trumpet length 5

[gio1]

Hello

An engine with inlet system consisting of individual trumpets for each cylinder develops peak power at a speed of X rpm. The total length of each inlect duct (from valve seat to trumpet edge) is L.

Assuming a new target engine speed Y>X for peak power I would like to (approximately) calculate the corresponding (shorter) length of each duct

Do you have any suggestions on how to proceed to have a quick estimate?

Thank you

Gio1

 

[SomptingGuy]

You could start by dividing X by Y.

 

[facty]

See the book below. Chapter 6 has a very good section on empirical guidance for intake and exhaust tuning

Title: "Design and Simulation of Four-Stroke Engines".
AUTHOR:Gordon P. Blair
ISBN Number: 0-7680-0440-3

 

[EngJW]

One should not assume that an intake system has been optimized for a certain speed. It may have been designed to other criteria and tuning may not have been one of them. Or the first one may have been a guess. I know of one where the length of the intake tubes was dictated by packaging constaints and didn't even come close to the tuned length.

An excellent book on this subject is "Scientific Design of Intake and Exhaust Systems" by Smith(?) if it is still in print.

 

[evelrod]

The "Tune to Win" series books are an excellent place to start. After reading the spots off Caroll Smith's books, I did all the math (multiple times, pre computer age) by the old slide rule, etc. Finally got my first Lotus engine on the dyno and found that, even though power was good, we were able to get a LOT closer to what we wanted with a little trial and error. I'm not saying the books are wrong but, rather, they are mostly theoretical studies and when applied to an unknown engine without the aid of previous experience leads to added difficulty. It seems after you have optimised the intake setup initially, all further adjustments are accomplished with much less effort.
One of the main problems that I had a hard time with is the measurments are taken from the opening of the "bell" to the CENTER of the intake valve...boys and girls, that center moves a bit! Also the Weber carbs I was using were not totally straight through with aux venturies, throttle plates, etc. in the airstream. Once I got something close on my first engine, all the engines since seem to be somewhat simpler to figure out.


Rod

 

[thundair]

I would think you are seeking a reversion wave to benifit your cam, rod ratio & RPM.

If this is so there is a thing (correct me if Iam wrong) called 1/2 wave reversion and it uses a shorter length to allow the wave to travel 2 times to an opening valve.
This will allow you to tune with some degradation, an acceptable inlet.

 

[GregLocock]

Rod, there's a tuning correction for end effects we use in acoustics, basically the tube is 0.6 D longer than you think it is. I'm sure a series of things like that would make a calculated tune slightly off.

OP, you might get some joy out of the free Lotus engine software at

http://www.lesoft.co.uk

It is a wave dynamics model, therefore it should be sensitive to plenum lengths etc. That also makes it rather cpu intensive.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Warpspeed]

Probably the single most important induction parameter is cross sectional flow area, versus cylinder displacement.

There will be an optimum flow velocity that produces a torque peak somewhere in the Rpm range. Below peak torque velocity cylinder filling poor. Above the peak, pressure drop will reduce cylinder filling.

This peak can be moved around somewhat by also tuning the runner length, but the peak cannot be moved very far doing this. If the valves and ports are far too small, the engine is not going to make massive top end power, no matter how short the runner length is made!!

Pressure waves in the intake runner can be used to shift the torque curve around, and the intake length can resonate at more than one Rpm. It can be tunes to a different harmonic.

Second harmonic tuning is the strongest where length in inches = 108,000 divided by Rpm.

Third harmonic tuning = 97,000 divided by Rpm
Fourth harmonic tuning = 74,000 divided by Rpm
Fifth harmonic tuning = 54,000 divided by Rpm.

While second harmonic tuning is the strongest, you can only get one peak, and the runner length can be inconveniently long. It works best on very peaky high Rpm engines.

On a sports car something much shorter that uses fourth and fifth harmonic tuning is better because it will give a better spread of torque and be less peaky. For instance a ten inch runner will peak at 5,400 7,400, 9,700, and 10,800 Rpm. The peaks get higher at higher Rpm.

Having a double hump at both 5,400 and 7,400 might work rather well on a road car. It might also work well on a motorcycle engine peaking at 10,800 Rpm.

But tuning the runner length is only going to be effective if the induction flow area is sufficient to support enough flow at that Rpm.

A third tuning parameter is plenum volume, which can create a helmholtz resonance in the intake runner, usually at a fairly low Rpm. So there is plenty to think about apart from just runner length.

 

[evelrod]

Yeah Greg. I knew of a couple corrections. My problems stemmed primarily from the 'taper' of the inlet port of the twincam head...from 45mm to 34mm on one head and from 45 to 29 on another. It's been 30+ years and I don't recall how I rectified the formulas to fit the model. I think and, don't take this too seriously, the overall length from bell to valve was something like 14 inches for 6500/7500 rpm (8500 red line) on the 1594cc engine at an output of 195/205 hp. Let's not forget the exhaust in calculating the optimum overall setup.
Sure wish the calcs would come somewhere in the park for my five port Mini Cooper head...so far it's at peak torque around 6066/6250 and hp at 6945 with something like 15 inches(from memory, I don't have my notes handy) overall inlet with a Weber...arrived at by seeing what others were doing and then a lot of 'cut and try'. With the help of the engineers and techs at APT, I've picked up ~13hp with wave tuning and anti reversion work since I started this project last June. Not to shabby for a supposedly optimized 1300cc engine, hey?

Rod

 

[SomptingGuy]

When modelling intake trumpet using 1D codes, the problem is always how to define the length of the trumpet if it has a bell mouth. The usual modelling approach for a bell mouth is a fixed Cd of 1.0, but how much of the bell mouth itself counts as trumpet length?

Acoustic end corrections depend on what kind of ambient the trumpet opens into. For an open-ended ambient, the correction is 0.6r, where r is the radius (note it's not 0.6 times the diameter as previously mentioned in this thread). And for a flanged ambient (one poking through a baffle board), it's 0.85r. So in real like, the correction is going to be somewhere between the two. Not important for long thin pipes (like exhausts), but VERY important for short fat ones (like intake trumpets)

 

[Warpspeed]

Don't forget that when the intake valve is open, the piston crown is the other end of the "tuned" length, and it is moving !

Too many variables and unknowns to work out an exact theoretical mechanical length. Final tuning really needs to be done on the dyno.

 

[facty]

General consensus in 1D engine modelling is that the discontinuity takes place at the valve seat and therefore the piston does not effect the tuned intake length. Discharge coefficient calculations and application is a giant field of study and is extremely important if 1D calculations are to prove accurate. We have studied this very carefully and CDs of 1 for all pressure ratios at a bellmouth are not an accurate description of the flow regime. Similarly CD maps of the intake and exhaust ports are equally important in 1D modelling.

On the topic of "end corrections", we have carried out tests in this very area using both tapered and straight pipes, with and without bellmouths, on a proven, validated single-pressure-pulse test rig and have seen the correction to be uniform over all the test specimens at approx 0.5D.

 

[EngJW]

Greg Locock- thanks for that link. Got it downloaded and just finished the tutorial.

 

[Warpspeed]

facty, you are quite right of course. The tuned length measurement is always measured from the back of the valve to the open end. But there are still a lot of extraneous factors that can influence the results of any simple length formula.

The resonant period or "round trip" time for the pressure waves are going to be influenced by piston movement and valve timing, and the actual pressure and velocity waves in the runner look nothing like the sinewaves that a simple harmonic resonance usually produces.

Some basis for an initial design is better than none, and the length formula given above will usually give a pretty good starting point. Only dyno results can tell for sure if it working as intended.

 

[facty]

Warpspeed - spot on! Simple length fomula might point you to the right ballpark. A well validated 1D unsteady gas dynamic approach will certainly get you even closer (in experienced hands!), but the dyno "must" be used in the development path.