Jump to content

catchyid

Members
  • Content count

    363
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won

    1

catchyid last won the day on November 20 2016

catchyid had the most liked content!

Community Reputation

19 Good

About catchyid

  • Rank
    Illusionist

Personal Information

  • Name
    Khaled Abdelhay
  • Location
    Montreal, QC

Recent Profile Visitors

1,451 profile views
  1. Hmm, usually I find solutions after I post Anyways, two workarounds that do work, either: use remesh or divide, and somehow automagically this fixes the problem... Actually No! The above solution works for simple geo, but for complex geometrical shapes it still fails. The solution that still works so far is convert to VDB then back to polygon, but this is tooo expensive and you need to have high vdb resolution to capture mesh details, so it's not practical!
  2. [Houdini 17.5.173] I have solid mesh with holes in it. I use RBD Material Fracture to break it (concrete mode), however : 1) If I don't enable "Detail" option, it works correctly 2) If I enable "Detail" option, I get bad geometry? Seems like a bug in Houdini! Has anyone managed to find a workaround? The only workaround it to fracture massively and glue pieces but this is a heavy solution? Thanks, fractureTest_DeatilsProblem.hip
  3. Thaaaaaaaaaaaaaaaaanks Tomas Even though the documentation talks about @airresist, I never paid attention to it! Just if someone is reading this, make sure you set your RBD DOP object to read @airresist and other values (i.e. @drag,...) , otherwise animating them won't have any effect on simulation... Once more, thanks a lot Tomas... Cheers
  4. Hi, I have a simple setup where I set v@targetv for couple frames to have impact-like initial velocity, then I want bullet to take over and compute speed due to gravity, other forces,etc. I have one problem here: Bullet "seems" to be always using the v@targetv for simulation, even after I delete the variable in SOP? It seems like once it's created it will be always used --> end result: since I delete or set v@targetv to (0,0,0), objects seems always trying to slow down!!! Any idea how to fix that? My end goal is to have impact like initial velocity then let the solver takes over, so I've tried @targetv as above but it has this problem. Other solution is to set v@force and see how it works. In any case, I don't wish to use collision geo (which is another option) as I want to control the direction and magnitude of the pieces easily and precisely... Thanks,
  5. Big Thanks Jasper I also added another control for angular velocity length(@w) and it works like charm, thanks again Just if someone is still reading this, I had another problem with my setup, I was setting f@accelmax and f@angaccelmax and from the documentation, they can cause jitter for collision (i.e. if you set them to low values, then solver won't be able to move pieces quickly to compensate for collision) Cheers,
  6. Hi, When RBD packed pieces hit the ground, they bounce and rotate a lot?! I am not sure how to damp this effect? I've used damp DOP node but it enhanced a little bit, my goal is to have the pieces hit the ground and instantly fall flat without any additional rotation and bouncing (think big concrete pieces)? Now, seems like they have extra energy that makes them "dance" on the floor...Also, is this behavior affected by bullet substeps or substeps in constraint iterations? Thanks
  7. Breaking Glue Constraints - Bullet?

    Yet another update if someone is reading: it's possible to break constraints using forces, basically use v@force at each packed primitive and if it's strong enough it will break the constraint. Note, initially I've used force dop nodes and they did not break the constraints, but when I replaced them with v@force constraints got broken nicely...
  8. New surface created from RBD Material Fracture node

    Update: I run the geo by RBD Boolean fracture node, and I disable it, the RBD Material Fracture runs correctly. Maybe RBD Boolean Fracture node introduces problems in my mesh?! I don't know...
  9. Hi, When I use RBD Material Fracture node (concrete mode), I get good results IF I do NOT enable Detail. If I enable "Detail", I get new surface generated??? I guess my input geo has some issues that confuses the fracture node, but I am not sure what issues I need to look at, is there any workaround? Plz see attached images... Thanks,
  10. Initially Overlapping Objects won't collide - Bullet

    Okay, I found why: Basically the solver initially checks for overlaps, if found it sets found_overlap to prevent collision... found_overlap Point Integer When the solver sees an object for the first time, it checks whether that object is initially overlapping with any other objects in the simulation and prevents any such pairs of objects from being forced apart. This value is used by the solver to determine whether it has performed this process for the object on a previous frame.
  11. Hi, I have two fractured boxes : A and B. Each box has independent glue constraints, i.e. no constraints between A and B pieces, and each has its own packed rbd DOP node. 1 ) If A and B are initially overlapping, the will fall together under gravity, i.e. Bullet won't separate them 2 ) If A and B are initially separated and have initial opposite velocity, the collide My question: What's the logic here? Is it like if objects are initially overlapping they won't collide Thanks...
  12. Breaking Glue Constraints - Bullet?

    Just a quick update, I can actually break constraints using targetv and pop force...it's just requires "big" values and constraints should be "somehow" easy to break (e.g. if I break some constraints, then the rest won't be able to withstand the force)...I guess, using a collision object is the safest and easiest way to break constraints... Seems like the only way to break constraints is either manually by placing them in broken group or a colliding object...
  13. Breaking Glue Constraints - Bullet?

    Hi, If you have glue constraints network (using Bullet), it's possible to break them using a colliding object (e.g. a moving sphere that is flagged as animated static). However, is it possible to break them using abstract forces (e.g. using field or fluid forces, or even uniform force)? I've tried many types of forces, but regardless of their strength they don't break constraints. I've noted that packed pieces don't have "F" attribute, neither the Glue primitives (i.e. impacts = 0)? I only get impacts if I use collision geo, and "F" I apply POP forces... Another related question: in addition to breaking constraints by colliding object, sop solver, and geometry wrangler, Is there another way to break glue constraints? Thanks
  14. Thanks Tomas! So one way to see pyro density as just a property advected by velocity field, and dissipation parameter removes percentage of it on each frame? As per divergence, can I think of it as a way to adding/removing the density property above? As per vusta, sorry, simple typo
  15. Hi, I am trying to understand the basic physics behind fluid simulation used in Pyro solver. All physics books say that density must be constant (otherwise the equations will be too hard to solve). However, when I do any smoke sim in Pyro, density field is not constant, i.e. one can compute max/min/average density for a Pyro smoke simulation. So my question: how come all books say that density must be constant and yet when I sim smoke in Houdini I have different densities? Thanks,
×