David Vizard's 128 cam selection for small block Chevys

[Warp Speed]

Aren't you a can designer with the best software available? I believe that is what you have told us over and over and over...

You just couldn't resist eh?!?

You neither eh?!?

I'm trying to stay out of it and watch the dumpster fire these always turn into. You'll make sure of that!
Tell us again was your field of expertise is?

[GARY C]

You just couldn't resist eh?!?

You neither eh?!?

I'm trying to stay out of it and watch the dumpster fire these always turn into. You'll make sure of that!
Tell us again was your field of expertise is?

Your a pro at staying out of it as you have shown here... My field of expertise is exposing those who claim to be experts!

[SchmidtMotorWorks]

Aren't you a can designer with the best software available? I believe that is what you have told us over and over and over...

I guess that you are referring to the cam design software I wrote.
It enables a designer to define valve motion with detailed control of the shape of the acceleration curve to include jerk and snap.
It is not involved with determining lift and duration specs or mechanical analysis.

For those goals working at a professional level, one would also need:
1D simulation
Dynamic Motion Simulation
CFD would be helpful

[GARY C]

Aren't you a can designer with the best software available? I believe that is what you have told us over and over and over...

I guess that you are referring to the cam design software I wrote.
It enables a designer to define valve motion with detailed control of the shape of the acceleration curve to include jerk and snap.
It is not involved with determining lift and duration specs or mechanical analysis.

For those goals working at a professional level, one would also need:
1D simulation
Dynamic Motion Simulation
CFD would be helpful

So what your saying is that what you have to offer for the average racer is useless and maybe what DV resents could be more beneficial to them?

[SchmidtMotorWorks]

So what your saying is that what you have to offer for the average racer is useless and maybe what DV resents could be more beneficial to them?

The average racer is not involved with cam design.
My software is involved with cam design.
What DV does isn't cam design.
Based on things he has posted, I don't think DV knows how to design a cam.

What DV does is analogous to choosing a 2br single floor house and saying that you architected the house.

[GARY C]

SchmidtMotorWorks wrote:
> [quote="GARY C" post_id=942248 time=1681136169 user_id=21117]
> So what your saying is that what you have to offer for the average racer is
> useless and maybe what DV resents could be more beneficial to them?
> [/quote]
>
> The average racer is not involved with cam design.
> My software is involved with cam design.
> What DV does isn't cam design.
> Based on things he has posted, I don't think DV knows how to design a cam.
>
> What DV does is analogous to choosing a 2br single floor house and saying
> that you architected the house.
What you do is talk!

[SchmidtMotorWorks]

Here is an image from the software for motion design that I wrote.
This is actual cam design, not just listing specs.
If you understand what all those curves mean, you will know that it enables extraordinary definition of motion.
This software can create any practical acceleration curve shape and solve the integration.
If you know of any software that does it better, I would like to know about it.

If you don't understand what those curves mean and why their shape matters, but are interested in learning, I will explain as time allows.

I eagerly await an example DVs cam design software.

CDS.png

[SchmidtMotorWorks]

What you do is talk!

I'm curious, what would you call making a video titled about Pro-Stock, that was nothing more than the most basic cam stuff that one could find in the 1950's?

[digger]

SchmidtMotorWorks wrote:
> Here is an image from the software for motion design that I wrote.
> This is actual cam design, not just listing specs.
> If you understand what all those curves mean, you will know that it enables
> extraordinary definition of motion.
> This software can create any practical acceleration curve shape and solve
> the integration.
> If you know of any software that does it better, I would like to know about
> it.
>
> If you don't understand what those curves mean and why their shape matters,
> but are interested in learning, I will explain as time allows.
>
> I eagerly await an example DVs cam design software.
>
> CDS.png

the average person doesn't care about all these curves its the cam grinders problem. I think a discussion on events is more useful.

Given the valve events are all coupled there is at least some merit in looking at centrelines but without a specific criterion its impossible to have a clear winner that cam A is better than cam B as there isn't a single answer as to what is the best cam.

Its fairly straight forward to determine which is more important for a given application as you start with optimum cam and change one variable at a time and look at sensitivity of the output. You could vary a single valve event or only a single lobe centreline (which involves 2 events and usually also the lobe area and lift change at same time in practical terms)

If you were use Engmod4t to iterate to a optimum cam what would be the method to determine start point and path in changing one or variables to arrive at the optimum solution as fast as possible?

[SchmidtMotorWorks]

the average person doesn't care about all these curves its the cam grinders problem. I think a discussion on events is more useful.

Given the valve events are all coupled there is at least some merit in looking at centrelines but without a specific criterion its impossible to have a clear winner that cam A is better than cam B as there isn't a single answer as to what is the best cam.

Its fairly straight forward to determine which is more important for a given application as you start with optimum cam and change one variable at a time and look at sensitivity of the output. You could vary a single valve event or only a single lobe centreline (which involves 2 events and usually also the lobe area and lift change at same time in practical terms)

If you were use Engmod4t to iterate to a optimum cam what would be the method to determine start point and path in changing one or variables to arrive at the optimum solution as fast as possible?

The video that I responded to was titled "Hyper advanced ProStock-plus, cam tech", if it were actually about what the title BS says then it is not for the average person.

For a Pro-Stock engine, events are limited by the capability of the mechanism, there is no point in calling for events that the mechanism cannot achieve with the required durability.

The ordinary way to optimise for multiple input variables is:
1. Do a sensitivity study
2. Optimize the most sensitive variables first
3. Check that variables are not stuck in local minima/maxima
4. Repeat

There are many methods to generate new variable dimensions, we used to license one called Hyperopt that was basically a black box that had a few optional methods. On some problems it was amazing, even when I knew the path to the answer, it could get there in the same number of steps with no knowledge of the problem, just the inputs and goal.

If I were to do that today, I would work in Python because there are so many people working on optimization code. That said, with so many people doing it, there is no money in it so it is a no-go zone (other than using others work) as interesting as it might be.

For engine stuff where the variables have very different scales, if one wanted to just do something that was fast and easy to implement and debug, I think it would be good enough just to use a finite differences method in a predetermined sequence.

[GARY C]

SchmidtMotorWorks wrote:
> [quote="GARY C" post_id=942274 time=1681148650 user_id=21117]
> What you do is talk!
> [/quote]
>
> I'm curious, what would you call making a video titled about Pro-Stock,
> that was nothing more than the most basic cam stuff that one could find in
> the 1950's?
More than you have ever offered to a cam thread.

[Erland Cox]

Jon! Is your cam program hard to use?
Is it available to the general public?
I would like to learn more about cams and mechanisms.

Erland