• Registered: 2011-01-04
  • Posts: 16

Topic: Cyclic loading - monitoring loads

Hello to all!

I am currently trying to model a cyclic test for a column. Up to a point, I have to follow a loading controlled test, after which I have to make a displacement control test.

Is there any procedure in which I can apply a stop condition, in the way that I can apply load up to the point where the prescribed displacement is reached? As far as I can see, I can only apply a prescribed displacement, without knowing the force necessary to be applied in order to reach that prescribed displacement.

I am using Atena2D.

Best regards,
Cosmin DAESCU
Politehnica University of Timisoara, Romania

Quote Post 1

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Cyclic loading - monitoring loads

Hello, I am not sure if I understand your problem - it sounds like you know the displacements (as applied in the real experiment), and just need to set up a monitoring point for the reaction (to find out the force, as recorded based on hydraulic pressure in the experiment) to record the column response.

We always recommend to use prescribed displacements whenever possible. However, if the load cycles are prescribed by forces (machine pressure), you have to use force loading.

You speak about switching from load control to displ. control - are there some load controlled cycles (e.g., 10x from -force to +force and back), followed by additional cycles with prescribed displacement (-distance to +distance)? If so, how is the machine switched during the experiment?

dpryl's Website

Quote Post 2

3 Reply by daescuc (edited by daescuc 2011-01-04 15:21:14)

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

Hello again.

I have to use several cycles for loading. Initially, as you well understood, I have to apply several load steps, in load control, up to a certain lateral displacement (24mm in my case). From this point on (24mm is considered the yield limit for my column) I have to apply the cycles in a displacement control test: 3 cycles at 24mm displacement, 3 cycles at 1.5*24=36mm displacement, 3 cycles at 2*24=48mm displacement, and so on.

During the experiments, the machine was manually controlled in order not to overcome the limit displacements corresponding to each cycle.

Hope I said it right this time. Thank you.


[But in your answer I got a hint - I may use a monitoring point for the horizontal reaction at the base of the column. I will have to check on that indirect measuring.]

Edit 1: I have an additional question: is there any way in which I can monitor a reaction line? (my support is a line, not a point)?

Edit 2: I changed the line support to multiple point support and I have defined monitoring points in all of them. it seems that the sum of all these values will give me exactly the applied force. Consequently, I will use this approach to determine the equivalent applied force that should correspond to a given displacement. Thank you.

Quote Post 3

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Cyclic loading - monitoring loads

Hello,
if the cycle is defined by displacement amplitude, I definitely rcommend to simply use displacement control.

We recommend to apply the load to a single point of an elastic loading plate if possible. In case of displacement load, you can then monitor the reaction at the load point to get the force value. If you monitor a line or surface or multiple points, you have to sum the values outside of ATENA (boring additional work).

dpryl's Website

Quote Post 4

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

I am trying to monitor the reaction at the point of the application of the external force and it seems it is not recording the force applied (records 0, actually,all the time). I need some advice, please. Thank you!

Quote Post 5

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Cyclic loading - monitoring loads

Hello, I guess the solution is relatively simple: if you load by external force, you have to monitor external force.
If you load by prescribed displacement, you have to monitor reaction.

dpryl's Website

Quote Post 6

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

As I have previously mentioned, in the displacement control phase I have tried to monitor reaction, both in the point of the application of the prescribed external displacement and in the neighboring point to it.
Maybe I am doing something wrong, but in both cases the monitored value was 0 all the way to the end.

In conclusion, I am not able to monitor reaction in the node where the prescribed displacement is applied. Do you have any suggestions, please?

Quote Post 7

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

I still have the problem. This message is just to update the previous message. Thank you.

Quote Post 8

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Cyclic loading - monitoring loads

If you still can not find the reason, the only option I see is to send us your model for inspection.

dpryl's Website

Quote Post 9

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

I think I found something. i just need a confirmation.

It seems that a model is able to deliver reactions if loaded ONLY in displacement control, from the start. In my model I started in a load-controlled test, followed by a displacement control.

I started a new model only in displacement control and it manages to record reactions so far.

Another problem arises when I try to use the displacement control in an Arc-length integration solution. It seems that, under Win7 environment, it blocks at some point. Do you have any knowledge about this type of issue?

Thank you,

Quote Post 10

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: Cyclic loading - monitoring loads

Arc length makes no sense with displacement control. Always use Newton-Raphson with displacement control. With force control, you can use N-R as long as you are not close to peak load. The only situation you have to use AL is force load near or after peak.

dpryl's Website

Quote Post 11

  • Registered: 2011-01-04
  • Posts: 16

Re: Cyclic loading - monitoring loads

I've modified the model as you suggested and it is fine so far. Thank you.

Quote Post 12