The dynamics of robotic arms are fairly complex, especially when there are more than three joints to consider. The problem is that the movement of each joint moves all the links beyond it, which can induce torques at other joints. You have to consider how the movement of all links affects each individual joint.
There is software that can automatically compute the dynamics based on the geometry you provide (MATLAB + SimMechanics comes to mind, although that is an expensive option). I don't have any knowledge of free simulators, but I'm sure they exist. You may have to search around to see what simulator is right for your application.
Another option is to program your own simulation. The mathematics behind the dynamics of serial link robots are covered in many different textbooks. In order to do this part from scratch you would need to be familiar with using matrices, plus some understanding of differential equations. I can give some of the basics here, to give you an idea of what you would have to program.
You can calculate the torque required to move a serial-link robotic arm using the following nonlinear matrix equation:
$\boldsymbol{\tau} = \boldsymbol{M}(\boldsymbol{q})\ddot{\boldsymbol{q}} + \boldsymbol{C}(\boldsymbol{q},\dot{\boldsymbol{q}})\dot{\boldsymbol{q}} + \boldsymbol{G}(\boldsymbol{q})$
where
$\boldsymbol{\tau}$ is the (6x1) vector of joint torques
$\boldsymbol{q}$ is a (6x1) vector of joint angles
$\dot{\boldsymbol{q}}, \ddot{\boldsymbol{q}}$ are the (6x1) joint angular velocity vector and joint angular acceleration vector, respectively
$\boldsymbol{M}(\boldsymbol{q})$ is the (6x6) inertia matrix of the robot, which is a function of the joint angles
$\boldsymbol{C}(\boldsymbol{q},\dot{\boldsymbol{q}})$ is the (6x6) centrifugal and Coriolis effect matrix, which is a function of the joint angles and angular velocities
$\boldsymbol{G}(\boldsymbol{q})$ is the (6x1) gravitational torque vector, which is a function of joint position
Notice that this equation includes both the dynamic and static loads on each joint. If you set the joint angular velocity and acceleration to zero, you would have the static equation by itself.
You may be able to find expressions for the $M$, $C$, and $G$ matrices online, but you can also try to derive them for your application using the Euler-Lagrange method.
So, the steps you would have to take are:
- Derive the dynamic matrices associated with your application.
- Derive the equations that convert the motion (position, velocity, acceleration) of the end-effector to the motion of the joints (the inverse kinematics).
- Write a program that
- Converts an end-effector trajectory to a set of joint trajectories (position, velocity, and acceleration).
- Computes the required torque using the dynamic equation for each point along that trajectory.
This is all possible (in fact I have done this myself for a 3 DOF manipulator) but it might be too much work for an enthusiast.