All that Helmholtz business

[Jersey Tom]

So a lot of people kick around this Helmholtz resonator talk with respect to intake design, limitations, talk about its accuracy becoming questionable above 4000rpm, yatta yatta yatta. Though interestingly enough according to Cornell's 1990 paper it proved successful enough to generate +34% torque over the stock unrestricted engine (albeit at the expense of a lot of high end).

But what exactly is happening when you reach those resonant frequencies that magically makes your engine perform better? Is this something simple I'm missing, or are we just plugging some crap into a spreadsheet and getting some engine RPMs out of it, and calling it good?

The concept of ramcharging and exhaust scavenging makes sense. Design your primary lengths such that your pressure wave returns to your valve just before it closes, either sucking out more exhaust or dumping in more intake charge. But that's a function of the thermodynamic properties of your working fluid, your manifold geometry, and the cam profile of the engine.

So what the hell.

[Wright D]

So, what exactly is your question?

Or are you just ranting? http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Jersey Tom]

Ha. What exactly is the mechanism by which the Helmholtz resonant frequencies is increasing engine performance. You have some peak RPM points.. but when you reach them, what is happening and why? It would seem the valve timing would come into play and be fairly important.

[BeaverGuy]

If this is an actual question and not just, a what the hell rant. Then the idea is the same for the Hemholtz tuning and ramcharing methodology, and it is really just a more complicated version involving more of the manifold. The way they take cam timing out of almost all these models is include some arbitrary constant that is supposed to match the helmholtz frequency with the induction frequency. Even some of the simple ramcharging equations ignore cam timing and use a different arbitrary constant that is likely to change with the engine.

Now, if you want to find a model that will incorporate a lot more of the geometry go look at the paper that Sam Zimmerman posted on this forum or get a copy of Design and Simulation of 4-stroke engines by Gordon Blair. But if you are going to deal with Blair's model you might as well just hand write the engine simulation program because that is really what his model is about

[Chuck Dean]

Here is a link to one of Zimmerman's posts. His paper on using an impedance transfer model is available here.

http://fsae.com/eve/forums/a/tpc/f/125607348/m/5531...10270621#52110270621 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/55310666121/r/52110270621#52110270621)

I encourage anybody interested in intake design and acoustic theory as it relates to engine performance to check out this paper.

From my experience using the model, it has been very accurate and much more useful than your standard Helmholtz model.

[Cody the Genius]

I think they were actually asking about the physical mechanism itself. The best way to explain it is with an open jar:

If you get a fast moving stream of air going into the jar you can get a 5 liter jar to hold 6 liters. But then when it tries to go back to equilibrium it spits out 1.5 liters instead 1 L due to inertial effects so the jar now holds 4.5 L. It tries to reach equilibrium again but sucks in 0.75L instead of .5L and the vibration contiues on lessening more and more each time.

So you can use this phenomenon to cram more air into the combustion chamber and get more exhaust out of it if you can make everything vibrate at the right frequency in the right phase. Helmholts resonance in air is what makes a coke bottle sing and speaker boxes work better.

Forgive me if this seems remarkably basic but I thought it was what you wanted.

[Homemade WRX]

Originally posted by Cody the Genius:
I think they were actually asking about the physical mechanism itself. The best way to explain it is with an open jar:

If you get a fast moving stream of air going into the jar you can get a 5 liter jar to hold 6 liters. But then when it tries to go back to equilibrium it spits out 1.5 liters instead 1 L due to inertial effects so the jar now holds 4.5 L. It tries to reach equilibrium again but sucks in 0.75L instead of .5L and the vibration contiues on lessening more and more each time.

that is one of the most interesting and best "visual" explanation that I have heard...

[Sam Zimmerman]

Here is my little piece on acoustic design of intakes. Feel free to disagree. I would especially like to hear from Charlie.

Helmholtz resonator models simplify acoustic analysis. It is not a different phenomenon, as some might suggest, but a different set of assumptions. The goal is to drop the impedance at a certain RPM(s) so you can manipulate the volumetric efficiency curve to your benefit. Whether you use an organ pipe method, Helmholtz method, impedance transfer method, finite element method, etc. you will come up with about the same primary resonant frequency. The Helmholtz method is a more refined method because Engelmann studied it extensively where as the "ram charging" method was not.

One of my main points in my research was that the organ pipe and Helmholtz methods are antiquated, especially in FSAE applications. They were both developed for 4000 RPM engines and are both horrible at determining secondary resonances. They both also ignore anti-resonances. If you want to really understand what your intake is doing across your 10K RPM band you need FEA, impedance transform, dual port networks or another more exhaustive method of analysis.

btw, I think the 1990 Cornell paper has some incorrect assumptions, even if the results are positive. Even Winterbone has some writings where he confuses (at least in his words if not in his mind) the difference between fluid flow and acoustics. Intake design just hasn't been discussed and published the way chassis and suspension design has and is often looked at as though it is black magic.

Charlie, where am I wrong? You are the professional; I have now turned to the dark side (thermo and fluids) and haven't looked at this silly stuff in a while.

[Jersey Tom]

It seems to me then the most practical thing to do would be to just prototype some manifolds out of simple materials and with a speaker and mic, get a plot of the actual resonant and anti-resonant frequencies of the thing. I believe thats what Sam did in his paper for validation, no?

I'd think you could do more or less the same thing with exhaust manifolds, and just take into consideration the difference in speed of sound of air at room temperature and exhaust at 1200F!

[Sam Zimmerman]

I agree with the prototyping approach.

It is notable that you bring up the exhaust side. If your exhaust is crap, you will see little gains from your intake. I mentored a lab experiment regarding 2 sets of headers a few years ago. One set worked with the intake, the other set didn't. The difference between the two was astonishing. If your intake and exhaust are working against each other (one in resonance while the other in anti-resoncance), it doesn't really matter how good each are individually. They will both look like poo on the dyno.

[fart can]

Hey guys- I'm totally new to this. Aside from the already mentioned link and book, are there any other good papers or out dated books regarding 'all that hemholtz business'? Any nudge in the right direction would be appreciated.

Sam- with the whole exhaust/intake working against each other concept, are there any physical signs that this may be occuring (other than poo on the dyno), like improper tube lengths, poor welds, cross over pipes on the headers? Or do you have to determine the resonance frequencies experimentally?

[Sam Zimmerman]

fart can,

Look at the paper linked above, skip all the other sections, go to the reference section, then order all the references via inter-library loan. Then read and re-read #'s 1-5. Once that is done, read the appropriate portions of 6-10. If that isn't enough, start getting the sources cited in those papers.

[fart can]

Thanks Sam-

[Sam Zimmerman]

I have been asked about some of the values used in the paper that Matt linked above, so I offer you the following:

c=343 m/s
rho=1.21 kg/m^3
k=w/c + j*alpha where alpha is 0.01 (alpha is an acoustic absorption
factor, 0.01 is an approximation)

the previous won't affect your resonance frequencies much, unless you
start using off the wall values.

w = 2 * pi * frequency
There is some disagreement in the literature with what number to use here, though.
Basically if your wave is at your valve at opening, you want your wave
to travel to your piston and back to your valve right at closing.
Therefore, the frequency that Engelman used was twice the engine
speed. This is something that isn't agreed upon nor is it well
explained in any of the research I have done, but I believe it is what
I used in the model that is shown on the graph.

There is one other peculiar convention, and that is the stroke length.
Engelman uses 1/2 the stroke for his calculations. I have all
Engelman's papers (including his ph. d. paper) and Eberhard's master's
thesis and none of them explain the reason for this.

If intake design is your thing, please feel free to object to any of the above (as though many of you need an invitation http://fsae.com/groupee_common/emoticons/icon_wink.gif ). Intake and exhaust has never really been discussed and bantered about the way other systems have and it would be great to see where people differ in their thoughts on this.

[Jersey Tom]

I'd love to if I had more of a clue what I was talking about.

It would be so nice if our school had some sort of automotive engineering cirriculum, where even introductory-level stuff was taught. As opposed to having to try to learn all of this on the side.

[Garlic]

This is pretty specialized stuff, not what you'd find in any automotive engineering curriculum I would suspect.

The thing about automotive engineering is, it's just application of standard engineering ideas. It's not an entirely different subject.

If you want to learn about this stuff, I'm sure your school has an Acoustics class.

[iolair]

Would it not be interesting if Heimholtz was simply a BS alternative explanation as it relates to the VE of an operating cylinder that simply mimics/dovetails symbolic vs phyical properties of what is actually taking place during an real time physical induction event......you know, like a built in mental block? It's not like pre-conceived ideas and pratices have not been circumvented in the past......hmmmm, maybe there is another explanation as to how a N/A cylinder is over pressured!

John

[BeaverGuy]

Originally posted by Garlic:
This is pretty specialized stuff, not what you'd find in any automotive engineering curriculum I would suspect.

The thing about automotive engineering is, it's just application of standard engineering ideas. It's not an entirely different subject.

If you want to learn about this stuff, I'm sure your school has an Acoustics class.

Probably not most automotive curriculum but I imagine Queens's University of Belfast includes it. That is where Gordon P. Blair was a proffesor and he used his research and teaching material to write the Design and Simulation of 2-stroke and 4-stroke books and the simulation program that became Virtual Engines. I would say that it should probably be covered in all of the graduate level automotive curriculum that focus on powertrain design. As far as at the undergraduate level, the unsteady gas dynamics that underly Blair's work and Sam's paper is beyond what 90% of undergraduates are prepared or willing to understand, even those interested in automotive engineering.

[iolair]

Let's see......timing a reflective wave to pack a cylinder!

Why not simply determine the swept volume of an operating cylinder and construct a port volume that will generate sufficient kinectic energy at a *specific* RPM to ensure that a vertical or horizontal vortices is created within the cylinder that will sustain itself as the piston rounds BDC. The rotating vortices would generate local lower pressure regions in the operating cylinder even as the piston ascends in the cylinder....drawing in addition air molecules until the intake valve closes! A cyclone in a bottle....it's fairly easy to understand....Ricardo understood it in 1919 and a reflective wave need not apply!

John

[Bam Bam]

Seriously dude,
1.25" 16 guage tubing fits nicely into 1 3/8" 18 gauge tubing
Bam adjustable exhaust!

Home depot has some cheap vacuum rated hose.
Bam adjustable intake!

Give yourself 2 hours on the on the dyno and
Bam wicked torque curve!

How much time are you gunna waste devising, debating and calculating models that are known to be inaccurate?

Let's face it you gotta build a car and carry a full course load in one year, effective time management will determine your engineering success.

[Jersey Tom]

Originally posted by iolair:
Let's see......timing a reflective wave to pack a cylinder!

Why not simply determine the swept volume of an operating cylinder and construct a port volume that will generate sufficient kinectic energy at a *specific* RPM to ensure that a vertical or horizontal vortices is created within the cylinder that will sustain itself as the piston rounds BDC. The rotating vortices would generate local lower pressure regions in the operating cylinder even as the piston ascends in the cylinder....drawing in addition air molecules until the intake valve closes! A cyclone in a bottle....it's fairly easy to understand....Ricardo understood it in 1919 and a reflective wave need not apply!

John

Is there time for a vortex to really develop and do much in a 5-20ms time frame?

[Sam Zimmerman]

Seriously dude,
1.25" 16 gauge tubing fits nicely into 1 3/8" 18 gauge tubing
Bam reflected wave at an unintended joint!

Home depot has some cheap vacuum rated hose.
Bam reflection at head!

Give yourself 2 hours on the on the dyno and
Bam not enough time to decide whether your changes are intake/exhaust based or fuel/spark map based!

How much time are you gunna waste changing, experimenting, and farting around on a dyno without understanding the effects, sensitivities, or potential of the changes you've made?

Let's face it you gotta build a car and carry a full course load in one year, effective time management will determine your engineering success. Trying to farmer systems that you don't want to take the time to understand will lead to wasted time, a lack of learning, and a habit of spending all your time building and testing new prototypes rather than using engineering principles to minimize the number of prototypes. Shop time, lab time, and your time do not come cheap.

These ideas are great prototyping tools; they should not be the primary design tools.

[Jersey Tom]

It would be nice to be able to model a manifold (or a 1-d approximation or what have you) and get a continuous result for resonance amplitude versus frequency.

There's an SAE paper out there by FIAT on the design of one of their ITC engines.. all about intake and exhaust pipe lengths, diameters, etc.

[VFR750R]

Diesels have intake ports that are designed to develope large amounts of swirl in the chamber. This is required to get enough turbulence at TDC to constantly mix new air with the fuel being squirted in. Just like downforce on a wing, it comes with a penalty...drag. So to produce large amounts of swirl, there is large flow losses associated with it. Plus what your thinking would work ok on a chamber where the intake port is centered in the chamber and is small compared to the bore, but with two intake valves off center, as much of the valve sees the higher pressure air at the edges of the cylinder as the low pressure in the center. And even if it did work, it won't have nearly the effect of the momentum of a tube with air at half the speed of sound. There is of course flow losses associated with speeding air up to that speed, but it has a more direct influence on filling the cylinder with additional air.

[Sam Zimmerman]

VFR,

As a student, it was (and still is) my position that to get the most bang for the buck, an intake should be designed with acoustics in mind first. Get the lengths and areas correct, then go after improving the fluid dynamics at the transitional regions. Improving the fluid dynamics at the transitional joints will lower the reflection coefficient but improve the volumetric efficiency over the entire RPM range.

Would you agree with this approach? What are the differences in the approach between FI and NA intakes?

[VFR750R]

Sam, I would agree with that. In an NA engine, getting the right size and length intake and exhaust runners should take priority. In a turbo application, these things are less important but not unimportant. Other factors such as turbo response, heat retention and powerband width have an effect. You'll find that a NA car must run higher average rpm to keep close to avg hp that a turbo car has, this obvioulsy has an effect on runner length. You'll end up with different goals.

One intake idea never discussed here is tapered runners. You see tapered runners in all forms of racing, which will reduce the flow losses measurably. Something that many engineers step over is the fact that the air in a runner isn't steady state, but starts from zero speed every cycle, so anything you can do to assist the acceleration of the stagnant air in the plenum up to max speed in the throat of the port can reduce those losses. You can look up the optimum angle, which I think is close to 1.5 degrees. Being careful not to decrease it so fast to diffuse the waves in the runners. It is obviously not as easy to calculate optimum runner length, but it would be worth it to do some physical testing.

[Damo]

Just wondering if any teams have experimented with / or use dual intake plenums. The basic theory is isolating pairs of runners to eliminate the effects of overlapping intake events. I haven't seen any information about them on here before, although I have read a couple of technical papers from Lawrence Technological University.

I Hope no one minds me bringing up a slightly off-topic question but i thought it was a good opportunity while all the intake & exhaust guys are in one place.

cheers
Damo

[Nathan C]

If I understand what you're talking about, I think Winterbone goes into that a little as well. It hasn't been tried to my knowledge yet though.

[Cody the Genius]

One intake idea never discussed here is tapered runners.

It may have never been discussed but we have run tapered runners since I joined the team in 2002. The tapered runners do just as you say and we also used the increased turbulence through accelerating the air down the runner to help atomize the fuel. This was done in part because we were injecting at the top of the runner. IMO every intake should have tapered runners;I can not think of a reason not to. I also think that the "optimal 1.5 degrees" should be more like ~1.5 degrees as there are no real "magic numbers" to which every engine works the best.

[iolair]

Diesels have intake ports that are designed to develope large amounts of swirl in the chamber. This is required to get enough turbulence at TDC to constantly mix new air with the fuel being squirted in. Just like downforce on a wing, it comes with a penalty...drag. So to produce large amounts of swirl, there is large flow losses associated with it. Plus what your thinking would work ok on a chamber where the intake port is centered in the chamber and is small compared to the bore, but with two intake valves off center, as much of the valve sees the higher pressure air at the edges of the cylinder as the low pressure in the center. And even if it did work, it won't have nearly the effect of the momentum of a tube with air at half the speed of sound. There is of course flow losses associated with speeding air up to that speed, but it has a more direct influence on filling the cylinder with additional air.

'engine and turbo guy'
Cornell 02-03


If a motor has a poppet valve in conjuction with a cylinder that is decending, there is a 100% chance that some kind of vortices is going to be generated. You simply cannot generalize what vortices is going to be generated without knowing the intent of the application. Since the vortices is generated within the operating cylinder, there is zero flow loss as it relates to port flow.

Since you are mr turbo guy....why is there no mention of a variable geometry induction system to spool up the turbine.....you know, the motor is normally asprirated before the transistion?

John

[Sam Zimmerman]

By the same token, the induction system cannot exist without acoustical effects. It is just a matter of whether or not one chooses to use this phenomenon to manipulate the volumetric efficiency of the engine. I only have one of Ricardo's papers, but his obviously empirical equations are very similar to Platner's organ pipe equations. Are you certain he was observing the effects of vortices or acoustics? Also, his runner length equations in his aircraft engine patent application, if applied to an FSAE engine, would require runners to be several feet long.

I would love to read more of this. Do you have any specific Ricardo papers you could direct me to?

[iolair]

By the same token, the induction system cannot exist without acoustical effects. It is just a matter of whether or not one chooses to use this phenomenon to manipulate the volumetric efficiency of the engine.

Sam....I think you are a good guy.....I don't think any of the acoustical stuff is valid!

John

[raska]

Why iolair?

[jaimin]

Sam....I think you are a good guy.....I don't think any of the acoustical stuff is valid!



John

You do know that Sam wrote a SAE paper on this stuff right? Us and a few other teams have intake designs that are based on and reference his work. considering that our team capt, and 2 others were/are band geeks, all his acoustical work makes perfect sense to them!

[VFR750R]

If a motor has a poppet valve in conjuction with a cylinder that is decending, there is a 100% chance that some kind of vortices is going to be generated. You simply cannot generalize what vortices is going to be generated without knowing the intent of the application. Since the vortices is generated within the operating cylinder, there is zero flow loss as it relates to port flow.

Since you are mr turbo guy....why is there no mention of a variable geometry induction system to spool up the turbine.....you know, the motor is normally asprirated before the transistion?

John[/QUOTE]

You don't think the in cylinder votice can have upstream effects? All I'm saying is every book I've read on the subject did not cover what you're talking about, so either you're smarter then the rest of us or there is a flaw in your theory. But, we don't have to agree with you, you can do the calculations you talk about, design a port and intake and prove us wrong. Then you can sell the patent to a major engine manufacturer and give us the finger while you drive by in your Bently.


As far as runner length spooling the turbo, stop by the Cornell tent this year and take a look at the car. And no I didn't realize the engine was NA before the turbo spooled, I guess I never thought of it like that http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

[Sam Zimmerman]

Sam....I think you are a good guy.....I don't think any of the acoustical stuff is valid!

John

That's cool. Perhaps it isn't. Although I like to research this topic, I would never consider myself to be in the league of professional Indy car/cup/F1 power train engineers. Please, though, explain to me why an engineer can calculate (using acoustics) where the resonant frequencies in an intake are going to be, build an intake and measure where the intake resonates, and then run the engine on the dyno with an anemometer measuring air flow and have the volumetric efficiency peaks match up w/in 10% of the calculated value and the value measured with a microphone?

Even if the timing of acoustical waves is invalid, wouldn't having a lower impedance to flow at some frequencies increase volumetric efficiencies?

Why can I model the intake of the diesel engine using Platner's, Engelman's, my method, or a 2 port network and calculate his volumetric efficiency peaks pretty close?

As I said, I am not a pro at this. If the stuff is invalid, though, please point us to the research. You obviously know your research so please share so we can all be enlightened.

And just for the record, I'm not that good of a guy. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Ben Inkster]

Sam....I think you are a good guy.....I don't think any of the acoustical stuff is valid!

John

John,
do you disagree with the mathematics of the impedance transfer models that sam developed in his paper or the idea of pressure oscillation in general?!

I'm almost a little shocked that someone could simply write off a well established theory of oscillating gas columns in the intake and exhaust that has been studied and demonstrated by so many! and so commonly experienced by Fsae'ers!

Is iolair a new psuedonym for John Bucknell? I'm not trying to expose you because i disagree, but I'm just interested to know if this is the opinion of a design judge?

Cheers
-Ben

[iolair]

Yep, IMO....therein lies the problem of how the relatively simplistic/chameleon world of symbolism can be magically molded or adapted to fit the complex physical/real world and simply mask and/or dance around a true dynamic solution. A set of strictly adhered to equations only tend to provide a comfortable/convenient means of providing a built in mental block that severely limits the range of options and operational data points.

The bottom line/goal for rather obvious reasons, is to utilize a very short/large intake port that does not circumvent low speed efficiency..... especially useful as it is applied to a turbocharged motor.....a variable geometry induction system. Since low speed efficiency is not compromised, increased low speed mass flow allows the use of a larger turbine A/R, increased cam timing and the large/short ports allow the compressor discharge pressure to match more closely to cylinder pressure as the intake valve closes. This is all good! BTW, telescoping intake trumpets is not exactly my idea of a high tech solution.

So anyway....kitty is hungry and lives in an operating cylinder. Kitty looks ideally like a twin-counter rotating vortex and must be fed over a broad operating range. You have to figure out how to generate X amount of kinetic energy within the intake port that is required to energize kitty and fill or over fill the cylinder at multiple data points. Please do not limit your thinking to pre-conceived concepts that are narrow in scope and application. Get a reprint of Sir Harry's book and read it.....pay attention to kitty!

If you do not keep an open mind you will simply end up like the professor that confidently predicted the impossibility of exceeding 150 mph in the quarter mile.....jeeze, how embarrassing!

BTW, here are a couple of examples that were made 20 years ago to validate low speed N/A vortices manipulation and were eventually used in turbo motors.

BTW2, I'm not really interested in debating this crap.....it really is old news. I'm retired now and my only job is to finish my '56 Chevy post with C4 chassis, twin-turbo 427 SBC and drink Budweiser before I croak......do/think what you want!

http://www.pbase.com/image/58684495.jpg

http://www.pbase.com/image/58684496.jpg

http://www.pbase.com/image/58684497.jpg

http://www.pbase.com/image/58684499.jpg

http://www.pbase.com/image/58551256.jpg

[Sam Zimmerman]

Originally posted by iolair:
If you do not keep an open mind you will simply end up like the professor that confidently predicted the impossibility of exceeding 150 mph in the quarter mile.....jeeze, how embarrassing!

Let me make sure I am reading this right. The ones who are asking questions and trying to expand their knowledge are closed minded. The one who dismisses theory without presenting data, references, or any other proof of any kind is open minded.



BTW2, I'm not really interested in debating this crap.....it really is old news.

Of course you aren't interested in debating. You dismiss others without cause and then when they ask you to point them to the research papers that will help them learn more you ignore the requests. Nobody has even attempted to debate you, they are simply asking for your analysis/data so they can learn from it. Believe it or not, many people who frequent these forums enjoy learning.

Nothing is "old news". If you believe that anything involving engine performance is "old news" and doesn't justify further scrutiny and development then I guess we can just call intake analysis finished work and quit studying it. There are obviously no more gains to be made here. hmmm, what was that you said about someone predicting the impossibility of 150 mph in the quarter.

BTW...since you don't wish to learn anything new like the rest of us, why do you come here?

BTW2...Nice car.

[Ben Inkster]

I must appologise to John Bucknell for associating you with iolair (aka John), sorry!

iolair,
leave your cryptic rubish to somewhere else and put some substance in posts. I'm interested in what you are trying to say but you don't seem to respect anyone else here?

Cheers
-Ben

[iolair]

Sorry for the misunderstanding.....the crap I was referring to is mine, not yours and in fact it is old news to me, since

it was done 20 to 25 years ago!

quote:

>>If you do not keep an open mind you will simply end up like the professor that confidently predicted the impossibility

of exceeding 150 mph in the quarter mile.....jeeze, how embarrassing!<<

Again, I was referring to the understanding/acceptance of my stuff, not yours!

I read your paper, and it's good stuff! http://www.pbase.com/image/57341663.jpg
http://www.pbase.com/image/57341664.jpg http://www.pbase.com/image/57341665.jpg


Outside in as opposed to inside out! If you bottom line it, induction science is

based on intelligent restrictions per rpm data points.....if not, a 55 gallon drum would be an appropriate port size for

all induction systems

>>BTW2...Nice car.<<

Thanks Sam, I appreaciate that! Here is a photo of the donor car!

http://www.pbase.com/image/60787188.jpg

John

[Nishant Jain]

I'm looking to implement an Acoustic Impedance Transform test into our intake design system, having read Sam's paper.

But I'm quite weak in my acoustics knowledge, and am confused about how to proceed with the testing.

Do I simply measure p/p ratio or voltage output against different frequencies to find resonance? I dont know how to go about calculating the effective acoustic impedance of the system, so I cant make the impedance vs rpm graphs that are shown in the paper?

can someone point me in the correct direction? i know it must sound pretty basic after the extensive conversations i've read up on the forum about this topic, but i think everyone happened to skip over this one bit!

[Jersey Tom]

Wow. Takin' it back to the old school with this one.

It has been a while since I looked over Sam's paper, and I don't have it in front of me, but I could have sworn it went through and showed how to calculate the impedance of the system vs RPM?

[Nishant Jain]

Yeah, it has equations for closed and open ended pipes. We're supposed to construct the full intake system from that.

The problem is, I do not know what kind of values to put in some of the variables. For example, length (L) would be in feet, inches, mm? Also, some of the constants I do not know the value to.

Again then, reading the posts here, nobody seems to ask about such things. So I can only presume that they know already! Or, as I was wondering, maybe it isn't needed at all. As I see it, measuring only for the resonant frequency is not a very accurate determination, as compared to having impedance values.

[Sam Zimmerman]

Originally posted by Nishant Jain:
I'm looking to implement an Acoustic Impedance Transform test into our intake design system, having read Sam's paper.

But I'm quite weak in my acoustics knowledge, and am confused about how to proceed with the testing.

Do I simply measure p/p ratio or voltage output against different frequencies to find resonance? I dont know how to go about calculating the effective acoustic impedance of the system, so I cant make the impedance vs rpm graphs that are shown in the paper?

can someone point me in the correct direction? i know it must sound pretty basic after the extensive conversations i've read up on the forum about this topic, but i think everyone happened to skip over this one bit!

In the paper I described the DVT method. I haven't worked in this field for about three years, but here it is from memory. You need a prototype intake, a large loudspeaker, and an acoustic microphone. You do the test twice, once with the microphone in front of the loudspeaker and once with the microphone in the same location but with an intake positioned so that the microphone will measure the intake resonate frequencies.

Record sound power vs. frequency for both cases. Plot the ratio of the two sound powers (with and without the intake) as a function of frequency and you will get your resonate frequencies.

If you design a plenum with plungers in the ends and runners that allow for their length to be changed easily you can work through several geometries rather quickly.

I hope this is vauge enough that you have to do your own planning and thinking but detailed enough for you to figure it out. If you are still struggling, take it to your acoustics professor and have him/her help you out. If that doesn't work, have a beer and work on something else for a while.

Good luck.

[Nishant Jain]

@Sam: Thanks a lot, Sam. I was thinking along similar lines, and it is a great help to know how to go about it. I'm figuring out the impedances now.

I'll buzz you again if there are hitches.

Thanks again.

Nishant Jain.

[Nishant Jain]

So I set up the test. I'm using a freeware off the internet to record sound-waves from my microphone. And it gives me the graphical amplitude for one frequency at a time, so I have to run it over and over for the different frequencies.

The problem: the amplitudes don't seem to vary at all! Over a large frequency stretch, they only get higher and higher to peak into a square shape and then don't change.

I cant locate any resonance OR anti-resonance!

Is this a microphone problem maybe? Or would it just be better to use a multimeter/CRO tapped on the microphone instead of a freeware? Or is it something else altogether?

http://i205.photobucket.com/albums/bb230/nishantjn/225.jpg

[Drew Price]

It looks like the signal input to your computer is maxing out. Maybe try some kind of attenuator on the end of your microphone (foam, like a wind-shield for the mics, or some cloth or something) or try moving it farther away? Or turn the gain or input levels down, or try a different mic and/or soundcard.



Best,
Drew

[Jevon]

Drew's got it. It looks like you are saturating your input. You said the amplitude gets higher and higher and then peaks. Well, It's still going higher you just aren't measuring it.

You can easily adjust this in software. I've never used the program you are using but it looks like you will need to turn down the volume slider in your "Realtime Analyzer RAL" window. Also make sure you properly adjust your "Line-In" volume for your sound card in the control panel. Conversely, you could turn down your output levels.

Adjust the levels so that the maximum amplitude you intend to record does not get cut off.

Regards,

Jevon

[Mikey Antonakakis]

Man, this was a good thread, worth the three year bump!

To anyone else who might read this thread, I highly recommend Ricardo WAVE, as long as you give yourself a few months to learn the software. Once you get the hang of it, you can start from scratch with intake and exhaust models and have torque curves and just about any other data you'd ever want in a matter of hours. I have yet to prove the simulations for myself, but once we get our dyno running, I'll be sure to start a thread on here (although I'll really only have one set of intake and exhaust geometries to work with).

[Nishant Jain]

Thanks guys. I'll be sure to try that out today.

But a major concern is why the amplitude just continously rises! It's supposed to show anti-resonance at one frequency or another. So there should also be a significant dip at one frequency and a significant jump at another frequency (due to resonance). All I'm getting is a continuous steady rise.

Why would amplitude continue to rise, if all I'm changing is frequency? I'm not an electrical student, so this is just a shot in the dark - could it be a low sampling rate causing constructive interference beyond a certain frequency?

Or alternatively, could this mean that the effects of resonance are not being registered at all? I don't think that would be right, unless I have significant noise that the system is picking up. FYI I use an ordinary microphone (that comes attached with the big headphones) and one left channel speaker out of a 2.1 system. Any changes needed here?

@mikey: I've been doing WAVE myself since January. After these many months, sometimes I still think its only playing games with me! But its great to get the torque curves and other valuable data.

Since we haven't dyno tested yet (although we're going to in May), can you tell me what kind of friction parameters you consider in your engine block? And does everyone use the same duct temperature and pressure values as shown in the tutorial? Because I don't know any other values to go by.

Also, I'm planning to hook my WAVE model up with Simulink for ECU Control Design. I feel the urgent need to have variable A/F Ratios and a changing throttle angle. Any idea how much time that would take? Or is there another way around to manage these and get a more 'dynamic' engine model?

[vengeance]

hey which is this software and form where did u downloaded it ? are u using this software for tuning exhaust?

[exFSAE]

Originally posted by vengeance:
hey which is this software and form where did u downloaded it ? are u using this software for tuning exhaust?

You gotta be kiddin me, dude. Look at the screenshot. The window says what software it is. Google it. The FIRST link even...

[Jon @ Electromotive, Inc.]

I'm not really sure how low a frequency you're starting with, but you might consider the frequency response for the speaker (and michrophone for that matter) at the frequencies you're interested in. Cheap comptuer speakers tend to have somewhat poor performance at lower frequencies (<200Hz) that ramps up somewhat quickly. You might need to find a speaker better suited to the lower frequncies.

[Kirk Feldkamp]

Rather than doing individual frequencies... wouldn't it make more sense to use white noise as your input to your speaker, and then do an FFT on the microphone output? That's going to show you your resonant frequencies a LOT faster than doing individual passes, right?

-Kirk

[Grant Mahler]

Originally posted by Kirk Feldkamp:
Rather than doing individual frequencies... wouldn't it make more sense to use white noise as your input to your speaker, and then do an FFT on the microphone output? That's going to show you your resonant frequencies a LOT faster than doing individual passes, right?

-Kirk

No one likes FFTs...

[Wesley]

Yeah, but with data and matlab, it should be a breeze.

[Zac]

Originally posted by Kirk Feldkamp:
Rather than doing individual frequencies... wouldn't it make more sense to use white noise as your input to your speaker, and then do an FFT on the microphone output? That's going to show you your resonant frequencies a LOT faster than doing individual passes, right?

-Kirk

I think it all depends on what equipment you have available. If you have access to semi-legitimate acoustics lab gear, than yes, by all means use white noise, a sine sweep, a pulse, etc. to get the complete frequency response in a timely manor. But if you're using cheap PC speakers (and have to worry about their ability to accurately reproduce the input waveform) and shareware for the analysis I think you're best off keeping it as simple as possible.