Restraints of crane

[Laserlassie]

Hello!
I am new to the forum, but need some help.

I am currently working on a project where i am designing a movable container crane. The design is fairly simple and resembles a straddle carrier or mobile boat lift. I am currently working on the structural analysis and for one of the load cases i am simulating a crash. I have the energy balance and spring constant down but i am having some trouble defining the restraints of the "wheels". The case is that the crane is colliding with one leg. My proposition per now is this:

3 legs free in x,y and locked in z. 1 leg locked in x,y,z and rotation. I have a feeling this is not the best solution but as i said i am having some trouble coming up with something that works better.

 

[btrueblood]

Simulating a crash - i.e. one leg/wheel hitting a solid object? Not sure about the physical reality of locking z-axis motion of all the wheels, in real collisions the vehicle can/may/will rotate about its center of mass and some or all the wheels may lift as a result. Why not model the collision with the center of mass of the vehicle fixed?

 

[rb1957]

I don't think you've constrained rigid body motion; I think you need a second x- or y- to react the Mz (multiple z- react Mx and My). The other thing to watch for is tension loads on the wheels.

I'd suggest supporting each wheel on 3 rods of some stiffness and constraining the far ends of the rods; to give the wheels finite stiffness. Then ask yourself "could some of these wheels move?".

another day in paradise, or is paradise one day closer ?

 

[Laserlassie]

I am working in SAP2000 and having trouble finding a way to restrain the center of mass (even tho that probably would have been ideal. I am looking for worst case scenario, rb1957 do any of you have any experience working in SAP2000 and have any suggestion on how to do this? I can upload the file if you want.

 

[rb1957]

no, sorry.

another day in paradise, or is paradise one day closer ?

 

[jlnsol]

laserlassie,

i suggest to leave that 1 leg that hits an object free to rotate in whatever direction it wants, even torsion so to say

 

[Laserlassie]

How would you restrain the rest then? I can add a torsion load equal to that of the friction torsion.

 

[rb1957]

first you have to constrain the rigid body freedoms, 6 degrees of freedom. As posted you are only constraining 5.

another day in paradise, or is paradise one day closer ?

 

[Laserlassie]

You mean constraining the entire body in 6 degrees? Or just the one leg? Because i thought i had it restrained in all 6, X,Y,Z and R1, R2, R3

 

[jlnsol]

yes, that's the way i simulate such situations: constrain the entire body and than add a force to the leg against the direction the body moved
then simulate and find out xyz deflections and 3 rotations of the leg

So look at this problem as if the crane is in rest and a object hits that one leg

 

[rb1957]

well that's not the way I do it. Your approach (as I understand it, fixing the four supports) over constrains the model and builds into your answer fictitious redundant (self-cancelling) forces. To properly (IMHO) restrain rigid body motion you need to remove 6 degrees of freedom, one point have x,y,z; a 2nd x,y; and a 3rd x ... or some combination depending on your co-ord system.

Constraining all 4 points in z is an acceptable compromise (saying the flat base remains flat) and constrains 3 freedoms (Fz, Rx, Ry). So you need 2 x and 1 y (or 1 x and 2 y) to take out the other 3 freedoms (Fx, Fy, and Mz).

Better than hard (infinitely stiff) nodal constraints is to apply a finite stiffness at each support. My suggestion is three orthogonal rods at each support point, and hard constrain the far ends of these rods.

another day in paradise, or is paradise one day closer ?

 

[jlnsol]

No, fixing all four supports is not what I mean. I suggest fixing the crane frame in a location as close to its C.O.G.
Then only add a collision force at one leg.
Then simulate and conclude what deflections and stresses that one leg will show.

Your method of adding orthogonal rods will also work by the way.

 

[rb1957]

ok, I see that ... just turns the problem on it's head.

another day in paradise, or is paradise one day closer ?

 

[Laserlassie]

How far is this from my idea of restraining one leg in 6 freedoms and dividing the collision force on the other 3? I really appreciate the help, i will look into your suggestions tomorrow!