Piezo injectors and hydro blow

I believe many of you have heard, that you have to be very careful from the situation if the piezo injector stucks in opened position – the engine can be irreversibly damaged. Fuel, leaked in the combustion chamber (which cannot be compressed as air or correct fuel mixture) can create hydro blow – connecting rod can be damaged, the piston, possibly, can be broken, the walls of the engine block can be damaged etc.

It’s clear, that the damages of such type can be very hard, with high possibility – the engine is irreversibly damaged. So the topic has to be taken seriously.

To keep the intrigue, I will tell about my experience at the end of this entry. Till then – I will very make very simple calculations, is such hydro blow possible or not.

 

In the image: the structure of “regular” injector

 

In the image: the structure of piezo injector

 

1. Piezo injector stucks in opened position – is such situation possible?

Yes, it’s possible. “Old type” injectors in a closed position were held by the spring. The spring was comparingly “strong”, even wear of the moving parts or a micropart, stuck in the injector, were not able to keep the injector open. Piezo injectors, intended for petrol engines, don’t have the spring anymore. The injectors are “closed” by the piezo element itself, changing its size. Additionally, the piezo element “feels” only changes of the voltage, which means – if the injector receives signal to open, it will stay open, until the signal to close will be received! For this reason, it’s not allowed to disconnect injectors, while the engine works; problems can also appear, if suddenly the supply voltage, for example for DME, disappears when the engine runs.

Additionally, the engines, in which piezo injectors are used, has much higher rail pressure, also the flow-rate of the injectors – higher. If we take in account, that, for example, for N52 series engines (with rail pressure around 3,5 bar) the opening times of injectors in idle are around 2.5 ms, but an N43/53 series engine with rail pressure around 200 bar: only 0.25 ms. Accordingly – continuously opened piezo injector during time X (for example, 1 second) will inject 10 times more fuel than injector of “old type”! Accordingly – there is some reason to worry.

 

2. What is the amount of fuel, which has to leak in the combustion chamber, will damage the engine?

It’s clear, that in a situation when the combustion chamber is full with fuel, the hydro blow is inevitable. From the other side, if the combustion chamber is filled for 50%, compression will be around 20:1 – that could be considered as max allowed leakage. The volume of each cylinder of N43/53 is around 0,5 liter, the min volume of the combustion chamber (compression 12:1) is around 40 ml. Accordingly, around 20 ml is considered as critical fuel amount, which can damage the engine.

 

3. Can the piezo injector, stuck in opened position, be the cause of hydro blow? Let’s calculate!

a) lets start with the situation – the engine is started and runs. Idle is the most dangerous mode, because in this mode the time, when the injector has to fill the combustion chamber, is the longest. By 600 RPM the engine performs 10 turns per second, each turn lasts 0.1 seconds or 100 ms. Accordingly – after each 100 ms (actually – even shorter time, because we have to take in account the time when the valve is opened) the leaked fuel will be pressed (sequentially) in the inlet or exhaust.

In regular mode, the piezo injector in idle is opened for around 0.3 ms and injects around 6 ml of fuel. So – during 100 ms the injector will inject around 1.8 g of fuel. Or around 2.6 ml, if we take into account the density of the fuel. Such fuel amount is not enough to initiate the hydro blow. It makes no sense to make more exact calculations because the fuel amount has to be for around 10 times higher, to cause the problem. Instead, when the RPM of the engine drops to 60 (the injector would have enough time, to inject necessary amount of fuel), the flow-rate of HPFP will not be enough to ensure pressure in rail to consider the possibility of a problem (by significant fuel consumption).

b) second situation – the engine is turned off. As the HPFP supply is mechanical, the pump will not work in such situation, but fuel from all system (under pressure – pumped before) can leak in one combustion chamber if the injector is opened. Theoretically – yes. Practically – the volume of the high-pressure contour is not large, already after leaking of some ml, the pressure in the rail system drops till pressure of low-pressure system (several bar) and the speed of leakage drops down to a tiny one. The injector leaks slowly, the fuel succeeds to spill in the crankcase. Additionally, for the average N43/53 series engine (with wear around/more than 100’000 km), when it’s turned off, the rail pressure drops to 4 .. 5 bar during several seconds – obviously, some injector is leaking. Such leaks, of course, don’t cause the hydro blow!

 

4. Have the BMW AG engineers thought about such defect? Yes, they have! DME “knows”, how to calculate a used amount of fuel. If DME will detect inadequately high fuel consumption, as one of the test algorithms – from standard pressure ( 150 .. 200 bar) switches to emergency pressure (100 bar). If the defect is caused by the opened injector, the fuel consumption will be proportional to rail pressure – this defect can be recognized. If inadequate consumption is confirmed (due to a defect of the injector), HPFP is disconnected and the rail pressure drops to 5 bar. Comparing to “standard” pressure, now the rail pressure is reduced 40 times. Accordingly – the impact of the defect of leaking injector is also reduced for 40 times. Of course, the cylinder of the damaged injector will not work, but you will be able to drive to the roadside (a safe place to stop) because 5 or 3 cylinders (according to N53 or N43 engine) will still work.

 

5. Is such defect – opened injector – encountered? Yes, it is. Happens, that the injectors are partly open (strongly leaking) and also fully open (stuck). The last scenario – comparingly rare, and it tends to “heal itself” after a short time (several til several tenths of seconds).

 

6. How to recognize this defect? INPA will help with it.

a) strongly leaking injector can be recognized by permanent vibration in idle and also when driving. INPA data ../F5/Shift+F6/F5 (data: Stratified charge/idle) – leaking injector “drops” out the allowed corridor (applied correction – strongly negative), also: /F5/Shift+F6/F1 (live data) In Homogeneous and Stratified charge – the opening time of damaged injector is strongly reduced (for 50% or even more – to 0.1 ms in idle);

b) in case of opened injector: strong vibration of the engine, reduced power. Damaged cylinder – switched off. INPA data: inadequate (below 1.0) Lambda of the corresponding bank (in addition with -30% or 0.0% integrator value, negative overwritten or 0.00 mg/stk – switched off offset type LTFT); damaged cylinder switched off (..F5/Shift+F6/F1 live data confirm 0.00 ms opening time of injector); ../F5/F7 confirms 0 mechanical efficiency of damaged cylinder; the error messages regarding Lambda probes, group misfires in damaged cylinder, HPFP (plausibility, lowered pressure) are possible; mode of 100 bar or even 5 bar rail pressure is possible.

 

7. My personal experience. My own car with N53B30 series engine several times experienced leaking of injectors. Two times the leaking was thorough and didn’t stop for several hours, both times the injector was replaced. I have to admit, that the remaining 4 injectors have served for more than 300’000 km without any trouble. I’ve seen several similar cases of leaking injectors on the spot. Quite often the leaking stopped after a moment – obviously, some micro part in the fuel has gotten in the injector.

Several times (every time other) injectors have stuck in opened position for a moment. With a full range of symptoms – switched off HPFP, inadequate Lambda (for example, at Stratified charge suddenly both – regulating and control probe – reported rich fuel mixture), the error messages regarding HPFP (plausibility, reduced pressure), group misfires of the cylinder, damaged cylinder – switched off. Both times – the defect disappeared after a moment and didn’t appear (for current injector) during 100 .. 200. 00 km again.

Collecting the data about several hundreds of N43/N53 series engines – no real confirmation, that the injector, stuck in opened position, would cause the hydro blow, was found. Obviously, this is a current urban legend, which is not true thou (and simple calculations, mentioned before in this entry, justify it).