Potential Piston Problem
Discussion
Pumaracing said:
I very much doubt if det was your main problem, or only problem. It certainly increases loads on the bearings but it's almost impossible for it to kill those before it damages the piston noticeably. The OP's det is mild and has gone nowhere near pre-ignition. I suspect his crank and bearings are just fine.
Your issues could have been many things including bearing clearance, oil viscosity, oil pressure and yes perhaps mild det was some part of the problem but unlikely to be the only one. Head gaskets can die for many reasons and poor surface finish on head and block is a main one. Many engine machinists manage barely better than a ploughed field finish, which is fine, well for ploughed fields.
I probably didn't write it clearly or correctly. The main issues I had were bearing and or crank failures, did you know a Jaguar XK engine will run quite well with a broken crank, it just rattles your fillings a bit. This would start with major overheating problems when I pushed the engine over 5K revs, under that it was fine. This went on for quite a few years with various engines, even a brand new VSE engine ended up the same way. It turned out to be purely down to timing, I'd set it at between 36 and 39 deg btdc at max revs, this was what various engine builders had recommended, even the Des Hammill or Vizard tuning books said the same for an XK engine. Finally it was explained to me that at such a big advance the spark was firing as the piston was only just starting the compression stroke (exaggerated) thus the flame front was hammering the piston back down before it could reach tdc which was causing the bearing failure and in hindsight this was in common with the type of damage they were sustaining, all pretty logical when you know. Since I've set the timing to something more reasonable (26-29 BTDC) I've not had a problem in three or four season's racing with the same engine. 123 ignition do two different electronic distributors to replace 6 cylinder Lucas versions, one a generic 6 cyl model and one specifically for the XK engine, the latter has these massive advance curves which are probably fine for pottering about in a 2 ton saloon but useless for a 1 ton race car, but we learn by our mistakes. Your issues could have been many things including bearing clearance, oil viscosity, oil pressure and yes perhaps mild det was some part of the problem but unlikely to be the only one. Head gaskets can die for many reasons and poor surface finish on head and block is a main one. Many engine machinists manage barely better than a ploughed field finish, which is fine, well for ploughed fields.
The head gasket thing was a different matter, I bought a second hand race engine, discussed everything with the seller such as CR etc. and was told he had used normal 98/99 octane pump fuel. The head gasket went at it's weakest point on the second outing and when I stripped the head off and measured the piston/combustion chamber volumes it turned out it was running almost 15:1 CR. This could well have been caused by poor machining but I'm sure the high CR didn't do me any favours. I then fitted a thick shimmed gasket to bring the CR down to a more respectable 10:1 and never had a problem again until one of the modified spark plug threads stripped, it was using 8mm bike plugs to clear the 2 inch valves he'd fitted. After fitting a more standard head it has run fine ever since at 9.5:1 CR.
Hemi combustion chamber engines like the XK, Ford CVH, Fiat Twin Cam have very specific issues. By the time a domed piston crown is up inside the chamber what you're trying to ignite when the plug fires is something akin to a sheet of orange peel. It's about the worst shape for fast combustion it's possible to envisage. As such it's tempting to assume that lots of ignition advance will be needed and technically it is, however with no effective squish area there's also very little resistance to detonation.
So what becomes important with race engines is the combination, as with so many things. The right balance of CR to try and get long duration cams to work reasonably well and ignition advance that doesn't lead to det. With a compact chambered 2v engine, flat top piston and plenty of squish area you'll run mid 11s or even 12 to 1 CR without a care in the world on race duration cams and pump fuel. With hemi style heads I limit this to 10.5 CR and ignition advance of 30-32 degrees.
If you drop much lower on the CR you might as well not bother with long duration cams. They'll run like crap. If you push the advance much further you'll probably see power go up on the rollers but you'll also be disappointed by the number of bits that subsequently fall out of the engine. This is to be avoided because if you run over one of these you can puncture a tyre and that's your race finished!
The advantage of the hemi design is in the large valve sizes and high port flow it can provide. Sensible people realise that the trade off is in the low CR and conservative ignition advance needed to take safe advantage of this.
A man's gotta know his limitations.
So what becomes important with race engines is the combination, as with so many things. The right balance of CR to try and get long duration cams to work reasonably well and ignition advance that doesn't lead to det. With a compact chambered 2v engine, flat top piston and plenty of squish area you'll run mid 11s or even 12 to 1 CR without a care in the world on race duration cams and pump fuel. With hemi style heads I limit this to 10.5 CR and ignition advance of 30-32 degrees.
If you drop much lower on the CR you might as well not bother with long duration cams. They'll run like crap. If you push the advance much further you'll probably see power go up on the rollers but you'll also be disappointed by the number of bits that subsequently fall out of the engine. This is to be avoided because if you run over one of these you can puncture a tyre and that's your race finished!
The advantage of the hemi design is in the large valve sizes and high port flow it can provide. Sensible people realise that the trade off is in the low CR and conservative ignition advance needed to take safe advantage of this.
A man's gotta know his limitations.
jagracer said:
It turned out to be purely down to timing, I'd set it at between 36 and 39 deg btdc at max revs, this was what various engine builders had recommended, even the Des Hammill or Vizard tuning books said the same for an XK engine. Finally it was explained to me that at such a big advance the spark was firing as the piston was only just starting the compression stroke (exaggerated) thus the flame front was hammering the piston back down before it could reach tdc which was causing the bearing failure...
Dunno who explained that to you but essentially it's gibberish.The SOLE purpose of igniting the spark plug some number of degrees before TDC is because combustion takes a certain amount of time to complete. What we want in a theoretically ideal world is for all the combustion to take place exactly at TDC so all of the fuel energy goes into pushing the piston down after TDC. However an instantaneous burn like that is called an explosion and that's bad n'kay?
In the actual world we're forced to live in we find that what the engine actually wants for best power is for peak cylinder pressure to occur a few degrees after TDC. Usually between 10 and 15.
So if we take an engine that is not knock limited, i.e. the CR is low enough that we can do pretty much whatever we want with the ignition advance without anything going pop then what we find is this. At a given rpm and load setting, as we gradually advance the ignition timing the power will rise, it will peak and then with more advance it will start to fall again. Too much combustion is now taking place on the upward part of the piston motion.
That optimum point is called MBT - Mean Best Torque. It is purely a function of how long combustion takes in that particular engine design. A slow burning chamber will want more advance, a fast burning one less advance. Simples.
However not all engines are knock unlimited i.e. they start to detonate before we have reached the ideal ignition advance setting. In such a case we can't run at MBT. We have to limit ignition advance purely to stave off det.
We can solve this by increasing fuel octane or reducing CR if those options are available (or by designing a better combustion chamber!).
Ignition advance is actually a bad thing n'kay? I am constantly mystified as to why humans seem to like it so much and think that more of it is better. The best engines have fast burning chambers and low ignition advance values. The best possible theoretical engine has nil!
Either your engines were detonating or they weren't. If they weren't then all that happens with excessive ignition advance is power goes down a bit from what it could ideally be. Things don't magically break. So it's very likely they were actually detonating mildly and long crankshafts like straight six ones have torsional vibration issues which make them sensitive to such pulses. A more torsionally rigid straight four would probably have been fine in similar circumstances.
I've been re-reading what I wrote above and there's a lovely bit of engine theory in it that maybe someone will spot and wonder if there's a contradiction. I said that in an ideal world all the fuel would burn instantaneously at TDC but later I said that real engines want peak cylinder pressure to occur after TDC. Why? If all the fuel burned instantaneously at TDC then peak cylinder pressure would also by definition be at TDC. Why would the engine really want it later?
Well this is actually another function of the need for ignition advance in real engines because fuel/air mixtures don't actually burn instantaneously. At TDC the conrod is vertical and the combustion pressure is doing absolutely nothing to turn the engine and generate power. All it's doing is trying to push the conrod out through the sump. It's not until the piston has moved down a bit and there's some conrod angularity that the pressure can do useful work.
If we set ignition advance to generate peak cylinder pressure at TDC then a lot of the fuel is actually burning on the upstroke and subtracting from useful work. Due to the way conrod / crank systems operate the piston spends quite a lot of time dwelling close to TDC without moving much. It's not until a good few degrees after TDC that it starts to move down the bore, generates some conrod angularity and lets the pressure turn the crank.
So we reach a compromise. Delay the ignition point a bit, less of the fuel burns on the upstroke, more burns on the downstroke and the net balance of negative work versus positive work on the crank reaches its optimum.
This leads us to an inescapable conclusion. The faster the burn the closer to TDC will be the optimum point of peak cylinder pressure. A fast burning 4v engine with only 22 degrees ideal ignition advance might reach peak cylinder pressure say 10 degrees after TDC for best power. A slower burning chamber needing more ignition advance might reach peak cylinder pressure 15 degrees after TDC for best power. A theoretically perfect instantaneously burning chamber would have everything happen exactly at TDC. There would be no point in delaying it.
As with everything engine related the more you delve into the theory the more levels of complexity you encounter and every supposedly simple topic turns into a much more complex and interesting one.
Well this is actually another function of the need for ignition advance in real engines because fuel/air mixtures don't actually burn instantaneously. At TDC the conrod is vertical and the combustion pressure is doing absolutely nothing to turn the engine and generate power. All it's doing is trying to push the conrod out through the sump. It's not until the piston has moved down a bit and there's some conrod angularity that the pressure can do useful work.
If we set ignition advance to generate peak cylinder pressure at TDC then a lot of the fuel is actually burning on the upstroke and subtracting from useful work. Due to the way conrod / crank systems operate the piston spends quite a lot of time dwelling close to TDC without moving much. It's not until a good few degrees after TDC that it starts to move down the bore, generates some conrod angularity and lets the pressure turn the crank.
So we reach a compromise. Delay the ignition point a bit, less of the fuel burns on the upstroke, more burns on the downstroke and the net balance of negative work versus positive work on the crank reaches its optimum.
This leads us to an inescapable conclusion. The faster the burn the closer to TDC will be the optimum point of peak cylinder pressure. A fast burning 4v engine with only 22 degrees ideal ignition advance might reach peak cylinder pressure say 10 degrees after TDC for best power. A slower burning chamber needing more ignition advance might reach peak cylinder pressure 15 degrees after TDC for best power. A theoretically perfect instantaneously burning chamber would have everything happen exactly at TDC. There would be no point in delaying it.
As with everything engine related the more you delve into the theory the more levels of complexity you encounter and every supposedly simple topic turns into a much more complex and interesting one.
Pumaracing said:
I've been re-reading what I wrote above and there's a lovely bit of engine theory in it that maybe someone will spot and wonder if there's a contradiction. I said that in an ideal world all the fuel would burn instantaneously at TDC but later I said that real engines want peak cylinder pressure to occur after TDC. Why? If all the fuel burned instantaneously at TDC then peak cylinder pressure would also by definition be at TDC. Why would the engine really want it later?
Well this is actually another function of the need for ignition advance in real engines because fuel/air mixtures don't actually burn instantaneously. At TDC the conrod is vertical and the combustion pressure is doing absolutely nothing to turn the engine and generate power. All it's doing is trying to push the conrod out through the sump. It's not until the piston has moved down a bit and there's some conrod angularity that the pressure can do useful work.
If we set ignition advance to generate peak cylinder pressure at TDC then a lot of the fuel is actually burning on the upstroke and subtracting from useful work. Due to the way conrod / crank systems operate the piston spends quite a lot of time dwelling close to TDC without moving much. It's not until a good few degrees after TDC that it starts to move down the bore, generates some conrod angularity and lets the pressure turn the crank.
So we reach a compromise. Delay the ignition point a bit, less of the fuel burns on the upstroke, more burns on the downstroke and the net balance of negative work versus positive work on the crank reaches its optimum.
This leads us to an inescapable conclusion. The faster the burn the closer to TDC will be the optimum point of peak cylinder pressure. A fast burning 4v engine with only 22 degrees ideal ignition advance might reach peak cylinder pressure say 10 degrees after TDC for best power. A slower burning chamber needing more ignition advance might reach peak cylinder pressure 15 degrees after TDC for best power. A theoretically perfect instantaneously burning chamber would have everything happen exactly at TDC. There would be no point in delaying it.
As with everything engine related the more you delve into the theory the more levels of complexity you encounter and every supposedly simple topic turns into a much more complex and interesting one.
Dave I think the bolded part is what Jagracer is alluding to. The advice he was give resulted in peak pressure occuring too close to TDC and therefore the force of combustion had nowhere to go and was over stressin the big ends.Well this is actually another function of the need for ignition advance in real engines because fuel/air mixtures don't actually burn instantaneously. At TDC the conrod is vertical and the combustion pressure is doing absolutely nothing to turn the engine and generate power. All it's doing is trying to push the conrod out through the sump. It's not until the piston has moved down a bit and there's some conrod angularity that the pressure can do useful work.
If we set ignition advance to generate peak cylinder pressure at TDC then a lot of the fuel is actually burning on the upstroke and subtracting from useful work. Due to the way conrod / crank systems operate the piston spends quite a lot of time dwelling close to TDC without moving much. It's not until a good few degrees after TDC that it starts to move down the bore, generates some conrod angularity and lets the pressure turn the crank.
So we reach a compromise. Delay the ignition point a bit, less of the fuel burns on the upstroke, more burns on the downstroke and the net balance of negative work versus positive work on the crank reaches its optimum.
This leads us to an inescapable conclusion. The faster the burn the closer to TDC will be the optimum point of peak cylinder pressure. A fast burning 4v engine with only 22 degrees ideal ignition advance might reach peak cylinder pressure say 10 degrees after TDC for best power. A slower burning chamber needing more ignition advance might reach peak cylinder pressure 15 degrees after TDC for best power. A theoretically perfect instantaneously burning chamber would have everything happen exactly at TDC. There would be no point in delaying it.
As with everything engine related the more you delve into the theory the more levels of complexity you encounter and every supposedly simple topic turns into a much more complex and interesting one.
I suppose the rod ratio will also have an unfluence on how where we would like the peak pressure to occur? The longer the rod the later the ideal peak pressure point?
stevesingo said:
Dave I think the bolded part is what Jagracer is alluding to. The advice he was give resulted in peak pressure occuring too close to TDC and therefore the force of combustion had nowhere to go and was over stressin the big ends.
Whether the pressure is turning the crank or not doesn't actually increase the stress on the big ends. It just determines whether that pressure produces useful work. However, increasing the pressure at TDC which you do with advanced ignition timing, will have some adverse effect. But it's very small in comparison with the increase in pressures that occur with detonation.stevesingo said:
I suppose the rod ratio will also have an unfluence on how where we would like the peak pressure to occur? The longer the rod the later the ideal peak pressure point?
Indeed and an excellently well observed point. The longer the rod ratio the slower the piston moves away from TDC. For normal rod ratios the acceleration away from TDC is considerably greater than the acceleration away from BDC. With an infinitely long rod the piston motion is symmetrical. So for longer rod ratios the piston dwells at TDC for longer and ignition can be retarded a bit to take advantage of this and put more of the burn on the downstroke and less on the upstroke.Evoluzione said:
I don't think that not investing in the correct and readily available equipment, charging so much that people can't afford your time and not understanding compensation values are a very good excuse for a shoddy job!
It doesn't take too long to get an engine to a drivable stage and making good power, but fully mapping an engine, to cover all the corner cases that Max mentioned is an extremely time consuming and expensive business.Starting from scratch with mapping would take 'forever' as you say. The dealings I have had with mapping is I will only work with folk who have 'done it all themselves', can play with all the software parameters, got it running on a proprietory map then set the'dangerous' bits on the rollers to make sure the engine is making best power at full throttle settings whilst being safe in terms of minimum advance to make the power, they then go out again and 'mess' about with part throttle settings or run the systems in 'learn' mode. We have seen the learn mode make some interesting differences on the rollers on some systems.
I do not see it could all be done in one session on the rollers to sort a car the owner has not done all the input to start with.
A guy who worked at Triumph bikes talked about engine dyno time, rolling road time, road time, rolling road time etc etc, a long drawn out process.
On the plus side for 'have-a-goers' as I said earlier, the generic maps are getting better and better as starting points requiring less tweaking.
Still no excuse for mappers to over advance under the power conditions that would cause detonation damage though!
Peter
I do not see it could all be done in one session on the rollers to sort a car the owner has not done all the input to start with.
A guy who worked at Triumph bikes talked about engine dyno time, rolling road time, road time, rolling road time etc etc, a long drawn out process.
On the plus side for 'have-a-goers' as I said earlier, the generic maps are getting better and better as starting points requiring less tweaking.
Still no excuse for mappers to over advance under the power conditions that would cause detonation damage though!
Peter
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff