Hi all,

What is the RPM limit for the stock 20vt engine? (can handle without issues)
And what about the modified forged inrenals?
What about the crank shaft?
How to increase the RPM limit, which modifications needed?

Thanks for help.
Regards,
Mostafa

pm Flea, he probably knows more than anyone about this sort of thing

i know, but is mechanical components limit the revs.

This mean another light and balanced crank would increase the rev limit, but the question is. what is the limit and what modifications needed to maximize it.

Also, what is the max record for this engine rev before?

6,800 is the stock limit. Flea will raise this to around 7,500 for cars with balanced internals.

Highest I know of was JohnS at about 8,000

After this, you run the risk of valve float

Jari has raised the limiter to 7800 rpm. His engine is completely rebuilt with forged internals, C&B cams etc.

I think Flea can raise the limiter to 7200 rpm on a standard engine but only if you really need it (i.e. for high rpm power). Most of us hit max power at 6000 rpm, so no use of a higher limiter.

Using the stock rods and bolts there is also the possibility of the rod bolts letting go at high rpm lift-off. This is going to be more likely the higher rpm you run.

On a stock engine running safe power (i.e. bigger than stock turbo, but not a top end monster) you are unlikely to be still getting more power over 7,000 rpm, so there is no real need to increase the rev limit that much.

Thought the stock limiter was at 7125 revs?

Nigel, What was the power figures when JohnS reach 8000 rpm?

Umm, what about evo cossworth 2.2L storker kit reaching 9000 rpm ?

If we get a very well balanced internals in the 20vt, can we reach such revs ?!! ( 9000 rpm )

And is there a real benefits from that?

Regards,
Mostafa

Even though peak power is achieved lower, there is still a benefit to higher RPM, as the rate of acceleration is higher in lower gears, so being able to "hold on" to a lower gear for an extra 1000RPM will benefit the acceleration, otherwise it'd be faster to change gear at 5750RPM when most of our cars reach peak power, but we all know that's not the case.

About 560bhp on pump fuel and 616bhp on race fuel

I dont believe there is an evo cosworth kit for the coupe?

It does not matter how balanced the internals are - you'll get valve issues well before 9,000 rpm regardless as Nigel has already mentioned.

As far as benefits are concerned - you can run a big turbo with a monster rev limit and still have a useable powerband, even if it does not spin up until 5,000 rpm. If you plan on using such a car on the road it will be a pain to drive....

Extending the rev limit more than 500 rpm over the peak power point is going to be counter productive IMO. You'll get 'higher' up the next gear, but that only helps up to a certain point.

No cosworth kits for the coupe, i meant the Mitsubishi evolution stoker kit that cosworth build.

I know about valve floatation, but there are other engines that don’t suffer such issue at 9000 rpm. Take for example Honda SIR engine. It reaches 9200 rpm smoothly.

I prefer mid range power with excellent drivability than drag monsters, but i wonder why our engine doesn't rev as high as Jap engines while nothing is missing it .

Honda engines are built to rev - that's where they get power from rather than mid range turbo grunt (stock 20vt).

The Stock 20vt engine was not designed as a power monster - rather a GT, and the rev limit would have been selected in accordance with this. There was no need to make the engine rev harder.

The amount of development that has gone into Evo tuning is staggering compared to coupes - it's all been done before. I dont know too much about the engines, however I do know that some of the big turbo cars use solid lifters - they also have forged rods as standard, and run far bigger turbos as a stock car.

Besides issues like valve float, there are a couple of things I can think of which limit an engines ability to rev.

One is mechanical strength - the ability of components to handle the stresses which increase with rising revs - and the other is the ability of the engine to breath enough air to get a reasonable level of volumetric efficiency.

Mechanical strength is largely down to the materials used, any treatment of the materials to improve strength and the fundamental design of the components, but the stresses are also a factor of stroke length, rod length and component weights.

Most modern engines are expected to cope with mean piston speeds of about 18 metres per second without stressing any component. For a 2 litre Coupe with standard everything that piston speed is reached at just over 7000 rpm. However, there may be a component in the Fiat engine that cannot take this level of stress, so please do not assume that by what I have said that 7000 rpm is a safe limit! Instead be guided by others on here with experience of failures on this particular engine. If mean piston speeds exceed 21 metres per second on a regular basis you can expect to suffer regular component failures on any standard production engine. Which component fails first is down to the relative strengths of each, but rods, rod bolts, piston rings and cranks as well as valve train components are all in the line of fire.

For the 20VT this means that you shouldn't consider 8000rpm as a safe rev limit without upgrading internal components.

Having made the engine 'safe' to rev higher, whether the engine will make any power at such high revs is down to the ability of the engine to breath effectively. One primary component that limits this ability is valve area (of course there are others but I am trying to make this as simple as possible).

The 20 valve engine has a valve area of about 6835mm^2. The 16VT has a larger valve area - about 7480mm^2 or about 9.5% larger. Other engines mentioned also have larger valve areas. The standard Mitsubishi Evo has 7263mm^2 or about 6.3% more valve area than the Coupe and the Honda S2000 8143mm^2 which is about 19% more valve area.

Therefore, as standard, these other engines all have more potential breathing ability inherent in their design and therefore the ability to rev higher than an un-modified Fiat 5 pot engine.

Great post ^^^

Is the valve area calculation just a total of the area of each of the 20 valves (ie Pi-r-squared for each valve), or does it also take into account the valve lift? A quick Google search suggests the former

The reason I ask is that if the Coupe is giving a bigger valve lift, it will make up some of the deficit from an engine with a bigger valve area

Guy Croft was very complimentary about the basic design of the 20v head (although he still managed a 14% flow increase with some porting)

I was simply using valve area as the limiting factor. Of course lift is important to total flow and a lower-than-optimum lift cam would cause restrictions, but I was comparing maximum potential flow area across the engines as I was trying to explain why one engine type can rev higher than another.

Fiat seem to have been designing great heads for a long time and I'm sure that this is another good example though I have not worked on one myself. However, if it had been designed with competition and revs in mind, as some prevoius Fiat heads have been, I think the inlet valves would be around 33mm as standard and without hydraulic lifters. Mind you, I don't know if there is enough meat in the head to allow for the larger ports that would be required. 33mm valves would allow the engine to breath to 9000rpm+ by my estimation (with all other appropriate mods).

To me it looks like the 2.4 head on a 2 litre engine would be a good change to make as I believe the 2.4 engine has 1mm larger inlet valves than the 2 litre - has anyone tried this?

The simplest solution is to use the VIS head and replace the exhaust valves with sodium-filled

I thought the valve were already sodium filled nigel?

A great read from Lancia - thanks for taking the time to write all that.

Not on the VIS, I seem to recall that they are just ordinary valves - only the turbo has sodium-filled

Yep, thats what JohnS told me too when I was sorting my head.

If I remember correctly from a post from Y E A R S ago, one of the main bearings (big end?) is officially rated by the supplier at less than what the engine is allowed to rev to as standard already, so yes there will be parts that cannot take the extra stress of higher revs.

That said my Alfa 147 gearbox is rated at 305 Nm and that one's been coping with 370Nm for 200.000 kilometers now without problems.

Piston speed is critical here though. The S2000 at max. revs (9.000 is the fuel cut point) has an average piston speed of just about 25 meters per second which is about the same as an F1 engine at 19.000 rpm. But that's all average speeds, the maximum speed is over 40 meters per second and it's the maximum speed that can cause some serious disruption in the engine, not only in terms of heat / wear but also air flow, vacuum forming in the sump and even the pistons have to be really really tough as they can just collapse due to the acceleration.

Errr... that's a bit from memory when I was talking to someone who was trying to make a 1.4 Punto GT engine rev to 10.000 rpm. I was pretty quite myself during that conversation.

As I said, I was trying to keep it simple! But peak piston speak and peak piston acceleration loads are other critical numbers.

What is quite interesting is that using the production stress levels of mean piston speed <= 18 metres per second, the S2000 should not be revving higher than about 6400 rpm. As it is built to survive 9000rpm all the components were obviously designed and manufactured to reach such high loads, so well above the standards of a 'normal' production engine.

However, piston acceleration/deceleration loads depend on stroke and rod length. The 90mm stroke with a 145mm rod on the 16 valve engine makes it safe in this respect up to about 7700 rpm. The shorter stroke of the 5 cylinder engine means acceleration forces are safe up to about 7850 rpm. The S2000 has a stroke in between the two at 84mm, but due to a longer rod, piston acceleration at less than 7500 rpm exceeds generally recognised safe limit of 2400G for production engines - yes, you did read it right - 2400G!

At 9000 rpm the mean piston speed on the S2000 (as said above) is 25.2 metres per second, maximum piston speed 41 metres per second and maximum acceleration/deceleration is nearly 3500G!!! That's 46% higher than most production engines are expected to survive.

Using lighter rods and lighter pistons is one way to reduce the damaging forces due to acceleration, as the amount of force is equal to the mass x acceleration.

Sorry - was not thinking straight, thought that we were talking about the turbo engine.

Hi,

I have 34 mm inlet valves and 28 mm exhaust valves in use in my Coupe.
Also forged internals, which are totally 100 g lighter (piston + conrod) than OE one.

Merry Christmas from Finland.

-Jari-

Isn't the S2000 engine low volume, produced in "human - intense" process rather than robot assembled? Along with high quality materials and engineering, that hardly makes it comparable to mass market engines... as 16vt (Lampredi TC) or 20vt (part of the SuperFIRE engine family - FIRE as "Fully Integrated Robotised Engine")?

I think the main limitation on the 20VT engine for RPM is the ability for the hydraulic lifters to cope without getting valve bounce at high rpm. I didn't get any even at 8000rpm. That was a bit of an arbitrary limit but I think they did test it above that and started to get some issues when it was being mapped. On some other cars hydraulic lifters like the VAG ones you can get issues of them locking out which will destroy the engine which is why people are wary of them. You cannot compare an Mitsubishi Evo or the like as they did not have hydraulic lifters (not sure if they do now or not)
For a stock 20VT engine I wouldn't want to run it much above stock as the rods are notoriously unbalanced with Fiat tolerances being +/-10g. One of the plug-in chips did a 7300rpm limit if I remember rightly.

Hi John - long time, no see

How can you tell the safe rev limit for an angine?

I'm now running lighter rods and pistons, and they are all balanced to within less than half a gram of each other. I'm currently running a 7,300 limiter, but it reaches this with ease and doesn't feel at all strained (in fact it starts to sound REALLY nice just before the limiter kicks in).

I've always wondered if I've been a little conservative with 7,300 or whether this safe-ish limit is what has contributed to the engine's reliability

Well there is a bunch of maths around piston speed but the 20VT is ok for that as standard, wheras the 2.4 really shouldn't go much above 8500rpm anyway if I remember rightly (it was a long time ago I did the maths). Whether you would get any benefit from it is easy to tell - if your power chart is dropping already at 7300rpm then you are likely to be at the limit of something else.
I also looked at the Exhaust gas temps to be sure that I wasn't getting a massive amount of heat build up due to the restrictiveness of the exhaust, manifold, turbo etc.

My power band was such that peak power was at 7700rpm with road fuel and was still climbing with race fuel. But when I was running a more restrictive manifold and turbo setup it dived down much earlier such that it really was pointless to go higher than 7300rpm. That was corroborated by the EGTs being at the higher side of what I wanted (in the 900C+ range)

I think where max power is in the rev range is another good measure of how good the whole setup is - if power is still rising at your rev limiter then the match of components isn't perfect, mine as described above had peak power at the high end of what is acceptable I would say and it might have been more beneficial for it to be at 7500rpm with an 8000rpm limiter rather than 7700rpm

I see - in which case, it looks like it wouldn't do me much good



At lower boost levels (runs 1 & 2 - 1.0 bar and 1.2 bar respectively) the power is still climbing at the 7k point. However, on the two high-boost runs (3 & 4 - 1.5 bar and 1.7 bar) the power is just beginning to tail off

Almost certainly a consequence of my 0.64AR turbine housing, but seeing as I will not tolerate any more lag, it looks like I may be stuck with it (until I can find a turbo with variable turbines)