[Matt Leuck]

I'm a Mech. Engineering Technology major at Texas Tech, and one of our 4xxx level classes is Mechanisms of Machinery (cams, linkages, and gears). We're getting into cams now, and I had a question.

So, there are many types of profiles... Parabolic, cycloidal, polynomial, etc.

Are automotive cams are polynomial cams, or are they something more complicated?



And I dont know if you can answer this, but I'll throw it out there. My prof. couldnt give me a good answer...

I know you must be concerned with position , velocity (v), acceleration (a), and jerk (j) when designins cams.
The 3rd order 3-4-5 Polynomial cam has the boundary conditions that when the follower leaves the base circle, the y, v, and a are all 0.
So if you move to a higher order polynomial, and start taking the derivatives, the jerk would go to zero also, right? And once you get the jerk to zero, why is there a need to continue to even higher order polynomials? If the derivative of the jerk does not even have a name, then why bother with it?


It's late, I should stop thinking :clonk:


- Matt

 

[Elusive_R]

Yeah, I'd stop thinking My head hurts a bit now, too.

I'm taking a Internal Combustion Engine class right now, and we just got to cam design. I'm not sure how in-depth we're actually going, but I'll see if I can get some answers for you.

Ryan

 

[Rowdy]

What the....

I think you just gave me a brain enema...I don't think I like it either.

You should post some kind of warning, before subjecting innocent readers to that sorta stuff.

 

[71454Chevelle]

You ought to see if you can bring in UDHarold for "Show and Tell"

 

[Guest]

Yes, UDHarold has posted about just what you are talking about here if you want to do a search.

It's way above most of our heads due to the math sorta deal when you get deep into it, imho.. (And I'm an ME!)...

He made Ultradyne Cams as well as the great old CC 268HE and others throughout his career if not mistaken.

pdq67

Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

 

[69LS1]

Matt,

I am not a cam expert but there are at least 3 dirivitives past Jerk.... According to Harvey Crane they are " Snap " Crackle " and " Pop ". I dont have any idea what difference it would make when talking about numbers in the millionths of an inch range but back in the 70's Harvey Crane must have felt there was something there as he designed a series of Drag Race Rollers lobes called " Hi Low No Pop "..... The " No Pop " refering to these lobes having very low to zero Pop numbers.

 

[UDHarold]

I really should charge for this, but........
When designing a cam, you must first decide whether to make it symmetrical, or unsymmetrical--opening and closing sides different. I have done unsymmetrical cam design only since 1977, although some of my symmetrical cam designs of the early 1970s are still being made today.
Each side needs at least 2 seperate equations, and I have used as many as 5 or 6, per side...... A design needs a ramp, and a flank, and all my work at Competition Cams was of this sort, what I call '4-Equation' cams, opening ramp, opening flank, closing flank, closing ramp.
Each equation has to have starting and ending conditions, at least 4 of each, and they are: Displacement, Velocity, Acceleration, and Jerk. Where 2 equations meet, they must meet in ALL 4 conditions, or you get infinite spikes in Jerk, at the least. This means the equation has to have 8 exponents, although some exponents may be 0, and some of the coefficients may be 0.
Exponents may vary from 0 to as high as 99, although I've never used any that high. Picking the right exponents is one of the major factors in good cam design. It is learned only through trial and a lot of error. As a point of interest, Displacement, ie---cam lift, always has the exponent 0.
Acceleration at the nose means there has to be an exponent of 2 present, to give a real 2nd derivative.
I have equations written to help design cams as far as the 4th derivative, Harvey's 'Snap'. Once you get into 5 or 6 0s before the 1st signifigant digit, ie---.000001", one-millionth of an inch, I wonder if there is any noticable benefit in further pursuit of derivatives. It's a very good cam grinder indeed that is more accurate than .0001".......
That's enough to get you thinking for the next year.....

UDHarold

BTW, I am self-taught in cam design, although I did get a B in Differential Equations in Jr. college....

 

[69boo307]

Let me see if I understand this...
aside from the jerking and snapping, what you guys are basically talking about is the equation that describes the curvature of the cam lobe at a given point, right?

Say you simplify things so that the area where the lifter is contacting the lobe is a single point, travelling around the curvature of the lobe at a given speed, then the first derivative of the curve is the velocity of that point, the second derivative would be the accelleration at that given moment in time, and the 3d derivative is what you're calling 'jerk'. Then from what Harold describes, there are 4 basic curvatures on the lobe, each described by an equation.
Does that summarize it?

My question would be, what is happening at the points where you are 'changing' curves, the transition points? How do you locate and describe those points mathematically?

 

[wizard67]

*falls on floor, bleeding from the ears*

 

[Matt Leuck]

Matt,

I am not a cam expert but there are at least 3 dirivitives past Jerk.... According to Harvey Crane they are " Snap " Crackle " and " Pop ".

I like his names for them

I really should charge for this, but........

And Harold, thanks a ton! Feel free to charge me... though I can't guarantee when you'll see the money
I'm sure I'll read it over and over trying to totally comphrehend it all.

THANK YOU! :beers:


- Matt

 

[UDHarold]

Matt,

You're welcome......

69Boo307,

The end of one equation is also the start of the other. They have the lift, and at least the 1st 3 derivatives, identical. If the jerk, or rate-of-change-of-acceleration, is identical, then the acceleration is identical, etc, etc, etc.
In real life, small variations, in the millionths of an inch, are unnoticed, as they are under the level of manufacturing tolerances.
When masters or models are cut out, it is not uncommon for there to be 360 very tiny ridges, only a few 100-thousands in height(.000012", for instance). These are normally polished out by a very light hand, real craft-work.
This is all based on the design and the designer being good, and the manufacture of that design being good, etc, etc, etc. A bad design by a bad designer is, well, bad or very bad.....

UDHarold

 

[kirkwoodken]

Not to sound negative, but the Stanley brothers of steam car fame said that internal combustion engines have so many variables that no one will ever figure out how to make them run right. It's now 100 years later, and we are still improving on the performance of the IC engine. To give some idea of the Stanley's point of reference, the Doble steam car built in the 20's made 1100 foot pounds of torque! Howard Hughes family drove nothing but Dobles. Howard made money in college by winning races against gas cars with his Doble. A Doble could maintain a speed of 90 MPH. Abner Doble was a pretty smart guy! There have probably been billions of dollars spent on IC research and we still don't have them running right.

People like Harvey Crane, Harold Brookshire, Ed winfield should have their statues in front of every automotive engineering school. The basic prerequsites to make an engine run are:compression, ignitable air/fuel mixture, and a properly timed spark. To improve HP, you need only to reduce friction, raise the Brake Mean Effective Pressure, or increase the RPM limit. Achieving those simple goals requires a knowledge of thermodynamics, physical sciences, fluid dynamics, metallurgy, chemistry, electronics; almost the whole sum of human knowledge. And then it's still cut and try.

Here is a Harvey Crane link:http://www.harveycrane.com

 

[rich1978]

Hey Matt, my wife and I both got engineering degrees from Tech, classes '02 and '03. Nice looking 71 you've got, how much are you asking for it?

 

[Jblack]

People like Harvey Crane, Harold Brookshire, Ed winfield should have their statues in front of every automotive engineering school. The basic prerequsites to make an engine run are:compression, ignitable air/fuel mixture, and a properly time spark.
[/url]

So your saying there's more to it than Suck-Squeeze-Bang-Blow ???????

 

[Matt Leuck]

Hey Matt, my wife and I both got engineering degrees from Tech, classes '02 and '03. Nice looking 71 you've got, how much are you asking for it?

Just sent you a private message


- Matt

 

[Rowdy]

UDHAROLD QUOTE ; BTW, I am self-taught in cam design, although I did get a B in Differential Equations in Jr. college....




What grade did you get in Home Ec ?