Discussion of Trebuchet Optimization
Case 1: Stationary Trebuchet
Now, I will attempt to come up with a qualitative discussion of what the best possible (most efficient) trebuchet would be. First, I will discuss the example above, to see what happens when we increase or decrease each variable. To start, given the equation that the angular acceleration is equal to the torque divided by the moment of inertia, and the higher the angular acceleration, the higher the range will be, it becomes apparent that we want to maximize the torque while minimizing the angular acceleration.
l1- given that the torque increases based on how far away the mass m1 is from the rotation point, and the fact that the mass of the beam increases as l1 increases, we have several reasons to want l1 to be long. However, when we increase l1, the moment of inertia also increases due to the additional distance from the center of motion and due to the added mass of the beam. Given that the moment of inertia varies based on l1^2, and the torque varies based on l1, it becomes obvious from the expressions above that there will be a maximum value for alpha achieved at a certain length for l1.
l2- When l2 increases, the
total torque, which varies based on
-m2l2, will decrease, and the moment of inertia, which
varies based on m2l2^2, will increase, both effects which
are detrimental to the total angular acceleration.
However, the final velocity of the projectile will vary based on l2 multiplied
by the angular velocity. Since we
once again have 2 effects which act to counter each other, there will be a
maximum value achieved for the range at some value of l2
m1 and m2 – these two variables are pretty straightforward. As m1 increases, the potential energy of the trebuchet at the beginning, which varies based on m1gy where y is the initial starting height, will increase, thereby increasing the energy that the trebuchet can give to the projectile. The only concerns here are that eventually the weight of the mass will become so heavy that it will begin to distort and eventually break the arm of the trebuchet, and that the more massive the object, the more it will cost. For m2, we must be concerned that the more massive m2 is, then the more energy will be taken in accelerating it, which will cause the velocity to decrease. Therefore, it is usually advantageous for purposes of range to keep m2 small to increase distances. However, since the final momentum of the projectile also varies based on the mass, if we make the mass to light, then the projectile will bounce harmlessly off its target. Therefore, the mass of the projectile m2 really depends on how massive you’re willing to make the counterweight, which increases the energy available to put into the projectile, and how strong the target is.
M – As M increases, the torque due to the right side increases, which is a helpful affect. However, the torque due to the left side also increases, as does the moment of inertia of the beam, both detrimental affects. Therefore, it is advantageous to keep the mass of the beam as low as possible, without making it so light that it is too weak to sustain the forces due to the masses on the end.
q – As the angle q increases, the potential energy of m1 also increases. The maximum for the potential energy occurs at q = 180º, since this is the highest possible point.
Release angle – Normally, when an object is thrown at an angle above the horizontal with gravity acting downwards, the most advantageous angle to fire the projectile at would be 45º above the horizontal (q also equal to 45º). However, in this case, because there continues to be a torque on the counterweight causing continued acceleration after this angle, it becomes advantageous to release the projectile where q is less than 45º. Since the torque goes to zero as theta decreases, and the range also decreases as q decreases, after only a few additional degrees the decreased angle will start to overwhelm the increased velocity, but it is necessary to note this fact.
Case 2: Trebuchet allowed to slide on a frictionless surface
Allowing the Trebuchet to slide on a frictionless surface produces some interesting results, since as the counterweight falls, the force of gravity produces both a torque acting perpendicular to the radius, and a force aimed into the trebuchet along the arm. While the projectile is falling, this force goes to zero as the torque approaches a maximum, a state that occurs at q equals 90º. When q is greater than 90º, the force will be directed into the trebuchet, causing it to accelerate backwards. When q is less than 90º, the force will be directed forwards, causing it to accelerate in the direction of the throw.
Since the torque on the object is not affected by allowing the trebuchet to slide, the only advantage gained by allowing the trebuchet to slide will be found by having an additional velocity in the x direction gained due to the movement of the entire assembly. Because the force pulling on the trebuchet is proportional to sin(2q), the higher qo is above 90º, the more the trebuchet will be pushed backwards, a detrimental affect. However, once q goes below 90º, the force on the trebuchet will begin to accelerate the assembly in the opposite direction. Due to the relationship of the force and q, for every degree above 90º that q is initially, it will take the same number of degrees below 90 to decelerate the trebuchet and allow it to accelerate in a beneficial direction. For example, if theta is initially 100º, the assembly will accelerate backwards until theta = 90º, and will then accelerate forwards until theta = 80º, at which point the assembly will not be moving. From theta = 80º until the projectile is released, the assembly will proceed to accelerate in a positive direction, increasing Vx and therefore giving the projectile an increased range.
Given these facts shown above, it becomes apparent that it is no longer advantageous to have theta initially equal to 180º. At this point, the release angle would have to be 0º for the treb to not be moving backwards when the projectile is thrown. In this case, it becomes advantageous to have theta initially less than 180º. In fact, because the sin(2q) is a maximum at 45º and a minimum at 135º, it becomes advantageous to start q at a value of less than 135º in order to maximize both the velocity of the trebuchet in the x direction and the angular velocity at the release point. In comparison to the previous example, due to an additional force pulling on the assembly after the release angle passes the value that it was with a fixed trebuchet, it becomes advantageous to set the release angle to be even less than it was in the last example, in order to get more force added in the x direction. The other variables have very little affect on this situation, since allowing the treb to slide does not particularly change any of them. However, it also becomes more advantageous in this case to have a light assembly, since the lighter the assembly, the greater the acceleration due to the force will be. Also, the heavier the counterweight is in this case, the greater the acceleration will be. Therefore, there can now be a predictable value for the mass of the counterweight that will cause maximum range, since adding mass to it increases the force placed on the assembly while at the same time decreases the total acceleration of the assembly.
Case 3: Sling added for projectile
There are several reasons why placing the projectile in a sling which slides along the frictionless surface at the bottom and swings around to the release point will produce a more efficient trebuchet. First, the sling will increase the distance of the projectile from the center of rotation. While this does increase the moment of inertia and reduce the torque, this fact is substantially overcome by the fact that the increased distance from the center increases the velocity that the particle travels at, due to the fact that velocity at the release point is proportional to the radius and the angular velocity. For this reason alone, the longer the sling is, the faster the particle would seem to move upon release. In addition, in the first circumstance, it was very difficult to get the trebuchet to release at the correct angle, due to the fact that the projectile moves through it so fast. With the sling, the margin of error is increased substantially, allowing for a less precise release mechanism to achieve the same results. Finally, given the fact that the sling takes longer to move around in a circle than just the projectile and bar alone; the addition of the sling allows a maximum range to be achieved at a q less than in all the previous examples, allowing the counterweight to reach a lower position and giving the projectile more energy available to use.
As stated earlier, it at first seems advantageous to try to
get as long of a sling as possible in order to increase the final velocity that
the particle will have. However,
since the torque and the moment of inertia are both affected by the addition of
a sling, there does exist a point where a maximum in the efficiency will be
achieved. This intuitively makes
sense. If the sling gets too long,
then the efficiency will drop off to zero, since there will be a point where it
is so long that the trebuchet is unable to lift the projectile off of the
ground. Moreover, in this case we
have to worry about the counterweight swinging to the point where q
becomes negative, and the force of gravity on the counterweight begins acting
against the angular velocity of the trebuchet arm.
This effect will also decrease efficiency. Therefore, the length of the
sling will at first cause an increase in efficiency, before eventually dropping
off to zero.
Now, we will consider a sling attached to a trebuchet that is allowed to slide on a frictionless surface. Recall that it is advantageous to us to minimize the amount of acceleration in the negative direction when considering a sliding trebuchet without a sling because it decreases the final velocity in the x direction. The interesting thing in this case is that, for this situation, some of the initial backwards motion of the trebuchet becomes useful! Consider the sling to be initially stretched out under the trebuchet in the positive direction from the point where m2 was originally placed. In order for the projectile to move around in a circle, it has to accelerate in the negative x direction at some point. While the projectile is sliding backwards, there is a tension applied through the sling onto the projectile, pulling it back and up. During the time when the sling is being pulled along the surface, the tension accelerates the projectile in the negative direction. If at the same time the entire trebuchet is moving backwards, there will be an increased amount of force applied on the projectile over the same distance, thereby causing an increase in it’s kinetic energy as it leaves the surface and begins to swing around. Due to the centripetal acceleration and its dependence on velocity, it is advantageous to have a high kinetic energy at this point. This will increase the total centripetal force applied on the projectile as it swings.
After the projectile leaves the surface, it begins to swing around due to its negative initial velocity in the x direction. Since it is fixed around a center, it feels a centripetal acceleration around the center, in addition to the force applied by gravity on the projectile and the force applied by the moving of the trebuchet arm. Once the sling passes a position where it’s angle with the horizontal is 90º, it then becomes disadvantageous to have the assembly moving backwards, because after this point, the centripetal acceleration will begin applying a force on the projectile that will possess a component in the positive x direction. Instead, it is possible to gain energy by having the trebuchet begin to accelerate in the positive direction. From this point on, the acceleration of the assembly in the positive x direction will apply a force that will once again increase the final x velocity.
These facts suggest that there will be an optimum value for the efficiency affected by qo, and since we can now use some of the acceleration in the negative x direction, it becomes apparent that it will be advantageous to increase qo over the value used without the sling. There will certainly be an optimum value for q here, since it is advantageous to get as much acceleration as possible in the negative x direction before the sling makes a 90º with the surface, and it is desirable to get as much acceleration in the positive direction as possible after this point. Therefore, one would try to set the length of the sling and the value for q to achieve the value so that the assembly stops and begins accelerating in the positive direction at approximately the same point where the sling makes a 90º with the surface.
For a printer-friendly version of the discussion on Trebuchet Mechanics and Optimization, see the word file below:
Mechanics and Optimization.doc