Engine Dynamic Simulation

[willgysi]

Hi,

I'm trying to do a Dynamic Simulation of an engine that's currently simplified down to just piston, con-rod, cam, case, and gears.  

When I run the simulation, the imposed motion on the crankshaft (rpm) does not translate to motion to the camshaft correctly.  

How can I fix this?

Here is a clip of what I'm talking about.  Notice how the camshaft is oscillating, not rotating.

Video

Thanks for the help!

[JDMather]

Attach the assembly here.

[willgysi]

Here are the files that you should need.  I did a pack 'n go and just included the .ipj, .ipt, and .iam for it.

https://www.dropbox.com/sh/bsfzymvmnarsv25/KkdLJPyaxr

[JDMather]

Your gears are not "welded" to the shafts.
I was able to fix on my end, but the files are too large to post here and I don't use 3rd party sites.

[willgysi]

Interesting...what constraint or joint did you place to weld the gears to the shafts? I had tried a number of things and everything looked okay in the assembly with a DoF check as well as rotating it all manually.

A screenshot of the model tree in the dynamic simulation may be very helpful.

Thanks!

[JDMather]

I ended up rebuilding your entire assembly, but as a quick check you could

deselect automatic coversion of assembly constraints to joints (with the option to keep the joint already created) then Ctrl click the shaft and the gear and select Weld.

I'm not at my Inventor machine at the moment - so I'm going a bit by memory.

[willgysi]

Ok so I treid as you said and when building the welded joint I get an error stating impossible to assemble the mechanism. Please check joints' nature and geometry.

I tried selecting the actual origin and x axis for both the gear and shaft from the model tree but that does not work.  I tried selecting whatever faces i could and get the same result.

[henderh]

I was able to get the motion going in DS by turning off automatic constraint conversion (retaining the joints) and adding two cylindrical joints manually (instead of welding).  The adjusted assembly file is attached.

I suspect that the redundancy may be due to the fact that the Design Accelerator gearset flexible sub-assembly doesn't  allow the gears to slide relative to each other (in the axial direction). 

Hope this helps..

[willgysi]

Thank you for posting this!

This definitely helps but the overconstraining by 8 DOFs is pretty concerning.  I'll try what you did to a non Design Accelerator gear set.

[willgysi]

Sorry for double posting, but I looked into your assembly a little more and I have a few thoughts...

I increased the run time to 10s as that is what I set the full rpm range to be.  When viewing the simulation, you can see that the gears slow down and change directions.  It looks as if the cam and crank turn as they should to each other though.

Is this a screen refresh issue or something like that?  I don't think it is as the cam and crank look to be blazing by.  

I looked at the graphs as that should tell a better story than visuals. The input velocity between the crank case and crankshaft looks as it should and the camshaft follows suit accordingly at roughly double the cam's velocity as it should (but isn't exact so this is an issue to work out).  

The piston in reference to the crank case velocity looks as expected with an increasing in magnitude sine wave.

The revolution joint between the spur gear and camshaft (7th order polynomial) are where I start to worry.  There is no change in position, velocity, or acceleration as expected but there is this noisy graph for force of about a .5 lb.  Where could this force be coming from?  There shouldn't be anything acting here dynamically.

Also, the two cylindrical joints give strange results I think as well.

[henderh]

HI willgysi,

  I was able to adjust the assembly model slightly to remove the redundant DOF's (attached).

  I removed the Design Accelerator sub-assembly gearset as a quick and dirty way to accomplish this, but still using those same components.  Now DS will be able to calculate unique reaction forces.

To answer your questions:

<<I increased the run time to 10s as that is what I set the full rpm range to be.  When viewing the simulation, you can see that the gears slow down and change directions.  It looks as if the cam and crank turn as they should to each other though.

Is this a screen refresh issue or something like that?  I don't think it is as the cam and crank look to be blazing by.>>

  Yes, it was an optical illusion, much like the classic wagon wheel in old-time western movies appearing to rotate backwards, due to the frame rate of image captures.

<<The revolution joint between the spur gear and camshaft (7th order polynomial) are where I start to worry.  There is no change in position, velocity, or acceleration as expected but there is this noisy graph for force of about a .5 lb.  Where could this force be coming from?  There shouldn't be anything acting here dynamically

Also, the two cylindrical joints give strange results I think as well..>>

  This may be from the redunancies in the previous version of the simulation.  It was a statically indeterminite solution, so the forces may not been unique (i.e. false).   Now these two components are in a welded group, so they are considered as one body / mass without any type of internal loads, with respect from one component to another.

  I noticed the student banner opening the model, and thought you might be interested in viewing the crank-piston DS model for verification if you haven't done so already:  Theoretical cases for validation

  As always, please let us know if you have any additional questions, comments or suggestions.

Best regards, -Hugh