141 replies to this topic

[Calorus]

I've tried to make sense of this several time...

With Vee engines many race applications sem to utilise flat plan cranks. Is this for reasons of strength, mass or inertia, or firing characteristics? Any help welcomed, thanks...

[alexbiker]

Flat plane crank v-8's are mildly out of balance, and will vibrate. Racers couldn't care less about this, provided it doesn't damage anything, but it would tend to mitigate against you buying the car, so road cars tend to be cross-plane, leading to a smoother ride.

However, cross-plane cranks require counterweights to spin smoothly, which equals rotating mass. Add mass, limit rev-ceiling and engine responsiveness. With today's capacity limited, naturally aspirated formulae, the limiting factor is getting oxygen into the engine, which is essentially limited by revs. So, a higher rev-limit can lead to a higher-power engine.

So, mass and inertia.

There's also a minor issue in the firing order of a cross-plane design, in that the order by bank is LRLLRLRR, which means the perfect exhaust system would have a cross-over from left to right for maximum power - very tricky to package, especially in a single seater.

Alex

[Calorus]

Originally posted by alexbiker
Flat plane crank v-8's are mildly out of balance, and will vibrate. Racers couldn't care less about this, provided it doesn't damage anything, but it would tend to mitigate against you buying the car, so road cars tend to be cross-plane, leading to a smoother ride.

However, cross-plane cranks require counterweights to spin smoothly, which equals rotating mass. Add mass, limit rev-ceiling and engine responsiveness. With today's capacity limited, naturally aspirated formulae, the limiting factor is getting oxygen into the engine, which is essentially limited by revs. So, a higher rev-limit can lead to a higher-power engine.

So, mass and inertia.

There's also a minor issue in the firing order of a cross-plane design, in that the order by bank is LRLLRLRR, which means the perfect exhaust system would have a cross-over from left to right for maximum power - very tricky to package, especially in a single seater.

Alex

Masterful, many thanks!

[m9a3r5i7o2n]

Quote from alexbiker;


However, cross-plane cranks require counterweights to spin smoothly, which equals rotating mass. Add mass, limit rev-ceiling and engine responsiveness. With today's capacity limited, naturally aspirated formulae, the limiting factor is getting oxygen into the engine, which is essentially limited by revs. So, a higher rev-limit can lead to a higher-power engine.

The above answer somewhat implies that a 180 degree crankshaft does not require any counterweighing for the primary out of balance. If this were so then the 1930-31-32 Oakland-Pontiac V-8 180 degree crankshaft engine did not need the primary counterweighing however they did have full counterweighting for the lower connecting rods primary shake.
M.L. Anderson

[alexbiker]

However, cross-plane cranks require more counterweights to spin smoothly

There. Fixed.

By the way, has anyone ever accused you of being just a bit of an engine geek?

No? Just me, obviously.

[GSX-R]

Hi,

I think inertia/weight and simplicity are not the main reasons to choose a 180° V8 :

I think you forgot : gas acoustics. It's more easy to make good manifolds for trapping and exhaust with flat-planes V8 (crossplanes require long and complex pipes), and especially shorts ones for high revs goals. 1/2 wave and 1/4 wave settings are more easy to achieve with 180°s. Flat plane V8 have very similar acoustics settings than L4 engines. For crossplane V8s it's another story. That's essentially, in addition, the reason why crossplanes V8s have a very different sound than flatplanes V8s.

Two more notes :

-with flatplanes V8 like inline 4 if you have too second order forces vibration, you can also install as long as you can longer rods to reduce them.

-with flatplanes V8 you can slighty change the angle of the V. It will change engine smoothness and will have quite limited effect on the balacing. It can be very interesting for the engine fitness into the chassis and aerodynamics. On the other side, try to change the angle of 90° of the crossplane V8 and you will soon have balacing problem that will remove it main interest against flatplanes : its perfect second order balancing.

[ray b]

has anyone used a flatplane crank in a chevy v8?
have they had racing success with it?
I googled that but got no hits

[m9a3r5i7o2n]

Some one in Australia I believe tried it but it takes a lot of dyno work to make it really work in any 180 degree crank engine. Just regrinding the camshaft with the almost identical camshaft to the 90 degree isn't all that easy as the timing of the impulses is highly critical. Also how many Chevy engines are in rear engine cars, not many as compared to the ones in front engine “stock” cars? Getting a cam grinder to grind a special cam for a 180 crank might be very difficult to do. All he wants to do is develop cams for the most customers he possibly can, Money talks we all know! The magneto or the distributor would be very easy to solve.
I once called a crankshaft manufacturer about a crank for a Hispano-Suiza to see what they would charge and if I remember correctly it was $10,000 USA dollars for each in a minimum lot of ten. At $100,000 I immediately said no thanks.
I think Ray Bell might know something of this Australian crankshaft.
M.L. Anderson

[cosworth bdg]

Originally posted by Calorus
I've tried to make sense of this several time...

With Vee engines many race applications sem to utilise flat plan cranks. Is this for reasons of strength, mass or inertia, or firing characteristics? Any help welcomed, thanks...

Firing order .............

[dead_eye]

out of interest whats meant by first and second order vibrations? when do they occur and whats the difference?

[GSX-R]

when you calculate the piston acceleration, speed and position of a traditionnel piston, you have to resolve an equation. Because of the con rod, the motion is not sinusoidal.

To resolve, you have to use a Fourier Transform. That gives an equation with an infinity of coefficients. The Fourier transform breaks-up the acceleration function to an infinite harmonic single vibrations succession.

-> first order vibration (frequency is the same as the piston engine) + second order (twice then engine REV frequency) etc... Generally we keep only the two firsts (sometimes the third) because the bigger the order is, the less the amplitude is. So we usually neglict from the third order.

http://www.ducati.co...cle&part=global

That is for a single piston. When you have an arrangement of piston/cylinders you have to use spacial vectorial analysis. You can if you want simulate all the arrangements using the excellent Manolis program : http://www.pattakon....duc/balance.exe. I have also made an excel spreadsheet for main architectures. The resulting vibration is generally expressed in 3 differents vibrations forces, and 3 differents vibrations moments (each one for first and second order).

Hope it helps a little.
Feel free to ask ..

For the firing order, it's acoustics (more even complex than balancing to achieve and calculate (when possible and especially at exhaust when it often need experimentation)). It deals with resonance properties of gaz and manifolds layout/size/length/connections to correctly fill and scavenge the cylinders.

[dead_eye]

Thank you gsxr a brilliant post and some very interesting links!

To this day it still amazes me that no matter how much you learn about engines theres always a hundred more topics and theories to come!

[GSX-R]

Maybe because there almost so much physics in the engines. This is really a case of study for the person that likes to learn physics and elements and to change his mind of current life worries.

For balancing, it generally doesn't impassion crowd. You can find very little elements on the net about that. Surely because vectorial analysis and Fourier Transform are not very exiting for most of folks. That explains why for instance 3 inline cylinders are more difficults to handle with and quite unsual on cars, why inline 4 doesn't generally exceed 2.2 liters without lanchester balancing shafts, etc etc..

For your information null first and second order forces and moments without specific counterweights on the crankcase can be achieved on : Inline/Flat 6, I/F8, F/V12. Using counterweights, the V8 crossplane achieves it also. V10 for instance doesn't achieve it, V6 neither and of course V6/V4/V2.

But thank to silent-blocks, we do not need so good balanced engines to run our usual engines at usual speeds and quite almost configuration is possible according to some needs and within limits !

[Greg Locock]

Bear in mind that there is very little you can do about second order vibration, once you have a given engine configuration and piston mass. Lanchester shafts are a very expensive solution.

[cosworth bdg]

Originally posted by Greg Locock
Bear in mind that there is very little you can do about second order vibration, once you have a given engine configuration and piston mass. Lanchester shafts are a very expensive solution.

So Mitsubishi have found out the hard way..?

[Ray Bell]

Originally posted by m9a3r5i7o2n
.....I think Ray Bell might know something of this Australian crankshaft.

Niel Allen ran one, according to a post by Graham Howard, that was made by Peter Molloy on Merv Waggott's gear...

It vibrated badly.

I spoke to Kevin Bartlett after I'd found out about this, because I knew they dabbled with flat planes in the Chev engines in the latter years of F5000. He told me that after he found they vibrated like crazy, the talked to one of the famous engine builders in America, who provided him with a camshaft that changed the firing order and made it much better.

[Greg Locock]

Roughly speaking a Lanchester shaft is about the same complexity (cost) as a second camshaft, so if you have the option a SOHC V6 starts to look attractive compared with a DOHC I4 with a balancer shaft, once you get above 2 litres.

For many years I struggled with mounting I4 engines without balancer shafts, and saw at least two cars that were delayed for YEARS as they struggled with an appropriate layout.

[dead_eye]

GSX-R thankyou very much for the information, are there any indepth books you can recommend on the subject?
So do i take it that any straight 6 can be 'balanced' to a negligable degree by changing the firing order and the crank patern? or is a case of any un-counterwieghted crank and the correct firing order will be naturally balanced by the engines design?

for give my ignorance, im just trying to set my head straight here but assuming the vibration comes from the change in direction of the piston why isnt an inline 4 perfectly balanced? after all you have one piston de-accelerating upwards while one is de-accelerating downwards?

Also why would a crank of the maximum allowable points (i.e a v8= 360deg/8=45deg crank) not be better balanced without counter weights as your maximising the overlap between power strokes and in theory getting a constant force running through the engine than the stop start of a 180deg crank?



I have to agree theres so little information out there on the more advanced subjects and without a financial gain to be had i.e just for my own interest and hobby a university course is massivley expensive way to learn!

[VAR1016]

Another interesting thread on this perennial topic.

I have asked in the past why, if flat-plan V8s vibrate, have they been fitted to production Ferraris since the 1970s? Last time I rode in one it seemed very pleasant to me. Or is it just superior engine mounts?

PdeRL

[GSX-R]

Originally posted by dead_eye
So do i take it that any straight 6 can be 'balanced' to a negligable degree by changing the firing order and the crank patern? or is a case of any un-counterwieghted crank and the correct firing order will be naturally balanced by the engines design?

One important thing to note : We usually calculate free inertial rotation, not a "charged" rotation. The firing order by itself doesn't have any influence on the free inertial balancing. Firing order is determined by engine architecture, crankpins arrangement and, gas acoustics when choice beetween 2 cylinders that are in phase. We do not take into accounts internal gas pressures and thus torque vibration due to the engine torque against transmission and fixing. That means we have to imagine the engine spinning by itself with no gas, no fuel and the spark plug are removed, so ideally the engine in the vacuum ! This is not a bad approach (to neglict internal gas pressures) because at high RPM we assumes that pistons, con rod and crankshaft [forces and moments] are much more big than the internal gas pressures due to the strokes of the engine. Forces and moments or pistons/rods values are square functions of RPM, internal gas pressures are supposed to be almost the same during the RPM range, so we simply decide to neglict them at top RPM. Torques vibrations due to the torque engine and transmited to the fixing are not taken in accounts and have to be calculated separately for other good reasons : we ignore the instant torque shape and value of a specific engine. The purpose is to define a general rule for one architecture and not for one engine. In application, if we know the instant torque curve, we can always add it to the free rotation vibration component and then we'll have finally the resultant "torque" vibration against the chassis's engine fixing. In any case, real case can be very complex so don't forget this is a "basic" calculation. It's possible engineers use more sophisticated software simulations and finally experiment to finalize their work, as usual !
Don't forget that assumptions. If not, you will quicky be lost.

To my knowledge, there's only one I6 design for crankshaft and it's always the same firing order. The crankshaft is naturally balanced and the engine by itself until the second order. Sorry for english books about balancing. I don't know them. I've just believe to remember an olf french book for that kind of matter. All those arrangements are known for approximately the piston engine was invented so that kind of stuff is rarely prone to revision. Very old stuff.

why isnt an inline 4 perfectly balanced?

The ascending pistons have almost never the same (opposed) acceleration that downwards pistons. Take for instance a glance on balance.exe from M. Pattakon or any else animation of I4 you will easily see that, for example a piston spend more time at BDC than at TDC. Take also a look at the piston acceleration/speed to make your mind.

Also why would a crank of the maximum allowable points (i.e a v8= 360deg/8=45deg crank) not be better balanced without counter weights as your maximising the overlap between power strokes and in theory getting a constant force running through the engine than the stop start of a 180deg crank?

Not sure to understand. The engine is 4 strokes so you need to dispatch only 8/2 power strokes around the one revolution. So a flatplane (180°) crank and two rows of 90° of cylinder can achieve the smoother power strokes for a 8 cylinders.

[GSX-R]

Originally posted by VAR1016
Another interesting thread on this perennial topic.

I have asked in the past why, if flat-plan V8s vibrate, have they been fitted to production Ferraris since the 1970s? Last time I rode in one it seemed very pleasant to me. Or is it just superior engine mounts?

PdeRL

A V8 flat plane with a limited unit displacement, light pistons/conrods, good attenuations mounts is quite good. V6 dont' have even first order moment balancing and fortunately do not vibrate too much. The danger is especially in the resonance of the vibrations or when the mounts cannot handle anymore the vibration amplitude.

Very difficult to imagine a Ferrari with a crossplane. Very good english cars TVR also use 180° flatplanes. BMW could use a flatplane V8 for its M3 to make it more wild. But...

[VAR1016]

Originally posted by GSX-R

A V8 flat plane with a limited unit displacement, light pistons/conrods, good attenuations mounts is quite good. V6 dont' have even first order moment balancing and fortunately do not vibrate too much. The danger is especially in the resonance of the vibrations or when the mounts cannot handle anymore the vibration amplitude.

Very difficult to imagine a Ferrari with a crossplane. Very good english cars TVR also use 180° flatplanes. BMW could use a flatplane V8 for its M3 to make it more wild. But...

Many thanks for that - I should have guessed really as I know that vibration is exacerbated by mass, the Vanwall 2.5 litre engine being a good example.

Also I recall that the 1.5 litre Ferrari F1 engines were greatly improved when they changed from the "offset" 65 deg V6 to a 120 deg engine.

PdeRL

[GSX-R]

Generally best balancing angle for a V6 is around 65°. It depends of the con rods/stroke ratio. (If con rod would have a infinite lengh, 60° would be the ideal angle). This value is generally interesting because then the moment vibration amplitude is spread almost equally around 2 axes and then generally easy to handle.

Because compactness or aerodynamics considerations or fixing considerations, other arrangement can be selected. Because two con rods of a V of a V6 usually (except for 120°V) don't share the same pin (contrary to flat or cross V8, or V12), the engine smoothness is not impacted (same instant torque curve).

The 120°V6 common crankpins is pretty bad balanced but it permits to share the same crankpin for two rods and is also quite flat.

[BarryD]

Fascinating stuff. Thanks GSX-R.
If this is a bit off topic, sorry.
Can you refer me to anywhere I can find out about these vibration issues relating to vee twins?
I'd be fascinated to hear what you think goes on inside my vintage 50 degree JAP vee twins, which have forked and blade rods ie.a common crank centre within circular flywheels (ie. Huge conterweights.)
I have one long-stroke racing engine which vibrates like mad, and a short stroke "square" engine with "wobbly" (eccentric) flywheels which revs beautifully smoothly.
Barry D.

[jpf]

dead_eye and others -- I asked about the second order discrepancy in I4 engines way back in 2001. Imaginesix emailed me a nice description that really clarified it at the time:

From imaginesix via email:
The asymmetry is geometric, and is due to the fact that the con rod must move sideways as it rotates.

If the sideways movement were perpendicular to the cylinder (so disregarding the up/down portion of the rotation movewment), the piston would still be forced to move up or down within the cylinder.

If you combine the lateral movement contribution to the up/down movement, then as the crankshaft moves through the upper half of it's circular motion, the lateral motion ADDS to the up/down movement, causing the piston to move more per degree of revolution, meaning faster, meaning with greater acceleration.

As the crankshaft moves through the lower half of it's circular motion, the lateral movement SUBSTRACTS from the up/down movement.

As the connecting rod approaches infinite length, the lateral motion has less and less impact on the total movement of the piston, causing it to follow more and more a perfect sinusoidal path.

Neat huh?

So, neat, huh?

The other thing, for VAR1016 about flat crank V8s, is that their second order vibrations come from the I4s that make up each bank. These vibrations are perpendicular to the plane of the bank, so when they are in a V configuration and added together, they partially cancel each other. I believe in a 90deg V, they should cancel to sqrt(2). So a 4l 90deg V8 should have only about 40% more vibration than a 2l I4, which is not the end of the world.

That last paragraph came out of my ass.

[edit]: typo (bank not blank)

[GSX-R]

Originally posted by BarryD
Can you refer me to anywhere I can find out about these vibration issues relating to vee twins?

Don't hesitate to open a thread. Same for other than V8 architecture. Otherwise one is likely to speak about too much case

[GSX-R]

Originally posted by jpf
That last paragraph came out of my ass.

Right ass

[toggy]

Look guys, the reason they do this is simply to get the max bhp out of the most compa
ct configuration. It has nothing to do with balance and everything to do with the exhaust plumbing....

[cheapracer]

Originally posted by Ray Bell


Niel Allen ran one, according to a post by Graham Howard, that was made by Peter Molloy on Merv Waggott's gear...

It vibrated badly.

I spoke to Kevin Bartlett after I'd found out about this, because I knew they dabbled with flat planes in the Chev engines in the latter years of F5000. He told me that after he found they vibrated like crazy, the talked to one of the famous engine builders in America, who provided him with a camshaft that changed the firing order and made it much better.

In the late years of F5000, John Edmond's car ran one and went thru the camshaft thing as well.

[TDIMeister]

A type of scotch-yoke cranktrain could eliminate second-order forces, since the motion is sinusoidal and the "connecting rod" (not to be confused with the articulating one in a conventional engine) is rigidly attached to the piston. Many people have tried to develop engines based on this principle with variations, but several fundamental problems exist. I recall discussion in the past of a scotch-yoke engine in a motorsport application, but not for any particular balancing benefit.

[Moon Tricky]

Nissan have devised a variable compression ratio engine with a multi-link crankshaft that has better balancing characteristics than a plain crankshaft. Essentially, the big end of the piston rod moves in an ellipse rather than a circle, the smaller sideways motion resulting in a vertical motion that is closer to sinusoidal. They are claiming an I4 with this system balances better than a V6. And it's got variable compression ratio as well, which is even better. I guess it's quite expensive though with all those extra parts, not to mention increased friction and potential for breakdown. It's all trade-offs in this game.

As for the balancing of crankshafts...

If you've got a crankshaft for N cylinders, with each piston pin equally offset from the last at an angle of 360°/N, linear motion will cancel to order N-1. However, there will also be a vibrational moment about the axis - imagine the case of a two cylinder engine, you've essentially got a rod with a weight on the top left and another at bottom right. When you spin that round it's going to shake. There are two ways to fix that - balancing shaft, or, double the cylinders up in a symmetrical arrangement so the shake of one half is exactly opposed to the shake of the other half. So a straight four flatplane actually has the linear vibrations of a two cylinder engine, whereas the crossplane version balances all the way up to 3rd order. Putting things in a 90° vee also helps because each pair of cylinders balances to first order on its own, so that also gets rid of the vibrational moment which is why a crossplane V8 doesn't need a balancing shaft but a V6 might do. This is also why you can't change the angle of a crossplane without getting those moments back, whereas for a flatplane each bank has no vibrational moment to begin with.

Why would a flatplane need counterweights? Well in theory it doesn't, but in practice it's better for each cylinder to balance as well as it can otherwise you have to transmit the force through the crankshaft from one crank to the next, which I'm guessing necessitates a stronger crankshaft and also puts more stresses on the journals.

[phantom II]

Masserati and Ferrari use virtually identical V8 engines save for 90 and 180 cranks. The Massa with the 90' crank is smoother and hence the 'luxury' application.

Originally posted by toggy
Look guys, the reason they do this is simply to get the max bhp out of the most compa
ct configuration. It has nothing to do with balance and everything to do with the exhaust plumbing....

[Moon Tricky]

Of course any power that's going into shaking the car about isn't going into driving it forwards, so presumably there's some other factor involved in the "trade off". Presumably, the reduction in exhaust flow restriction in the flatplane outweighs the power loss from the increased vibrations. In a luxury car, you care more about the smoothness than that extra couple of hp.

I've become quite enamoured by the inline 5 configuration lately. Linear vibrations better than an inline 6 or V6, vibrational moment no worse than a V6, power delivery only marginally less even. Or have I gone mad?

[GSX-R]

A good 5 cyl engine will be better than a 'bad/old designed' I6.
A I6 is naturally a lot better balanced than a I5. But with rubber and a good suspension you can make miracles on odd arrangements (V6s for instance).

[Moon Tricky]

An I5 needs a balance shaft to cancel the vibrational moment. But it linearly balances to 4th order, whereas an I6 balances linearly "only" to 2nd order.

[GSX-R]

Countershaft on I3 and I5 half cancel the vibration on one plan to translate the other half amplitude in another plan. They're pretty bad balanced, especially the I3.

But actually, we don't care after 2nd order, the amplitude is quite neglictible for those orders.

[Greg Locock]

"An I5 needs a balance shaft to cancel the vibrational moment. But it linearly balances to 4th order, whereas an I6 balances linearly "only" to 2nd order.
"

I'd love to see the maths on that.

Also the power absorbed in shaking an engine is tiny- of the order of less than 100W, so whatever else a flat plane crank is doing it is not wasting much energy in vibration.

[Moon Tricky]

Intuitively:
sin(θ)+sin(θ+2π/5)+sin(θ+2*2π/5)+sin(θ+3*2π/5)+sin(θ+4*2π/5)=0
just because it's cyclic and symmetric...
only when you get to substitute 5θ for θ do the sine waves add up instead of cancel out, because then you've got
sin(5θ)+sin(5θ+2π)+sin(5θ+2*2π)+sin(5θ+3*2π)+sin(5θ+4*2π)=5sin(5θ)

What you've got is, if this makes it any simper, the "offets" for each cylinder in multiples of 2π/5:
first order: 0,1,2,3,4
second order: 0,2,4,6,8 = 0,2,4,1,3 = 0,1,2,3,4
third order: 0,3,6,9,12 = 0,3,1,4,2 = 0,1,2,3,4
fourth order: 0,4,8,12,16 = 0,4,3,2,1 = 0,1,2,3,4

but for fifth order: 0,5,10,15,20 = 0,0,0,0,0

[Moon Tricky]

I've had half a bottle of wine. Let me come back to it tomorrow.

[TDIMeister]

Well, this thread has developed to a discussion about several different topics off the original of flat-plane cranks... I will only add to this... :

Almost every VCR concept involving some newfangled crank mechanism effects the change in the CR by changing the relationship of the clearance volume (i.e. raising or lowering the actual piston TDC location). This has an obvious deleterious effect on squish and cylinder turbulence.

For all the complexity and added reciprocating bulk of most VCR mechanisms, I'm not a terribly big fan of the idea. It is beneficial in a gasoline engine IF you want to achieve both good part-load efficiency (high CR for good indicated efficiency) and high specific-power in a downsized, forced-induction engine (low CR for knock resistance).

This does not need a large CR adjustment range and IMO can be effectively addressed strictly from valve timing (early intake valve closing, i.e. Atkinson cycle for low-speed, efficiency-sensitive operating modes; and late IVC, i.e. Miller Cycle for high-speed, full-load operation that requires a reduced effective compression ratio for knock).

This can be realised in a very simple intake cam phaser (for DOHC applications), already in widespread use for gasoline engines.

Back to flat-crank V8s. I am privy to some simulations in GT-Power investigating different firing orders in exactly a flat-crank V8 and cross-plane V8, and particularly their effects on BMEP and cylinder-to-cylinder CoV of volumetric efficiency. The project goal was not strictly to maximize VE, but to make sure the cylinder-to-cylinder variation in VE was minimized, since this would have an effect on the AFR of each cylinder and at each cycle, thereby affecting the composition of the exhaust gas that the catalyst downstream in the exhaust "sees". Alternating pulses of "rich" and "lean" exhaust streams can deleteriously affect the catalytic conversion efficiency.

The GT-Power models were created with separate specific designs for the intake- and exhaust manifolds for the 90- and 180-degree cranks, respectively, and the pipe lengths were optimized to be tuned at the same peak power RPM, but all other geometric parameters remained the same. The intake manifolds were of a plenum-type of no particularly special geometry, and the exhaust headers were of a 4-into-1 type. For each type of crankshaft, both manifolds remained the same in every investigated case, (i.e. shape, volume and pipe lengths not re-optimized)

We found that there was a detectable difference in VE and BMEP between the 90- and 180-degree cranks, but the magnitude of the difference of the peak values was small. Varying the firing order in the flat-crank design also showed insignificant differences in BMEP as well. From these simulations, it was concluded that: 1) the different crank layouts do not affect overall BMEP by as much as some people have speculated; and 2) The effect of firing order on BMEP is even less (although non-conventional firing orders were not investigated, only even-firing ones). Optimizing the manifold design for each case could yield greater gains, but this was beyond the scope and time constraints of the project.

The big eye opener was to find an astounding difference in CoV in both VE and AFR between the 90- and 180-degree crank models. The 180-degree crank gave (predictably) a much lower CoV in both VE and AFR in some cases. A suggestion was extrapolated that could explain a performance angle between the different designs. Some cylinders could be running quite rich, and therefore much unburnt fuel wasted out the exhaust, while other cylinders could simultaneously be running lean and therefore knock limited. A tell-tale sign that this might happening might be an engine that runs without excessive CRs and seeming to have it all together, but experiencing detonation despite a great deal of fuel enrichment. Reducing the CoV of VE and AFR will mean that all cylinders see roughly the same AFR, which means that the global AFR must not be enriched so much to make up for the lean cylinders.

[GSX-R]

The big difference is Inertia and compactness and probably a higher possible RPM on the flat-plane conf. More vibrations on the flat-plane sure but on a sport engine this is not bad for feelings.

It could be interesting to tune especially the exhaust system on both configurations to see if it's easier and more efficient to tune/design the exhaust manifods on the Flatplane configuration than on the cross-plane configuration, especially on the kadenacy/(1/2) wave effect. But this could be an real expert and a hard job. On the cross-plane conf, the fact that 2 Cylinders neighbours sometimes fire successively give automatically an odd exhaust pressure and flow in the mannifolds. I imagine this is harder to manage and to study also. A flat-plane conf don't. We can imagine back pressure at exhaust is tehn easier to manage and so to tune on the flat.

And maybe the exhaust efficiency simulation, even with GT Suite, is still a simulation. There is i agree less debate on inlet.

I could be interested if you can send me the GT datafiles.

Regards.

GS

[TDIMeister]

At each major step in the simulations, we validated the results in actual fired engine tests with the same configurations, so "still only simulations" were reproduced in the real-world with accepted norms in R^2. We collected CVS emission data, fast gas sampling in the cylinder and at various points in the exhaust and pressure traces, again in-cylinder and at various points in the exhaust, plus also this time in the intake ports.

So I'd say our "still only simulations" correlated well with real world tests.

I obviously cannot give out the GT files, but maybe I can post a screenshot to show I'm not making this up.

[TDIMeister]

Actually, the work was presented as a paper at the GT Users Conference in Frankfurt last October. http://www.gtisoft.c...onfarchive.html

Specifically: http://www.gtisoft.c...8-Breathing.zip

[GSX-R]

Thank you for those documents. I like charts. It seems, except for vibrations and, maybe, for the sound (subjective) the Flat plane is more homogenous and easier to tune for the gas. To summarize a more even engine than a crossplane and a wilder engine.

Very interesting the charts for adjacent firing cylinders and the related problem of the AFR disparity..

Have you also done any simulations with supercharged (especially turbocharged) V8 configurations ? It could also be very interesting. (Nowadays, i don't see any advantage for naturally aspirated engines except for their "vintage" and conservative aspect)

I'm gonne read the other documents in the library. I haven't made so much test in GT. I'm an amateur and then rather a beginner with it.

P.S : by the way, the first V8 on a car was built in 1903 by Clement Ader

[Peter Leversedge]

I was at Ascot Park in I thimk it was late Aug 1981,there was a new sprintcar racing with a Keith Black Chyrsler based engine and I was told it had a flat crank. It did not sound like a V8, more like a F1 car. I understand it was third time out and it was one of the fastest car there. If I remember correctly it was 5th fastest in the time trials.

[TDIMeister]

Originally posted by GSX-R
Have you also done any simulations with supercharged (especially turbocharged) V8 configurations ? It could also be very interesting. (Nowadays, i don't see any advantage for naturally aspirated engines except for their "vintage" and conservative aspect)

Turbocharged gasoline engines no, but I have done simulations of a V8 turbodiesel; primary objective was to investigate staged turbocharging.

P.S : by the way, the first V8 on a car was built in 1903 by Clement Ader

I'll be sure to point out the error. Thanks.

[Gerald Ryan]

If you would like to run a two plane crankshaft in a V-8 at high rpm (more than 10,000rpm) try this.

Start by building a crankshaft with reduced counterweighting mass. The best method is to reduce the amount of counterweight adjacent main journals 1 and 5. It is necessary to try and make the counterweights more nearly uniform in mass. Two things occur. The first is that the position or phasing of the counterweights will be "unusual". The second is that you will end up reducing your cylinder block included angle to ~80 degrees. Sure, the engine will uneven fire but it will be very smooth (better than the 90 degree design). Best part is that all the mains will be loaded more uniformly and peak load excursions will be reduced. Crank bending and torsionals are improved. Now you can reduce diameters some.

Deal with the exhaust manifold issue using "reversine flow" cylinder heads. The exhaust ports open out into the valley so that it is convenient to use a 180 degree exhaust design. Use individual stacks on the inlets.

Regards

Gerald

[TDIMeister]

Originally posted by Gerald Ryan
If you would like to run a two plane crankshaft in a V-8 at high rpm (more than 10,000rpm) try this.

Start by building a crankshaft with reduced counterweighting mass. The best method is to reduce the amount of counterweight adjacent main journals 1 and 5. It is necessary to try and make the counterweights more nearly uniform in mass. Two things occur. The first is that the position or phasing of the counterweights will be "unusual". The second is that you will end up reducing your cylinder block included angle to ~80 degrees. Sure, the engine will uneven fire but it will be very smooth (better than the 90 degree design). Best part is that all the mains will be loaded more uniformly and peak load excursions will be reduced. Crank bending and torsionals are improved. Now you can reduce diameters some.

Deal with the exhaust manifold issue using "reversine flow" cylinder heads. The exhaust ports open out into the valley so that it is convenient to use a 180 degree exhaust design. Use individual stacks on the inlets.

Gerald, I'm afraid I'm confused here. How can you change the block included angle to, as you say, ~80 degrees, simply by reconfiguration of the counterweights? Are you proposing to do this on an existing engine, or are you proposing a ground-up design? And are you including the crank pins when talking about reconfiguration of the counterweights?

If, having an uneven-firing, non-90-degree bank angle V8 does anything, I suspect that you are considering the gas forces and not only mass forces, since a fully-counterweighted cross-plane 90-degree V8 IS in fact mass-balanced for forces and moments in the first- and second order. However, I could see that if you play with the counterweighting and intentionally introduce some mass imbalance, when you superposition the gas forces, there may be periods where the gas force excitations cancel out the mass forces, but this would be harmonic in nature, with some RPMs and their harmonics being better for vibrations and bearings loads while being worse in others.

Also, I come out with a blank for "reversine flow" cylinder heads. I know about exhaust ports that open up in the valley. Is that all there is to "reversine flow" heads? Can you point me to some information about it?

[TDIMeister]

Further to what Gerald stated about non-conventional crank-pin and bank angle layouts, I remember what my Dynamics of Machines professor said on the first day of the course and repeated it right to exam day: If you forget everything else in the course, remember only one thing -- only mass forces can balance mass forces. I run away when people start talking about using springs or gas forces in the combustion chamber for mass balancing purposes.

Now I'm certain Gerald does not imply that his proposed setup gives better mass-balancing than a fully-counterweighted cross-plane 90-degree V8, but this is for the others who may misunderstand what is being implied here. You can reduce bearing loads and total excitations, but the mass imbalance from what Gerald proposes will still be there. And I reiterate that the phenomena are harmonic in nature -- what might be better at a specific frequency (read RPM) and their harmonic multiples will be worse at other frequencies and their respective harmonic multiples.

By the way, OE multi-cylinder engines for automotive engines, especially V8s, are rarely fully counterweighted anyway. Most go to between 75-90 percent before the added weight and inertia overtakes the marginal improvement in vibrations.

[GSX-R]

I am also sceptical about this arrangement. That way, you'll break the precious balancing of the crossplane V8 that is the reason of being of the crossplane V8.

Did you study and make figures for this arrangement or is it more a kind of empiric methods and a real engine that has already run a bench ?

TDImeister, do you think the gas pressures could really take part in balancing at high revs ?

Mecanic Forces and moments are functions of square of revs. Gas forces, I imagine are more linear functions, at least, if we stay under the speed of the sound in the exhaust mannifolds. So, for me and until now, at high revs, gas forces are neglictible in the figure of balancing.

For the rest, if you intend to make a V8 to spin at 10,000 rpm, I think it could be best to change the crankshaft for a flat-plane one and idem for the ad hoc camshaft.

From a more general aspect, does a crossplane V8 can rev as fast as a flatplane configuration ? How much the crank-pins, the crankcase and journals need to be big to sustain high-rev like a flatplane ?

In the general automotive industry, how expensive is a flatplane vs a crossplane crankshaft ?

Regards