Camshaft Design Torque 1

[michaelb7]

thread404-121103

I'm trying to find the torque to drive an existing cam (the calculations were lost ages ago). I think I'm going to recommend that we buy some software for it, because it seems like the calculation that Zekeman did would have to be repeated MANY times around the surface of the cam.

In Zekeman's calculation, couldn't the normal force to follower (Fc) be calculated as:
Normal Force To Follower = (Weight to lift load)*cos(pressure angle).

Then to get torque:
Torque = (Normal Force to Follower)(Radius)
Where the radius is obtained as Zeke described.

This isn't a burning question I need to know. I'd just like to have a rudimentary understanding of the situation before I start randomly plugging numbers into software.

 

[MintJulep]

There is a chapter on cam design in every mechanical engineering design text ever written. Everything you need will be there.

Your cam profile should be described my a nice continuous function. Which means you can write the derivative of that function and identify the points of maximum and minimum whatever.

We were designing cams long before computers were invented. (In fact early mechanical computers were cams)

 

[dvd]

Although it won't do calculations for you, here is an oldie, but goodie, courtesy of Camco -

http://www.camcoindex.com/svcman/moonbook.pdf

 

[MikeHalloran]

Or, you could just assume a terrible pressure angle, the stiffest spring that will fit, and a more or less arbitrary torsional deflection/inaccuracy that you are willing to tolerate, given the maximum torque that the cam and spring can generate and the length and size of the camshaft.



Mike Halloran
Pembroke Pines, FL, USA

 

[Occupant]

Or this -->

 

[michaelb7]

@MintJulep

I didn't say it couldn't be done without a computer, it would just be very time consuming. I have a BSME from a good University and I don't remember any time being spent on cam design, so forgive my ignorance. My question was just a conceptual question to see if that particular equation could be look at in a different light.

 

[michaelb7]

@dvd

Equations 12 and 13 in section G helped. When I combined those two equations I get the exact same equation I thought it would be.

My only complaint is I don't agree how they draw r in Fig G-1. I figured r would be defined as the distance starting at the center of the cam, and drawn perpendicular to the normal force on the cam follower, ending where it intersects with the extension of the normal force line.

 

[michaelb7]

Ok last post on this topic.

I convinced myself that the book's equation was correct, they just drew the diagram differently. Thanks for the help all!

 

[tbuelna]

The first thing you need to do is establish the characteristics of each component in the system, including the cam profile, follower, spring, etc. The geometric relationship of all the components. The force/displacement characteristics in the system. And the friction characteristics at every contact in the system.

You can reverse engineer the cam profile using something like Cam Doctor. However, establishing an accurate value for torque about the cam axis will be difficult due to the large variation in frictions that will occur over the range of speeds/loads you will likely experience.

 

[michaelb7]

@tubelna

Cam doctor looks like an interesting device. I don't need to physical measure anything since I have a 3D model. Ideally there would be some program where I could drop the cam profile in, and it would spit out the pressure angle and radii in one degree increments, that would be good enough for a rough estimate. (I realize it could be done by hand...if only my company would hire an intern...)

 

[Occupant]

Camtrax from Camnetics does that. It'll even give you the CNC-Code.

 

[robyengIT]

read first the attached file and if you like go to

http://www.turci.biz

and download the add-in to excel (but this will be in italian) : up to You

 

[robyengIT]

and this too (in english)