Is Studying Planar Multibody Dynamics Essential for Mechanical Engineers?

[zoltrix]

Hello

I am reading the book :

" Planar Multybody dynamics, Formulation, Programming and Application"

An interesting lecture , however is it worthwhile studying these methods for kinematic and dynamic analysis of mechanical mechanisms?
I mean
There are many softwares on the market which make the job for you, for example "multibody" by MatLab ( ex Simmechanics)

what did mechanical engineers actually use on job ?

maybe none of them just the experience ?

 

[jrmichler]

Few practicing mechanical engineers ever perform analytical analyses of multibody dynamic mechanisms. Most dynamic mechanisms simplify down to a servomotor or crankshaft moving a load. Most of those cases are solved by first calculating the position, velocity, and acceleration vs time. Components are sized to meet requirements of stiffness, inertia, and strength. Peak drive torque is calculated, and the drive is sized. More advanced problems require an understanding of torsional natural frequency, and calculating an appropriate motion profile. Dynamic mechanism typically also require calculating the total system inertia reflected to the drive motor.

I had one job where the best solution used a four bar linkage driven by a servomotor. I wrote a Matlab program to plot the output motion, and used that program to iteratively design the linkage. The peak acceleration was twice the peak acceleration of a similarly sized crankshaft, which was all the necessary information to size the parts. After sizing the servomotor, it was built, started up, and it worked. Durability testing resulted in several redesigns to improve bearings and mounting of change parts.

Simulating motion is the easy part. The challenge is in designing and building something that actually works in the hands of a customer, and keeps working.

 

[FEAnalyst]

There are many softwares on the market which make the job for you, for example "multibody" by MatLab ( ex Simmechanics)

what did mechanical engineers actually use on job ?

In practice, advanced MBD software such as Simpack and MSC Adams is used to analyze the motion of complex mechanisms and machines. Basic motion studies can also be performed in proper modules of CAD software like Autodesk Inventor, Fusion 360 or SolidWorks. They all offer quite advanced motion studies with evaluation of displacements, velocities, accelerations and forces, among the others. If you don't have a license for such software but would like to try something like that, check the open-source solutions. For example, you can find the add-on "Dynamics" module for FreeCAD which uses MBDyn as a solver. Its capabilities are quite impressive:



Here's a GitLab repository of this module: https://gitlab.com/josegegas/freecad-mbdyn-dynamics-workbench

However, I still think that theory can be useful. It's like with Finite Element Analysis where a software can do almost everything for you but you won't be a good analyst without the knowledge about the theory behind this method and behind solid mechanics or whatever physics are involved in your analysis. For example, you can't interpret the results properly if you don't fully understand the meaning of each output variable and don't know how your model should behave under specific loading. Setting up the analysis also requires understanding of various aspects such as material models and boundary conditions. And when something fails, you have to dive deep into the problem from the perspective of a solver. You have to understand the Newton-Raphson method to know why the analysis is divided into increments and iterations and so on. Not to even mention the verification with analytical solutions which is very often necessary in the case of FEA. Multibody dynamics simulations are different but many of these remarks still apply in some ways.