inertial load

[jakub999]

Hello, Does anyone know how inertial load works in Patran? I thought that if I apply gravity on some body that it's like I would apply nodal force to every node. Value of that force woul be (mass * g)/ (number of nodes) but it seems like I get diffrent results.

 

[Erik Panos Kostson]

There are different ways of applying body loads (e.g., gravity) and element loads in FEA (consistent and lumped). What you are trying to describe is the lumped approach.

For more info about these 2 different load options in MSC see: Link

It could be that the gravity load is applied in MSC by multiplying the mass matrix with g (that is stated in the manual). Then how the load vector becomes depends on how one defines the mass matrix (again here it can either be lumped or consistent mass matrix, at least in Strand7 which I use).

Look in the msc manuals that are referred to in the above link, and you will find it.

 

[jakub999]

Well I want to apply inertial force to piston but I'm wondering what is the proper boundary condition? To define acceleratin like gravity or do it by this lumped approach?

 

[Erik Panos Kostson]

One would not do it by hand manually applying all the loads on the nodes (especially consistent would take some time to do.
The software do have the capability to do this automatically thus making life easier for you, hence why you pay many to buy them .

In Nastran there should be a proper definition of gravity and accelerations (see manual, think the card is GRAV and ACCEL, and could also be used from the GUI possibly called inertial loads, but not sure since I have not used MSC Nastran).

Since Nastran uses the mass matrix (consistent or lumped), to define inertial loads, one must have density defined in the material properties.

 

[jakub999]

Well, when I want to apply gravity I will do it with Inertial load and it works fine but It is defined to all model. I want to apply acceleration only on some parts. So I wanted to use this lumped approach through many little forces, but when I get diffrent solution then what is right?

Then I had idea to use function nodal accleration but when I apply it for example on some fixed body I get zero strees and zero displacement. I looked to the manual but didn't find any proper way or real situation when to apply acceleration
_____________________________________________________________________________________________

So now I applied Total load with magnitude mass*g and it spreads that force through every element and now I got more or less the same results so I guess that is the answer on my first question. But still I wounder about that acceleration

 

[jakub999]

I also tried apply anglular velocity but still got 0 displacement, stress. I dont understand why.