easy way to change cam design? 1

[FrenchCAD]

I am currently analysing an assembly in wich a cylindrical is used to transmit reciprocating movement of an incoming shaft to an exhaust shaft perpendicular to first one. Then, there is a connecting rod going from this second shaft to another sub-assembly to get final movement. The groove's design of the cylindrical cam is known.

Matter is that I do not like this assembly design. It doesn't seem efficient to me compared to the whole system and what is expected from it.

My thought here is then to replace the cylindrical cam by a gear/worm connection and add egg-shaped cams on the exhaust shaft so that I will be able to work with a more efficient cam/follower assembly.

Now, is there an easy way to "transfer" the cylindrical cam's groove's design to the egg-shapped cam to get the same movement at the final stage?

Cyril Guichard
Mechanical Engineer Consultant
France

 

[MintJulep]

Just plot displacement vs. rotation of the input shaft.

 

[FrenchCAD]

basically, I need to round my current groove profile around me exhaust shaft?

Cyril Guichard
Mechanical Engineer Consultant
France

 

[MintJulep]

Pretty much. You need to take the lever ratios of the connecting rod and "another sub-assembly" and any gear ratios into account.

 

[FrenchCAD]

I was pretty sure about that. That's why I intend to use a gear ratio of 1:1

I have no idea of lever ratio yet though.

Cyril Guichard
Mechanical Engineer Consultant
France

 

[zekeman]

I don't get the problem. Do you mean that you have a shaft reciprocating along its axis ( not oscillating) at a prescribed rate and you want an outut shaft at right angles to reciprocate at some other motion? Or what? A drawing or better description might help.

 

[FrenchCAD]

I'll upload some picture of the mechanism this weekend

Cyril Guichard
Mechanical Engineer Consultant
France

 

[unclesyd]

You might want to take a look around this site.

http://www.cunningham-ind.com/index.htm

 

[nbucska]

it may be worth to check if an electronic solution
may not be cheaper or easier.

Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[solider]

Typically, the motion of the cam follower is designed with a standard cam law, such as modified trapezoidal, mod-sine, etc. Then you use more links, each transmiting motion from one to the next, until you get to the output motion that does the action. The problem is, the linkage between the cam and the output distorts the motion from the standard motion law at the cam, so that the motion at the output (where you want it) isn't the same any more..

If you know the exact displacement/ timing (motion) of the output arm of the cylindrical cam (Barrel cam), then what you asking for is not too much, but you will have to put that motion and linkage configuration into a cam design package such as "Camlinks"(

http://www.camlinks.com).

As an aside, what is preferrrable with cam design is to design the motion at the place you want it (at the output), (with a timing diagram to put in dwells, rise, return dwell or something like that), and then do the inverse kinematics to derive the cam. In that way, the motion where you want it never changes no matter how complicated the linkage is.

 

[FrenchCAD]

for any of you curious, I finally uploaded pics :

current design

http://www.forum-des-ingenieurs.com/upload/img01.jpg

desired design

http://www.forum-des-ingenieurs.com/upload/img02.jpg

cylindrical cam way design

http://www.forum-des-ingenieurs.com/upload/img03.jpg

Cyril Guichard
Mechanical Engineer Consultant
France

 

[solider]

Hell FrenchCAD,
I've taken the liberty to model you a cam in 2D with the motion (timing diagram) you have given. Hopefully the images appear!!).
I can not quite make out what the mechanism is in your images - I have taken a guess and modelled it kinematically - ie only the centres of the links without the body of the parts. The solid bodies make no difference to the cam data.
I have guessed you have a vertical slider driven by the cam via a rocker and a roller that slides across a horizontal plate at the top of the vertical slider. I have put the motion (in mm even the timing diagram says deg) on the vertical slider and derived the cam via the linkage.

I have shown the angular displacement, velocity, and acceleration on the timing diagram. It is not the best motion to start with - the type of motions you get with radii as given on your timing diagram drawing are very poor as you get step changes in acceleration and that will excite vibrations needlessly. The last segment in the motion diagram has been designed with continuous acceleration (3-4-5 polynomial) as a better example.

Before I could do a full model, I would need to the mechanism details. Perhaps you could post a side view of the 3D image with a section showing all the parts. And then I would need to know the detailed design data: for example - distance between Cam centre and Rocker centre, the length of the rocker, the diameter of the roller, and the direction the cam is rotating.etc

But anyway, the cam cutting data can easily be saved once you have all this data. If you let me know the full details I could post you the cam cutting data in a suitable format for your CNC.

Hope this helps.

Solider

 

[FrenchCAD]

wow, I wasn't hoping so much help, that's great!

I have very poor information myself as well (only geometrical information - no velocity, no torque) but I'm fighting with people here to get those.


Cyril Guichard
Mechanical Engineer Consultant
France

 

[solider]

Dear FrenchCad,
If you do not have the geometric data, then you will not be able to generate a Cam profile!

If the geometric data is still flexible then you have an opportunity to arrange your mechanism to ensure it runs well at your design speeds. The final design should consider: cam pressure angle, optimising the motion design, linkage stiffness, cam meterial and hardness, cam shaft stiffness (length and diameter), backlash, connection of cams and gears to shaft (fretting), motor speed fluctuations (drive stiffness) etc, etc.

It is quite easy to arrange the mechanism to run badly!!!


Good Luck
Regards
Soliders

 

[FrenchCAD]

thanks for the enlightment solider

we decided to not change design as it would be much more complicated than find another solution.

Cyril Guichard
Mechanical Engineer Consultant
France