Engine simulation programs ??

[mk e]

For the love of god don't ask about engine oil.

LOL

[mk e]

the answer is quite simple, id use the sim to work out what parts to test as this will cut down on dyno time/cost and provide some validation to the simulation at the same time. then you go and test those parts. rinse repeat

That sure sounds right and I know sim developers collect model/dyno sets to help get the sims closer.

I've sent a few bug reports into DM over the years so for sure mistakes in the code happen....but everyone one I've sent in has been addressed promptly and they asked if they could retain the frankenferrari model as I optimized the bejesus out of it so it's pretty sensitive to ever small changes. Hopefully I'll get my garage finished so I can fix the engine and get it to the dyno this winter to see how good the sims are with something on the limits.

[engineguyBill]

I like Allen Lockheed's engine simulator program, but haven't heard much about him or the product lately. Don't know if it is still available.

[Stan Weiss]

One problem with most dyno testing is the method used. For how long or where on the track or at what engine RPM is the engine actually seeing lets say a 600 RPM / Sec acceleration rate?

Stan

Have you ever used a dyno? Operated one? Been to a racetrack? Honest questions

A variable acceleration rate (fix load) dyno I think is a much more real world test. Some even use computer control load to simulate different gear ratios or track up hill or down hill loads.

Stan

Would you answer me please..

Sure just like you never answered mine and just asked questions. I will follow suit. Why should I?

Stan

[4vpc]

Let me ask you a question, I'd like a little more power in my pulling truck engine, would you run it on a sim and make the changes it shows to be better, and just go truck pulling again assuming it's making more power, or would you pull the engine and take it to an actual dyno and try some things on it ? Keep in mind in truck pulling there's no way to validate 10-20 HP difference by the way the truck pulls. Too many variables.

Knuckledraggers pulling truck team !

Randy

You didn't ask me, but I have a question:

Do you now or have you considered using some of the modern data logging capabilities on your son's truck, or would it be cost prohibitive?

I'd use what my sim suggested (above my own personal experience) at the track as long as I could compare the data logs and weed out the variables.

We don't currently have data logging on the truck, but may in the furture. Some of the other trucks have it. But, I don't know what you could log that would tell you if you made more power after a change to the engine. The variables at the track are too many. It's different track and different sleds with different set ups in the sleds every event. The sled speeds and distances pulled change too much to know if there was a small change in power. All the front running engines have been on the dyno and most of them several times. It's not like a drag strip where you can judge a lot by et slip.

As for me, I'll stick to what I've seen on the dyno till I get the chance to dyno again, even if a sim disagrees with it.
Maybe if I had a high end 3d sim, but I don't.

Randy

What kind of dyno, chassis or engine?

[user-30563]

I sense a disturbance in the force concerning the direction threads have taken lately. Not pretty.

[randy331]

Randy,
Since pullers, mudders and such never get a chance to make test laps I have often thought one of those rear end mounted torque sensors might be a worthy investment. Real time Dyno Truck.

Dave,
Not sure you'd be able to see a change on a tq meter on the driveshafts.
For one, the total axle tq is over 14000 lb in high gear, more in low, so it's take a tq meter that's 10,000 lb capable and you'd need one on both driveshafts. Even then I'd say the loading and un loading it goes through during a pull would be several hundered pounds. The load would change from front to back during a pull etc. Don't think you'd ever be able to see a 10lb change to the engine that way.


Randy

[randy331]

What kind of dyno, chassis or engine?

Stuska engine dyno.

Randy

[David Redszus]

No matter what the dyno may read, accurate or not, the power produced while on the track cannot exceed the force
transmitted to the ground. Who cares how much power your engine makes. How much gets to the ground?

Without data logging you don't know nuttin.

[mekilljoydammit]

The turn this thread took is really weird. Why wouldn't you use every tool you have at your disposal?

So, I'm an engineer... I call engineering the science of laziness. All engineering is, is a bunch of tools to not have to build a real version of whatever to see if it will work. So everything, flowbench, cam checker, dyno, software simulation, whatever... it's all tools to try to do better without having to run down the track. Have a dyno and a 1d sim package? Use one to refine the other! A dyno is less of an abstraction, 1d sims can tell you more of what's going on and let you try more things more cheaply. Neither substitutes completely for the other without a pile of money and time.

[Frankshaft]

The turn this thread took is really weird. Why wouldn't you use every tool you have at your disposal?

So, I'm an engineer... I call engineering the science of laziness. All engineering is, is a bunch of tools to not have to build a real version of whatever to see if it will work. So everything, flowbench, cam checker, dyno, software simulation, whatever... it's all tools to try to do better without having to run down the track. Have a dyno and a 1d sim package? Use one to refine the other! A dyno is less of an abstraction, 1d sims can tell you more of what's going on and let you try more things more cheaply. Neither substitutes completely for the other without a pile of money and time.

Summed up well.

[GARY C]

After reading all of this and years worth of SIM talk it appears the only way to confirm the sim data to be correct is to test it so there is no way it would reduce testing time, it could help you decide what to test but that would only be assuming that all the others you didn't choose were worse than the one you chose but without testing them all it's only an assumption.

Many people will tell you about making their best power with their lowest flowing heads but that is only because they chose to test the head that was assumed to be worse.

Yes you can model 10 engines on a sim and choose one to test but that doesn't confirm that it was the best choice.

[mekilljoydammit]

If you get the sim to where it represents one configuration pretty well, you can try a bunch of variations and see if you can improve on it, even stuff that's counterintuitive and seems like it would be a waste of time, or weird stuff where you don't want to pay for one-off parts until you have a pretty good idea it's going to work.

Not everything is as well proven out as American V8s - not every engine are you going to be able to talk to someone who did it already and find something off the shelf that gets you the result that you want. It's nice to be able to narrow the list down from thousands of potential things to try to the best half dozen or so.

[GARY C]

If you get the sim to where it represents one configuration pretty well, you can try a bunch of variations and see if you can improve on it, even stuff that's counterintuitive and seems like it would be a waste of time, or weird stuff where you don't want to pay for one-off parts until you have a pretty good idea it's going to work.

Not everything is as well proven out as American V8s - not every engine are you going to be able to talk to someone who did it already and find something off the shelf that gets you the result that you want. It's nice to be able to narrow the list down from thousands of potential things to try to the best half dozen or so.

I understand that but you still have to choose what models your going to test and they have to be tested but it still doesn't confirm that any of those half dozen ideas are better than the other thousand unless you test them all against each other.

If you get wrapped up in believing the model knows best you may never test what is actually better because the model says it wasn't, you would need to test best and worst models and some in between to baseline your models and could actually increase the amount of testing needed.

Regardless of the name plate on the engine you are still dealing with the same dynamics.

[hoffman900]

If you get the sim to where it represents one configuration pretty well, you can try a bunch of variations and see if you can improve on it, even stuff that's counterintuitive and seems like it would be a waste of time, or weird stuff where you don't want to pay for one-off parts until you have a pretty good idea it's going to work.

Not everything is as well proven out as American V8s - not every engine are you going to be able to talk to someone who did it already and find something off the shelf that gets you the result that you want. It's nice to be able to narrow the list down from thousands of potential things to try to the best half dozen or so.

I understand that but you still have to choose what models your going to test and they have to be tested but it still doesn't confirm that any of those half dozen ideas are better than the other thousand unless you test them all against each other.

If you get wrapped up in believing the model knows best you may never test what is actually better because the model says it wasn't, you would need to test best and worst models and some in between to baseline your models and could actually increase the amount of testing needed.

Regardless of the name plate on the engine you are still dealing with the same dynamics.

Do you know you made the correct change on the dyno though?

If I change a camshaft and it does worst, was it the cam? Maybe I have to change three other parameters and then I'll realize a gain with that camshaft that I couldn't realize before.

Without giving away details, an engine builder I chat with told me he changed a camshaft and it did nothing but burn more fuel and made about the same power. He developed a whole new intake set-up, and the cam made a quite a bit more power than what the other two could provide (and he tested them all with both intake set-ups).

Kevin Cameron wrote a great editorial recently about the fallacy of testing and changing one thing at a time. When you have a validated simulation, you can change things at a much greater speed (and less expense) and get a snapshot into the engine that may make you realize what needed to be changed in addition to, to realize the full potential of that change.

Everyone needs to watch this: https://www.youtube.com/watch?v=rBZCnG1HwDM&t=1369s

If you've watched it already, then watch it a hundred times. Listen to what he's telling you about how to think about engine development. Go listen to what he says at 26:00. Simulators help with realizing the "modern way".

I've posted many links over and over again, showing simulation used to help realize development in Formula 1, FSAE, and Grand Am competition. I don't know if people read them or what, but everyone can search out my posts.