What i've done to my Car this week!

[toofast2race]

Yep road train and trail of dawdlers came past - no doubt with huge mirth - TBQH I was more worried about where it had failed - leaving it wasn't an option- reckon it would have been torched by nightfall - amazing how you remember the wiring methodology when required - glad mine made some sense!

[Triumph-V8]

The basic issue is the Volumetric Efficiency isn't linear - ie the ability to pump air at a constant rate in relation to revs . Where my engine has the best VE it will also need the most fuel - this is very obvious when fuel mapping an EFI engine - peak fuel delivery ( injector pulse width) is at peak torque. Obviously other factors come into play at high revs where inefficiencies demand more fuel just to cope with incomplete burn in a crap combustion chamber.

 

That is exactely the point and as this is the main topic to use EFI

I would like to show this in a picture to make that clear:

 

 

This is a fuel map as maybe used with MegaSquirt.

Every point where an engine might work is defined by its rpm (X-axis)

and its pressure in the inlet manifold (Y-axis)

Defining the fuel requirement in this way is for the so called speed density system.

 

90 and above on the y-axis is outside pressure and is more or less full throttle.

52 or somewhat in that area is low pressure at idle.

Lower figures are achieved when car is at higher revs and throttle is closed.

 

This gives full throttle at the 90 line where EFI pics up the figure for each rev

the proper amount of fuel e.g. is 86 at 2700 rpm.

It is a virtual number that is used to calculate the fuel amount

 

If we start at 1700 with acceleration we need 77 of fuel

and at 5600 we need 98 of fuel.

The amount of fuel injected into the engine at full throttle acceleration

follows precisely the red line above.

 

The PI system can not devide between different rpms.

It has to be setup like the black line at 2700 rpm.

 

That means it provides the engine always at full throttle with 86 amount of fuel!!!!

It does that from idle where it is terribly overfueled up to high revs where the engine

runs lean and gives not full power.

 

As one can easily see it is like a too short blanket:

If you make the 86 bigger what can easily done with the cold start lever

or setting up the bolt underneath you overfuel at idle where we would need only 69

but get proper high revs performance.

 

If we are happy at low revs with about 82 like at 2200 rpm

we go quite lean at 5600 where we would like to have 98.

 

Hopefully this makes everybody understand why I always give a smile when reading

that the PI outperforms the EFI in general.

 

If you are on the race track you might overfuel the engine at low revs because you want it

with best performance and will not go below 3500 rpm or so.

 

But if you drive an alldays engine you will burn fuel as hell with that setup and have the engine

stumble at low revs or with leaner setup suffer from loosing some hp at high revs.

 

-always-

 

That is not the ability of somebody to setup

but it is the principle of the system being used!

[GT]

I don't think that Pi necessarily has lean areas in the rev range, but it may lack the 'accelerator pump' that a carb has...

 

Kastner had his own adaptation, that used a cam connected to the throttle cable, but even that could not have provided an accelerator pump

 

John what on earth are you on about?

For goodness sakes read a bit about the subject and stop quoting Kastner.

 

Kastner knows zilch about fuel injection KINSLER does.

He is the US expert on lucas injection, and has maintained it on high powered engines for years.

Why?

Because it delivers the goods!

 

Cosworth fitted Lucas injection on their DA, FV and DF engines for years, and those engines don't have any of the issues you are quoting or V8.

It's utter nonsense to say EFI can come anywhere near beating Lucas Pi.

It can't, because the Lucas system uses much higher pressures, which are a massive advantage to get high powers at high RPM.

 

It's pretty simple, you make the engine have a LINEAR fuel demand, and bingo it makes the power.

To do that trick on a 2V engine is not so easy, because you have to be very careful with camshafts, exhausts and head flow.

 

On a 4V engine it's somewhat easier, but you still have to be careful what you are doing when it's off cam.

For this reason, ALL cosworth engines disposed with vac sensing and ran the the things at 160PSI direct off a control cam, like you put in the photo but much better.

On a cross flow engine it's easier because you don't have stuff getting in the way!

 

It's much more difficult to do it all on a long stroke 2V engine, but it's not impossible.

 

The fact is, the Lucas Pi versions of the 2L engines are amazingly fuel efficient and don't have any of the issues you describe because they are SHORT STROKE.

[GT6Steve]

I don't know anything about the PI beyond getting the spelling correct.

 

But, I'm pretty sure Kas said his cam was simply an enlarged version of the Lucas to allow for a detailed tweaking of the fuel at various throttle openings. Simply to make it more accurate. He didn't claim or intend to reinvent anything.

[GT]

That is exactely the point and as this is the main topic to use EFI

 

This is a fuel map as maybe used with MegaSquirt.

at 1700 with acceleration we need 77 of fuel

and at 5600 we need 98 of fuel.

 

That means it provides the engine always at full throttle with 86 amount of fuel!!!!

It does that from idle where it is terribly overfueled up to high revs where the engine

runs lean and gives not full power.

 

I'm afraid you have the whole world on it's head.

What you have described is a situation of utter nonsense,and it shows your engine is way out of the right window for correct running.

 

If you took the time to plot it out, you would find the Volumetric efficiency of an engine is NOT at its best at peak torque, neither is the fuel efficiency.

(You could read up some stuff from Riccardo about that, as well as aero engines which have been specially calculated to run in the best high efficiemcy window because they have to be able to LIFT their own fuel off the ground!).

 

The volumetric efficiency of a long stroke engine like yours is ALWAYS at low revs, because it has TIME ON IT'S SIDE.

Therefore your presumption that Lucas injection overfuels the engine low down is totally erroneous.

 

Normally speaking if anything, as VE drops like a stone at higher revs (unless something particularly clever has been done), the system will run high HC and high emissions high up.

This I presume is why people mistakenly believe 40DCOE webers give better performance on a TR6.

This is totally impossible, as it compromises the VE even more, because of the need for a restriction, or it won't work as a carb.

(That is why you need an accelerator pump, and injection doesn't)

 

This is more than adequately compensated for by the extra demands needed for cooling caused by the immense frictional loss components of a long stroke engine.

 

If you may remember 85% of all the fuel going into the engine just goes into keeping it cool!!!

[JohnD]

I quote Kastner, because he writes about Triumphs, from his experience racing Triumphs, as team manager and chief enghineer of a professional racing team.

I've quoted you too.

 

I am aware of Mr.Kinsler, but did not know of his writing on Lucas Pi.

Where may I read his work?

 

I know that Cosworth and Ford knew the value of the Lucas Pi system, because I've read about that too. I also know that Maserati fitted to their road cars.

Me, I am a great fan of Lucas Pi, which from my experience (save recently) is reliable and provides the power.

 

But then, if I know nothing, teach me! Don't just tell me I'm an idiot, because I know that.

For instance, how does Lucas Pi provide for richening of the mixture under acceleration?

 

John

Edited June 13, 2013 by JohnD

[Triumph-V8]

There a lot of unproven statements after my explanation of EFI versus PI.

It would take pages to sort that out with real facts and not myths,

what might bore most of the readers.

 

But in short because I can not stop :-)

 

The above fuel-table is not a TR6 table

It is simply taken as an example from the web as it was a jpg to copy.

But I promise my real TR6 Table is also not even in those areas.

 

If somebody wants to read more about the work that has to be done on EFI and PI

compared to run on an engine I recommend what I wrote at the Kastner Forum

for the old article of Kastner where he modified the PI system.

Each step of PI is compared with the step you do with EFI.

 

But to be honest I never could see an engine making a torque flat line

indicating 100% filling of cylinders over a wide area.

 

That is why we speak of a "torque curve" in Germany.

 

If you take a look a Kastners booklet he compares several stages of his engines.

What can be seen is that the curves meet more or less in the area of 2500rpm.

The hotter cam engines rise faster and drop more beyond that point.

That means also the torque is not flat but especially at low revs drops

and the engine does not need that much fuel!

 

The reason is that the ramming effect of the air mass in the inlet

and the swinging effects in the engine do not meet the revs any more

and the real suction and breathing of the engine overweights.

Than the late inlet close of the valve allows to pull some air back into the inlet manifold

and that is missed for godd filling.

 

That is absolutely logical to anybody who is familiar with race engines:

The hotter the cam the higher can be the performance

and the more the torque curve rises with the revs and drops after max.

The reason is that the hot cam is bad at low revs spoiling everything

and makes a ramming effect at high revs adding something air

and the both points in the torque curve have to be connected.

That is pretty easy basics of tuning!

 

One last point please let me state because it is so often bad interpreted:

It is right that PI works on about 100psi and the race engines employed

a mechanical pump after start to rise the pressure.

 

But the Maserati and the Triumph only workes on electric pump

and the reason is simply that you have to accelerate the fuel mass

from the metering unit on through the pipes up to the injector.

It is a standing mass that has than 2-4 milliseconds time to start to run and inject.

If you rise rpm you need more pressure because you have less time

but the street engines worked on 100psi

and from that could save the cost of the mechanical pump.

 

The atomisation is nowadays no topic anymore.

Early race engines had the injectors with butterflies in the trumpet,

allowing more way to atomize towards inlet valve

later versions in the 80s already employed them in the manifold

because modern injectors work pretty well.

 

If one looks at modern cars and pollution and E5 standard

nobody can be lazy not to optimize injectors in modern cars

for perfect atomization to get both nice fuel consumption and emission.

 

We have one beam to four beam sprays and a lot of nice things in the injectors

and they are fabricated in a perfect way that was in the 60s not availiable.

Holley for example uses only 1 bar for his injectors and they work good

on American sports engines.

 

 

@John:

As you already know the Pi has no provision for ACCEL fuel.

The reason is what I always point out:

It is a race injection and for low revs you must have a lot of provisions

that might be neglected if you do not touch low rev area.

At high revs you do not need that urgent!

 

Also the Megasquirt can switch off or reduce ACCEL at higher revs

because sometimes you do not need it.

If you are above 4000 rpm the missing of ACCEL

is no longer felt and one can save fuel not pumping additional fuel.

 

The PI is lucky to cure one disadvantage with the other:

If you set the engine right for top end performance it is too rich at low revs.

 

But that is exactely the area the metering unit looks after if you press the pedal at idle.

You jump into that overrich area and although the engine would like some additional fuel

it is not noticed because need for more fuel is compensated with overrich fuel amount offer from the metering unit

and for a short time it is closer to perfection than constant driving in that rich area.

[GT]

For instance, how does Lucas Pi provide for richening of the mixture under acceleration?

There's several things here, most of them misunderstood.

 

1/ You don't NEED enrichening of the mixture under acceleration.

It is already calculated to be exactly right to give optimum power at full throttle.

As you back off the throttle, there is a very steep cam>injection amount curve which cuts the fuelling down very fast from a vacuum of as little as 0.9>1.1" depression. This makes it behave quite well at part throttle or closer to stoic.

If your engine fuel demand is linear, you have a virtuous curve engine which on the ones I have tested, made a power band of some 5000rpm.

 

You have to do that (enrich) on a Weber type fixed jet carburettor for the simple reason that unless you include an accelerator pump it goes momentarily very lean as you go from one circuit to the other.

You also have to provide some form of high speed enrichening.

The behaviour of a weber is highly non linear, because of the venturi. This is why the thing is really an elaborate bodge to match the different fuel:air weights, velocities and air flows present in a carb.

As you snap open the throttle the air is lighter than the fuel, so you have to enrich to prevent it stumbling.

 

Fuel injection saves all this embarrassment in one hit, simply by ignoring the need for a vaccum (or a vaccum signal as they like to call it). That vaccum signal is also highly dependent of the state of tune of the engine, which can easily be affected by exhaust shape, camshaft timing, exhaust temp, overlap and who knows what else.

 

All we need full engine vac for, is for load sensing which = mass airflow.

 

Webers even hate stand off, which again is ignored by Pi.

SU carbs also often generate bucket loads of stand-off because of the basic design defect of sticking the jet in exactly the wrong place, making them work very badly at large openings.

 

Fuel injection just ignores it all, and part throttle just takes an average value and injects all the cylinders exactly identical amounts.

We don't have to worry about the bane of carbs, which is ultra low temperatures or high underbonnet temperatures, although I DO admit the Lucas system can't compensate for altitude or temp very well.

 

2/ Let's have a look at these statements...

 

"It is right that PI works on about 100psi and the race engines employed

a mechanical pump after start to rise the pressure.

 

But the Maserati and the Triumph only workes on electric pump

and the reason is simply that you have to accelerate the fuel mass

from the metering unit on through the pipes up to the injector.

 

It is a standing mass that has than 2-4 milliseconds time to start to run and inject.

If you rise rpm you need more pressure because you have less time

but the street engines worked on 100psi.

 

The atomisation is nowadays no topic anymore.

Early race engines had the injectors with butterflies in the trumpet,

allowing more way to atomize towards inlet valve

later versions in the 80s already employed them in the manifold

because modern injectors work pretty well.

 

If one looks at modern cars and pollution and E5 standard

nobody can be lazy not to optimize injectors in modern cars

for perfect atomization to get both nice fuel consumption and emission.

 

We have one beam to four beam sprays and a lot of nice things in the injectors

and they are fabricated in a perfect way that was in the 60s not availiable.

Holley for example uses only 1 bar for his injectors and they work good

on American sports engines."

--

 

Firstly, the position of the injectors is irrelevant to atomisation etc.

Placing them in the trumpets has absolutely nothing to do with atomisation whatsoever.

(I remember posing exactly the same question to John Reid at Holbay, as they made my trumpets, and it was the first time, those injectors had ever been put in the trumpets before on a Triumph).

 

If you notice in a modern F1 engine, they also put them very far out from the inlet valve right up in the mouth of the trumpet...

No "dear boy" it's to get the charge temperature down as much as possible, nothing to do with mixture quality.

 

Also they use rather exotic electronic injectors in race engines, which don't run at 1bar like you say.

 

Now,

3/

The idea that uncompressible fuel is somehow taking time to get from the fuel rail

"a standing mass that has than 2-4 milliseconds time to start to run and inject"

This is complete nonsense.

 

The pressure in the injector is always a static 45psi, although I do recommend 60psi for high revving engines to get the droplet size down.

The pulse width at full load is always going to be the same, (unlike an electronic injector), but as revs rise, effectively speaking the synchronous pulse width is widening at the engine (because the engine is going from TDC to full valve opening much much quicker).

The syncronous mechanical injection system therefore has a big advantage again in terms of fuel loss, as less and less time is available to spill it out of the exhaust port on overlap.It may take as long as 100 degrees of crankshaft rotation to inject at full revs, whereas at 2000rpm it could be 20-30.

....this is a virtuous trend.

 

Now we come back to the "how it works in practice bit".

 

Here we do see time and again, (NICK showed it only a few months ago), how it's perfectly possible to get an almost entirely FLAT torque curve from the 2.5pi engine using the early 308778 camshaft.

Although this amounts to quite a serious detune of the potential torque of the engine, in practice, the virtuous part of the torque curve is hitting absolute max as low as 2000rpm, very much like a diesel car.

 

I have seen loads of these 2.5L engines making extremely flat torque curves even with somewhat more aggressive cam timing, quite simply because the 6-3-1 style manifold, makes up for the deficiencies in engine VE & efficiency as revs rise.

 

Steve showed one recently with parts from TRE, that made an amazingly flat strong torque curve to right up to 5000rpm.

That would be a very fast road car.

 

Where you invariably start to get a high risk of pinking on those engines, is around 2000-2300rpm proving without a shred of doubt that the cylinder is really being filled to the maximum possible at those revs.

 

If you don't believe me, you will see that exhaust valving and efficiency has a FAR greater effect on the VE of an engine that modifiying the inlet valve will ever do.

 

That sounds crazy, because generations of tuners have been taught to believe the opposite.

However, - If you plot it out, you effectively find that increasing the flow of the exhaust port increases the VE of the engine at precisely the point where it normally tapers off very quickly.

 

The effect of modifying exhaust port breathing on a small bore 2V engine is roughly 2-3x more effective at raising VE than the same modification to the inlet port anything above 3500rpm, and it affects the pumping losses so strongly that you actually decrease engine friction and drag.

 

Again these things can't be denied, they have been the subject of innumerable scientific papers.

 

You, of course get the deniers (many of them "TR experts")....saying you can't possibly get more power and torque from nowhere.

Well yes you can, if you avoid wasting it & the fuel that is being dumped in there 85% for nothing, and use it to increase the volumetric efficiency of a long stroke engine at precisely the point it's hitting a brick wall in terms of dropping mechanical efficiency.

This is why the 6-3-1 exhaust was so successful and alone could generate a power increase of 20bhp+ doing nothing else.

 

How come? you may ask.

Well, again in serious scientific studies where they drove engines with an electric motor, instead of with gas, there was almost NO difference between the power losses observed with or without being driven with fuel, all of which leads you to realise the frictional losses (in say a modern F1 engine) are so gigantic, that at one point they actually outweigh the power being made at the flywheel.

 

FYI this point around 18 000 rpm in a F1 engine is where you are LOSING around 800bhp in frictional losses against 800bhp in flywheel bhp, and all you are doing is heating up the environment and the water as a result.

 

In practice this is why they used to have to have pitstops to refuel constantly, but nowadays they asked for a different strategy to "make racing less predictable"

Indeed F1 races are won almost entirely on aerodynamic drag/down force and fuel efficiency nowadays.

 

On a Triumph TR6 engine, you have to be making about 200bhp of frictional losses to be making 180bhp of flywheel power.

Which is why the fuel consumption is phenomenally bad, compared with the short stroke 2L engine making identical power.

 

(I measured roughly a 40-50% fuel consumption penality, because I have actually been able to compare the 2 in test conditions btw). Nice thought isn't it?

If the Lucas Pi were so bad, how come it's relatively banal to get a 2L version of the same engine to make 160-180bhp as a road car and still be able to return over 40mpg when driven sensibly?

[Triumph-V8]

I am not a teacher or a priest.

All I can say is that the things are somewhat different

and I wonder how two people with fully different basic knowledge

both can make engines that work.

 

Anyway, some of the physics behind the need to add fuel under acceleration

and cut fuel at decel for the fuel consumption and emissions

can be read here:

 

http://www.msextra.com/doc/ms1extra/MS_Extra_Software_Manual_files/AETxtFile.txt

[GT]

The difference of approach is simply between open loop, and closed loop.

 

With open loop systems you have to assume a predictive analysis method using integration of real world empirical figures.

This entails measuring everything and redesigning things that work less well than ideal.

It costs a lot more to get right and takes time to get valid results you can duplicate on a larger scale.

 

With a closed loop system you don't have to know anything, you just control what you have got in front of you.

It costs nothing to do, because you can measure it with AFR.

Anything less than ideal,doesn't really matter (such as cam duration and lift), because you can always compensate for it.

 

The first one requires you to create an operational window, and optimise everything to converge in there.

(eg. 6 individual throttle bodies)

 

The 2nd one with its inherently higher losses makes up for them by running tighter tolerances on the whole operation.

(eg. a single inlet butterfly)

 

I think the hybrid approach using computer controlled ignition timing, but mechanical sequential injection has a lot to be said for it.

[James]

How about computer controlled mechanical injection hybrid

[Nick Jones]

Been done with Bosch KE-jet..... done for emissions and only really a "trim" function to improve CAT function and life. Bosch K-jet had pretty good control anyway - my A90 had a CAT ("1 way", no closed loop) with straight K-jet and CAT lasted 140k. Not really an ultimate performance devisce though as continuous injection (also called CIS) and has that great fuel metering flap in the intake.

 

Nick

[Mother Teresa]

Worked on the car today- phew!

Washed and serviced it. Kind of forgot how quick the car is for an old girl when I took it out for a spin after. Looking to spend wednesday doing more in readiness to sell so someone else will enjoy using it as it was intended and not languishing in my garage. It's quite cathartic don't you know!

[Nick Jones]

You'll miss it when it's gone...... make sure it goes to a good home - it deserves it!

 

Nick

[JohnD]

Spent today puting new bearings in the old 2.5 engine, as being the quickest back to the grid.

This engine was only pensioned off because

a/ The CSCC said that cars had to have original engine size (they didn't want Fords wtith Cosworths in, as Swinging 60s is a low cost series)

b/ It was making an odd noise

and

c/ CSCC have now chnaged their minds, and my Vitesse can run a 2.5 (Grrrrrrrrr!)

 

On dismantling, the mains were hardly worn and the big ends a bit more, but two showed scoring, just one 'stroke' each. Hardly enough to cause 'funny noises', but it made me wonder what had got into them. Then, to get the new shells in, I had to remove the timing cover to give the front of the crank some clearnace. And Lo! The chain tensioning spring had snapped!

 

It had gone at its pivot, just snapped across. Could some spring steel fragments have got through the filter??

Replaced it with a used, OE spring I had in the box of bits. Found that the broken spring was narrower at the pivot, so much so that I had fitted two washers inside the pivot bracket to stop it flopping around. The OE one is wider and needs no washers.

Think I've heard of this before, with repro springs.

 

Now, I need an undisturbed weekend to change engines.

Previous week was a washout, as half my most often used tools were stolen on Monday. Teach me to lock up properly.

 

Oh, and the swallows' nest in the second garage means I have to ask nicely if 'er indoors will move her car outside and let me use the main garage as a workshop. Doesn't reality get in the way of real life?

But who cares about that? Have you ever had a swallow actually touch you as it flies into your garage? Thrilling!

JOhn

[JMH]

John, I don't know with the 2ltr engines, but on the 4 cylinder ones I have dismantled, I have always found one washers on both sides of the chain tensioner.

[JohnD]

My OE GT6 manual agrees with you, JMH, but this is a duplex chain and spring, but-but, my OE TR6 manual for an engine versio hat had an OE duplex chain, also shows washers!

The old worn, but intact one was just too wide at the pivot to take any washers.

 

John

[JMH]

Aha, sorry, didn't notice that ... what joy!

Couldn't find any s/h usable parts for my own duplex conversation (4 cyl), so still waiting for my made-in-Tyrkey-repro chain tensioner to disintegrate into my completely rebuild engine

Edited June 17, 2013 by JMH

[Mother Teresa]

Fitted new master cylinder seals on the clutch and brakes and ran through new fluid. Even booked the car in for an MOT this saturday!

 

[Nick Jones]

Blimey - real action..... You could always try driving it round again when you have the MoT..... you might even like it?

 

Nick

[mark spit]

Not done much on mine recently, was in a mates garage Dec to Feb while I sorted some house stuff out.

Got it back on his trailer, dropped it off and got an MOT pass with advisories on the brakes due to standing.

Drove it back from the MOT centre and though, diffs noisy, top up and still noisy really noisy :-(

 

Stripped everything out to remove and the long bolt seized to a bush inner and wouldn't budge. Finally out with assistance of a reciprocating saw last week, new one on order. That last sentence spans 2 months............

 

Mark

[AJ.Lintern]

Mine passed it's MOT last Friday Nearly left me stranded at work when the choke linkage jammed open and prevented it from running properly - luckily noticed the jets sticking out, gave them a wiggle and they sprung back in again - ran ok then

[Mother Teresa]

Hmmm... Unlikely Nick

Saw a big over restored MGC yesterday at the traffic lights(too tall on the suspension and the bloody hood wad up! No wonder he looked hot). Wouldn't tick over properly as only running on 5 cylinders and burning oil.The realities of classic car ownership :/

[Mother Teresa]

MoT'd this morning and passed wth flying colours- brakes all balanced and wheel bearings as they should be. Need to put some fresh fuel in before we go any further as it is perillously low!

 

 

 

 

[Nick Jones]

She's alive...... well done!

 

Nick