DamageTC3d - snapback error

[AntoMarinelli]

Dear STKO,
I'm modelling a prestressed beam made of high performance fiber reinforced concrete with fcm=120MPa. I set the model as what I have done during the calibration phase but the analysis, even with a mesh of 1mm, keeps giving the snap back error.
I attached the model.
Thank you
Antonella

[marafini.f]

In the model you sent mesh size is around 15mm. Anyway the reasoning is always the same. If you use the autoregularization you need to normalize the value you get from the stress and strain experimental relationship, in reference to the length of your specimen, so that STKO can relate it to the mesh size.

By the way be careful cause you are using Norm Unbalance with the Penalty method, that cannot be done.

[AntoMarinelli]

In the model you sent mesh size is around 15mm. Anyway the reasoning is always the same. If you use the autoregularization you need to normalize the value you get from the stress and strain experimental relationship, in reference to the length of your specimen, so that STKO can relate it to the mesh size.

I don't understand this. I was thinking that with the autoregularization the software made the normalization by itself.

By the way be careful cause you are using Norm Unbalance with the Penalty method, that cannot be done.

I had always use this combination. What is the better one?

Thanks you
Antonella

[marafini.f]

Norm Displacement Increment

Why don't you check these two webinars?

https://asdeasoft.net/?get-started-webinars
#W9 Analyses Command (Part 1 and 2)
#W10 Analyses Command (Part 2 and 3)

[AntoMarinelli]

Yes, I will do it.

The problem with snap back still remains. I don't understand the regolarization because in precedent post you told me that setting autoregolarization I don't have to adjust them with mesh size but just pay attention to units.
Thanks for your support
Antonella

[marafini.f]

I am going to cite a post from this post from Diego Talledo,
viewtopic.php?f=36&t=2018&p=3961&hilit= ... tion#p3961
If you just explore a bit our forum and website you will find the answers you need.

Fracture energy is defined as an energy per unit area. So basically it is [Force]*[Length]/[Length]^2. Thats why in Model code you find that in N/m (or equivalently in J/m2).
The area under a stress-strain curve is the specific fracture energy, so it is in [Force]/[Length]^2 (for instance N/m2) and it is equal to the fracture energy divided by the characteristic length.
When you use auto-regularization you give the fracture energy and the material calculates on his own the specific fracture energy. If you don't use auto-regularization, you need to give directly the specific fracture energy.

Just to make an example: fracture energy of a medium-good concrete is 100 N/m (see e.g. model-code 2010). If you work in N and mm, it corresponds to 0.1 N/mm. If you use auto-regularization you input 0.1 N/mm. Suppose you element size is 100 mm, the specific fracture energy is 0.1/100 = 0.001 N/mm2. If you don't use auto-regularization you input then 0.001 N/mm2.