Comparing the simulation results of the base setup and improved shell element properties

AIM

Comparing the simulation results of the base setup and improved shell element properties

OBJECTIVES

1.Using the crash beam file , change the run time to 55 ms.

2. Change the number of animation steps during simulation to a minimum of 25 and maximum of 60.

3. Run the base simulation without any modification to element properties.

4. At the end of the simulation, do the energy error and mass error checks and determine whether results would be acceptable.

5. If acceptable, Plot rigid wall forces, internal energy, hourglass energy, contact energy, total energy of the simulations.

6. Change all the shell elements properties to recommended properties, save this as separate, rad files and run the simulation.

7. Follow steps 5 and 6.

8. Comparing all the results for both simulations

PROCEDURE

First we need to import the Crash beam RAD file into the hypermesh using the import solver deck option.In the first case there is no need to look at the properties.

Fig 1.1 Imported crash beam Fig 1.2 Material card Fig 1.3 Close out panel Fig 1.4 Hat section property CArd

property card

In the property cards we can see the values of some parameters are spcified as in DEF shell,Which is already set in this card and is called a Default shell card.When we give input as Zero for parameters ,Then the radioss will assign the default value which is assigned in the default shell.We need to set the run time to 55 ms.And also We need to obtain the animation in between 25 ms and 60 ms.

To set the run time to 55 ms ,We need to open the RUN CARD and set Tstop as 55 ms.To set the animation steps Open the ANIM/DT CARD and set Tstart time as 25 ms and set Tfreq as 3 to get 10 animation frames.Then we can set the print value which the Frequency at which the number of cycles are written in the engine out file.

Fig 1.5 ANIM/DT CARD Fig 1.6 RUN CARD Fig 1.7 PRINT CARD

Then we need to run the analysis,for that Place the radioo button in analysis option >select Radioss>choose the directory to which the we want the hypermesh to store the files produced after analysis>Type in the option area _nt 4 for specify the number of CPU cores we need to use for analysis>Run.THis will store the analysis files in the directory we specified.

What we should do to check whether the analysis is acceptable or not?For that we need to check the values of the Energy Error and Mass Error.This can be viewed in the engine output file which is of having name in the format 0001.out. If the values of the energy error is below the 15% then we are good to go.For a good model this value can be maximum upto +/- 5% .Negative values of the energy error means the energy has be dissipated from the system and possitive value means the system has absorbed the energy .While deformation is happening,the strain energy gone to high but this doesnt cause the energy error,beacuse the one form of energy is gets converted to another energy form,Here there is no loss of energy occurs so the energy error must be zero.But in the some cases like where the Hourglass formation occurs there is a good amount of chance for the Energy error to occur.

Hourglass is the zero energy modes occurs while deformation occurs.Here the deformation of the component occurs so the the nodes are actually moving ,But the stress value at those regions are not gets recorded and so those nodes movement has no effect on the nearby nodes tht causes unstable analysis.This will gave us the unrealistic simulations.This is most happens in case of under reduced integration elements.The shell elemets like QEPH (Quadrilateral ElastoPlastic Physical Hourglass Control)) and Q4 elements are under intergrated shell elements.But each of these element uses some methods to Reduce the hourglass formations.

Two methods used to reduce the Hourglass formation are

1) Perturbation Method

in this method , a viscous force will be applied when there is hourglass formation occurs. When there is a hourglass formation gets detected this ,method provides some viscous damping to the elements and so that the deformation gets stablized.Viscous forces generate hourglass forces proportional to components of nodal velocity contributing to hourglass modes.The force applied is depended upon the stiffness of the material.So this force which is the viscous damping force ,It is an artificial energy created just because to stablize the perfformance of unintegrated elements.So this creation of additional energy can cause possitive energy error.This energy is called HOURGLASS ENERGY.This is most effectively used by Q4 elements where the Ishell value is 1,2,3, or 4.This method is not useful in case of large rotation and distortion in case of elatic loading -unloading happens.But it can be used in case of high strain applications like in case of explosives.

2)Physical stablization Method

In this method when the hourglass is detected ,a stiffness is added to the material level itself.When the stiffnes is improved ,deformation gets stablized. Stiffness forms generate hourglass forces proportional to components of nodal displacement contributing to hourglass modes. It is able to reduce total accumulated hourglassing deformation .This method is preferred for lower rate problems, including crash simulation.Stiffness forms may artificially stiffen response. Therefore, reduce hourglass coefficient, say from .10 to .03, to minimize stiffening effect.So with adjusting the hourglass coefficient we can use the QEPH elements for most of the applications.

Fig 1.8 Engine output File

Energy error as w can see from the output file is about -10.3 %.THis is an acceptable value as the values upto 15% is acceptable even though a good model should only have 5 %.This much energy error occurs because of the hourglass energy used by the solver to reduce hourglass deformation.That energy is an artificial energy creayed by solver.This happens only because we used the Q4 shell elements which is a reduced integrated elements with viscious mode of hourglass stablization.So that cause the energy error.Mass error is also is also very very less value and is about .1695x10^-3.So the mass error and energy errors are under the limits so the analysis is acceptable and we can proceed to extract the energy curves from the time history file.

Open the hypergraph2D >openb the relevant folder containg the T01 file of the analysis we done above >select Global Variables>Select internal energy,Kinetic energy,Hourglass energy,contact energy>select MAG>Apply..This will gave us the energy curves with respect to time.Similarly select the rigid wall>select Resultant normal force.resultant tangential force,Total resultant force>Select MAG>APPLY.This will gave us the rigid wall forces.

Fig 1.9 Energy Graphs Fig 1.10 Contact forces graph

From the energy graph it is evident that the total energy is not constant. But according to the law of conservation of energy the total energy must be constant.So what made this reduction is the Hourglass energy and contact energy.As the velocity is applied to one end of the crash beam ,we are doing an external work,This external work is getting converted to strain energy when the crash beam gets deformed.That leads to increase in Internal energy.As the strain increases the strain energy also increases and therby the internal energy also increases.So the true cause of energy error is perfectly perfectly figure out from the graph itself.

As we pressed the crash beam against the rigid wall with velocity ,the full slave nodes of the crash beam is having the contact with the rigid wall surface by using Type 2 tied contact.So the normal force devoloped is very high at the instant we are pressing the crash beam against wall.But very quickly due to high velocity the elements gets deformed and the slave node gets disconnected from contact ,So the contact area reduces. That leads to generation of small normal force.Then again the velocity is applying into it so the crash beam gets buckled and it gets kinked with the deformed elements.That created a little more elements comes in contact with rigid wall and that again increases the normal force little much.Due to these contact of elements happens ,some amount of tangential force also generated depending on the area of contact because of the friction between the contacts.Resultant Normal force + Resultant Tangential force = Total Resultant Force.

CASE 2

In case 2 we are improving the properties of the elements use with the recommended properties.Recommended properties are given below:

Parameters		Comment
Ishell =24		QEPH 4 nodes shells are best combination of cost and accuracy.
Ismstr=2		Full geometric nonlinearities (default)with possible small strain formulation activation in RADIOSS Engine.
Ish3n=2		Standard 3 noded shell (C0)with modification for large rotation(default)
N=5		Number of integration points set to 5 for accuracy bending.
Ithick=1		Thickness change is taken into account for accuracy.
Iplas=1		Iterative plasticity for good accuracy.

Ishell =24:This is used for the different formulation of shell elements.Earlier we used the Q4 element which is having theIshell value os 1,2,3,or 4.here in this case we are using I shell = 24 it is representing the QEPH elements.These elements are improved elements and they are hourglass stablized by physical stablization through stiffness method even though they are underintegrated.

Ismstr = 2: Ismstr is the stress train formulation.It can be of values 1 or 2 .For value of 1,solver will compute the the strain value based on the intial length.For Ismstr =2 means the strain value is computed bases on the current lenth of elements.

Isch3n= 2 :Ish3n is used for tria elements. ISh3n =2 : It is the formulation os strandard triangyukar elements with modification for large rotation and Hour glass stablized.

N=5 : N is the number of integration points.here 5 integration points are distributed like 2 on the both ides of opposite faces of shell elements and one at the middle.It helps in increasing accuracy.

Ithick =1 :Ithick is the formulation to consider thickness change or not.When I thick is 1 thickness change is taken into account.this is need iin case of plastic deformation.If Ithick = 2 ,Thickness change is not taken into account.

Iplas = 1 : Iplas is used to choose the method to calculate the post yield stress.When I plas =1 solver use the iterative projections with three Newton iterations.And I plas =2 for Radial return.I plas is actually like a switch when plastic region comes.

So we need to make improvements for these properties and then we need to run the analysis.

Fig 2.1 Cross rail property Card Fig 2.2 Hat section property card

So we changed the Property card paramters to recommended values.Then again run the analysis and obtain the graphs.

Fig 1.3 Hourglass stablized

The simulation frame itself shows the hourglass stablization due to the improved parametrs and QEPH elements

Fig 1.4 Engine output file after Improved properties

The engine output file shows the energy error is near to zero.This is beacause of the hourglass stablization by the QEPH elements.The mass error is same as above.mass error occured because there may be addition of any mas at any nodes when the time steps drops below the critical time step.Since the mass error and energy error are acceptable we can proceed for analysis.

Fig 1.5 Energy Graphs Fig 1.6 Contact force graphs

We can see that the total energy is here constant due to the zero hourglass energy due to stablization by improved properties of shells.we can clearly say that the her Internal energy + kinetic energy = Total energy.Some amount of kinetic energy is actually converted to internal energy when the crash beam is deformed.In the contact Force graph we can see like as usual at the time of impact the normal force and after that the normal force is reduced to a low level .But after the 50 th milli second we can see there is a peak in the curve ,This peak is formed because after the 50 th second the crash beam is kinks and buckled in such a manner that the more elements of the crash beam is comes in contact with the rigid wall.So as the contact area increases contact energy as well as tangential force.Tangential force increases because of increased friction because of increased area of contact.

Fig 1.7 more lements are in contact with rigid wall

CONCLUSION

It is quite evident that the improved properties will gave us more realistic and acceoptable simulation results.QEPH shells ability to stablize the hourglass formation without causing the energy error is clearly visible in second case.

LINK FOR FILES

https://drive.google.com/drive/folders/1faeV1pwTI6_cY7HTNnTilOgryyA03ojW?usp=sharing