Applying Angle velocity to a component in Radioss
Hello Everyone,
i am a newbie with Radioss and have a question regarding Applying an angular velocity to a cylindrical components.
so i already set the BCS for both of the cylinders (free in z rotating axis) and try to apply the angular velocity using /IMPVEL. However i dont get any result at the end of the simulation (The cylinder did not even rotate a bit).
I am not sure if i use the correct parameter in the IMPVEL or maybe i missing something in the Boundary conditions. Below i attach a screenshot of my model.
Can someone please help me?
Thank you
Greiffith
Best Answer
-
greiffith tanzal_21385 said:
Hi Paul,
Thank you for the modified .hm model. I already studied the model and i try to built a similar model. 1 Cylinder and also trying to rotate it.
After the simulation is done, i got a warning (error) that said "Incompatible kinematic condition". As far as i know, this means that the rotational DOF is accidentally constrained but I did not constraint it.
I upload the new .hm model, can you please check my model and give me some input? would appreciate it.
The incompatible kinematic condition warning is because the BCS and IMPVEL are not using the same skew/frame (The ZZ of the IMPVEL is therefore not necessarily aligned with the free DOF Wz of the BCS since the 2 skews could be aligned differently), but because in this case they are the same orientation it is ok to ignore the warning. The reason your cylinder is not spinning is not related to the warning, it is because it was still set using /MAT/RIGID, you should use /MAT/ELAST instead (the /RBODY is doing the job of making it rigid). The final thing is that your chosen timestep 3e-3 is too large, you will see the cylinder 'grow' significantly at that timestep due to linearisation in the increments, here I reduced to 1e-4 (I wouldn't recommend any bigger for this model, ideally smaller, but once you add deformable elements it will be smaller anyway and likely no need for the /DTIX), as a final point, you say you are new to HM, but you are using a 5 yr old version (and 4 releases ago, 2017 isn't even supported officially today) and you are using the 'old' HM interface. I would recommend updating to 2022 (2022.1 is the latest version)
0
Altair EmployeeAnswers
-
Hi, /IMPVEL is an imposed velocity (like a moving boundary condition) where you prescribe the velocity (in this case rotational velocity) vs time.
I can't see from your screenshots what you have applied it to (what is in node groups WO and WU?), but it would only make sense to do this for this model at the main node of an /RBODY for your cylinders, and it doesn't look like you have any /RBODY defined? This might explain why you have no motion in results.
What is the aim of your analysis?
It might be that what you are looking for is /INIVEL/AXIS rather than /IMPVEL?
/INIVEL/AXIS is an initial condition that you could set on all the nodes of the cylinder and each node would have it's own specific velocity vs the axis of rotation (the rotational velocity is set at the start of the analysis and after that the behaviour comes down to the physics of the problem).
If you can share your model I can have a look to see if there are any other issues and set it up for you both ways (with /RBODY and /IMPVEL or with /INIVEL/AXIS)
0 -
Hi @Paul Sharp
Thank you for your reply, I used /IMPVEL exactly because the angle velocity is changing with time. I assume with /INIVEL/AXIS the angle velocity is constant?
I already try to add the rigid body but the result still not good, the cylinder still not rotating.
The aim of my analysis is a simple roll forming of a plate, for that i decided to first built a simple similar model.
yes, of course i can share my model. I would appreciate any input from you, because it is my first experience dealing with Radioss.
0 -
greiffith tanzal_21385 said:
Hi @Paul Sharp
Thank you for your reply, I used /IMPVEL exactly because the angle velocity is changing with time. I assume with /INIVEL/AXIS the angle velocity is constant?
I already try to add the rigid body but the result still not good, the cylinder still not rotating.
The aim of my analysis is a simple roll forming of a plate, for that i decided to first built a simple similar model.
yes, of course i can share my model. I would appreciate any input from you, because it is my first experience dealing with Radioss.
Hi, yes, ok, /INIVEL/AXIS isn't even constant, it is just instantaneous (initialising the velocity)
You had used /MAT/RIGID this is deprecated and not often used, you should put everything in an /RBODY instead
I modified your .hm
In the attached, the 2 Cylinders are in an RBODY each, their main nodes are constrained in DOFs 1-5
The Main Nodes have /IMPVEL referencing functions (WO -20,0,-20 and WU 10,0,10) and frame fixes are defined at the centers of the RBODY to give a reference system
The Nodes of the cylinders are also in /INIVEL/AXIS (WO -20, WU 10) per your original model
Because everything is Rigid, I added a /DTIX card in the engine fixing timestep at 5e-6
One of the example models has a similar set up, you might want to check that out too?
0 -
Paul Sharp_21301 said:
Hi, yes, ok, /INIVEL/AXIS isn't even constant, it is just instantaneous (initialising the velocity)
You had used /MAT/RIGID this is deprecated and not often used, you should put everything in an /RBODY instead
I modified your .hm
In the attached, the 2 Cylinders are in an RBODY each, their main nodes are constrained in DOFs 1-5
The Main Nodes have /IMPVEL referencing functions (WO -20,0,-20 and WU 10,0,10) and frame fixes are defined at the centers of the RBODY to give a reference system
The Nodes of the cylinders are also in /INIVEL/AXIS (WO -20, WU 10) per your original model
Because everything is Rigid, I added a /DTIX card in the engine fixing timestep at 5e-6
One of the example models has a similar set up, you might want to check that out too?
Hi Paul,
Thank you for the modified .hm model. I already studied the model and i try to built a similar model. 1 Cylinder and also trying to rotate it.
After the simulation is done, i got a warning (error) that said "Incompatible kinematic condition". As far as i know, this means that the rotational DOF is accidentally constrained but I did not constraint it.
I upload the new .hm model, can you please check my model and give me some input? would appreciate it.
0 -
greiffith tanzal_21385 said:
Hi Paul,
Thank you for the modified .hm model. I already studied the model and i try to built a similar model. 1 Cylinder and also trying to rotate it.
After the simulation is done, i got a warning (error) that said "Incompatible kinematic condition". As far as i know, this means that the rotational DOF is accidentally constrained but I did not constraint it.
I upload the new .hm model, can you please check my model and give me some input? would appreciate it.
The incompatible kinematic condition warning is because the BCS and IMPVEL are not using the same skew/frame (The ZZ of the IMPVEL is therefore not necessarily aligned with the free DOF Wz of the BCS since the 2 skews could be aligned differently), but because in this case they are the same orientation it is ok to ignore the warning. The reason your cylinder is not spinning is not related to the warning, it is because it was still set using /MAT/RIGID, you should use /MAT/ELAST instead (the /RBODY is doing the job of making it rigid). The final thing is that your chosen timestep 3e-3 is too large, you will see the cylinder 'grow' significantly at that timestep due to linearisation in the increments, here I reduced to 1e-4 (I wouldn't recommend any bigger for this model, ideally smaller, but once you add deformable elements it will be smaller anyway and likely no need for the /DTIX), as a final point, you say you are new to HM, but you are using a 5 yr old version (and 4 releases ago, 2017 isn't even supported officially today) and you are using the 'old' HM interface. I would recommend updating to 2022 (2022.1 is the latest version)
0 -
Paul Sharp_21301 said:
The incompatible kinematic condition warning is because the BCS and IMPVEL are not using the same skew/frame (The ZZ of the IMPVEL is therefore not necessarily aligned with the free DOF Wz of the BCS since the 2 skews could be aligned differently), but because in this case they are the same orientation it is ok to ignore the warning. The reason your cylinder is not spinning is not related to the warning, it is because it was still set using /MAT/RIGID, you should use /MAT/ELAST instead (the /RBODY is doing the job of making it rigid). The final thing is that your chosen timestep 3e-3 is too large, you will see the cylinder 'grow' significantly at that timestep due to linearisation in the increments, here I reduced to 1e-4 (I wouldn't recommend any bigger for this model, ideally smaller, but once you add deformable elements it will be smaller anyway and likely no need for the /DTIX), as a final point, you say you are new to HM, but you are using a 5 yr old version (and 4 releases ago, 2017 isn't even supported officially today) and you are using the 'old' HM interface. I would recommend updating to 2022 (2022.1 is the latest version)
Hi Paul,
Thank you for the explanation. My model is now working.
yes i am new to HM because usually i use LS-Dyna. I will upgrade the software as per your recomendation.
0 -
No problem, another thing you might be Interested in if you are familiar with LS-DYNA input is that you can run Radioss with Dyna keyword input format in the latest versions (recommend 2022.1) the supported input is in the manual. It might be easier to build the models in the format you are more familiar with and just run them in Radioss?greiffith tanzal_21385 said:Hi Paul,
Thank you for the explanation. My model is now working.
yes i am new to HM because usually i use LS-Dyna. I will upgrade the software as per your recomendation.
0