Week - 8 Mass Scaling

MASS SCALING USING LS-DYNA

 

AIM:

 

1. To use mass scaling to reduce the runtime of a model and ensure the stability of mass scaling.
2. To reduce the runtime required to run the analysis of the given model using mass scaling technique and stability has to be completely intact.
3. To plot a histogram to compare the runtime with mass scaling trials.
4. To optimize both the DT and TSSFAC parameters.
5. To run the same model using implicit solver with necessary changes and compare implicit and explicit runtime

Note: The hard limit on mass scaling is 8%.

INTRODUCTION:

1. Mass-scaling is a term that is used for the process of scaling the element’s mass in explicit simulations to adjust its timestep. The primary motivation is to change the global compute timestep which is limited by the Courant’s stability criteria.

 

2. LS-DYNA allows two different types of mass-scaling using the DT2MS parameter from *CONTROL_TIMESTEP with the default set to no mass-scaling.

 

3. When DT2MS is less than zero, LS-DYNA adds mass of each element whose timestep is below abs (DT2MS) such that the element’s updated DT is equal to abs (DT2MS).

 

4. When DT2MS is greater than zero, LS-DYNA adds mass to elements whose DT is below abs (DT2MS) and “removes” mass from elements whose DT is greater than zero.

 

5. DT2MS>0 is seldom used while DTM2<0 is frequently used for overcoming the smallest computed timestep.

 

6. Care must be taken when using DT2MS<0 to ensure that the added mass does not have an adverse effect on the simulation accuracy. It is common practice to limit the percentage of added mass to less than 5% (at part level) in dynamic simulations.LS-DYNA outputs the percentage of added mass at component level which is a better indicator of amount of added mass due to mass-scaling.

Here we are provided with the fully prepared deck with the control time step card. Control_timestep card has a parameter called DT2MSwhich is the mass addition to prevent the lower timestep which happens because of the lower element characteristic length or due to the high deformation happens to the elements while it undergoing large stress.So inorder to prevent the simulation from error terminated or became too slow due to the low time step we are inputting this control_timstep card.DT2MS have accepted the values in negative as well as in positive values.DT2MS is basically the time step below which the mass will be added so as to increase the timestep of that particular element .This increasing the time step happens to the DT2MS value which is the lowest time step acceptable in simulation. This increasing time step because the mass is added to the nodes of the element So basically the density of the element at that area is inncreasing ,So when density increases the speed of the wave through it will reduce .This will increase the timestep. Increasing the timestep will finish the run in less amount of time as iterations will be reduced. But when there is lot of non-linearities involved in the problem,this usage of higher time steps may skip those non linearities and that will lead to inappropriate results.

PROCEDURE:

In this project, the runtime has to be reduced using mass scaling technique by varying the values of DT2MS and TSSFAC with percentage increase of mass within 8% using Explicit and Implicit analysis.

Explicit Analysis

 The given simulation file is made to run by keeping the given values for DT2MS = -3.5E-5 and TSSFAC = 0.9 to know the computational time required to run the simulation.

Trial 1: DT2MS = -3.5E-5 TSSFAC= 0.8

TSSFAC= 0.9

 

Trial 2: DT2MS = -5.5E-5 TSSFAC= 0.8

 

 

 TSSFAC= 0.9

For, DT2MS = -5.5E-5, the estimated clock time to complete the simulation for TSSFAC= 0.9 is 57 hrs. 32 mins whereas for

TSSFAC= 0.8 is 64 hrs. 44 mins. The percentage increase in mass for both TSSFAC equal 0.010034%.

 Trial 3: DT2MS = -7.5E-5 TSSFAC= 0.8

 

 

For, DT2MS = -7.5E-5, the estimated clock time to complete the simulation for TSSFAC= 0.9 is 31 hrs 39 mins whereas for TSSFAC= 0.8 is 47 hrs 28 mins. The percentage increase in mass for both TSSFAC equal to 0.9 and 0.8 is 0.3728%. The runtime has reduced whereas the % increase in mass has increased compared to trial 2.

 

Trial 4: DT2MS = -9.5E-5 TSSFAC= 0.8

 

 

For, DT2MS = -9.5E-5, the estimated clock time to complete the simulation for TSSFAC= 0.9 is 30 hrs. 28 mins whereas for TSSFAC= 0.8 is 34 hrs. 17 mins. The percentage increase in mass for both TSSFAC equal to 0.9 and 0.8 is 4.3136%. The runtime has reduced whereas the % increase in mass has increased compared to trial 3. The iteration for DT2MS value is continued till the percentage increase in mass is within the hard limit of 8%.

 

Trial 5: DT2MS = -1.0E-4 TSSFAC= 0.8

Trial 6: DT2MS = -1.1E-4 TSSFAC= 0.8

 

 

TSSFAC= 0.9

For, DT2MS = -1.1E-4, the estimated clock time to complete the simulation for TSSFAC= 0.9 is reduced to 19 hrs.7 mins whereas the percentage increase in mass is 20.393% which is beyond the acceptable limit. Hence, the iteration is continued to get the optimized value of DT2MS till the percentage increase in mass is within the acceptable limit of 8%.

 

Trial 7: DT2MS = -1.08 E-4 TSSFAC= 0.8

 

TSSFAC= 0.9

Trial 8: DT2MS = -1.08 E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 9: DT2MS = -1.04E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 10: DT2MS = -1.02E-4 TSSFAC= 0.8

TSSFAC= 0.9

For, DT2MS = -1.02E-4, the percentage increase in mass for both TSSFAC equal to 0.9 and 0.8 is 7.4771% which is within the acceptable limit but the iteration is continued to optimize the value of DT2MS to get a lowest possible runtime.

Trial 11: DT2MS = -1.022E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 12: DT2MS = -1.024E-4 TSSFAC= 0.8

TSSFAC= 0.9

 

Trial 13: DT2MS = -1.026E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 14: DT2MS = -1.028E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 15: DT2MS = -1.0282E-4 TSSFAC= 0.8

TSSFAC= 0.9

 

Trial 16: DT2MS = -1.0284E-4 TSSFAC= 0.8

 

TSSFAC= 0.9

 

Trial 17: DT2MS = -1.0286E-4 TSSFAC= 0.8

 

TSSFAC= 0.9

.

Trial 18: DT2MS = -1.0288E-4 TSSFAC= 0.8

TSSFAC= 0.9

Trial 19: DT2MS = -1.0290E-4 TSSFAC= 0.8

 

 

TSSFAC= 0.9

Trial 20: DT2MS = -1.0289E-4 TSSFAC= 0.8

TSSFAC= 0.9

 

For, DT2MS = -1.0289E-4, the estimated clock time to complete the simulation for TSSFAC= 0.9 is 17 hrs. 50 mins whereas for TSSFAC= 0.8 is 20 hrs. 14 mins. The percentage increase in mass for both TSSFAC equal to 0.9 and 0.8 is 7.9990%. Hence the value of percentage increase in mass is optimized within the acceptable limit of 8%.

Explicit Analysis
TRIAL	DT2MS	TSSFAC = 0.8			TSSFAC = 0.9			REMARKS
		CLOCK SPEED IN HRS AND MIN	CLOCK SPEED IN SEC	% MASS INCREASE	CLOCK SPEED IN HRS AND MIN	CLOCK SPEED IN HRS AND MIN	% MASS INCREASE
1	-3.50000E-05	136222	48 Hrs. 22 Min	0	174147	37 Hrs. 50 Min	0
2	-5.50000E-05	233058	64 Hrs. 44 Min	0.010034	207162	57 Hrs. 32 Min	0.010034
3	-7.50000E-05	170909	47 Hrs. 28 Min	0.3728	113939	31 Hrs. 39 Min	0.3728
4	-9.50000E-05	123444	34 Hrs. 17 Min	4.3136	109728	30 Hrs. 28 Min	4.3136
5	-1.00000E-04	138408	38 Hrs. 26 Min	6.4233	84848	23 Hrs. 34 Min	6.4233
6	-1.10000E-04	87396	24 Hrs. 16 Min	20.393	68870	19 Hrs. 07 Min	20.393	APPLIED FILTER FOR %MASS ABOVE 8% - CAN NOT BE USED
7	-1.08000E-04	108055	32 Hrs. 47 Min	16.477	61728	17 Hrs. 8 Min	16.477
8	-1.06000E-04	10986	28 Hrs. 03 Min	12.639	80616	22 Hrs. 23 Min	12.639
9	-1.04000E-04	72115	20 Hrs. 01 Min	8.8759	54778	15 Hrs. 12 Min	8.8759
10	-1.02000E-04	73529	20 Hrs. 25 Min	7.4771	101604	28 Hrs. 13 Min	7.4771
11	-1.02200E-04	62711	17 Hrs. 25 Min	7.5915	65231	18 Hrs. 07 Min	7.5915
12	-1.02400E-04	62588	17 Hrs. 23 Min	7.7075	73981	20 Hrs. 33 Min	7.7075
13	-1.02600E-04	104332	28 Hrs. 58 Min	7.8251	83288	23 Hrs. 08 Min	7.8251
14	-1.02800E-04	113415	31 Hrs. 30 Min	7.9443	64261	17 Hrs. 51 Min	7.9443	APPLIED FILTER FOR %MASS BETWEEN 7.9 AND 8%
15	-1.02820E-04	103446	28 Hrs. 44 Min	7.9566	64248	17 Hrs. 50 Min	7.9566
16	-1.02840E-04	72928	20 Hrs. 15 Min	7.9687	55396	15 Hrs. 23 Min	7.9687
17	-1.02860E-04	72914	20 Hrs. 15 Min	7.9808	64223	17 Hrs. 50 Min	7.9808
18	-1.02880E-04	62296	17 Hrs. 18 Min	7.9929	64211	17 Hrs. 50 Min	7.9929
19	-1.02900E-04	103365	28 Hrs. 42 Min	8.005	54775	15 Hrs. 12 Min	8.005	APPLIED FILTER FOR %MASS ABOVE 8% - CAN NOT BE USED
20	-1.02890E-04	72893	20 Hrs. 14 Min	7.999	64204	17 Hrs. 50 Min	7.999	APPLIED FILTER FOR %MASS BETWEEN 7.9 AND 8%

 

 

Histogram:

 

• From the graph, it is observed that the runtime required to complete the simulation for DT2MS = -3.5E-5 is more compared to DT2MS = -1.1E-4 whereas the percentage increase in mass is beyond the hard limit of 8% for DT2MS = -1.1E-4. Hence, the optimum value of DT2MS is -1.0289E-04 which has an estimated runtime of 17 hrs. 50 mins.

 

• The runtime required to complete the simulation for all the iteration with TSSFAC=0.8 is not stable as compared to TSSFAC=0.9. Hence the ideal value for TSSFAC is 0.9 for this simulation.

 

 

Implicit Analysis

 

For implicit analysis, keywords like CONTROL_IMPLICIT_AUTO, CONTROL_IMPLICIT_GENERAL, CONTROL_IMPLICIT_SOLUTION and CONTROL_IMPLICIT_SOLVER are added with necessary inputs to the given keyword file.

 

 

Then we run the same model in LS Program Manager. With the ITEOPT default value of 11, we found that the runtime is pretty much reduced as compared to the explicit case. Since the implicit can move with high time steps if the iterations required is lesser than the ITEOPT and ITEWIN values speedy solving of problem is a capability of implicit solver. Since the problem is not so complex, there may not be anytime step reductions happens. IF there is so much complexities occurs in model, then may be implicit may take so much time than the explicit.

 

 

Then we add the implicit cards such as implicit general, implicit auto for clobbering the simulation to implicit analysis (by Imflag =1).

 

 

 

 

Trial 1: DT2MS = -3.5E-5 TSSFAC= 0.9

 

 

 

 

 

COMPARISION IN EXPLICIT AND IMPLICIT ANALYSIS

 

COMPARISION IN EXPLICIT AND IMPLICIT ANALYSIS
TRIAL	DT2MS	TSSFAC = 0.9
		EXPLICIT	IMPLICIT
		TIME IN SEC	TIME IN SEC
1	-3.50000E-05	174147	6
2	-1.02890E-04	64204	5

 

In Implicit analysis, each time step has to converge, but we can set pretty long-time steps and it is based on iterations.

Explicit on the other hand doesn’t have to converge each time step, but for the solution to be accurate time steps must be super small.

Though the running time for implicit analysis is shorter than the explicit analysis method because the timestep is defined larger in implicit compare to explicit analysis and we have used constant timestep formulation

 

From the table it is observed that the runtime required to complete the simulation varies drastically for different values of DT2MS using explicit analysis whereas for implicit analysis

the variation in runtime for different values of DT2MS is negligible. Hence the parameters like DT2MS and TSSFAC does not affect the runtime of implicit analysis.

CONCLUSION:

 

1. The objective of mass scaling is achieved for the given model by varying the values of DT2MS and TSSFAC using trial and error method.
2. By considering the hard limit on mass scaling being 8%, the optimized value of DT2MS = -1.0289E-4 and TSSFAC= 0.9 with lowest possible runtime required to complete the simulation is 17 hrs. 50 mins.
3. The values of DT2MS and TSSFAC does not affect the runtime required to complete the simulation for implicit analysis.
4. After, that we have run the analysis using implicit control cards which gives us the result that the implicit analysis is based on

         iterations and doesn't depend on mass scaling. The concept of mass scaling and its necessity in explicit analysis.

5. As the dt2ms value increases the timestep will increases with the addition of the mass in the model. After a particular limit we cannot accept that mass addition as it will leads to destruction of the physical sense of the problem. Within a limit this can be used to speed up the simulation with the use of the TSSFAC which is a cautionary parameter to stabilize the model.