The segment of the main crack behind the intersection point seems blocked

[neclipse]

Problem description

• The scenario is that we have two normally crossing cracks: [0,30; 0.15, 30] and [0.105,29.95; 0.106,30.05].
• For simplicity, these two existing cracks are labeled as traction-free, or "perforated".
• The injection mouth is at (0, 30), belonging to the main crack.

Issue description

• The segment of the main crack behind the crossing point suddenly turns to negative pressure.
• However, the aperture profile of the whole main crack seems smooth, gradually decreasing from the mouth to the tip.
• The crossing crack has negative crack opening displacements and pore pressures.

Suspect causes

• The definition of the normal vector, ntaud, may be wrong.
  I checked again. It was all right.

• The calculation of phi and phishift may be incompatible with the normal vector.
  I doubled check the calulation of sign

• The enrichment of the element with two cracks may be wrong
  BUG 1: When the HF intersects the NF lying ahead, the gaussian points were regenerated and somehow the enrichment info for the existing crack is gone.
  1. The EnrichGauss (for the domain) totally miss the enrichment info for the existing crack
  2. The LineGaussDict for the later one miss the enrichment info for the existing crack
  3. This problem seems not present for the initially intersected cracks.
  BUG 2: The enrichment of NuenrPlus and Minus for the cohesive gauss points by the other enrichitem are wrong for the current enrichitem.

• The problem with two intersecting cracks were not proper, like the traction-free setting, or the relative location of two cracks.
  1. This may be one of the reasons. The existing HF was too long and the injection of fluid to the crack mouth generated too much negative pore pressure far from the crack mouth.
  2. Now the configuration changed to one short HF (0.08m) and a short NF (perforated = false) lying ahead (at 0.16m).
  87eee0e
  3. The difference between minaperture and perfaperture was also too big. 89e943e
  4. Increase the kmat_crack and use consistent small biot_mod. 9c082a5

• The pore pressure profile is not smooth. The pressure far from the crack mouth has a very big negative pressure or we will have a very big fluid lag. The lag is there and the more the crack mouth opens, the larger the negative pressure gets as long as the dissipation of pressure has not arrived.

[neclipse]

When the intersecting fracture (NF) closes, the fracture pressure along the main HF after the intersection point stoped increasing.

[neclipse]

New test

I tested another interaction scenario where the two cracks will have a much smaller intersection angle when they get crossed.

Results:

1. The existing crack did not close as the previous normal crack.
2. The main crack keeps building up fracture pressure but the pressure could not propagate to the crossed existing crack when they get crossed.
3. Although the fracture pressure gets very high in the intersection element, the fracture pressure suddenly drop to a negative value and keeps decreasing as time goes on, the existing crack starts propagating. The diversion occurs.