prova sperimentale BB1prl BB2prl Physical

9 Adams model Adopted Solution older release FE beam model for shaft 3D flexible model for pitch control beams Introduction of Master nodes along the shaft for defining contact points (discretized contact) (Vector forces) From 3D to beams Developed interpolation routine for smoothing the transition between discretized contact For D(node,ring) = F = FMAX For D(node,ring) = F = FMAX For D(node,ring) = D(node,node) F = For D(node,ring) = D(node,node) F = For D(node,ring) < D(node,node) F=f(D) For D(node,ring) < D(node,node) F=f(D)

10 Adams model Flex2Flex Adams model built from same flexible bodies used for SOL6 shaft pitch control beam bushing Ring Introduction of Master nodes for constraints and concentrated loads and cards for SOL13 and MNF generation Kinematic model: Primitive joints Fixed joints Motion Two contacts regions defined: between the shaft and the bushing between the shaft and the ring Output: Displacement Loads on constraints Nodal loads (FEMDATA) Fixed Primitive Motion

11 Adams/SOL6 correlation Workflow Submodels for tuning the contact parameters with SOL6: A submodel shaft + bushing A submodel shaft + ring Test Crociera - Spostamento in direzione Y - SOL6 Test Crociera - Spostamento in direzione Y - Adams Tuning on the basis of the main OUTPUTs, i.e. displacements and forces Spostamento [UMS] Node F1 Node F2 -.8 Node F3 Node F Spostamento [UMS] Node F1 Node F2 Node F3 Node F Accurate tuning on Adams contact parameters (K, damp, exp) using DOE analysis and comparison with SOL6

12 Adams/SOL6 correlation Submodels for tuning the contact parameters with SOL6: A submodel shaft + bushing A submodel shaft + ring

13 Adams/SOL6 correlation Workflow -2 Procedure applied to each contact to tune all parameters Evaluation of shaft deformation Evaluation of nodal loads and reactions Evaluation of displacements

14 Adams/SOL6 correlation Workflow -3 Adams > SOL6: CPU time comparison: big advantage for Adams Adams allows easy parameterization and DOE analysis SOL6 > Adams: Adams needs contact parameters tuning and/or identification Adams needs solver parameters tuning DOE analysis has been useful Developed a procedure, using SOL11, for parameter contact identification when SOL6 results not available: Simple Nastran models Disp vs Applied Load Contact parameters estimation and Solver parameters tuning Parameters database linked to material Adams becomes a tool for prediction not validation only

15 Adams/SOL6 correlation Workflow -4 Adams > SOL6: Adams allows high complexity of the model Adams allows transient analysis Full Model: Once tuned ADAMS contact parameters on simpler submodels, the full assembly has been simulated Same outputs used for correlation with experimental results: pitch beam displacement and bending moments Other outputs for correlation are : shaft deformation shape and reaction forces To achieve a good correlation, equivalent modelization rules have been implemented into Adams and Sol6: rigid elements for the joint between the shaft and the pitch control beam (instead of a glue contact, to match experimental results as well) As before, two configurations of the control chain have been considered: 1. Max pitch 2. Min pitch

16 Adams advanced analysis Transient analysis: Two additional cases studied in ADAMS only to assess the robustness of the described results: 1. Time-constant loads with rotating shaft 2. Time-variable loads (performed in lab tests) Both cases with continuos change of configuration (from maximum to minimum) CPU time analogous to previous simpler analysis

17 Conclusion Scope of the work: Establish feasibility of SOL6 to evaluate/correlate experimental test in a analysis of contact btw flexible bodies Establish feasibility of Adams Flexible Contact feature to evaluate/correlate the same SOL6 test case Main results: SOL6 vs. Experimental test: very good correlation on a number of significant measured OUTPUTs SOL6 vs. Adams: very good correlation Achievements: Adams can be adopted as a predictive and design tool together with a database of good and reliable contact parameters Robustness of Adams solution has been confirmed by time-variable runs Time to complete an Adams run, without friction, is 1/1 of equivalent SOL6 one

18 Further development/investigation Repeat correlation with MDNastran SOL4 Exploring friction effects in the contacts Extend to more materials the SOL11 method adopted for the characterisation of Adams contact parameters Evaluation on more complex models: introduction of flexible blades, aerodynamic forces,

19 Thank you