Belt Pulley Simulation
Altair EmployeeHi ,
I had a query I have modelled a rope pulley sought of arrangement. I had few queries in regards to these cases.
1) Is the modelling of the rope with truss element correct. As rope doesn't need bending and all. Are there any other available elements or Is this correct?
2) I have given a type 7 Contact. Is my method to provide contact correct wherein I have selected both the components?
3) In the simulation results of the H3d File a separation of the rope from the pulley is observed. Why does such a phenomena occur?
yours sincerely,
kartik
Find more posts tagged with
Altair EmployeeHi Kartik,
I recommend you to try modelling the rope with spring elements. And use the /INTER/TYPE7 (node to surf) and maybe need also use the TYPE11 (edge to edge) as contact interface.
In your case as one end is fixed and you are pulling the other end the model is behaving like this. Try with spring elements and let's review the results.
By node to surf do you mean selecting all the nodes in the rope and creating a surf set of the pulley ?
Altair EmployeeHi Kartik,
Yes, you can try the same. For pulley you can try to create a contact surface and can recall in the master segment for the interface.
Altair EmployeeHi,
I have given the contact, as specified you can check. But I am getting a error.
I have changed the property to P4 Spring and elements using line mesh option spring. Could you tell what is wrong.
ERROR ID : 60
yours sincerely
kartik
Altair EmployeeHi Kartik,
In pulley component collector you have spring elements. Delete those spring elements from pulley collector.
And please note that no material will be assigned for springs. Remove the material from rope component collector.
Altair EmployeeHi George P Johnson,
I have updated the file you can check. I also added a point mass of 0.5 Tons at the end. I am still getting that weird looping effect which was seen in the truss element. But it has reduced considerably.
Also i did the simulation without the point mass it was exactly the same.
I have new question below...
1) In the h3D file if you observe values of stresses, forces etc. have not been generated. I do not know why in this case. Is it because of the spring element?
The Hypergraph file seems to have the data.<?xml version="1.0" encoding="UTF-8"?>
Also for info
MESSAGE ID : 259
** WARNING ZERO OR NEGATIVE VOLUME : 3D-ELEMENT ID : 837
ELEMENT IS SWITCHED TO SMALL STRAIN OPTION
yours sincerely,
kartik
Altair EmployeeHi Kartik,
For spring elements you will not get stress output.
If you have a exact spring stiffness value you can simulate it without much unstable model behavior.
I varied the spring stiffness value and imposed displacement was provided. Attaching the same file for your reference.
** WARNING ZERO OR NEGATIVE VOLUME : 3D-ELEMENT ID : 837
This error happens when the solid elements are very deformed and their characteristic length goes to 0. For large strain formulation the time step of an element goes to 0 when the element is compressed. From RADIOSS 2017 version those elements will be automatically switched to small strain.
Altair EmployeeHi George P Johnson,
The file isn't supported you need to give me a block 120 radioss file .
<?xml version="1.0" encoding="UTF-8"?>
In the case of spring elements then how are the failure cases observed. The values I gave were relative to the material like (K=EA/L) or
(mass of spring = density*volume of the element)
yours sincerely,
kartik
Altair EmployeeHi Kartik,
Please find the file attached.
If you want specifically the rope failure cases you can try modelling it with beam elements.
Altair EmployeeHi Kartik,
Please find the file attached.
If you want specifically the rope failure cases you can try modelling it with beam elements.
Hi George P Johnson,
I had already modelled with beam elements stresses become very high due to the presence of a bend in the middle and bending stress is generated. But in a rope it is not true.
Also Like in our cases which we see in the h3d file .... the rope moves our weirdly as observed in the truss and the spring elements. (Similar is observed in case of beam elements)
yours sincerely,
kartik
Altair EmployeeHi .
I just changed the stiffness and the initial velocity case also works correctly.
I had a query how is the stiffness of the spring element calculated?
I used this (K=EA/L) where E is youngs modulus, A is area of Cross section and L is length of each spring element
(mass of spring = density*volume of the element)
I couldn't find any material on how is spring stiffness given.
In the reference guide of radioss spring stiffness has unit of N/s what is the time variable which we are dividing it with ?
<?xml version="1.0" encoding="UTF-8"?>
yours sincerely,
kartik
Altair EmployeeHi ,
In regards to the previous post i tried to simulate all of the same using truss element with a point mass.
But i am getting an error and the program gets stopped abruptly.
Could you tell me what the problem is?
yours sincerely,
kartik
Altair EmployeeI couldn't find any material on how is spring stiffness given.
In the reference guide of radioss spring stiffness has unit of N/s what is the time variable which we are dividing it with ?
Please go through the RADIOSS user guide ( http://forum.altairhyperworks.com/index.php?/topic/20652-radioss-2017-reference-manual-documents-user-guide-and-tutorials/) where the stiffness formulation is shown.
This error message on 'Infinite domain detected' may appear during sorting of contact interfaces when the distance between some nodes of the model becomes infinite. One has to look at the model behavior to understand why some node coordinates are infinite.One common explanation is 'flying' nodes after failure of elements belonging to interface surfaces. In this case using Idel =2 in contact interfaces should fix this problem.
Altair EmployeeHi ,
I simulated again with IDEL =2 still i am getting the below error.
Also I was looking in the tutorials of hypermesh for estimation of K but every where they have used it directly in the case of Type4 Springs. Is there anywhere we have modeled such a case with spring elements.
Altair EmployeeHi Kartik,
If you check you can see that the model is behaving weird when initial velocity is provided. SInce one end is fixed and applying initial velocity it will behave like this,
In the above model set up I changed the loading condition to imposed and ran.
<?xml version="1.0" encoding="UTF-8"?>
Hi Kartik,
I recommend you to try modelling the rope with spring elements. And use the /INTER/TYPE7 (node to surf) and maybe need also use the TYPE11 (edge to edge) as contact interface.
In your case as one end is fixed and you are pulling the other end the model is behaving like this. Try with spring elements and let's review the results.