Inspired by Vern and John's tinkering I wanted to play a little with Newton too! This is a simple catapult throwing a stone towards a brick tower, nothing fancy

I have four towers and four walls, enough to build an small castle, however, it looks like its to much for A:M to handle. I'll do a few more test before reporting

Click to view attachment

That was very cool!!!!
I am a medieval buff and love it!
A bit more weight on the weighted end will give you a harder and higher hit.

Inspiring! How is A:M letting you down? In my tests I had hundreds of falling Newton objects...

I'm not sure whats wrong. Testing the towers and walls separately works great, but as soon they are assembled I get an warning triangle while simulating (no message), then A:M quits. However, with all four walls and towers there is a total of 1120 bricks

Nice test ,looks like you can do endless stuff with newton

Very nice test!

Michel

That looks great!

I saw a documentary on war machines and they had found that catapults, which typically were on wheels, actually got more range if the base were allowed to roll free when firing rather than being immobilized. It's counterintuitive but there was some sort of mechanics reason for it.

I wonder if Newton would bear that out?

Great test, Stian! How much RAM are you working with?

Interesting. I'll see if I find some time to test it out, if I figure out how

Thanks. Im running XP 64, Quad Core, 8Gig RAM

Oooooo! Not a RAM issue then.

If you were to make the ball (or projectile) smaller, would it change the trajectory, i.e., would the cannon ball go higher or faster towards the wall?

The Empty Warning Triangle of Doom®

In my work with newton, 9 times out of ten it is some weird thing in the model that I forgot to do. For example, if you have an explode object and you have extra groups or mismatched groups and bones. If you have "Newton_" groups that contain no mesh. That other 10%... have no clue. It could be too many objects in the simulation. Remember it's that initial "start" that has the most collisions. Every single object is colliding at the same time even if they aren't moving. It's after they get hit and fall around that the collisions drop off.

When you say you "put them all together" are these breakable explode objects? Or are they assembled in the chor? I have a heck of a time with breakable explode objects causing that error usually due to forgetting an explode bone or group or having too many of them after deleting something. If assembled in the chor is it all created from one brick model? Does the brick model have a "Newton_" group? I've found adding that newton group even to a small simple model can help.

Basically I've found that that error in my case is usually tied to some "mistake" I made with the set up. I also sometimes mix and match my newton property settings. Setting some in the model and some on the shortcut.

I spent at least an hour or more with that stupid triangle error on my newton car project... only to discover it was because I deleted parts of a mesh and had "empty" newton groups in the model.

-vern

Newton appears to cause the mind to tinker or is that think error .... great stuff

The models are Dynamic Explodeobjects. The tower is one model using Explode_ groups and bones, same with the wall. So the assembling is just dragging and dropping four towers and four walls into the chor and arrange them. I think I have tried every setting there is, but I might have missed something. Also, the crash happens a few frames into the impact of the ball

good pass the ball candidate.

Aha! I hadLOADS of trouble with breakable/explode objects. Did you "build" the explode model by hand? Like... putting in all the groups and bones etc? Or did you export the model from a chor after building it there first? It's no problem if you want me to look it over. I am almost 99% sure I can locate the trouble. The "wall" in my "Change your Pants" animation was done the exact same way... I had bunches of those errors before I got it right. It might even have something to do with a conflict with multiple instances of the same model but I can't be sure of that.

-vern

Thanks Vern. I'll do a few more test before I'll upload it


Haha! Not only did it just work, it worked very well! test attached

Click to view attachment

You know what's coming. The writing is on the wall...

http://www.youtube.com/watch?v=9V7zbWNznbs

-vern

So, we should see a certain rag-doll character thrown by a catapult pretty soon, Vern?

Great tests, Stian!

Love that clip!!!!

Here is the projectfile with the two catapults. Feel free to use it

Click to view attachment

That was very impressive. I guess the moving mass of the counter weight is what gives it that extra snap.

Thanks for the model

On that catapult project... don't forget to turn of "Create model for bounding hull" newton plugin property for the catapult model. This will speed up the simulation a lot. That setting is only used for checking the shape of the newton simulation collision hull to see how it matches the splines. It has no effect on the simulation. Plus it creates a new model with a TON of freaking splines each time you simulate.

I'm having trouble convincing Thom it's perfectly safe to sit in the catapult. I may have to blind fold him first.

-vern

Click to view attachment

I can't do one of those wonderful AO renders right now... I'm just too underpowered (about 8 hours to render). I want to do a different angle to show Thom better. I might stick the camera on him as well.

I love that catapult setup!

-vern

Haha! Hilarious belly landing!

That is hilarious!

The one thing that is bothering me is that there isn't an arc to the throw...it's just throwing downward.

have Thom flying towards the camera.

Thom is quite a bit heavier than the "ball" so he doesn't go as far (I simulated both catapults originally and thom didn't go nearly as far as the static catapult ball). I think too the ball kind of rolls along the inside of the catapult during the arc of the catapult arm before it is released giving it some upward force. Thom is kind of stuck in there because his arms and legs hang over the edges. He also needs more bounce on the hips. I forgot to turn the bounce back up from the previous project I used him for falling on a crushable cloth "box". If he bounced too much it didn't look right. Now when he hits the ground his hips kind of "stick" to the ground and he slides along. His head bounces up quite painfully though .

This is still my original "static" rag doll Thom. I haven't finished the poseable version. With that one I can put him in a sitting position inside the catapult with his legs hanging over the edge BEFORE the simulation so he might have a higher trajectory. Right now he starts out "flat" and "falls" into the catapult a little bit before it comes up. This impact might also be "slowing down" the catapult.

I wonder if a trebuchet could be built with Newton? Might have to "fake" the rope aspect with a series of newton chains.

-vern

Yikes!

I'm going to adjust the density of the catapult parts.

Rag doll Thom requires "Scale World" turned on or his arms and legs "pop off" when he hits the ground.
Scale world turned on in the chor properties changes all of the physical properties. Makes things "heavier". Things are quite different than with it off. So the catapult gets "heavier" and doesn't have that nice "bounce" or movement because the overall density of everything keeps it stuck to the ground. This is actually good because the limit on density is 30 and that is the density of the counter weight on the catapult arm. I can lower the density of everything and have more range. I was trying to make Thom go higher and further by increasing the density of the counter weight but couldn't because it was maxed out. When I turned on scale world the catapult barely bounced and Thom ended up... uh... on mars somewhere. I think he's still going actually... I hope he brought some survival gear...



-vern

haha

I have been thinking about that too, faking it with chains that is, but haven't had time to test it out


Haha! made me laugh really hard! Nice catch though, didn't think of that

Quick note on the catapult project file. It has that link in the project setting for the "troll.txt" extended newton material. Click on the project thingy all the way at the top in the PWS, open the newton plugin triangle and set extended material to "off". It would cause that window to pop up asking for the missing "troll.txt" file when you open the project. Be careful with that setting because it can spread... unless it was fixed already. I had a problem with that during my CYP project. All of my models and projects linked to the nonexistent troll.txt material.

-vern

This is so much fun to watch. Simply amazing.

that was aaaawesome