Modeling a mechanical power system

[Flaxbeard]

I'm working on modeling a basic mechanical power system for a video game project. The transfer system consists of meshing gears, belts, and axles/shafts which can have basic clutches that engage/disengage. There are also power producers (hand crank, windmills, water wheels, turbines) and consumers (millstones, stamp mills, etc).

I'd like to ignore transmission losses. I'm wondering the best way to model the system. Right now, I am considering a system having a total torque and angular velocity. What would be the best way to model the system as a whole, using discrete time steps? Each power producer has its own angular velocity (a fast-spinning turbine vs a slow-spinning water wheel) and torque, and certain power consumers need requisite amounts of torque to spin (a millstone, for example). I'd also like to incorporate storage in the form of torsion springs and flywheels. Specifically, I'm wondering how to model interactions when there are multiple producers or consumers hooked into the same network (ie, a turbine and water wheel together). Would it be easier to model the system based on forces?

 

[tech99]

I'm working on modeling a basic mechanical power system for a video game project. The transfer system consists of meshing gears, belts, and axles/shafts which can have basic clutches that engage/disengage. There are also power producers (hand crank, windmills, water wheels, turbines) and consumers (millstones, stamp mills, etc).

I'd like to ignore transmission losses. I'm wondering the best way to model the system. Right now, I am considering a system having a total torque and angular velocity. What would be the best way to model the system as a whole, using discrete time steps? Each power producer has its own angular velocity (a fast-spinning turbine vs a slow-spinning water wheel) and torque, and certain power consumers need requisite amounts of torque to spin (a millstone, for example). Specifically, I'm wondering how to model interactions when there are multiple producers or consumers hooked into the same network (ie, a turbine and water wheel together). Would it be easier to model the system based on forces?

I suggest using the electrical circuit analogy throughout as it is so well known and understood. I suggest using power as the main commodity being supplied, and use just a DC representation, but pretend you have transformers for it, to represent gearboxes.

 

[Dr.D]

You also have the option to write the actual equations of motion for each item, if that interests you.

 

[Flaxbeard]

You also have the option to write the actual equations of motion for each item, if that interests you.

I'd like to treat the system as a whole, if possible. The system is statically constructed (no new gears etc) after creation, so if each item's properties could be combined into a whole that could work.

 

[berkeman]

I'm working on modeling a basic mechanical power system for a video game project. The transfer system consists of meshing gears, belts, and axles/shafts which can have basic clutches that engage/disengage. There are also power producers (hand crank, windmills, water wheels, turbines) and consumers (millstones, stamp mills, etc).

What language are you coding int? Are you familiar with basic Mechanics, including Kinematic Equations, Summing of Forces and Moments, etc.?

http://hyperphysics.phy-astr.gsu.edu/hbase/index.html#mechcon

 

[Flaxbeard]

What language are you coding int? Are you familiar with basic Mechanics, including Kinematic Equations, Summing of Forces and Moments, etc.?

http://hyperphysics.phy-astr.gsu.edu/hbase/index.html#mechcon

I'm using Java. I'm a uni student who has taken Kinematics & Mechanics and will take Electricity and Magnetism next semester.

 

[Dr.D]

I'd like to treat the system as a whole, if possible. The system is statically constructed (no new gears etc) after creation, so if each item's properties could be combined into a whole that could work.

I'm not sure what you mean when you say "the system is statically constructed..." Most machines are constructed in a static state, but surely that is not your point, so this does not tell me very much.

I see no reason why you cannot write equations of motion for the system as a whole. Are you acquainted with the Lagrange formulation? This should make it pretty easy, I would think.

 

[Flaxbeard]

I'm not sure what you mean when you say "the system is statically constructed..." Most machines are constructed in a static state, but surely that is not your point, so this does not tell me very much.

I see no reason why you cannot write equations of motion for the system as a whole. Are you acquainted with the Lagrange formulation? This should make it pretty easy, I would think.

It would be inefficient to model the state of every individual component in the system - it would be more performant to treat the system as one entity.

 

[berkeman]

It would be inefficient to model the state of every individual component in the system - it would be more performant to treat the system as one entity.

Can you give some concrete examples to help us our? How would you model these in your technique?

• Car automatic transmission
• Car manual transmission
  
• Glider aircraft
• Motorized Combat Tank with main gun and coaxial machine gun
• Exoskeleton equipped infantry soldier

 

[Dr.D]

Oh, I'm sorry. I thought you want to model that system as defined, not some other system thought to be in some undefined manner equivalent.

By the way, what does "performant" mean? This is a new word for me.