Please do not spam this site with highly similar questions to ones you already asked. As already mentioned in the answer to your earlier question, multiple simulators like gazebo or V-REP can be used to provide a simulation environment for motion planning experiments. Note that these simulators only provide a simulation of the robot and environment, but no features for potential field based planning. That part you'd have to care about yourself, independent of the simulator used.

As noted before, classic potential field methods for obstacle avoidance of manipulators are prone to getting stuck in local minima. Most current state of the art planners employ other approaches and this is the reason why no implementation of classic potential field based planning for a manipulator is available as a ROS package (as far as I'm aware). Instead, planning is often performed using sampling (OMPL), search (SBPL) or optimization based (CHOMP, STOMP) planning approaches.