MARS - A Software Dedicated to Advanced Concrete Modeling

For some applications it would be convenient to have also a virtual extensometer on top of the virtual gage with the following properties:
- definition of it's endpoints instead of center and length
- 1 segment only measuring the secant between the specified points
- reporting the actual length change instead of the strain

Is there some progress on this feature ?
It would be very helpful to have, especially in connection with the new PID controller.
Thanks
Roman

Feature request is implemented in version 2016.7.45: using end points to define gages. Please see manual for details.

Thank you for implementing the option to define a gage by it's endpoints.
However, this feature still returns strains and not opening information.

For the comparison with experimental data it would be convenient to have an option to obtain Delta L instead of strains.
Otherwise, the length has to be determined.

How can I determine the strain (or better length change) on a curved surface?
This is relevant for the post-peak control (simulated or experimental) of cylinder compression tests which are controlled by circumferential expansion.
If we want to use our new PID controller for this type of tests, we need the option to define a strain gage by a sequence of nodes or an edgelist.

Is it possible to set a gage between two different tetlists, e.g. two loading blocks ?
This is the case for notched three point bending. Another example would be the fixtures of an LVDT sensor.
For our anchor tests we set an LVDT between concrete surface and some steel support.

However, this feature still returns strains and not opening information.

As explained in manual, you need to output the length using command:

Code: Select all

TimeHistoryList HIST {
  sg-GageName length
}


You need to subtract the initial length to get the change, though.

How can I determine the strain (or better length change) on a curved surface?

You can choose two end points on the curved surface, set Segments keyword to be the resolution you want, MARS will discretize the two end points by number of segments, then project the middle points onto the curved surface. In this case, the length of the gage is on the curved surface. This feature works in version >= 2016.9.05.

Is it possible to set a gage between two different tetlists, e.g. two loading blocks ?

No, it has to be inside the same tet list.

Thank you for the clarification.

The main point of having an additional extensometer (on of the strain gauge) was to mimic experimental setups in which typically the sensor is sitting between different structural elements and returning relative deformations.
- CMOD, LVDTs, ....

A modification of the strain gage function is not sufficient and not even the easiest approach.

I suggest having a very simple separate function that
finds the closest nodes each to two given locations and calculates the relative displacement directly as \Delta L which is equivalent to the experimental data.

As discussed in other posts this is essential for the PID controller to work correctly.

Can you please send / upload to gitlab an example of the gage subroutine.
Then I will create the extensometerUD.