Cam follower

[marcofa]

How can I constrain cam with follower?
I understoud that in Assembly is not possible to constrain two surfaces.
How can I simulate in DMU Kinematic?
Thanks
Marco

 

[Azrael]

make a section on the cam (in line with the center of the lifter)and create a point where the lifter meets the cam. Then create a point-curve joint, as a driver you can set a revolute on the cam

 

[marcofa]

Is not right, because, when cam turn, contact point on the follower move along a circle that is round head section.
Marco.

 

[itsmyjob]

I would use a sliding curve kinematic joint.

Eric N.
indocti discant et ament meminisse periti ​

 

[marcofa]

How constrain two curves coincident, in assembly?
I make for both, intersection and extraction but I cannot constrain
Marco

 

[itsmyjob]

Create another file with just 1 point and make coincidence curve 1 / point and curve 2 / point.

Technically you don't need assembly constraints for Kinematics

Eric N.
indocti discant et ament meminisse periti ​

 

[marcofa]

But you need to put in contact two lines, if not in PRECISE contact, when you call sliding lines joint in DMU you are refused because there is a 0,0000012 mm gap in between.
So you need anyway to constrain two lines.
I discovered that creating a point on follower you can constrain with cam surface. When you apply sliding joint that constrain will be ignored.
Another idea?
Marco

 

[marcofa]

What you means for "another file"?
Marco

 

[itsmyjob]

1 catpart A for curve 1
1 catpart B for curve 2
1 catpart C for one point

coincidence A,C
coincidence B,C

sliding curve A,B

Eric N.
indocti discant et ament meminisse periti ​

 

[marcofa]

How have to make the point in PartC? In a sketch or 3D?
I tryed both, but when I link point to curve, Point go away and geometry disappear from design space.
Marco

 

[MRSSPOCK]

This is the best I can manage to achieve.

It isn't 100% accurate, but it could be made to be accurate.

There is probably a very easy way to do this, but I never used kinematics until today, so I really just did it in probably a silly way.

 

[marcofa]

Can you post the files?
I also find a solution and I like to compare.
Marco

 

[MRSSPOCK]

I prefer not to post a bad solution. Once I get a more precise technique, I can post it.

In my example, I actually constrained the centre point of the follower's spherical end, i.e. like the centre of the ball, to a contour which was an offset from the cam profile, an offset equal to half the ball. I am still trying to understand if my example is good or not, or does it include a mistake.

 

[marcofa]

Your solution is good only if head of follower is spherical, because you revert to Point Curve joint.
If head of follower have parabolic profile your system is not valid. Picture of your solution enclosed.
I developped procedure, using Sliding curve joint, that work with any profile coupling, but is a bit complicated in case of multiple followers i.e 8 cylinder engine (32 followers): for any joint I have to constrain tangent first couple, go to DMU to create a joint, revert to Assembly, rotate the cam spindle to do tangent second couple then constrain, go back to DMU to create a joint ......... 32 time.

Any other solution USING "SLIDING CURVE" JOINT ???
Marco

 

[MRSSPOCK]

The title of the thread below suggests it refers to tutorials, including one on cams possibly? If you want to wade through all those links to find if there is any, then good luck. There are too many for me. I couldn't be bothered. It's a pity there wasn't an index explaining what each one relates to.

http://www.eng-tips.com/viewthread.cfm?qid=284100&page=1