6 Replies Latest reply on Nov 2, 2018 9:13 AM by Glenn Schroeder

# hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to create

hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to create such situation ? i saw some time curve options under force and i tried doing that. Can anyone guide me what does the values in x and y show?

Also, can anyone guide on how to create stress-strain graph, to decide whether my model fails or not for non linear analysis.

i have attached an image showing my model.

Thanks.

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

Hi Sagar,

Glenn Schroeder is our residential guru on designing barriers for vehicles to destroy themselves on but I'm not really sure how far (if at all) he gets involved in the simulation and analysis side of things. Perhaps he can comment when he gets online later..........

Dave.

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

Dave Bear wrote:

Hi Sagar,

Glenn Schroeder is our residential guru on designing barriers for vehicles to destroy themselves on but I'm not really sure how far (if at all) he gets involved in the simulation and analysis side of things. Perhaps he can comment when he gets online later..........

Dave.

I'm not really involved in simulation at all, and anyway, they use LS Dyna here.  As I understand, and my understanding is minimal at best, it is much better since it takes into account the crush and collapse of the vehicle and not just a solid impact.

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

Typically this type of crash and impact analysis are typically done using a dynamic solver since we have accelerations and velocities and thus inertia is important. You could use the nonlinear transient dynamic solver. This link shows a nice tutorial of an impact analysis between two bodies: Non Linear Dynamic Contact with SolidWorks Simulation - YouTube . Hope this helps.

Have in mind though that if things are collapsing and failing too much then typically implicit solvers like the ones used by SW (as far as I can see there is no explicit one), are not ideal. Often were a crash scenario is simulated, explicit solvers are typically used.

Hope this helps. Of course as suggested it might be better and mandatory/necessary to do tests since modelling and simulation needs to be verified anyhow, and one cannot just rely on simulations.

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

You can get  a reasonable approximation with a NL static analysis. You need to get the load displacement curve at the impact location and the point on the curve that is of interest to you is when the area under this curve is equal to the vehicles energy. So apply some large force and let is run and then plot the displacement vs. Time and time is equal to the load fraction When you use 0 to 1 as the time. This approach is typically preferred as it is computationally quite efficient over explicit or implicit simulations that involve contact. The contact iterations in am implicit analysis are very expensive typically and 1.5 seconds is  an eternity in an explicit solve even though the solves are quite robust.

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

That sounds like a good way forward Bill (this assumes that the car is very stiff of course) since it would be difficult to model a car crash in SW I believe. The reason for that though is the fact that in an implicit solver, it needs to do NR iterations for every time step, thus if we have plasticity (you will have that in a crash that is for sure), GNL, and contacts it can require many iterations (and thus time) before the residuals are small, and we can move to the next time increment and so on. In an explicit solver no NR iterations are needed, just time marching, and that works fine as long as we have a good estimate of Delta_t which is often calculated by the solver, so it makes the solution stable (implicit is unconditional stable, delta_t is only to capture the motion of interest, thus a matter of sampling fast enough).

• ###### Re: hi, i have a problem in which a car weighing 4000kg hitting a barrier at 6km/h. Now i need to run simulations on barrier. i choose the non linear static analysis. i want to find impact on barrier after 1.5 seconds after car hits. is it possible to cre

it sounded like the idea here was to design the barrier so a rigid assumption for the car would be conservative, that is to say it doesn't absorb any energy in the crash. If the barrier is steel then the time step in an explicit analysis which is govern by the speed of sound in the structure and the smallest element size would likely be in the order of 10e-8 seconds with solid elements. This would require a lot of time steps to reach 1.5 sec and that assumes that 1.5 seconds is enough time to stop the vehicle or get to the point of interest which seems like a bit of guess from what has been provided thus far. SWX does have an explicit solver option. It is used in the drop test study but it is otherwise not available. Of course beam elements would make the job pretty straight forward in an explicit code if they were available but then again if you had proper beam element in the implicit code it would be equally advantageous to use them in the implicit NL static approach.