I'm attempting to destroy a bridge progressively and have cell fractured a mesh that is 80m long and then smashed through it with a cube (both are rigid bodies). All the cells in the fractured mesh are "Start Deactivated" with angular and transitional velocity set to .5m/s to activate. The bridge starts deactivated but the whole thing falls as soon as my cube touches the first end. How do I get around this? Image below shows how it starts and when cube hits.
I've had similar problems in the past, there are just a few simple things you need to do to avoid everything exploding. Just need to set the threshold for activating to be a little higher.
When making the cell fracture make sure that you have a decent margin so that the pieces aren't inside of each other. As well as make sure that the pieces have their origins recalculated to their centers. While the margins might not look good, you can always boolean the original uncracked object out of the way later during the animation.
The most important part is to make sure that after you've made all the rigid bodies, to go to the object menu, rigid bodies, then select connect. This will add a bunch of rigid body constraints that will help keep the bridge from exploding.
Just make sure that you change the setting in the constraints to breakable so that it can actually fall apart when given enough stress. You can also quickly copy the value to all of them by selecting all the constraints with one set as the main active. Then just set the breaking to on and then right click it, a small window should appear, all you have to do is click copy to selected.
I've arrived at a "workaround" which involves having the bridge split into progressive groups of cell fractured components. All the groups have their rigid body "animate" option keyframed as ON so the physics sims has no effect. As my collision object moves across the bridge I turn the "animate" option off for the group the collision object is approaching, thus kicking in the physics for that section but not touching the others. It gives a nice control and having small sections really speeds up the cell fracture calculations. It would be better to cell fracture the bridge as a whole and then split the bits into groups after so that there are no clean breaks between sections but you can't really tell even with rectangular sections.