ptuomov

It was just bull**** gibberish, so don't feel like a dinosaur. Or should it be "dynosaur" in this conext?

To the extent I had any point, it was the following. I think it would be cheaper and more effective (=more efficient?) to use some computer software tools to work out some kinks. I am thinking of the following algorithm:

1. Think
2. Simulate
3. Flow
4. Dyno
5. Go to 1 to rethink

It seems that step 2 could make going where no man has gone before a bit easier.

ptuomov

Let me make a prediction. If you commit to producing a model, the forum will come up with all the parameter values you need. Why don't you start by listing all the parameters that you don't yet have values for and see what you get.

I would suggest you make model for completely stock '87 s4 with manual transmission. You can later complicate it with an automatic. Once the base model is ready and matches the dyno results, then let's go to town modifying the beast!

People hanging out here are either insomniacs or lack social life (or come to think, probably both) so you'll get a lot of help as long as you are very specific with your requests.

Out of curiosity, which Ricardo software components you have? Engdyn, Pisdyn, Wave, etc. Or do you have everything from them?

ptuomov

jleidel

...assuming two things:
1.) the simulation has been verified mathematically correct with corresponding empirical testing/evaluation
2.) the time it takes one to run and evaluate and sufficiently large enough simulation that reasonable results can be derived is less than pure empirical testing... [dT_sim < dT_real]

I took a look at the Ricardo software namasgt outlined. Looks pretty interesting. I personally don't have any experience with it but I do with other flow/thermo solvers [fluent/cfx/mfix/openfoam/etc].

..my two thoughts

ptuomov

1) Simulations are never correct. If a simulation would be completely realistic, it wouldnät be a simulation, it would be the reality!
2) Applies to all steps, you shouldn't overthink either! ,-)

But you knew all this already. I am just bull****ting about stuff in a public forum.

In any case, since you have experience with simulations, which I don't, I could use a piece of advice. Should I try to design a sheetmetal intake manifold, which software would a) give reasonably accurate predictions, b) could be learned by a dumb-as-rocks business undergraduate like myself, and c) could be purchased without breaking the bank (a very small bank, in fact)?


Best,
ptuomov

namasgt

Im gona start making the model this weekend.

Hilton

I'm sure there are a bunch of interested people.. including me, although my interest is theoretical at this stage as I'm not quite ready to start modifying my car (or close to choose NA vs. boost etc).

There's been lots of posts about simulations using Engine Analyzer Pro, and other 3d fluid dynamics modelling lately - mostly around work on identifying intake parameters like optimal runner length etc.

You can also find the stock S4 intake dimensions in one of the recent threads - I'll see if I can dig it up (different runner lengths etc. plus the opening of the resonance flap).

Edit: Stock Intake thread - check post 25 for more dimensions