4.6 Rover V8 (Problems)

[Tom.H]

Hi

I am looking at buying a 4.6 rover v8 to use in one of my vehicles. I am planning on using a ZF4HP22EH4 auto box as well and have concluded that the cheapest and easiest way would be to buy a complete P38 (as along with the gearbox and engine I also need all the ancillaries)

The reason why I’ve looked at 4.6 instead of a 4L is because it creates quite a lot more torque (as it has a longer stroke). As a result I feel its a lot more suitable for what I am planning to fit it into.

I am not planning on tuning the engine and it will basically be completely standard.

I have heard, very vaguely that the 4.6 rover is known to have some problems…… but I don’t really know what they are and how likely they are to occur (could anyone spread some light )

Tom

[Dave W]

First off, don't panic ! There are a huge amount of 3.9/4.0/4.6 engines out there that have no problems. I have no idea what the failure rate is but it's a lot lewer than the impression some people give. It can, however, work to your advantage when it comes to negotiating a price !

The 3.9/4.0 and 4.6 block walls are a bit on the thin side compared to where they should be and this can lead to hot spots and then cracks in the block behind the cylinder liner. There are two possible results to this, the first is that you start losing coolant for no obvious reason, the second is that the liner can start to move/slip, normally accompanied by a knocking noise when the engine is running.

There is a theory that a lot of the engine's problems are primarily down to the management system, GEMS in particular runs the engine very weak in cruise areas of the map and the resulting heat build up meant that the GEMS 4.6 had the highest failure rate of them all. By trying to make the engine as economical as possible they basically over stressed it so you end up with a large number of pre-2000 4.6 P38s with replacement blocks.

When I fitted my first 3.9 engine about 12 years ago I was warned about the impending doom and was told it was a huge mistake to replace the existing 3.5 because of the 3.9 problems. That first 3.9 is still running in Pam's daily drive and has probably done around 160,000 miles now without any issues other than a cam and followers change and a head gasket change. The second 3.9 I fitted to my Range Rover approx. 10 years ago and I got it on the cheap because it was using water, which all the "gurus" said would mean a new block. I replaced the leaking valley gasket and that engine served me well until I replaced it 18 months ago with an LS1. In teh next few months that 3.9 will probably end up in Pam's motor as that first 3.9 now has a problem with the crank (for some reason the keyway for the crank pulley has worn allowing the crank pulley to move about 10 degrees.

The problems with these engines exist BUT they really are blown up out of all proportion because you never hear about the numerous good ones, only the bad ones. There are also a large number that get replaced by unscrupulous companies that will write an engine off without even looking at it properly. The assumption seems to be that using water = cracked block when in reality there are a lot of reasons for an engine to lose coolant, things that all engines can suffer from.

In terms of running the engine, just make sure it's well cooled, check the coolant regularly, change the oil every 6000 miles and there's a high probability that it'll never give you any problems.

[landybehr]

There is a theory that a lot of the engine's problems are primarily down to the management system, GEMS in particular runs the engine very weak in cruise areas of the map and the resulting heat build up meant that the GEMS 4.6 had the highest failure rate of them all. By trying to make the engine as economical as possible they basically over stressed it so you end up with a large number of pre-2000 4.6 P38s with replacement blocks

Hi Dave (what follows is not adressed to you personally. No intent to attack s/o),

that´s the theory, being repeatedly reported. There is close-to-none chance to have it happen - but I will only believe that argument when I am shown what´s really a original fuel/AFR map.

a) The car was manufactured in a time when fuel regulations were strict already. I hardly can believe that Rover allowed the engine to run at anything other than stoichiometric (ie. AirFuelRation of 14.7:1) unless at full throttle (where it is excused and obligatory). But in all other driving conditions the manufacturer probably must have made sure that the NOx, CO and HC levels did not exceed the regulations. And easiest way to do is to keep the catalysator as happy as possible - which is the AFR of 14.7:1; if doing else the car probably wouldn´t have been accepted by government. With today-cars, I heard (and do not know any better), even good control of the very short time transition from steady-state to acceleration via EAE or X-tau (as MS people use) is necessary to comply with modern Euro-5 classification - so then I don´t think that a rather "coarse" change like leaning really had taken place.

It sounds like very tempting for a manufacturer to lean some areas in favour of fuel consumption and neglect the emissions. But I doubt that THAT happened.

B) Other way: if it really is so easy to avoid the problems you have put into a right light, why didn´t do Rover ? Instead Rover bothered with measuring the wall thickness of cylinder walls and not taking too bad blocks into cars which did increase production and tooling costs.

c) ´Nother argument: it´s said that the cylinder walls suffer from temperature stress because the mixture is kept to lean in some conditons (I think there´s no doubt that cruise conditions are being ment, as in no other driving conditions a mixture change will have significant effect due to car being most of it´s time in cruise speeds). And I think that there is ONE BIG misunderstanding, well, should better say a too vague definition of WHERE the lean condition happens. The MS-forum guys from msefi.com have tought us/me, that the hottest combustion temps are at an AFR of 14.1:1 (any higher or any lower will get lower temps); this in turn is an AFR that you will see at 3/4 throttle (to be precise: MAP is ment) but not at cruise. Cruise, from an economical point of view means 15.5-16.5:1; these AFR´s you´ll only allow up to 1/2 throttle (resp. MAP). So there are arguments that a leaner mixture at cruise will give you lower combustion temperatures (remember: throttle is relatively closed therefore you only get a fraction of the possible air amount into the cylinder - this combined with less fuel .... where from should the heat arise ?). IF YOU lean the mixture at higher-load conditions or WOT then you´ll get damage, not doubt about that, but we hear that the cruise conditions shall have the bug. And, at last, when you buy a chip you are promised more power and better fuel economy that pays for itself over a certain mileage. I don´t know if "they" still advertise like this, but "they" did, once. Nobody has given me a better idea of how that could be done unless by richening WOT conditions and leaning cruise-conditions (which sounds somewhat contradictory).

Well, I nowhere am skilled or trained enough to prove/disprove anything. And more and more I start feeling like DonQuixotte by repeating my doubts like above. Just that I suspect the theory to be based on marketing rather than on research. I even have no better theory to offer (other than having read in DesHammill´s excellent book about the ultrasound thickness measurement of cyclinder walls during engine production); there are some additional ideas like "too lean at higher load", "to high coolant temps" (the P38 runs quite hot due to emissions, which could make it more prone to knock and could contribute to metal expansion issues) but these are just ideas on the level of rumour - I should emphasize that fact.

[Dave W]

The ideal cruise mix for economy is between 16.2 and 17.6:1 but as Land Rover only have very narrow band lambdas fitted it's difficult for them to be very precise in cruise conditions so all they can do is adjust the "too rich" to "too lean" fuelling so that it errs more on the side of too lean than too rich during cruise conditions. In other words if you take a one minute period rather than spending 30 seconds running too rich and 30 seconds too lean effectively centering around 14.7 they would run, say, 45 seconds too lean for 15 seconds too rich as a way of guestimating an overall 17:1 ratio with the sensors they have available. The real downside of this though is that the earlier plenum chambers (pre-Thor) are prone to leaching with preceding cylinders taking part of the fuel charge from others. This can lead to one or more cylinders running leaner than the others and as the lambda is measuring the combined output of four cylinders it's only ever a best guess.

The speculation (and that is all it is/was) that LR were running too lean at cruise in order to get better economy came about because after market maps that used a richer cruise mix set up around a 50:50 closed loop cruise for the engine or those that were set up on a rolling road to an accurate 16:1 mix at cruise seemed to be far less susceptible to block problems.

Personally I suspect it's "all of the above", the later engines had better quality control on the blocks and a richer mixture so were more reliable. The earlier engines with weak blocks were less likely to fail if you ran them at 14.7 in the cruise areas as decreasing the internal temperature reduced the stress on the weaker block.

Lean mixtures are dangerous for the engine because of the increase in internal temperature this causes. A richer mixture absorbs the heat and carries it out of the exhaust port, a leaner mixture will burn hotter than a richer mixture and less of the generated heat goes out of the exhaust port. In extreme cases temperatures inside the cylinder can get hot enough to melt aluminium, at which point the piston normally fails first. The point regarding EGT is valid up to a point because with a weaker mixture more of the heat stays inside the cylinder so you will get a decreased EGT whilst also getting an increased cylinder temperature.

I doubt Land Rover will ever admit the full reasons behind the failures and we'll probably never know for sure but like all good theories it's fun to speculate

[landybehr]

the earlier plenum chambers (pre-Thor) are prone to leaching with preceding cylinders taking part of the fuel charge from others. This can lead to one or more cylinders running leaner than the others and as the lambda is measuring the combined output of four cylinders it's only ever a best guess.

This interesting. I wouldn´t have considered that (I´ve read about carb engines with these manifold related problems) because I had taken for granted that the injectors (spraying towards/onto the inlet valve) would somewhat take care for feeding every cylinder to same rate of fuel.

I doubt Land Rover will ever admit the full reasons behind the failures and we'll probably never know for sure but like all good theories it's fun to speculate

Yeah. Fully agree. But it´s a pity as it leaves space for being misled.

Your argument about the closed-loop percentage thing sounds "sound". Ermmwell, still not fully convinced. I don´t see that a vehicle will pass a "approval procedure" (taking place when a new vehicle is about to be sold and the manufacturer has to prove that it will comply to all recent regulations) when there´s a closed-loop action that will allow lean mixtures that even are suspectible to misfires.

It resembles partly a conspiracy theory that Rover leaned things to fool the fuel-consumption classifications. Somehow like the Apollo Moon landing hoax theory: http://en.wikipedia.org/wiki/Apollo_moon_l...oax_accusations

But don´t take me too serious. My thoughts have little weight in that matter. I only cannot stop to be awkward and provoke people to falsificate my writing - this in order to get closer to truth.

[Steve Hiatt]

To summarise, fit a Chevy LS?

[landybehr]

Besides,

Dave, you´re right that the EGT is an unreliable indicator of combustion temperatures. With too little ignition advance the combustion process is not finished when the exhaust valve starts to open which in turn shoots still-hot gases past the EGT probe. That gives the impression of a hot = lean mixture. But in truth the leaner = slower burning mixture had desired more ignition advance.

I think the problem is the "mental picture" about things which differ in all of us when talking about the same thing. So better specification - when I talk about a lean mixture then this accors to low-load conditions (i.e. 40-75kPa of manifold absolute pressure). In load conditions above that (75-100kPa) I wouldn´t dare to risk something by running lean. Well, if I run at 50kPa ( roughly equals to relaxed running on level ground at 45mph) the engine cannot fill the cylinders with athmospherical air to full amount. Therefore the intake air can only get a faction of fuel compared to WOT. Say the cylinder only gets 50% of air and therfore only 50% of fuel compared to WOT (in reality even less fuel because we want to run lean in my example) then there can only be 1/2 the "mass"/"amount" of heat. Very much like putting 10 logs of wood into the chimney which will produce more warmth than 4 or 5 logs, even if the flame would be a little hotter with them. So in the lean condition you have heat but to a significant lesser amount and coolant pathways that are good enough to cope with full throttle heat conditons should survive that. In addition to that the msefi - people said that the combustion temperatures are lowered when you go from 14.1:1 towards lean.

So there are 2 factors:

a ) the total amount of fuel injected (because of exothermic reaction with O2 providing more or less heat "mass") and

b ) the influence of the mixture on combustion temps (Because even at idle you will make a difference if playing from 12.0 up to 16.0).

And both these factors influence the combustion temperature all time.

In the cruising / low load conditions I would talk about these two factors would both work towards lower temperatures when you lean.

In WOT condition this changes. Even if the mixture is leaner (factor b ) ), then there is more than enough heat from factor a) to damage things. Combine this with the lack of fuel (because of "lean")to catch every oxygen available and you can get reaction of aluminium with oxygen (which damage things in no time) and furthermore an excess of fuel (if being "rich")in this condition does cool the components on the intake stroke (evaporation, enthalpy or how it´s called) and furthermore you get higher levels of Carbonmonoxide (as not enough oxygen to produce pure CO2) and CO is a molecule that could work as fuel itself and therfore must still contain some energy (in other words: energy that could have been heat escaped into the exhaust in form of CO).

Hope, my weird mind managed to explain so that there´s some sense in it.

[landybehr]

To summarise, fit a Chevy LS?

BTW: given the effort and costs that are necessary to get anywhere near the power of the Chevy V8 (which is designed for that power and does handle it well) ...

I think the costs for a overhaul of my 4.2, with all the parts and work, even keeping it in standard trim throughout (except for flanged liners) had been considerably higher than buying a brandnew, fully assembled (with ancillaries) Chevy V8.

Just that the Rover V8 serves me well and with a different engine I would have some disadvantages because nobody will consider it a historical-value-car any more (which would reduce the running costs to next to nothing when the RRC gets 30years old).

[Dave W]

To summarise, fit a Chevy LS?

lol, too right, the LS engines are what the RV8 should have turned into, rather than trying to adapt obsolete manufacturing and design to make a more modern engine they should have done what GM did and design a new one from scratch based entirely on modern design and manufacture. I assume the decision to stick with adapting the RV8 came down to lack of investment. It's a shame really that it was Ford and not GM that bought them last time out maybe we'd have avoided the "transit van man" philosophy and got something a bit more exotic and interesting.

I still love my RV8s but I confess I love my LS1 a LOT more, it is well thought out, well put together, easy to work on (even the majority of the gaskets are reusable) engine that delivers more "bang for the buck" than an RV8 could get to in it's wildest dreams.. but we digress

[Ally V8]

All these arguements about the "lean at cruise" cause of Rover V8 failures would come rapidly to an end if people actually got off their ass and did some testing in real world situations,Set up a scope on the oxy sensors of a 14 CUX,Gems or Bosch engine and you will see what rubbish it is.It will also make clear how inaccurate even modern LPG sequential systems are compared to an old 14Cux working well.

Better still,hook up a pair of wide band UEGO's and really find out what is going on.You will soon realise that you need to look elsewhere for the cause of the problem.

My feelings are that the Buick block was fine with a bore of 87.5mm and pressed in tapered top liners,the 94mm ones become indifferent esp when you take ageing tooling,less than top notch QC and less metal to support the liner.The flanged top liner is the only reliable way to go with a 94mm or larger bore.

Making the sump smaller was a big help too......

[SteveRK]

Good poijt AllyV8 - its not too difficuly to monitor the Lambdas -

also, if an engine is running 'lean' then its obviously not running as efficiently as it might when running at optimum? which implies that is is not developing as much power at a given engine RPM, which obviously means less MPG surely?

[Dave W]

also, if an engine is running 'lean' then its obviously not running as efficiently as it might when running at optimum? which implies that is is not developing as much power at a given engine RPM, which obviously means less MPG surely?

Not at all, as has been stated several times above, all engines are designed to run lean at cruise (lean being defined as an AFR greater than 14.7:1) You neither get the best power at 14.7:1 nor do you get the best economy at 14.7:1. The ideal (most efficient) AFR at cruise, as already stated, is between 16.2 and 17.6:1 it gives you the best fuel economy under low load conditions.

The narrow band lambda sensors fitted to the engines have no way of telling what the mixture is other than when it's at 14.7:1 so they can only be used to guess the mixture at cruise.

All these arguements about the "lean at cruise" cause of Rover V8 failures would come rapidly to an end if people actually got off their ass and did some testing in real world situations,Set up a scope on the oxy sensors of a 14 CUX,Gems or Bosch engine and you will see what rubbish it is.It will also make clear how inaccurate even modern LPG sequential systems are compared to an old 14Cux working well.

My understanding is that people have done just that on rolling roads and have shown that the engines run lean at cruise. The real question is how consistent they are and at what point is "too lean". After all, a perfect engine will run without damage at 17.6:1 at cruise but will an engine that has a flaw in the block ? If you run the engine at 14.7:1 you increase fuel consumption but reduce temperatures in the cylinder and as a result reduce stress on a cylinder design that could well be considered "borderline".

[landybehr]

If you run the engine at 14.7:1 you increase fuel consumption but reduce temperatures (compared to leaner mixtures) in the cylinder and as a result reduce stress on a cylinder design that could well be considered "borderline".

Dave,

in this point (the temperature thing) I dare to disagree with you

Please! read here:

http://www.msefi.com/viewtopic.php?f=31&am...+behind#p190774

and, additionally here:

http://www.msefi.com/viewtopic.php?f=31&am...mixture#p212431

and if you want to know how cumbersome I am:

http://www.msefi.com/viewtopic.php?f=31&am...mixture#p209874

[FridgeFreezer]

Jeez what a lot of words. Buy it, fit it, megasquirt it, forget it. If it hasn't blown up by now it probably won't.

True an LS is better bang for buck but is ultimately more bucks. A complete P38 4.6 could probably be picked up for £1500 or less.

[Dave W]

Dave,

in this point (the temperature thing) I dare to disagree with you

Please! read here:

TBH having read those posts they seem to be concentrating on EGT and aeroplane mixture controls used to compensate for the lack of oxygen % and cooler air at altitude allowing the mixture to be leaned out based on EGT. Having melted pistons and seized engines due to running them too lean I'll take a lot of convincing that a lean engine runs cooler !

I can see I'm going to have to start reading some more recent tuning manuals though because there are some interesting theories in there that contradict some of the things I've been taught over the years.

[Ally V8]

TBH having read those posts they seem to be concentrating on EGT and aeroplane mixture controls used to compensate for the lack of oxygen % and cooler air at altitude allowing the mixture to be leaned out based on EGT. Having melted pistons and seized engines due to running them too lean I'll take a lot of convincing that a lean engine runs cooler !

I can see I'm going to have to start reading some more recent tuning manuals though because there are some interesting theories in there that contradict some of the things I've been taught over the years.

No,what you need to do is what I siad earlier - get out on the road with some monitoring gear,it would achieve several things.First off is to show you how close to stoich engines run when they are closed loop,second is how much of the time they are in closed loop,third is how willing they are to come out of closed loop when you put your foot down and demand power,forth how much harder you can push a Bosch engine than a Gems one before it will go open loop etc,etc.

But most of all it will make closed loop injection make a load more sense to you rather than just reading stuff and quoting figures.The wide band UEGO's are fascinating to watch - and when you get to truly understand whats going on and why you you will realise how good 14CUX,Gems and Bosch are.

I understand the vintage car race tuners are getting more out of carbed engines than ever before as a result of live data from UEGO's.

[Dave W]

No,what you need to do is what I siad earlier - get out on the road with some monitoring gear,it would achieve several things.First off is to show you how close to stoich engines run when they are closed loop,second is how much of the time they are in closed loop,third is how willing they are to come out of closed loop when you put your foot down and demand power,forth how much harder you can push a

It's all been done before, why do I need to go out there myself and prove the world isn't flat ? I'm more than happy to accept the findings from someone else and on my own RV8 maps I make sure the engine runs at or slightly below 14.7. On my LS1 however I run target AFRs and run at 16:1 in cruise as a compromise between economy and keeping everything sweet using a wide band lambda in closed loop.

On the more general discussion regarding combustion temps on lean mixtures, I found this explanation to why internal temps increase with a leaner fuel mixture...

You have to look at what is being heated in the combustion process. If the fuel / oxygen mix is "lean" (excess oxygen), the heat capacity of the gas in the cylinder consists mostly of nitrogen and CO2 -- all the hydrocarbon is already burned. This mixture of gases has a comparatively low heat capacity. As a result the temperature skyrockets (you are heating less gas and those gases have relatively low heat capacity (e.g.) nitrogen.

Not saying it's right, just that it's a better explanation of what I've been taught in the past but struggled to explain

[Ally V8]

The simple reason that you need to do it for yourself is that most of the carp that has been printed is from companies like RPI that have a financial interest in spouting it,not from just reporting fact.That they still sell single point LPG for GEMS cars proves it,if you ever tried tuning a single point system while monitoring via a scope or even Testbook you would realise what a waste of time it is.

Not having a go at you,just trying to encourage people to open their eyes and learn.

[landybehr]

You have to look at what is being heated in the combustion process. If the fuel / oxygen mix is "lean" (excess oxygen), the heat capacity of the gas in the cylinder consists mostly of nitrogen and CO2 -- all the hydrocarbon is already burned. This mixture of gases has a comparatively low heat capacity. As a result the temperature skyrockets (you are heating less gas and those gases have relatively low heat capacity (e.g.) nitrogen.

Hi Dave,

pity I have too little knowledge depth in physics. But when the major component of air is Nitrogen to 78% (and 21% oxygen) it would take a lot of difference in terms of heat carrying capacity for those molecules that emerge from reaction with oxygen. Still explanation is good and I cannot say it´s wrong.

AND

has anyone ever experienced an engine melting pistons when idling ? Would you fear to do so when idling with a mixture of 16.5 (given the engine still turns over nicely then) ?? Don´t think so. If the only relation were like: LEAN = always HOTTER then shouldn´t you expect damage even at idle ? I bet the damaged parts you mentioned had suffered at WOT conditions.

(and I really think one needs to consider the amount of heat. If you get an little ice cube out of the fridge at -20°C and hold it in your hand it would be cool but not do very much. If you step into a bath tub full of ice cubes that are 0°C cold then you will feel a difference. Even though there 20° more of comfort.

So may exhaust gas be hotter even by 100°C, it will have less impact than twice the amount of air being that 100°C colder. If you watch the VE-table which somehow represents the cylinder filling. I think it gets close to the cylinder carrying twice the amount of air at WOT than at partThrottle.)

[Dave W]

After all the compelling arguments on here I went away and did some more research/reading into low throttle/cruise fuel mapping and came up with some interesting results, particularly in the various race and road tuning books. While all agree that lean mixes at WOT are a bad thing the consensus is that this doesn't translate to low throttle/cruise areas of the map. Some of the tuning books even go as far as to suggest you should simply lean the mixture out as much as you can in the cruise positions while the engine runs smoothly.

So, contrary to the points I made earlier about lean mixtures at low throttle being harmful, I now think the whole lean at cruise causing liner/block problems should be consigned to urban legend. I have a tuning session planned in the not too distant future where I'll lean out the cruise areas and hopefully save some cash in the process

[landybehr]

superb. Not because "I always said" or "didn´t I tell you", but rather because the consensus is sth. one could work with. Another success-story of Megasquirt and it´s educational purpose. Without that we´d still believe that lean-theory as much as people once believed the earth is flat .. and the advertisement would still attract us like the worm on the hook does with the fish. Big step forward, all about: http://en.wikipedia.org/wiki/Truth

Once thing I´m not very certain about is the MAP as to where call it "cruise" area. Respectivly at what level of MAP the WOT begins.

Presently I keep the map "conventional" and rich above 80kPa.

[V8 Freak]

John (AKA Fridge) posted this in the Megasquirt thread in "tools n fab".....

Maybe this will help....

Failing that, take a partner in crime with you and bring up the Spark table or VE Table while "Cruising" and see what cells you are regularly using......

Neil

[steve_d]

Also your cruise will be different to both mine and FFs.

Have no idea what you are driving, mine is an auto on 33" tyres and god only knows what ratios FFs beast is running so all our cruising will happen at different revs and vacuum.

Best to start with your data log. Go out and do some cruising. When you review the log the steady state, almost straight lines, of throttle, revs etc will show you where your cruise falls in your VE table.

Steve

[FridgeFreezer]

Once thing I´m not very certain about is the MAP as to where call it "cruise" area. Respectivly at what level of MAP the WOT begins.

This is what datalogging is for