Why is the stress-strain response almost unchanged after adding fibers in a Constraint-Based Beam in Solid Ultra-High-Performance Fiber-Reinforced Concrete tensile model?

[ming fang]

Hello everyone,

I am working on a uniaxial tensile simulation of Ultra-High-Performance Fiber-Reinforced Concrete (UHPFRC) in LS-DYNA using the Constraint-Based Beam in Solid approach, and I would like to ask for some advice.

My main issue is that after adding steel fibers, the overall stress-strain response is almost unchanged compared with the plain matrix model.

My model setup is roughly as follows:
- Matrix: solid elements with a concrete material model
- Fibers: discrete beam elements
- Fiber-matrix interaction: Constraint-Based Beam in Solid
- Loading: quasi-static uniaxial tension by prescribed displacement

What I observed is:
- The beam fibers do carry axial force during loading
- However, the global stress-strain curve of the fiber-reinforced model is still very close to that of the plain concrete matrix model
- The peak stress increases only slightly, and the overall response shows little difference after adding fibers

So at the moment, the fiber contribution is not clearly reflected in the global tensile response.

I am wondering whether this may be caused by one or more of the following:
1. The reaction force extraction method is not capturing the fiber contribution correctly
2. The beam-to-solid interaction is not transferring load as expected
3. The loading or boundary condition setup makes the fiber effect difficult to appear
4. The specimen has not developed enough cracking or localization for the fibers to become effective
5. There may be an issue in my post-processing method for obtaining the overall stress-strain curve

My questions are:
1. In a Constraint-Based Beam in Solid tensile model of UHPFRC, what is the recommended way to obtain the correct global force-displacement or stress-strain response?
2. If the beam elements already develop axial force, why might the overall curve still remain close to the matrix-only result?
3. Is this more likely to be a post-processing issue, a load transfer issue, or a modeling issue?
4. Are there any common checks you would recommend for this type of fiber-reinforced tensile model?

Any suggestions would be greatly appreciated. Thank you very much.

[银星]

Hello, your observation is actually quite reasonable for UHPFRC.

In uniaxial tension, the main mechanical role of steel fibers is usually not to significantly increase the pre-cracking elastic response of the composite, but to provide crack-bridging resistance after the matrix starts to crack. In other words, fibers are most important in the post-cracking stage, where they transfer tensile stress across localized cracks through bond, debonding, slip, and pullout. Therefore, if the specimen in the simulation has not yet developed sufficient cracking or strain localization, the global stress-strain response may remain very close to that of the plain matrix, even if the beam fibers already carry axial force.

- Yin

[ming fang]

Dear Yin,

Thank you very much for your explanation. I think this is exactly the core issue in my current model.

In my simulation, the plain UHPC matrix is very brittle in tension. The matrix reaches a tensile stress of about 5.3 MPa at a strain of around 0.00017, and then the stress drops to nearly zero over a very small additional strain increment. In other words, once cracking starts, the matrix almost fails immediately.

Because of this, I feel that the fiber bridging effect cannot fully develop before the matrix has already lost most of its load-carrying capacity. As a result, the overall stress-strain response of the fiber-reinforced model remains very close to that of the plain matrix model.

I am attaching my current matrix stress-strain curve for reference.

In your experience, how is this problem usually handled in simulation? More specifically, after the matrix reaches its tensile peak stress and starts to crack, how can I model it so that it cracks but does not lose its carrying capacity too abruptly, allowing the fibers to develop their full bridging effect?

Do you think the key is mainly to:

make the post-peak softening of the matrix less abrupt,
avoid immediate element erosion or complete stress loss after cracking, or
improve the fiber-matrix interaction model, such as bond-slip or pullout behavior?

I would really appreciate any suggestions. Thank you again for your help.

Best regards,
Fang

[银星]

Hello, Fang
From a practical modeling viewpoint, I would focus on four points. First, the post-peak tensile softening of the matrix should be regularized, ideally with fracture energy and characteristic element length taken into account. Otherwise, the response may become overly brittle and mesh-dependent, and the stress can collapse too quickly after cracking. Second, immediate element erosion or complete stress loss after cracking should be avoided, because this can suppress the post-cracking stage where the fibers are supposed to work. Third, the fiber-matrix interaction should be improved, especially through a bond-slip law, since the key mechanism is not just fiber axial force but stress transfer through debonding and slip across cracks, and increase NCOUP. Fourth, mesh refinement can help, it is most useful after the matrix softening law has been regularized properly.

Best regards,
Yin

[ming fang]

Dear Yin,

Thank you for your previous advice.

I would like to ask one more question: in my UHPFRC uniaxial tensile model, how should I extract the total force, including the combined contribution of both the concrete matrix and the fibers?

Since I did not use SPC constraints in this model, SPCFORC is not applicable.

Best regards,
Fang