Tensional rupture appears when upscaling one_way.dat

Hi everyone!

I am facing some troubles with a simple one-way coupling problem, where hydrostatic pressures systematically trigger tensional rupture in faults.

To understand the problem, I used the 3DEC one_way.dat example from the documentation.
Joint Fluid Flow — 3DEC 7.0 documentation.

The default script highlights shear rupture along the fault plane.

Now, if I scale up the model dimensions from metric to kilometric, tensional rupture occurs at the pressure front.

I cannot figure out the physical problem behind this. Gravity would change something but is not activated here.
Am I missing something in 3DEC parameters?

Thanks a lot for your help and suggestions!

Regards,

Théophile

PS : the one_way.dat I slightly modified to have comparable mesh cell dimensions

model new
; one-way coupled simulation
model config fluid
model large-strain off

def parameters
    ; Default
    ; local edge = 1
    ; Upscaled
    local edge = 1000

    global xmin = -edge
    global xmax = edge
    global ymin = -edge
    global ymax = edge
    global zmin = -edge
    global zmax = edge
    
    global mesh_edge = edge/10.
end
[parameters]


block create brick [xmin], [xmax], [ymin], [ymax], [zmin], [zmax]
block cut joint-set dip 0 dip-direction 90

block zone generate edgelength [mesh_edge]
block zone cmodel assign elastic
block zone property density 2500 shear 2e9 bulk 5e9 

block contact jmodel assign mohr
block contact property stiffness-normal 1e10 stiffness-shear 1e10 

block fluid property bulk 3.8e5
block fluid property density 1000
block fluid property viscosity 1e-3

flowplane vertex property aperture-initial 1e-4 ...
                          aperture-minimum 1e-4 aperture-maximum 1e-4 

;
block gridpoint apply velocity-x 0 range position-x [xmin]
block gridpoint apply velocity-x 0 range position-x [xmax]
block gridpoint apply velocity-y 0 range position-y [ymin]
block gridpoint apply velocity-y 0 range position-y [ymax]
block gridpoint apply velocity-z 0 range position-z [zmin]
block gridpoint apply velocity-z 0 range position-z [zmax]
;
block insitu stress -1e7 -1e7 -1e7 0 0 0

model fluid active off
model mechanical active on
block fluid property bulk 0.0

model history mechanical unbalanced-maximum
model cycle 100

model fluid active on
model mechanical active off
;
flowknot apply pore-pressure 1.1e7 range position-x [xmin]
flowknot apply pore-pressure 0.0 range position-x [xmax]
;
block fluid property bulk 3.8e5

history delete
model his fluid time-total
flowknot history pore-pressure position 0 0 0
model cycle 1000
model fluid active off
model mechanical active on
block fluid property bulk 0.0
model cycle 1000
program return

Hi,
It’s because the pore pressure gradient along the x-direction is not the same.