How to develop cam specs

[SchmidtMotorWorks]

Jon do you consider yourself a high-functioning sociopath? Or is this an engineers attempt to torment the lesser educated by posting when he gets drunken bored after midnight? An engineering version of a prank phone call lol.

What do you recommend the other 99.76% of us do that don't have your level of Education to understand simulation isms

If you can write the above, and have the intelligence to safely operate a modern high performance engine, you can handle a 1D software.

[Stan Weiss]

I believe that the majority of members of this forum are V8 engine enthusiasts, so what follows is specifically for their interest.
Any 1D software package has an inherent problem when simulating such an engine - it simulates a static valve lift design, whereas the engine running at the speed of the sim will be experiencing a dynamic valve lift design.
Because there is considerable difference between the two, most notably in a delayed exhaust valve opening, unless this can be effectively made allowance for, the result of the simulation may well be far different when compared to the measured output from a running engine.

This can be very effectively handled by combining the results from program 10 in Blair's software 4StHead, using a feature which enables the user to save a dynamic valve lift file, and then import it into Vannik's EngMod4T software, into the Dat4T section.

Of particular note might be the result from the program 10 in Blair's 4StHead, which are extremely accurate when compared to those from a Spintron for the same engine.

The delayed exhaust valve opening mentioned above, can be as great as 20 crankshaft degrees, and possibly even more. This delay is almost completely absent in a DOHC engine, so in this latter case the dynamic valve lift file is not as important

Thanks for ALL of the great technical information / help.

Stan

[CamKing]

The most experienced person alive working in engine development that I know of; Hans Herman, does 1D engine simulation as a consultant job for leading OEMs and race teams all over the world.

I don't know how he's still BS'ing people to pay him.
Every project. where we both designed cams, he wasn't even in the same ballpark.

[pcnsd]

(actually most major dimensions)

1. Get a 1 D software and learn how to use it.

2. Design the best engine you can based on where the 1D software leads you

3. Build an engine based on #2

4. Plan a dyno session with the variety of parts that you can afford to test.
a. Cams
b. Cam index
c. Intake manifolds
d. Headers
Note, make the tests with both single-part changes and reasonable matching sets of parts.

5. Make 1D models for each combination planned in #4

6. Dyno test to the plan in #4:

7. Compare the dyno results to what the 1D models predicted

8. Understand the "simulationisms" (how the tests vary from the simulation)

Repeat steps 2-8 until you are satisfied.

This is an example of a industry standard optimization or continuous improvement loop. It is used the world over. It is what the Japanese applied, most notably at Toyota (Toyota Production system) to kick our industrial asses in the 1980's.
I train HS drop outs on its application and I have yet to find one that can't comprehend and apply it. It is also called the Deming cycle after its first advocate William Edward Deming, a US statistician. It's most basic format is Plan, Do, Check, Act...repeat.

[hoffman900]

I believe that the majority of members of this forum are V8 engine enthusiasts, so what follows is specifically for their interest.
Any 1D software package has an inherent problem when simulating such an engine - it simulates a static valve lift design, whereas the engine running at the speed of the sim will be experiencing a dynamic valve lift design.
Because there is considerable difference between the two, most notably in a delayed exhaust valve opening, unless this can be effectively made allowance for, the result of the simulation may well be far different when compared to the measured output from a running engine.

This can be very effectively handled by combining the results from program 10 in Blair's software 4StHead, using a feature which enables the user to save a dynamic valve lift file, and then import it into Vannik's EngMod4T software, into the Dat4T section.

Of particular note might be the result from the program 10 in Blair's 4StHead, which are extremely accurate when compared to those from a Spintron for the same engine.

The delayed exhaust valve opening mentioned above, can be as great as 20 crankshaft degrees, and possibly even more. This delay is almost completely absent in a DOHC engine, so in this latter case the dynamic valve lift file is not as important

Thanks for ALL of the great technical information / help.

Stan

Then people should focus on fitting larger cam cores, finding ways to run less valve spring, etc.

In all seriousness, Billy Godbold brought this issue up. He said what you can do, is you have Spintron data, is to input that into your model.

The user defines the parameters and must recognize the limitations of what they’re doing and be able to I repeat the information.

[mekilljoydammit]

Man, what's the intended dividing line between "advanced engine tech" and "engine tech" on this forum anyway? If you are happy just plugging one cam in it's "engine tech" if you want to test stuff rigorously it's "advanced"?

Kind of joking, but also not really. Maybe it's my age, but I've found dinking with some of the less sophisticated 1d sims (at least I'm pretty sure that's what EAP is anyway) at least as easy as the various rules of thumb like the Blair one laid out. Which reminds me I really need to cut Vannik a check next payday.

I think what people really need to do is apply design of experiments to this stuff, though that basically slots into what Jon is suggesting in the OP of this one.

[CamKing]

Man, what's the intended dividing line between "advanced engine tech" and "engine tech" on this forum anyway?

If the discussion uses terms like "3/4 Race", "Full Race", "Lopey Idle", "Sounds mean", or "max effort", it shouldn't be on the Advanced Engine Tech forum.

[SchmidtMotorWorks]

The most experienced person alive working in engine development that I know of; Hans Herman, does 1D engine simulation as a consultant job for leading OEMs and race teams all over the world.

I don't know how he's still BS'ing people to pay him.
Every project. where we both designed cams, he wasn't even in the same ballpark.

The project he did for us last year was as close as the two dynos it was tested on.

[Orr89rocz]

https://www.speed-talk.com/forum/viewtopic.php?f=1&t=32990

One method posted here

[blykins]

There are no rules or formulas.

The best bet is to gain experience with a specific engine, or talk to someone who has.

I've had custom camshaft grinders offer me cams that trailed by more than 25 hp compared to ones that I have spec'd. But 99% of my engines are SBF's or FE's.

The biggest issue I've found with most grinders is that they think that they have nailed it on the first lick. The best scenario is to grab a hand full of cams and swap them out on the same engine, on the same day, on the same dyno.

[Orr89rocz]

Another method i found for typical pushrod v8

Intake duration

Int dur = 148.1+ (3.8 * comp ratio) + (12 / (rpm at tq peak * stroke / 33002.4) ) + 0.007 * ( rpm at tq peak ^ 0.77) * (bore diam ^ 0.27) * ( stroke ^ 1.47)

Exhaust dur = int dur * ( 1 + ( ( 0.348 / rod stroke ratio)^1.99)) + 12

Now that 12 seems arbitrary for a drag race na engine for long powerband after peak but who knows

Let me dig up the lsa part

[hoffman900]

There are no rules or formulas.

The best bet is to gain experience with a specific engine, or talk to someone who has.

I've had custom camshaft grinders offer me cams that trailed by more than 25 hp compared to ones that I have spec'd. But 99% of my engines are SBF's or FE's.

The biggest issue I've found with most grinders is that they think that they have nailed it on the first lick. The best scenario is to grab a hand full of cams and swap them out on the same engine, on the same day, on the same dyno.

You’ve created a database in your head.

Create a working computer database, and you can reverse engineer why what you spec’ed worked better.

[Scotthatch]

Man, what's the intended dividing line between "advanced engine tech" and "engine tech" on this forum anyway?

If the discussion uses terms like "3/4 Race", "Full Race", "Lopey Idle", "Sounds mean", or "max effort", it shouldn't be on the Advanced Engine Tech forum.

Now that's funny

I may giggle about that line all day

[blykins]

I really don't need a computer to tell me that a port is big and slow, or an exhaust side is poor.

Too many variables for a database or formula to work correctly. May get you close, but there's sometimes a big difference between close and optimal.

[RevTheory]

Another method i found for typical pushrod v8

Intake duration

Int dur = 148.1+ (3.8 * comp ratio) + (12 / (rpm at tq peak * stroke / 33002.4) ) + 0.007 * ( rpm at tq peak ^ 0.77) * (bore diam ^ 0.27) * ( stroke ^ 1.47)

Exhaust dur = int dur * ( 1 + ( ( 0.348 / rod stroke ratio)^1.99)) + 12

Now that 12 seems arbitrary for a drag race na engine for long powerband after peak but who knows

Let me dig up the lsa part

Thank you!