Simple pendulum
Altair EmployeeHi
Do we have any load collector in Radioss to impose oscillatory motion on a body. For example I have a simple pendulum in which I have arrested the DOF of the centre node (except for rotation along X direction) and gave an intial velocity and gravity to the bob. If I do so I am getting oscillation in one direction only. However I would like to see a full oscillation. Could you specify if there is any option???
The velocity I have chosen is 5 m/s. Here is the simple pendulum model.
Answers
-
Hi Raghav,
Have you allowed it to run for enough time i.e more than time period of one oscillation?
0 -
Raghav,
There is no specific load collector for imposing oscillation. Please share the model file (_0000.rad) so that we will have a workaround on the same.
0 -
Hi,
The theoritical time period is 1.099sec which is too high for my simulation. My simulation has 120ms of run time. Could you suggest me a way?
0 -
Hi,
As the theoretical time is much higher than the run time, you will not be able to see on complete oscillation you run it for 1.099 sec.Generally , explicit simulations are of millisec. magnitude , but still if you want you can go for using AMS for the simulation.
Before that, it is important to understand the objective of the simulation, like if you want to study only motion kinematics considering pendulum bob as rigid body , motion solve is the right place. If you want to do any impact with the bob, instead of complete travel , use hand calculation to get the velocity at the position where impact will happen and then use initial velocity as load collector.
Thus, depending on the objective this can be defined.
Another option is to go for a transient analysis in Optistruct which being implicit analysis will not take longer times.
0 -
Hi Raghav,
As Gopal mentioned in the last reply the objective of your analysis is important. For kinematic study, use MotionSolve and we have an interactive video tutorial available for it. Please visit https://www.youtube.com/watch?v=ZPMbbrwWuaY for the video series of multibody simulation of a pendulum.
We have an example in Help Menu which simulates the oscillation and wave propagation of a group of pendulums, arranged in a line, when impacted at one end and it is done in RADIOSS. The example file is attached along for reference.
0 -
Hi
Thanks for the reply. My objective of study is to compare the results of Conventional mass scaling( /DT/NODA/CST) with AMS and no mass scaling load cases.
0 -
Hi Raghav,
It's interesting to see this comparative study and also excited to know that you are exploring more. In general , for a quasi-static process, mass scaling is generally acceptable because inertia effects should be small. On the other hand, if inertia/dynamic effects are important, user can still use mass-scaling but should make sure that the overall percentage increase in mass is very small.
Please refer this example in Help Menu where a similar comparative study is done: Example 44 - Blow Molding with AMS . Also refer http://altairatc.com/india/previous-events/2013/images/presentations/RNL-02_Study_of_different_Magna_Steyr.pdf for similar paper.
0 -
Hi,
The load case is this. I have a simple pendulum (3D body), I took the mid surface of it . I arrested the top center node in all directions except in Y (as my model is build in Z-X plane). Then I gave an initial velocity of 5 m/s to the bob. My time period with respect to the length of string is approx 1.099s as I mentioned before. I would like to have an animation with full oscillation.. I have tried doing it just more than half time period to see if it reaches the maximum point and start to trace back. But it's not happening.
Does running the simulation for the entire time period 1.099s gives me the result I want (full oscillation) or should I include anything in the model??
I have chosen this to mainly study the effect of rotational nodal mass at an element while employing mass scaling..
Thank you,
0 -
Hi Raghav,
With these loading conditions it doesn't seems to work as expected. Apply imposed displacement (/IMPDISP) loading condition (create a function to define the loading) and also apply gravity to the model and try simulating the same.
0 -
Hi,
I have tried the model with imposed displacment. But the results are not encouraging... Pl find the load collectors I am using
gravity, Imposed displacement . I have a created a function with one value 0.52 radians which is around 60 degrees. And I have the centre node free along x rotatation. I also tried with angular velocity as well.. creating a function with one value 5.71 rad/sec... but unfortunately no encouraing results. Can you suggest me on this..
0 -
Hi,
Try applying gravity also and update. Meanwhile please share the model file (imposed displacement load file) so that we can have a workaround with your model.
0 -
G'Day Experts,
I have a question regarding simple pendulum. I tried to simulate simple harmonic motion using a simple pendulum with an initial angular displacement of 30 degress. When I run my simulation and plot the contour results, plastic strain is being recorded on the model which is absurd. as there should be no deformation on the pendulum acting under gravity. I chose M_36 Johnson-cook material card. But the idea of chosing an elasto plastic material is justified as the pendulum is made up of steel/brass or whatever material which is not purely elastic. Can anyone tell me the reason why plastic strain is being recorded?? Surprisingly there is no deformation on the pendulum. Does plastic strain appear as a default feature when using elato plastic card?? By the way the plastic strain magnitude is significantly low..(0.00048)
Thanks in advance,
0 -
Hi Raghav,
Can you share a picture of the same showing the plastic strain in the model.
0 -
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi Raghav,
Can you replace the material with /MAT/LAW2 (elasto-plastic material) and check the results and update.
Also add /ANIM/SHELL/TENS/STRAIN/ALL (for shell elements) in the engine file.
0 -
Hi,
What is the mass of the bob? What are the parameters entered in material model , share screenshots for the same.
0 -
The overall mass is 326.7 g in which bob is 180. Parameters are E, poissons ratio and density, Used a stress-strain curve as function.
0 -
Hi,
How the link and the bob are connected, do we have a rigid element there or node to node connectivity. Both link and bob are made of steel?
Also any hand calculation is done to check the stress due to self weight, this is generally very low and thus neglected in simulations.Also is there any inconsistency in unit systems, check this also.
0 -
Hi,
Answer to the series of questions.
- The bob and the link are a single entity modeled in NX and hence it a single body.
- I made the bob exclusively rigid with 1D rigids. The intention behind this is to apply gravity to the master node instead of application to the entity set of bob. Also ideally there will not be any deformation on the bob in whatsoever case (either an elastic / elasto-plastic material card).
-Yes both bob and link are made of steel.
- I didn't check the stresses due to self weight because ideally there should not be any plastic strain meaning no stress should generate.
-No inconsistency in units.I have used mm,ms,MPa,g,N scale prescribed under RADIOSS.
The only possiblity I see is because of using a elasto plastc card...
0 -
Hi,
For plastic strain to come , the object has to go up to the stress which for steel is minimum 210 MPa and in this case this value cannot be reached at all but still plastic strain exists. The value is very less however expectation is it should be zero.
What is the curve used for stress-strain?
As we do not expect any plastic strain other option is to use only elastic material for this.
0 -
The curve used for M36 material card has some values of stress-strain beyond the elastic limit. I have the experiment data on USTM. As the data is really huge (there are 3152 values of gauge length Vs traction force). I scrutinized the data to 13 values and calculated stress-strain out of which 5 of them are in plastic range. Do you think this strain could be because of those values? If so then I guess the RADIOSS code generates the requested contours whatsoever may be the load case.
0 -
Hi Raghav,
This behavior seems to be because of the curve you defined.Assign a different material law (Mat Law2, elasto plastic law) and you can recheck whether plastic strain is developing in the model or not.
0 -
Hi,
What is the stress value at which you have given zero plastic strain. IN the MAT LAW 36 curve, the first reading should be like the following eg:
Stress Strain
240 0.0
248 0.06
and so on
What is the curve you have used?
0 -
Hi again,
I ran the simulation with M2 material card. Here is the plastic strain. BUT THE STRAIN TENSOR IS ZERO.
0 -
<?xml version="1.0" encoding="UTF-8"?>
0 -
0.00111 241 0.001585 350.1 0.002002 444.4 the values of strain versus stress just after the elastic limit.
0
