Projectile Launchers

If there are no specific requirements for long-range, you might be best to go with a compression or expansion spring of known qualities to achieve the consistency. It's much easier than trying to get identical results from successive trials with compressed air or explosive propellants. The whole thing can be assembled on a board hinged at one end to another one. Set your angle using a protractor and a wedge.

 

The simplest device is probably to take a pencil or some flat stick and place it through a number of similar rubber bands.

Then wind the pencil around and around until you have built up lots of potential energy in the elastic (so the elastic is twisted).

Of course you have to attach the ends of the rubber bands to something.

Working alongs these lines you should be able to knock something up pretty fast. And then as Danger says you just need to place it on a board with a hinge so you can measure the angle of test firing (of course you also need to attach a little basket to the end of the stick so that it can indeed launch things).

Or if you wanted to be fancy...

http://www.redstoneprojects.com/trebuchetstore/onager_plans.html

 

Use a slingshot and measure the distance in paces. The problem with the constant force devices is that when you change the angle the initial velocity will not necessarily stay the same, especially if you accelerate the object over some fixed distance (usually in a pipe) since the component of gravitational acceleration will vary as you accelerate the projectile. One could maybe use a friction coupling on a pump. The object then flies off at some fixed pressure. Test it horizontally with several test runs to see if it is reproducable.