Plotting Issues

[rework]

When graphing measures in Mechanism, I have noticed that the plotted values are not consistent with time. For instance, in a revolution, the first curves have a value in this case, 11,500, compared to a normal value of 10,000 for ending values. I am wondering if this is related to another issue I have noticed, that being the rotational speed of the model when it first starts the "Run" is rapid --- then it slows down to a consistent rpm. Strange. I find that increasing the frames per second affects this. Am I the only one who has had these issues?

Thanks for looking. --- rework

 

[rework]

lamon19 --- Thanks for your reply. Are you suggesting that the program is picking up parts of a previous release? I was hoping others had noticed these inconsistencies also, and had an explanation for them. I really don't care that the "Run" speed starts fast and then slows down. I am concerned that the graphed values don't seem to be consistent. It causes me to doubt their accuracy.View attachment 5448

rework

 

[Ray Ellender]

Mechanism Dynamics runs a transient analysis, which may settle down to a steady state after a period of time, or may not. If you start it with an instantaneous input, it is going to take a while to settle down, just like it would in the real world.I notice you have only plotted one revolution. Would you really expect a real system to settle down to a steady state within 1 revolution?


You may need to add some damping. In the real world some damping is nearly always present. On the other hand maybe your system really does produce an oscillating result, e.g if you have a Hookes joint running at an angle.
Edited by: Ray Ellender

 

[rework]

Ray --- Thanks for the analysis of my question. The reason I am showing only 1 revolution, is because I have entered 14 or so equations to define the forces on the pistons, tied to time for the single revolution. Going beyond that can certainly be done, but I would rather not, if I can with some assurance, ignore the first cycles. How would I add damping to an engine or piston of any kind, without effecting the forces?

Is the transient analysis also the reason Mech Dynamics runs the model fast at first? Is there a way around that, because the plotted points are also spaced farther apart at the start, sometimes.

Rework

 

[Ray Ellender]

Adding damping will affect the forces, but damping will exist in reality (e.g. oil film viscosity, friction) and these generate forces. For your purpose you probably need very light damping which would generate very small forces, just to help the system get to a steady state (which might be an oscillating 'steady' state) more quickly. You may have to try a number of different values to get the best response. You are in the process of discovering that getting a dynamic mechanism analysis to run well is as much an art as a science, because you are dealing with a system that is probably non-linear and has to go through a transient to get to a steady state, using an iterative numerical solver. If you have already written 14 equations, and your mechanism is relatively simple, e.g piston, conrod and crank, it may be quicker to write the equations for the mechanism (which you can find in any decent textbook on engine design) and not bother with MDO!


Good luck

 

[rework]

Ray -- Thanks for the suggestion, but I am playing with a wobble plate axial mechanism, either engine or pump/motor. I have attached a simple example that I made for the purpose of understanding how Pro-E treats a wobble plate. It is in WF5.0. The nutating wobble plate sees centripetal force as well as acceleration in two planes (or is it just 1 force -- I can't find much info on nutating bodies). On my model, I do not understand why the "Nosepin_Hor_Force" jumps position depending upon which body is selected for the force to be exerted upon. The value seems to stay the same. Also the "Mom_Vert_Plane" is calulated using some strange moment arm, unknown to me.

I am eventually needing to balance this beast. I can handle balancing of a rotating assembly by suspending it all from a single pin joint. In this case, I see no way to build it using a single pin joint. Can I sum the reactions of two pin joints? Thanks for any help --- this is a bit unusual. --- Rework

2011-12-03_174912_Export_1.rar

 

[rework]

In the process of working with this nutating body, I am having to verify the forces generated by the masses involved, by looking at the reactions at the connections. This is not easy with both forces and moments generated. It would be a big help if there was a way to see what forces are actually generated, as in thru the CG or radius of gyration. I suppose this would be more code to create in the program, since the code must solve the forces for each small mass.

Does anyone else ever question the dynamics results in the above manner?

Any discussion would be helpful ---- am I just a nut case!!

Rework