Pile of Lego bricks with particles?

Hi all, I’m trying to create a randomized pile of Lego bricks in the middle of a surface. Seems to me that it’d be easy enough to just create some particles and watch them drop onto a surface and poof, instant pile of randomized bricks. Blender seems to have other ideas though.

Even after cranking up all those damping/friction settings, with no random, I have a few particles hitting the surface and just flying off into outer space. What’s going on here?

How do you get the particles to pile into a mound? They just all lay flat on the surface, which is obviously useless.

Attempting to give the particle objects their own collision modifiers seems to result in them bouncing off into never never land in mid-air, presumably from hitting each other. What is going on here?

-Stefan-

I’ve seen “Bake Physics” used to get this kind of look before. If you search the forums for physics and baking, you may get some useful info. The skeleton of the solution is to set up a bunch of bricks in the air, apply Bullet physics properties to them (appropriate mass, etc), then run the simulation and save the results to keyframes.

I would like to apologize on behalf of Alterscape for his reply.

Why? Because it was useful?

Wow. I’m sorry if it was incorrect, but please at least tell me (and the guy who asked the original question) why it was incorrect and what a better approach would be?

Use physics baking as Alterscape suggests.
I have attached an example blend (the number of bricks was limited to enable attaching to post). This is for blender 2.5beta
In the Game Render settings I have set to record the animation. Play the physics sim with P, after you have run the animation go to the frame you want to freeze, on the timeline. Delete the keyframes in the graph editor with X
lego.blend (965 KB)
http://www.screencast.com/users/blenderwho/folders/Jing/media/64f198ec-385a-4398-aa32-56775d1d6908
A similar thread with the same solution http://blenderartists.org/forum/showthread.php?t=192702&highlight=bulb

Soo…just to share alter - I was able to test out your method, and yes looks like that’s what I want, but switching to triangle mesh collision so the pieces don’t just lie flat results in god-awful solve times…actually after three bricks the computer just dies.

Edit: Thanks Richard! will test.