Intake Design Exploration

[vex]

Originally Posted by NoDOHC

CFD software is outside my experience, but if it is capable of simulating the flow in the manifold accurately (assuming that it is given good data) you have an excellent opportunity to maximize your learning as you go, while using the resources that you have to their fullest potential.

As you said, you do not have the typical toolbox.

Helmholtz tuning works for any level of boost (it is basically the natural frequency of resonance of the fluid system). The speed of sound varies with density of the charge, so you will have to that into account on the Helmholtz equation.

You can write the Helmholtz equation most easily in terms of the resonance frequency (First equation) Solved for L gives the second equation. This expects a uniform cross-sectional area for the runner, as any changes in velocity will create additional and possibly conflicting pressure waves.

(a = speed of sound, V = velocity in the runner at time of wave excitation, L = Runner length from source to plenum, A = cross sectional area of runner, f = frequency of resonance (which is related to engine speed, obviously).
I hope this helps some.

With the tools at your disposal, this should be one awesome manifold.

The offset sinusoid expresses the chamber volume as a function of E-shaft rotation (the period is 270 degrees, the amplitude is 20 in3 (327 cc) (654 cc peak to peak). Taking the derivative with respect to time requires that the x axis be in time units (pick an engine rpm). This will give you the rate of change of chamber volume with respect to time, which should give you a good velocity characteristic (given port cross-sectional area). Hopefully this will be good enough input data to get a reasonable approximation of how the manifold will flow.

Keep up the good work!

Thanks! Those look much, much, better than what I was looking at a few minutes ago!

[vex]

Okay, I'm working with the formula's you gave me and I've got it narrowed down a little bit. I have found at least two parts where the pressure will affect the runner length:
The amount of air displaced (when calculating the velocity in the runners), and the speed of sound propagation (sqrt(gamma*R*T)). Temperature and Gamma may change with an increase in pressure. I'll post up the full formula I have when I've got the little details worked out.

[vex]

so... my original understanding of the formula was flawed so I did a little research, properly solved for L and then plotted it as a function of RPM. Here's a picture of the plot I was able to derive.


I also had the program pump out the minimum length at an 8k redline. Here's what I had the program pump out:

>> runnerlength()
Assumed redline of 8000
Min Runner Length: 0.162229 m

Consequently I think it will require a little tweeking in order to be extremely accurate. But assuming that there's about 10 cm or so inside the engine itself it makes absolute sense that a 11 cm intake manifold at redline is so small... but who knows.

So what do you guys think?

(also this is for a runner diameter of about 4 cm)

[vex]

I'm surprised... No critiques or suggestions on improvement? Is my math right? Do the results look accurate?

What runner diameter should I consider accurate? I've been leaning towards 3 to 4 cm in diameter, but if I wish to tune for 3500-5000RPM in peak torque I may run into problems actually fitting the amount of pipe for the runners.

I'll be fooling around with the code (different temperatures, narrowing down the diameter to use, etc) and see if I can't get a finalized result for everyone's benefit. This should be quite helpful for individuals (and vendors) who wish to design their own intake systems.

Consequently here's my next round of questions with regard to runners:
What are the benefits/drawbacks of having multiple diameter runners--Having primaries one diameter and secondaries another? What about tapering? Venturi/Velocity Stacks and their affect on pressure drops, and other flow conditions?

This should prove fairly interesting and thanks everyone for their help and support in this thread thus far! Hopefully others are learning as much as I am!

[scotty305]

Nice Matlab code, I used Matlab in school but never got around to using it for fun car stuff like this.

I have one comment and one question:
1. I knew some people who built a variable-volume intake plenum for dyno testing... it was essentially a box with fixed runners and a sliding lid (the throttle attached to the lid), sealing it was difficult but slightly easier than fabricating multiple plenums.

2. Is there a way of predicting the relative power gain or loss due to resonance tuning? For instance, if you could somehow predict that there would be a X% gain in power at the resonant RPM but a Y% loss due to some sort of antiresonance elsewhere, you might try runner lengths/diameters that not only maximize gains in your desired powerband but also minimize losses.

[vex]

Originally Posted by scotty305

Nice Matlab code, I used Matlab in school but never got around to using it for fun car stuff like this.

I have one comment and one question:
1. I knew some people who built a variable-volume intake plenum for dyno testing... it was essentially a box with fixed runners and a sliding lid (the throttle attached to the lid), sealing it was difficult but slightly easier than fabricating multiple plenums.

I didn't even think about doing a variable volume plenum. Honestly the requirements of such things for sealing it would be a little too much for positive pressure tuning. Consequently I have not seen any mathematical model to which I could digitally model the results for it.

2. Is there a way of predicting the relative power gain or loss due to resonance tuning? For instance, if you could somehow predict that there would be a X% gain in power at the resonant RPM but a Y% loss due to some sort of antiresonance elsewhere, you might try runner lengths/diameters that not only maximize gains in your desired powerband but also minimize losses.

I doubt very much you could hope to tell how much torque you'd gain at a specific RPM. Based off the 26B the torque increase is a rather small percentage of incorrect length intake. There's also too many variables to stat that intake runners and a plenum volume will have this effect on a car: The difference between one car and another in power really comes down to the whole shebang. Intake, Combustion quality, Exhaust, rotating mass of the drive train, and even atmospheric conditions. You'd honestly have to hold 4 of those things constant just to test the effect of one--and even then I'm not sure that holding them constant would give realistic results.

[scotty305]

Additionally, you might want to think about how you're going to test your setup. It seems it would be wise to use realistic ramp rates on the dyno (if the dyno software allows for this).

I fully appreciate the benefits of step testing with sweep testing at the actual rate of acceleration of the vehicle. This can show completely different issues, and allow the separate tuning of load/rpm tables from acceleration enrichment. It can also help to separate engine loads from drivetrain rotating mass loads - both of which alter tuning requirements (both fuel/spark for engine management and pulse tuning in intake and exhaust systems).

An example of different results - Using step testing, I can develop a plenum where a specific volume shows a very nice improvement in output at higher rpm over a 1000 rpm window of the power band - without hurting low rpm output. However a sweep test will show a tendency to hurt lower rpm response and not influence upper rpm output nearly as much as the step test. On the track, this plenum causes the engine to be lazy with poor transient response. And a plenum 20%-30% smaller is actually considerably quicker and faster on the track - dyno results be d@mned.

The same things hold true in exhaust design.


All that matters is the track or street results - RIGHT?
Dyno results (HP/TQ numbers) are irrelevant other than for impressing uninformed customers, internet bragging rights, and the Engine Masters Challenge. They don't win races.

IMHO, both step and sweep testing need to be performed together. And all atmospheric conditions need to be taken, and corrected for, at the point of air entry into the vehicle.

this is about actual results using several types of testing methods on both engine and chassis dynos, and comparing the results to actual performance on road or track, not theory.

On a quick accelerating vehicle, intake and exhaust pulse tuning does not have the same time to stabilize as it would with a step test spending 6-8 seconds at each rpm point. The difference can be very large. This is a well understood issue. Perform your own testing and see the results for yourself.

There is a big difference between tuning being "fine" and being right. To engine builders, racers, and customers trying to get the best results for their hard earned Dollar, this matters.

info from http://www.efi101.com/forum/viewtopic.php?t=4105

[scotty305]

Sorry if that wasn't clear... the plenum volume wasn't adjustable on-the-fly, the 'lid' position was changed between dyno tests and this involved clamps, bolts and probably some sort of sealant. In hindsight I should have taken a closer look at the setup.

[NoDOHC]

Good call on the math error! I should have just given the original equation. Apparently the algebra is a bit rusty (I didn't think it looked right, but I couldn't find the formula solved for L in the book).
The natural frequency formula is correct though, I just checked it.

Anyway, I like where you are going with this. Something to remember on the Plenum is that the two rotors are 180 degrees out of phase and have a 270 degree intake duration. This means that both rotors take in air at the same time.

This is looking good!

There are more advanced formulas for varying cross-sectional-area runners, but I don't have the ambition to enter them into paintbrush and you can probably find them online anyway without any errors.

If you want to do the math yourself, it is very simple to draw it up as a dynamic system and then find characteristic equations for it. I say this because you are most likely taking a course in dynamic systems right now or in the near future.

The Plenum has compliance as it acts as an accumulator. The runner has a resistance (dissipative) and Reluctance (inertial) element to it. The port closing is the disturbance function. The system is a lot harder with changing area.

Anyway, you will find that the wave intensity is a logarithmic decaying function and that the time constant is related to the runner smoothness (resistance) and the air velocity (inertia). I can't find the equation right now, maybe it is in a book at work.

I will try to find it and let you do the Algebra.

[vex]

Originally Posted by NoDOHC

Good call on the math error! I should have just given the original equation. Apparently the algebra is a bit rusty (I didn't think it looked right, but I couldn't find the formula solved for L in the book).
The natural frequency formula is correct though, I just checked it.

It happens to the best of us. I can't remember the times I've taken a test and screwed up some simple algebra and ended up with a horrible, horrible, horrible wrong answer.

Anyway, I like where you are going with this. Something to remember on the Plenum is that the two rotors are 180 degrees out of phase and have a 270 degree intake duration. This means that both rotors take in air at the same time.

That's actually really helpful, but I think we need to adjust the angle of intake and have it based off of port size, and what not as you're going to get a larger intake with a bridge port compared to a street port

Well my class is finished, time to get back to what I was doing.

This is looking good!

I like it a lot and I figure this will be an easy tool for other rotary (and heck, even them piston) guys to effectively make and design a proper manifold without the downfall of one too short or too long. If anyone is able I'd like to see someone make this into a web app that RCC can host on their page (I'll make a thread about it in lounge for those interested).

There are more advanced formulas for varying cross-sectional-area runners, but I don't have the ambition to enter them into paintbrush and you can probably find them online anyway without any errors.

I don't even know where to begin with that. Could you point me in the right direction?

If you want to do the math yourself, it is very simple to draw it up as a dynamic system and then find characteristic equations for it. I say this because you are most likely taking a course in dynamic systems right now or in the near future.

Actually I don't think I have any dynamic systems setup for my classes now or in the near future. This is why I'm doing a lot of it on my own. I can probably do something similar to what you're telling me in the schools CFD program (if I can ever get it running)

The Plenum has compliance as it acts as an accumulator. The runner has a resistance (dissipative) and Reluctance (inertial) element to it. The port closing is the disturbance function. The system is a lot harder with changing area.

You're going to have to go into more detail if you're going to make me want to understand what you're saying. I think I understand, but I'm not too sure.

Anyway, you will find that the wave intensity is a logarithmic decaying function and that the time constant is related to the runner smoothness (resistance) and the air velocity (inertia). I can't find the equation right now, maybe it is in a book at work.

If you ever come across it in the near future post it up (don't forget references).

I will try to find it and let you do the Algebra.

DEAL!

[vex]

Crap, I think I made have figured out a mistake. Should the volume displaced be 1.3L or should it be half that?

[vex]

Okay, just double checked. The original formula I posted is correct.

We're not looking at the .0013/2 m^3 volume for displacement volume since intake is a continuous displacement of 1.3L (both rotors are 180* out of phase)

[vex]

http://www.rwdi.com/cms/publications/26/t09.pdf

^Some helpful rules of thumb for diffuser designs.

[NoDOHC]

I think I need to have a look in the attic, I can't find that book anywhere. From your other threads, it looks like you have done a lot more research than I can remember from old textbooks.

The intake is looking good!

I was thinking that the rotary intake strokes come twice as often as the piston engine intake strokes, meaning that the air had half the time to travel. The intake runner length equation may be off. I will try to get a chance to draw the dynamic model and solve the characteristic equations to see if I can derive the resonance frequency formula properly for a piston engine (it helps to stretch the algebra muscles every so often anyway, as they obviously atrophy). If I can, I will draw the model for the rotary and compare.

[vex]

Originally Posted by NoDOHC

I think I need to have a look in the attic, I can't find that book anywhere. From your other threads, it looks like you have done a lot more research than I can remember from old textbooks.

The intake is looking good!

I was thinking that the rotary intake strokes come twice as often as the piston engine intake strokes, meaning that the air had half the time to travel. The intake runner length equation may be off. I will try to get a chance to draw the dynamic model and solve the characteristic equations to see if I can derive the resonance frequency formula properly for a piston engine (it helps to stretch the algebra muscles every so often anyway, as they obviously atrophy). If I can, I will draw the model for the rotary and compare.

Please do. I have since had that class and now understand what you're talking about. I honestly can't model a dynamic system like this without just retarding it down to a simple mass-spring-damper problem. I do believe the equation I have is correct, though I am of the same concern as you.

The way I looked at it is degrees of rotation compared one to the other. For 720 degrees of rotation of the crank you get 1 intake stroke (Just looking at a single piston), similarly for the rotary you get 1 intake stroke for 720 degrees of rotation (looking only at a single rotor).

I had a simple program at one point that showed the correlation between the intake stroke of the rotary and piston with respect to engine RPM. I'll see if I can dig it up and validate my mental experiment.