What is Engine Knock?
Many of us have heard the term engine knock or pinging and have possibly witnessed the final outcome of engine knock related failures. How then might we better understand this phenomenon? How might we also guard against it happening in our expensive road or race bred engines? These are good questions and a step towards getting a hold of this concept will be to start with an understanding of normal combustion, where engine knock is not evident.
To begin, we shall concentrate on the ignition phase of the 4-stroke engine cycle. I will assume you are familiar with the Otto-cycle of intake, compression, combustion and exhaust strokes common to most engines we come into daily contact with; however the general principles apply to 2 stroke engines as well.
At the point of ignition, the mixture of fuel and air has been compressed by the compression stroke to a point where it can be efficiently ignited, usually by a single spark at just the right time. When the spark ignites the mixture, it normally causes the advancement of a flame front that moves throughout the combustion chamber like the ripple caused by a stone dropped in a pond. This burning proceeds to envelope all of the fuel air mixture in the clearance volume at the top of the piston and creates heat and pressure. The key here is the orderly nature of flame propagation. It is not an explosion, which is an instant uncontrolled event. On the contrary, it is progressive, starting at one place (the central spark gap) and proceeding until all the available mixture is burnt. This takes a finite amount of time. In order to extract the most work (power output) from each combustion event, the ignition is advanced to a number of degrees before top dead centre to account for this delay. Exactly how many degrees of advance is optimum will depend on many factors, but the dominant one is engine speed. For a given (fixed) burn time, at high engine speeds, the ignition will have to be advanced by more degrees than at low speeds; all other factors being equal. There is more to that story, as we shall find out, however the extent of this explanation shall suffice for now.
Remember now, how we talked about the orderliness of the combustion event? This is the crucial factor. The ready-to-burn mixture is ideally in a state of turbulence for good mixing and the when the spark sets it off; most of the combustion takes place away from the metal parts. In an optimum world, only a smallish portion of the heat of combustion finds its way to the cylinder walls and the other metal components, such as the combustion head chamber or the valves. And it mostly does this where the burning gas just touches the metal surfaces (Remember this point).
So, stable and controlled combustion minimises heat loss to the water jacket and allows the metal components to live without much stress. This leaves a significant proportion of the heat left over to do what we want, and that is to expand the intake gas by making it hot and pressurised. Simple enough. The hot gas can then only become expanded when the piston moves down the bore, then generates mechanical work, turning the crankshaft and along the way wins you the race! The hot gas becomes cooler as it expands and the energy lost to the gas comes out at about 30% in exhaust heat, 30% in water jacket heat, and 30% (roughly) in energy to turn the crankshaft to push you forward and prepare any other cylinders on the crankshaft to fire.
So what of engine knock then? In simple terms engine knock occurs when the orderly combustion process referred to, breaks down. It turns out that petrol and most hydrocarbon fuels have distinct limits of pressure and temperature at which they will sustain orderly combustion. What happens when these limits are reached, is that the fuel mixture will self ignite (a bit like a diesel) so that there are pockets of self-ignition combustion as distinct from the intended spark-induced combustion.
This causes considerable disarray in the combustion event. The combustion is no longer ordered and stable. It often happens that a particular part of the combustion chamber will trigger a separate self-ignited flame origin. This causes a localised pressure area that pushes or distorts the remaining un-burnt mixture with such force it will wobble backwards and forwards in the combustion chamber and makes the combustion chamber contents audibly ring like a bell for an instant. While that ringing or washing backwards and forwards is happening, of course the mixture is continuing to burn. This swishing causes the burning mixture to contact much more of the surrounding metal parts than it would normally do (more metal surface area is exposed) and so it imparts far more heat by conduction to those metal surfaces. What results is a kind of runaway process. The super-hot component (valve edge, squish nose, poorly cooled patch of chamber roof or piston crown) will then likely cause another unstable combustion event, and that will add more heat to these susceptible areas. And before long, pinging/knock/detonation becomes the rule and engine destruction becomes inevitable if left alone on a highly stressed engine.
Other knock causes are also related to temperature input, such as exhaust gas contamination. Here exhaust back-pressure may cause hot exhaust gas to leak back into the combustion chamber during valve overlap and kick off premature ignition, causing knock. However poor cooling around the combustion chambers or high inlet charge are often the most sustained causes of engine knock.
If you can reduce point sources of high temperature in the combustion chamber (proper cooling methods), reduce the inlet charge temperature and reduce peak compression pressure, you can sustain higher ignition advance with a given fuel.
Engine knock and cylinder pressure
So we now know engine knock is bad and that it can lead to engine destruction, lets now look at how knock effects cylinder pressure. A few years ago we had the opportunity (and time I guess!) to install pressure sensors on each spark plug of an engine, then run it up on our engine dyno while plotting the results on a computer.
Utilising an aftermarket engine management system, we were able to induce knock on demand by altering air-fuel ratios, ignition timing and boost levels , etc. The graphs below demonstrate cylinder pressures recorded, take not of the graph with knock and how it demonstrates about 30% more cylinder pressure as a result.
Connecting Rod Failure Example
Imagine we have an engine producing 400hp, and it’s fitted with connecting rods rated up to 500hp. This would appear as an acceptable margin of overkill to spec into the engine build. If we introduce an additional 30% of stress due to knocking on that engine producing 400hp, you now have the pressure waves equivalent to 585hp put on that rod. This additional stress can result in bent connecting rods as shown below.