VIBES.academy

• Company introduction: who we are, where we come from, and the engineering principles behind our software
• VIBES Solutions: Source Characterisation, Test-based Modelling, Transfer Path Analysis, and Hybrid Modular Engineering
• Products: DIRAC, SOURCE and COUPLE, and how they connect through open data standards (testing and CAE)

Hybrid Modular Engineering replaces a sequence of physical prototypes with a chain of validated component models — measured, simulated, hybrid or a combination — assembled and modified virtually. The economic case is shared across OEMs and Tier 1 suppliers:

• Time: Make NVH concept decisions months earlier in the program while the design is still cheap to change; characterise components once and re-use the same dataset across receivers and vehicles.
• Cost: Fewer physical prototype iterations at full-vehicle level, and fewer re-tests when an interface, bracket, or target vehicle changes.
• Resources: Re-use test-based models and independent source characteristics across projects; less back-and-forth between OEMs and suppliers on what was delivered and how it was measured.

• TPA framework: classic vs. component-based, when each applies
• Source characterization strategies: in-situ blocked force estimation, interface definition, virtual point modeling
• Matrix inversion strategies in source characterization, including the patented X-DoF inverse
• Validation per ISO 20270: the quality indicators you can put in front of a customer
• Case study: EDU source characterization. Complete blocked-force identification on an electric drive unit, from interface definition to validated dataset.

• Understand the available source characterisation strategies and choose the most effective method for the component and the program
• Judge the implications of different test bench configurations on the validity of the result
• Run the end-to-end process robustly, with the right quality checks at every step
• Master the implications of the inverse, when it is reliable, when it is not, and how to act on the difference.
• Apply the 20270 ISO Standard.

• Theory of dynamic substructuring (Frequency Based Substructuring) and how components combine into assemblies
• Component models inside an assembly: measured FRFs, modal models, numerical models, parameterized models, and hybrid models
• Structural modification: modifying assembly dynamics by tuning or swapping a component (mount, bracket, subframe) inside the assembly
• Full-vehicle-as-model: using a measured full-vehicle baseline as the starting point for virtual design exploration, e.g. mount stiffness optimization
• Auralization: turning predicted responses into audible vehicle sound for design evaluation
• Case study: Virtual Mount Modification for road noise NVH refinement. Tuning mount transfer stiffness on a measured full-vehicle baseline to refine road-noise behaviour.
• Case study: Virtual integration of a steering gear into the vehicle. Balancing NVH targets with steering response when integrating a new component into an existing vehicle architecture.

• Create a full-vehicle assembly from a mix of measured, numerical, modal and parameterized component models
• Use numerical (modal) models from your CAE environment inside FBS assemblies
• Modify a measured full-vehicle baseline virtually, swap mounts, subframes, or modal models without rebuilding the assembly
• Use parameterized models to explore a large design space efficiently
• Auralize virtual assemblies to listen to design alternatives before any prototype is built

• Theory on hybrid model training based on the SEMM method
• How to build a hybrid model from a CAE model and measured FRFs in DIRAC
• Where hybrid models fit in the wider methodology: source characterization, component TPA, and dynamic substructuring
• Case study: Suspension optimization for road noise NVH. Using a hybrid model of the suspension within a full-vehicle dynamic substructuring assembly to evaluate design alternatives for road-noise refinement.

• Identify the practical cases where a hybrid model is the right choice
• Design measurements that suit a hybrid model, with far fewer sensors
• Train a hybrid model from measurement to application
• Apply hybrid models across source characterisation, component TPA and dynamic substructuring