I've never used any modeling software before and just wanted to know if what I'm envisioning can be even be modeled. I wasn't sure where to post this question, so if this is the wrong place, please let me know. I would be grateful for any advice you may have. What I'd like to do is:

  1. 3D modeling of flexible (soft body?) spheres in contact with one another. The spheres are just a surface with no internal points.

  2. There's an initial sphere (say layer 0) whose surface is covered as much as possible by non-overlapping spheres of the same size in layer 1. Because you can't fit an integral number of equal-sized spheres around an inner sphere of the same size, some amount of a few of the layer 1 spheres will be trying to occupy the same location. Because of this, these spheres won't be spheres but will be kind of hemispherical shapes meeting in the middle in a flat interface.

  3. Any of the outer surfaces of the layer 1 spheres and hemispheres that are not already covered by other spheres, will also be covered by non-overlapping spheres (layer 2). Some of the layer 2 spheres will overlap if there's not a surface area that can fit an integral number of them. Ideally, the appearance of the sphere layers (0, 1, 2, etc.) is a sequential process, but I don't think that's critical for the model. In other words, they could all be there initially.

  4. The overlapping hemispherical entities are trying to achieve their natural spherical shape and thus will exert pressure on each other. This is where the flexibility part comes in. As the overlapping hemispheres push against each other, they will change shape and push into neighboring spheres. All the spheres/hemispheres are trying to maintain a spherical shape, so if a sphere is bulged into, it will offer a small resistance to this.

    Is it possible to make a 3d simulation of what happens when you start this running to see how the pressure waves and forces move through the spheres?

Thank you to both of you for your help. Both your answers were very informative and
will help guide my work on my hobby!



2 Answers 2


There is nothing to model really, you can use blenders primitive spheres, and position them procedurally by using the centre of the parent sphere as the pivot point of the first layer of children and constraining them to the radius of the parent. This way they will appear on the surface of the parent.

The way you can create the spheres is creating a emitter sphere which would be at the origin of your parent sphere with a radius of: EmitterSpereRadius = ParentRad + childrenRad

Then create a particle system using the emitter sphere as the ... emitter, and creating the desired amount of particles which you will set to be emitted from the surface of the sphere rather than the volume. Turn off all forces acting upon the particles and they should stand still.

Now you can replace this particles with instance geometry (your children spheres) and you will have a parent sphere with it's children just about touching it's surface. (Note, that if you distributed your particles randomly, some children could be overlapping).

Then as John Burrill suggested apply softbody dynamics to all spheres, make them collidable with each other and their parent, and apply a strong force filed to the children from the parent which would allow them only to encompass the parent, squashing against him and themselves...

Repeat the same process for more layers.

This would get you some results. In the case that i haven't totally misunderstood the question, please ignore. :)

  • $\begingroup$ Thank you! I appreciate your feedback. I asked Mr. Burrill this, too, but could you roughly estimate how long it might take in my spare time (separate from my job) to become proficient enough at Blender to do what you're saying? Because this is kind of a hobby, I want to balance the time spent on modeling with time spent thinking about how to further my model. Thanks! $\endgroup$
    – Roger846
    Jul 25, 2014 at 5:33
  • $\begingroup$ It is very hard to say, cause there are many unknowns... Start from there, think about the problem you want to solve and split it into smaller tasks. Relate each task with a specific blender, feature that you want to use, and then allocate time for reading the about the feature (say softbody dynamics or particle systems) and also try to play around with it. Only after that, start combining features together and see how it works. $\endgroup$ Jul 25, 2014 at 8:48

You're talking about 2 different things: modelling the spheres and simulating the physics of applying forces to them. The answer is yes, you can model this kind of thing in Blender and simulate elasticity using softbody dynamics which can show you realistic deformation based on general parameters. Here's a quick clip showing an example of softbody dynamics:


Blender provides a convincing effect, but it doesn't incorporate real material properties like hardness, poisons ratio and modulus of elasticity and it doesn't do the kind of reporting needed to evaluate how the model would perform over time in the real world so it's not an analysis tool like ANSYS

  • $\begingroup$ Thank you! I appreciate your feedback. Starting from no knowledge, in order to become proficient enough at Blender to do what you're saying, can you roughly estimate how long this might take in my spare time (separate from my job)? Because this is kind of a hobby, I want to balance the time spent on modeling with time spent thinking about how to further my model. Thanks! $\endgroup$
    – Roger846
    Jul 25, 2014 at 5:32
  • $\begingroup$ I think you'll need to learn these topics: User Interface Mesh modeling/editing (including a couple of obscure tools like align faces, coordinate systems, dupilverts) modifiers constraints animation key framing rigid body simulation soft body simulation Since you're new to animation and 3D modelling, you're going to have to absorb a lot of knowledge and concepts. If you are aggressive with finding answers-going to tutorials, forums and the online documentation as questions arise-then you can probably attain the skills needed for your project in 60-100 hours over 3 months) $\endgroup$ Jul 25, 2014 at 15:51

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .