Understanding Design of Motion Graphs Like Cams

[Stan Weiss]

Jon,

I have in Dynomation 5 the camlobe files from compcams.
I asked Larry Atherton from Motion Software if he could change the valvelift numbers from two digits behind the comma to 4 digits.
He told me compcams would not give this detailed info for a reason I think.
So what we get is duration with valvelift numbers like 0,43” lift.
In my opinion this is not accurate enough for example:
0,43” lift = 10.922 mm
0.4359” lift = 11.07186 mm
Difference lift = 0.14986 mm

Meaning exporting those numbers in excel to analyze this profile make less sense because of accuracy.
Is there another way to do something with these profiles from compcams so we can use these profiles for design purposes ?

JS

This is the exact reason that they do not give more detail.

Stan

[SchmidtMotorWorks]

Jon,
I have in Dynomation 5 the camlobe files from compcams.
Difference lift = 0.14986 mm
JS

Dynomation isn't intended for analysis of cam dynamics.
0.15mm is a smaller deviation than the actual motion of a valve in a running push-rod engine.

I don't think the comp lobes library is based on the designed motion anyhow.
I think they just take a few key dimensions and fit a spline to it like I do in CamFlowRPM, that is close enough for flow if it is done correctly.

If it is real motion data, you could smooth it and get close.

[SchmidtMotorWorks]

I am a member of SAE so i get reduction but in this case do you think the paper is worth the money ?

JS

That depends on what you want out of it.

If you want to get a basic understanding oh how cams were first designed with polynomials and how dynamic response was computed in 1953 that will be a good start.
There are a couple of related papers that you will also need.

If you want to code it, you won't be able to do it in a spreadsheet.
It is fairly advanced coding, a few months concentrated hobby work for someone new to the subject.

[Stan Weiss]

This is an interesting read.

http://www.profblairandassociates.com/p ... 20Lift.pdf

Stan

[statsystems]

This is an interesting read.

http://www.profblairandassociates.com/p ... 20Lift.pdf

Stan

So I opened the link you posted Stan and read the FIRST paragraph. Damn it made me laugh. The FIRST thing they state is what Mike jones already published here on ST (and possibly other places)!

Sometimes we strain at a camel to swallow a gnat.

[vannik]

So I opened the link you posted Stan and read the FIRST paragraph. Damn it made me laugh. The FIRST thing they state is what Mike jones already published here on ST (and possibly other places)!

Sometimes we strain at a camel to swallow a gnat.

Except 4stHEAD predates SpeedTalk as far as I know, I was a beta tester for Prof Blair in 2000 and it stated it then already.

[statsystems]

So I opened the link you posted Stan and read the FIRST paragraph. Damn it made me laugh. The FIRST thing they state is what Mike jones already published here on ST (and possibly other places)!

Sometimes we strain at a camel to swallow a gnat.

Except 4stHEAD predates SpeedTalk as far as I know, I was a beta tester for Prof Blair in 2000 and it stated it then already.

I think if you look back in this thread...no there in another thread started by the OP where Mike stated how he developed a cam lobe. He basically said he didn't. He said that he determines VALVE MOTION first and then does the lobe. That is what it states in the first paragraph of the link posted.

Mike will please correct me if I misquoted you here.

[SchmidtMotorWorks]

Except 4stHEAD predates SpeedTalk as far as I know, I was a beta tester for Prof Blair in 2000 and it stated it then already.

Do you know if 4stHead made modifications to the acceleration curve to solve the integrated velocity and lift to ~0.0, or was that up to the user to adjust?

[SchmidtMotorWorks]

I think if you look back in this thread...no there in another thread started by the OP where Mike stated how he developed a cam lobe. He basically said he didn't. He said that he determines VALVE MOTION first and then does the lobe. That is what it states in the first paragraph of the link posted.

Mike will please correct me if I misquoted you here.

One of the points in Blairs paper is that it is a spiral process. The optimum design is a compromise between the desired valve motion and the mechanical limitations.

Jones and Blairs method of designing cams is very different.
If you think that the point that they are both designing valve motion makes them the same, you are missing some interesting stuff.

As far as I know Blair was the first person to make a commercial cam design application that formed the valve motion primarily through design of the acceleration curve.
Hans Hermans method did also design with an acceleration curve but it was very simple and offered little control of the shape.

[hoffman900]

I'll have to read Prof. Blair's article later. I've always found his stuff to be very good.

You can still access Harvey Crane's old site using Waybackmachine:

https://web.archive.org/web/20100112082 ... ecrets.htm

[JCR]

You can still access Harvey Crane's old site using Waybackmachine:

https://web.archive.org/web/20100112082 ... ecrets.htm

Thanks for posting that link.

[SchmidtMotorWorks]

Here is a zip file with the images from Harvey Cranes site.

cam_images.zip

[wyrmrider]

from cranes site re-referenced above
recommended reading
https://web.archive.org/web/20100115052 ... eading.htm

[Kevin Johnson]

... Hmm. I might find out decisively why the Mitsubishi 6G72 cams we used last time throws the rockers.. and even snaps the lifter heads off .. study time, I think.

http://www.stealth316.com/2-headinfo.htm

http://avto.pro/part-MD150266-CHRYSLER-235/

32.jpg

Check to make sure that the cam lobe does not have a slight taper. I think this hypothesis would also predict side loading of the valve stems which would act as a crude bending spring and transmit a side torque to the ball.

I am wondering if the rollers on the rockers are getting too much side loading.
If you can get the geometry, maybe you can graph that and see.

[SchmidtMotorWorks]

I should have been more clear about the direction of side loading; I meant from too extreme acceleration on the profile causing the contact angle to increase too much.
I have some of those rockers in my donor search box, the sockets do look rather shallow.