Problem with simulation, change of rotation after helmet impact

[Kimpan]

I am doing a finite element analysis using LS-DYNA of a helmeted head form being struck on an inclined plate.
The helmet and head is given an initial speed of 5.3 ms in the Z direction. As a result of the impact to the inclined plate, the helmeted head starts to rotate in the Y axis. The problem I have is that after a while, the helmeted head starts to rotate in the opposite direction for some reason. I can not understand the cause of this.
When I run the same simulation without the helmet I do not see this problem.
The helmet in these pictures is made up of elements that have been extended from the head form. It is 2cm thick and of the material crushable_foam. The outer part of the helmet is a linear elastic shell.
I have checked so that the centre of gravity of the head form is located in the centre of the head as it should be and that the inertia properties of the head are set.

 

[rb1957]

investigate the line of action of the initial impact force ... which side of the head CG is it ?

is the inclined plate fixed ?

 

[Kimpan]

Thanks for the response!

How do i check the line of action in LS-DYNA? Can't find how to display the direction of the reactionforce at the plate.

The inclined plate is fixed.

 

[rb1957]

i don't know LS-DYNA (though I'm sure she's a very nice girl). can you get a free-body output ? maybe the reaction on the plate would tell you ? the reaction force wouldn't need to be normal to the plate, though it probably is opposite to the motion of the head (at the impact site).

 

[corus]

With the C of G offset the natural behaviour of the head is to spin clockwise, offsetting the anticlockwise motion.. eventually?

corus

 

[BobM3]

Can you turn off gravity? If it doesn't spin much initially (so the c.g. stays to the right of the center of rotation) and stays in the air long enough I can see gravity eventually spinning it clockwise.

 

[GregLocock]

I agree, turn gravity off. Even so i'm surprised that it has a big effect in such a short time period.

Once the head starts rotating widdershins then it really should carry on doing that, in the absence of external forces. Mr Newton says so.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Kimpan]

I understand what you mean; that the rotation should change direction given the head is left in space for a long period of time and CG is to the right of the center of rotation.
This is under a very short period of time so the gravity effect should be negligible. I can confirm this as I have done experimental testing of this exact situation.
I turned the gravity off just in case the gravity was set wrong and somehow effected the rotation abnormally. The results remain the same.

Frustrating problem this.

 

[BobM3]

Is it possible to mark the center of mass so you can see how it is moving during the simulation? Even better, but probably not do-able, mark the center of rotation and the center of mass?

 

[corus]

Could it be that you have 2 forces acting on the head, friction and a direct reaction force. Friction will turn it anticlockwise, whereas the reaction force, normal to the inclined plane, acts to turn the head clockwise as it is offset in an opposite sense from the centre of mass. Why the friction force should disappear nd the reaction force dominate in a fraction of second I don't know. Just a thought.

corus

 

[rstupplebeen]

If the CG is off center and the head is in free fall with no wind resistance why would head rotate? If you drop a hammer sideways in a vacuum there would be no rotation. I may be missing something though. So after impact the head would rotate but after that the rotational speed should not change. I think that this is an indicator of numerical errors building up over such a long time period. I hope this helps.

Rob Stupplebeen

 

[corus]

Perhaps the head isn't solid and the inside sloshes about a bit so that the effective movement of the C of G causes opposite spin.
Good quiz question anyway.

corus

 

[Kimpan]

I marked the CG and it seems to coincide with the centre of rotation. What could be the cause of numerical errors in this case? I don’t see any errors in the message file created from the simulation.
I have also tried using a linear elastic material instead of the crushable foam as padding in case there was something wrong with the material model.
As I was using a more advanced model for the helmet I would get warning messages when I used a “tied surface to surface” connection between two shells separated too far from each other, the warning would say that there could be rotational constraints.
This model however does not need any defined contacts as the helmet shares the nodes with the head.
The inner part of the head is made of defined as rigid whereas the outer surrounding part is defined as viscoelastic. There would be no moving parts inside the head.

 

[Kimpan]

An update:
I remodelled the helmet and somehow the "ghostforce" is gone =)
I thought this "ghostforce" was the cause of my original problem but it seems it wasn't. I will post a new thread about that as it is a different problem.
Thanks for the help on this guys!

 

[corus]

It would have also been interesting to know what experimental work you had conducted to verify the model as the only similar situtation I could see a head striking an inclined plane would be on a guillotine, as was used in France.

I could only imagine that it would be the british Health and Safety Executive (HSE) that would insist that the condemned man wore a safety helmet.

corus

 

[rstupplebeen]

The test looks similar to the standard DOT motorcycle helmet requirements that have propagated to military helmets and bicycle helmets.