Hi Juli!

## What the CRM mu(I) model does at/near the free surface

The behavior you're seeing (upper bed disturbed, lower bed uniform) is 
consistent with how cohesion interacts with the low-pressure region of the bed, 
which is exactly the near-surface zone. The relevant model mechanics from 
`doxygen/documentation/manuals/fsi/fsi_mu_i_rheology.md` are:

- **Yield limit with cohesion:** `tau_max = mu * p_tr + c`. Cohesion `c` (your 
`cohesion_coeff`) "adds shear strength at low pressure." The near-surface 
material is precisely where pressure `p_tr` is lowest, so cohesion changes the 
constitutive response most strongly there.
- **Cohesive tension cutoff:** "if `p_tr < -c/mu_s`, stress is zeroed." With `c 
= 0`, this cutoff is at `p_tr < 0`. With `c = 299` and `mu_s = 0.61`, it moves 
to `p_tr < -490 Pa`, i.e. the model now permits tensile stress states before 
zeroing. This is a fundamentally different near-surface behavior between your 
two cases.
- **Free-surface treatment:** "if flagged (`nabla_r < free_surface_threshold`), 
stress and pressure are zeroed." So surface-flagged particles have their stress 
zeroed regardless of cohesion, while sub-surface particles now carry cohesive 
tensile strength. The transition between these two regimes sits right in your 
disturbed layer.
- **No EOS in CRM:** "EOS is not used in CRM; pressure comes from stress 
trace." So in a loose bed there's very little confining pressure near the top, 
amplifying the sensitivity to the two mechanisms above.

**So to your first two questions:** yes, a change in near-surface behavior when 
you introduce cohesion into a loose bed is an expected consequence of how 
cohesion couples with (a) the low-pressure free-surface region, (b) the tension 
cutoff `p_tr < -c/mu_s`, and (c) the stress/pressure zeroing at the flagged 
free surface. It does not by itself indicate that your material construct is 
wrong. It is genuinely an interaction between cohesion, loose packing (low 
pressure), and free-surface handling.

## Two things in your setup worth checking

1. **Young's modulus / step size.** Your `Young_modulus = 35.9e6` is ~36× the 
CRM baseline of `1e6`. The parameter guidance explicitly warns: "increase if 
response is too soft, but expect smaller stable step sizes when stiffness 
rises" (`fsi_mu_i_rheology.md`). A stiffness this high with an unchanged 
`step_size` can produce noisy/unstable behavior that would be most visible in 
the loosely-confined surface layer. I'd verify your step size is small enough 
for this stiffness before attributing all the disturbance to cohesion.

2. **How cohesion is applied.** The docs note cohesion can be set either 
through `mat_props.cohesion_coeff -> SetElasticSPH` *or* via `SetCohesionForce` 
(`fsi_mu_i_rheology.md`). Make sure you're only setting it one way and that 
`SetElasticSPH(mat_props)` is actually being called after you assign 
`cohesion_coeff`.

## Last thought: I encourage you to try the new MCC rheology in CRM, in Chrono 
10.0. It works well for soils when there needs to be consolidation. The mu(I) 
rheology is not great at doing that.

I hope this helps a bit.
Dan
------------------------------------------------
Bernard A. and Frances M. Weideman Professor
NVIDIA CUDA Fellow
Department of Mechanical & Aerospace Engineering
Department of Computer Science
University of Wisconsin - Madison
4150ME, 1513 University Avenue
Madison, WI 53706-1572
608 772 0914
http://sbel.wisc.edu/
http://projectchrono.org/
------------------------------------------------

From: [email protected] <[email protected]> On Behalf 
Of JüliMari
Sent: Friday, June 5, 2026 1:28 PM
To: ProjectChrono <[email protected]>
Subject: [chrono] CRM/SPH cone drop with LHS-1: unexpected behavior when adding 
a cohesion parameter

Hello Chrono community,

I am working with the CRM/SPH cone drop demo and adapting it for LHS-1 lunar 
regolith simulant. My research team would like to include cohesion in the LHS-1 
material model, but I am seeing behavior that makes me unsure whether I am 
applying the cohesion coefficient correctly, or whether there is something in 
the cone drop setup that affects cohesive cases.

I ran two loose LHS-1 cone drop cases with the same setup (other than 
render_fps, to capture as many snapshots as possible). The only change between 
cases was manually editing the LHS-1 material construct from cohesion_coeff = 0 
Pa to cohesion_coeff = 299 Pa. With 0 Pa, the loose LHS-1 bed behaves smoothly 
during the cone drop and as expected following other CRM research papers. With 
299 Pa, the upper region of the bed becomes highly disturbed, while the lower 
material remains comparatively uniform. I attached snapshots showing the 
difference between the two cases.

The LHS-1 material construct I implemented:

struct lhs1_material {
double density_max = 1740; // kg/m^3, dense LHS-1
double density_min = 1340; // kg/m^3, loose LHS-1
double Young_modulus = 35.9e6; // Pa
double Poisson_ratio = 0.3;
double mu_I0 = 0.08;
double mu_fric_s = 0.61;
double mu_fric_2 = 0.61;
double average_diam = 51e-6; // m
double cohesion_coeff = 0; // changed manually to 299 for cohesive case };

And for MU_OF_I rheology,

mat_props.rheology_model = RheologyCRM::MU_OF_I;
mat_props.mu_I0 = mu_i0;
mat_props.mu_fric_s = mu_s;
mat_props.mu_fric_2 = mu_2;
mat_props.average_diam = lhs1_mat.average_diam;
mat_props.cohesion_coeff = lhs1_mat.cohesion_coeff;

My questions are: is this disturbed free-surface behavior expected when adding 
cohesion to a loose granular bed, or does it suggest that my setup is 
inconsistent? Could this be caused by the interaction between cohesion, loose 
packing density, free-surface handling, and the initial 
pressure/preconsolidation setup? And is there any hard-coded assumption or 
scaling in the cone drop (or cone penetration) examples that could make 
cohesion behave unexpectedly in this case?

I can provide the full code file if you need to see it, though I wanted to 
check first whether this behavior is expected for CRM/SPH cohesion 
implementation.

Undergraduate Student
Aerospace Engineering

1 Aerospace Boulevard
Daytona Beach, FL 32114

Embry-Riddle Aeronautical University
Florida | Arizona | Worldwide
--
You received this message because you are subscribed to the Google Groups 
"ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to 
[email protected]<mailto:[email protected]>.
To view this discussion visit 
https://groups.google.com/d/msgid/projectchrono/fd428527-985e-4b60-a7db-0088bf43d482n%40googlegroups.com<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/projectchrono/fd428527-985e-4b60-a7db-0088bf43d482n*40googlegroups.com?utm_medium=email&utm_source=footer__;JQ!!Mak6IKo!KUCQyXEPDPbuUTMfjCsxxqdA2wyolDpLbA_gOkMouaSBQpaR1-eWpKmpFG1tXRMvcTkB1UuVt7It2XU$>.

-- 
You received this message because you are subscribed to the Google Groups 
"ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/projectchrono/DM8PR06MB770318D34AB21657A11B3EA9B1F62%40DM8PR06MB7703.namprd06.prod.outlook.com.

Reply via email to