event simulation is pretty unrealistic

[prefetch]

i was doing some simulation work and the results looked kind of funny, so i thought i do a super simple test.

i made a "box" of 1/2 inch thick structural steel going 15 mph towards a 1/2 inch thick "plank" of structural steel going 15 mph towards it.  i used the non-linear physical material option.

the simulation ran, and the results are kind of unbelievable.

i think i can toss steel at 15 mph, and if a friend of mine threw some steel and i threw some steel and they collided, i'm pretty sure it would NOT gash through 1/2 thick plate.  it's just not going to happen like this:

so how can i really trust the event simulator to get anything even close to reality if a super simple test like this gives ridiculous results??

has anyone else spent anytime with the event simulator?  thoughts?

[AndrewSears]

Can you attach your model so we can look at your setup?  I will start working on my own with the information you provided.  

Andy 

[prefetch]

attached.

i'm hoping i'm just doing something dumb, but maybe there is a problem with event simulation.

[AndrewSears · Accepted solution]

Correct me if I am wrong but you stated "i'm pretty sure it would gash through 1/2 thick plate."  I think you meant to say it would not, correct?  If that is the case then you need to turn off element deletion in the settings.  This is just another way for you to tell when something has plastic deformation or rupture of the structure could occurred.  

When you turn it off, you see more expected results.

I do have a suggestion for you.

15 mph = 264 in/sec

You set the parts 0.5 inches away from each other for a 0.003 second simulation.  264*.003= 0.8 inches.  Event Simulation works best with parts starting very close to each other.  You are wasting 2/3 (guessing) of the computation time by moving parts through space so far.  You already know the initial velocities so just start the parts off very close to each other.

Hope this helps.  Let me know if I missed something.

Andy 

[AndrewSears]

Correction:  I stated that you should turn of element deletion.  What you really need to do is use your judgement on keeping it on or turning it off.  Many simulations will provide more realistic results with element deletion on.  My suggestion is to run a simulation both ways, review the results and go from there.  They are both valid but just help you look at the results in another way.

Andy 

[prefetch]

yes, i fixed my type - i meant to say that i'm pretty sure it would not make a gash in the steel.

so it looks like element deletion leads to some pretty wacky results unless you realize that element deletion just means "some damage happened to this element so we're just going to delete it" (which is what i think you are saying.)

i'm not sure i see any advantage to deleting elements or how that would make a simulation more realistic, at least with the level of speeds and types of materials i'm using.  perhaps if i was simulating something that could create enough energy that it would vaporize the materials involved then deleting would be realistic?

so i'll switch element deletion off - and also, thanks for the tip about making the computation go faster by placing elements close.  that makes perfect sense.

[jeremy.wiesner]

Hi @prefetch,

The element deletion feature is a very important aspect of Event Simulation, but it must be used cautiously. This is the feature that allows you to test your model all the way to the point of failure, or simulate extreme events such a bullet impacting a piece of steel armor. 

At the present moment, you can only specify element deletion for Event Simulation based on maximum principal strain (we have plans to expand the element deletion options in the future). If you enable this option, the simulation solver will inspect the strain values in every individual finite element in your model at every time step to see if the strain exceeds the deletion threshold that you specified. If the threshold is exceeded, the element is deleted and removed from the simulation. One useful thing to note is that when elements are deleted and new surfaces are exposed in your model, Event Simulation will automatically handle these new contact surfaces for you assuming you had contact enabled in your study.

The problem that you're seeing in this model is that the default deletion criteria of 1% strain is too low for metal materials. Generally speaking, metal is a very ductile material that yields quite a bit before rupturing. Intuitively what you would expect to see in your simulation is the metal parts dent/bend/buckle/etc before you see a perforation develop. With a more realistic deletion criteria value, this is the behavior you will observe. To give you an idea how low the 1% value is, I noticed you chose the High-Strength Structural Steel model out of the Fusion 360 Nonlinear Material Library. This model is based on ASTM A992 Grade 50 structural steel, which has a typical elongation value in the range of 18-21%. I suggest giving your simulation another try and modify the deletion criteria option in the study settings to 0.2 (20%), instead of the default value of 0.01 (1%). 

If you haven't seen them already, I also suggest you check out the Event Simulation sample models called "Axe Chop", "Projectile Impact", and "Three Plate Impact". If you open the Data Panel, go to the "Simulation Samples" folder, then to "3- Event Simulation Samples" and you can find these models. These three models in particular provide great examples why element deletion is so important to Event Simulation. The key ingredient, however, is calibrating your deletion criteria so that the material you're trying to simulate is failing in the simulation in a way that is consistent with reality. This usually requires a bit of engineering judgment on behalf of the user, but it's something that we plan to better automate in future releases. Our goal is to embed realistic deletion criteria into the nonlinear material models, that way you won't have to worry about figuring out the correct strain-based deletion value(s) for your material.