The benefits of injecting nitrous oxide are well known and documented. Thanks to the Fast and the Furious now every teenager knows that if you press that red button on the steering wheel your face grows longer, your arms stretch as you hold on for dear life and then the floorpan breaks off as the inlet manifold bolts explode (or something like that).

The terms "NOS" and "Nitrous Injection" are commonly used interchangeably, although technically"NOS" is an American trademark for Nitrous Oxide Systems, as know-alls hasten to point out. Here I use it as an acronym for a Nitrous Oxide System. Next thing the Yanks will trademark the word water (along with our genes) and we'll have to find alternatives to those as well...

This particular system is British and belongs to TFS
The solenoids are all double, because this car runs two separate systems with their own bottles and controllers (not shown here). They can potentially flow gas with enough oxygen content for 600bhp (assuming there's enough fuel to match it up and strong enough engine internals!)
It's not a show-car, no half-naked models lurking around - it's more of a development vehicle, a test-bed.

Below are pictures of the solenoids, sitting in the engine bay's relay-box. By convention, red if fuel, blue is nitrous. (Click for larger pics)

Here is a close-up of the nozzles. Note the angle relative to the airstream. In this installation they are just before the 'top hat'

 

Another view of the nozzles and the fuel pickup. Note that fuel is taken directly from the fuel rail, tapping into the 'schraeder' valve. This is like a bicycle valve, usually sitting there with a protective black cap. It can be used to quickly depressurise the fuel system. Just press it and fuel squirts out, straight in your face. Also used to connect fuel pressure gauges. Here it's being put to more serious use.

 

From this angle we can see another addition to the end of the fuel rail. An extra fuel pump is used, along with its own fuel filter (not shown here)

Despite the ducktape/cabletie appearances, this engine pushed a 1.35 tonne car to a 12.8 at the Pod, while burning more than it's fair share of oil

The progressive controller. When running 50+bhp shots it's essential. This system can flow 150+ so it's a must. Rate of delivery, delay etc are all in the console.

A close-up of the throttle position sensor. Note the extra wire sticking out, it notifies the controller that full throttle is achieved and the fun may begin. This is the safe and easy way to activate nitrous. No fancy red buttons and computers shouting 'abort' and 'self destruct' (better leave those to Hollywood)

Last but not least, there is also the big bottle in the boot, perhaps hidden underneath the ICE. The bigger the bottle, the longer it lasts (naturally) but since it's only used during full throttle, even a few minutes of operation can seem enough (well, it's never enough, who are we kidding now?)

But how often have you kept full throttle for more than ten seconds? Be honest now...

 

The feeling you get is one of urgency. Off-boost and during low-throttle operation it's exactly the same as any other boosted LET. This is expected since everyday running has not been compromised in any way.

Lag is noted by its absence - it just isn't there. These engines are not laggy anyway, but there's always some delay in the power delivery which some people describe as lag (even if it isn't technically). NOS fills this gap and masks the rest. You hit full throttle and there's an immediate rush of power. It's not sudden or uncontrollable, but smooth and progressive. In cases where it would take the turbo a couple of seconds to spin, it only takes it about half that (the extra exhaust gases force the turbo to spin faster), and even during that second the nitrous fills in the gap rather nicely.

It's very hard to tell that there's nitrous just by driving such a car. Revving at high boost seems so effortless and immediate, it's like having a HUGE turbo that is as flexible as a tiny one though. Contradiction in terms, as you normally have to choose one or the other. There are no clicks and pops, no buttons to press - nothing at all. Just a flow of power during full throttle, exactly when you need it.

Thanks to The Free Spirit for the pics.

 

Nitrous-specific fuelling and ignition

Nitrous injection creates unique fuelling and ignition requirements, none of which can be 100% satisfactorily resolved, to the best of my knowledge.

Fuelling is the constant attempt to keep injecting adequate fuel to compliment the extra oxygen that will be released by the gas in the chambers. A main problem is that metering the flow of nitrous is not an exact science, least because the pressure does not stay constant as the bottle empties. Also gas bubbles may occasionally appear in nitrous lines that are excessively long, wide or hot. These can dramatically reduce the amount of extra oxygen, and if fuelling doesn't adapt accordingly (which it may not) then we end up with mixture that is far too rich. This can foul the plugs, creating misfires, that allow unburnt gases in the exhaust. The oxygen sensor sees this as excess oxygen, indicating 'LEAN'. This gotcha alone can throw off the uninitiated, who will think Ah, I'm running lean, that's why it misfires, let's throw in some more fuel

Even worse is the opposite scenario, where fuelling is restricted under full load, so there are not enough fuel molecules to pair with all the oxygen released by nitrous. Running lean under full load is a sure way to blow an engine. Again, maybe in simulated tests fuelling seems fine under load, but in real life it could be that the limits of the fuel pump are exceeded, or the fuel pipes, or the fuel rail, or the electric system that is suddenly taxed with running the injectors at full blast, plus NOS solenoids, plus any other auxiliary pumps and bells and whistles. Since fuel delivery is heavily dependent on fuelpump voltage, losing fuel flow at that moment is the most likely time to happen, and the one that will hurt the most too.

To make matters even worse, the accuracy of oxygen sensors (both narrowband and wideband) is questionable while lots of nitrous is being injected. Try getting the fuelling right when you can't even measure it properly!

Ignition can be yet trickier! Nitrous mixtures will typically be denser than normal (it's not called chemical supercharging for nothing!) Therefore even if AFR is kept continously at optimal levels (say 11 - 12:1), then the more nitrous is injected, the higher the speed of burn will be. Quicker burning means that we might end up pushing the piston downwards, while the crank is still pushing it upwards. Result: Ouch, that's gotta hurt.

The good news (in a sick sort of way) is that many nitrous installations have botched fuelling, compromised towards too rich or too lean. Either way, the suboptimal fuelling slows down the speed of burn, effectively cancelling out the accelerated burn that the increased density would create. Think about this: so many nitrous 'afficionados' get away without retarding their ignition, or retarding it by just a few degrees. Since they don't blow anything, they wrongly believe that the whole idea of NOS demands retard is an old wives' tale. So they add bigger gas shots (it's very addictive) and then even bigger, until they blow up the whole lot. Then they blame the pistons, the rods, the fuelling, what have you. In fact, the funny thing is that getting the fuelling right would have blown their engine way earlier.

On high-boost turbos we also have the added complication of the extraordinary charge-cooling effect of NOS, as it changes state from liquid to gas. This alone negates some of the need for retard, but people think that if some is good, more is better. Unfortunately this charge-cooling effect can become a liability on large shots where the charge temps are already down to ambient, hurting fuel atomisation and creating a lean condition. On n/a engines this reversal of the cooling benefit happens even sooner, especially in the winter months. Sub-zero air just will not hold the fuel droplets suspended.

Large nitrous shots (relative to the engine's capacity and activation revs) always demand a different ignition curve. The higher the density of the mixture the faster it burns - however also take into account the effect of AFR on the speed of burn. If you wonder how some people get away with running large shots of nitrous without retarding the stock ignition, take a look here

We're talking in the lines of extra 10 degrees at midrange and a bit less at high revs. In fact they also demand different spark profile, as well. It is a fallacy that you can have the bone-stock engine for everyday abuse, and add 100-200bhp nitrous shots whenever you feel like it. Either the ignition will have to be permanently retarded for NOS, compromising everyday full-throttle response, or the ignition will have to be stock and keep the nitrous shots small, so that the ignition retard needs are low and just 'eat away' the stock car's safety margins. As we saw above, just don't depend on the stock knock sensor to get you out of trouble because it may be deaf on this one.

The best compromise is to have a standalone management system with different ignition curves for n/a and nitrous operation, or at least an ECU interceptor retarding whenever you feel like going crazy with the gas. There is no one-size fits all, don't believe the hype from the resellers of nitrous systems, they merely try to make sales, and the easier a system looks, the more it will sell.

 

Health & Safety First:

These bottles are under extreme pressure inside. If they overheat (car left in the sun for example) this pressure can go well over 1000psi. The safety valves are not infallible either, so they can ruin your day

...ooops
ooops again
and again...
Yank article on the subject:
Bottle wide open. Scary.
Cleaning up the mess. Let's hope he learned his lesson...

Here is a good link explaining some nitrous facts. (Local copy here)

If money's a bit tight and you're feeling kinda lucky, here is someone DIYing the lot
(local copy here)

 

 

Nitrous oxide is thought to have been first used by the Germans during WWII
in their Messerschmitt Me-109F fighters. It was used to increase low altitude flight speeds on the Russian front but it was later determined that these superior planes didn't really need this sort of performance boost. The German airforce later shifted the development of nitrous oxide systems to reconnaissance aircraft, which needed it to fly fast at high altitudes over Britain. The British also used nitrous oxide on the Mosquito twin-engine light bomber whose only defence was not guns but acceleration. Officially, US planes did not experiment with nitrous oxide due to their overall superiority against all adversaries.

After the war, jet engines were at the forefront of aviation technology and there was no longer a need for continued development of aircraft nitrous oxide systems. Rumour had it that a few American pilots returning from Britain were also avid race car drivers, and a few of them experimented with nitrous in automobile engines. Unfortunately, nitrous oxide is very inefficient when it comes to power vs. mass, and a form of racing did not exist that could really make use of it-- until drag racing was born.
Granted, a few people like Smokey Yunick used it to post favourable qualifying times or track lap records, or during passing at a critical moment in the race, but nitrous proved to be so effective that they eventually banned its use in most forms of racing.

Nitrous produces horsepower in either three or four ways, depending on whether it enters the engine as a liquid or as a gas. Here's what happens:

1) The nitrous oxide is stored in a tank under pressures approaching 1,000 PSI to keep it in a liquid state. When the engine or driver triggers the nitrous, solenoids open up to provide the engine with a carefully metered amount of both additional fuel and nitrous oxide (called a wet system) or just nitrous oxide (called a dry system). The nitrous oxide flows from the tank into the engine's intake system, either near the throttle plates (via an injector plate) or directly to the intake ports a few inches from the intake valves (called a "fogger" set-up). Upon injection, the nitrous oxide undergoes a rapid decrease in pressure-- from 1,000 PSI to atmospheric pressure. As the liquid nitrous oxide warms up to -129 F it undergoes a phase change and converts to a gas (evaporation).
A substance that undergoes such a phase change releases a great amount of heat, just as water must absorb a great deal of heat energy to turn itself into steam). This is called "latent heat" because although energy is being released during the phase change, the temperature of the substance does not change. This cools the air/fuel charge even more.

2) The gaseous nitrous oxide is now at -129 F, while the incoming air/fuel charge is obviously still much warmer than that. Like step 1, the result is continued cooling of the incoming air/fuel charge.

FACT: steps 1 through 3 are responsible for almost HALF the horsepower gains on an engine equipped with nitrous oxide injection, with step 2 contributing the most.

3) Nitrous oxide consists of two nitrogen atoms and one oxygen atom, and is 36.35% oxygen by weight. As mentioned in step 1, the goal of any high performance engine design is to get as much air and fuel into the cylinders as possible. Turbos and superchargers do this by physically compressing atmospheric air and forcing it into the engine, while nitrous oxide carries the additional oxygen in initially as a liquid (which is about 300 times as efficient). A wet system will inject a carefully metered amount of additional gasoline into the air stream when the nitrous is triggered, thereby ensuring that the proper air/fuel ratio is maintained. A dry system accomplishes the same task by ordering the engine's fuel pressure regulator to increase fuel pressure, thereby increasing the flow of fuel. Wet systems are usually considered safer and more dependable that dry systems.

4) Even though nitrous oxide is 36.35% oxygen by weight, that oxygen is still chemically bonded to the nitrogen atoms. Chemical bonds are one of the strongest bonds in science. Lucky for us, however, the chemical bonds in nitrous oxide molecules are exothermic, which means that the bonds release energy as they are broken.
As the existing air/fuel mixture is ignited and begins to burn, the temperatures within the burning mixture exceed 572 F. At that point, the chemical bonds within the gaseous nitrous oxide break down. The released oxygen combines with the extra fuel that was injected with the nitrous oxide, while the nitrogen supposedly (?) helps to alleviate detonation. The energy released when the bonds break is in the form of heat-- heat = pressure, pressure = work, and work = horsepower. In layperson's terms, the extra released heat pushes down on the piston.

So there you have it. The injection of liquid nitrous oxide into an engine helps provide a substantial increase in horsepower by

1. cooling the intake charge as the nitrous oxide converts from a liquid to a gas,
2. continued cooling of the intake charge as the nitrous oxide warms up from its initial sub -129 F
3. adding extra heat to the combustion process within the combustion chamber as the nitrous oxide breaks down into nitrogen and oxygen
4. as the released oxygen combines with the excess fuel inside the combustion chamber.

 

If the pressure within the storage bottle is too low, only gaseous nitrous oxide will be released into the engine and the benefits of the phase change will not be experienced.
"Surging" may occur when the bottle is less than one-quarter full. The NOS/Fuel is unlikely to be correct then either.

We mentioned in step 4 that the released nitrogen supposedly acts as a deterrent to detonation. According to my references this hasn't been fully explained or understood (theory would state that the opposite should be true). One idea suggests that the free nitrogen slows down the flame front inside the combustion chamber, providing a "smoother" burn that is much easier on pistons, rings, bearings, etc. The nitrogen is also thought to lower exhaust valve temperatures-- a benefit in any engine. Both of these theories have not been proven as such, although engine teardowns have revealed bearings and rings that were still it fine shape, while EGT gauges have detected a 75 F drop in exhaust valve temperatures. Likewise, testing has shown that engines have experienced power increases of over 40% before detonation sets in.

There is also some speculation as to where the nitrous oxide should be introduced in the intake system. By injecting it near the base of the carburettor or throttle body, the nitrous oxide spends more time suspended in the intake charge, thereby cooling it to a greater degree. However, tests have shown that this method often cools the fuel on a carburetted or TBI set-up so much that it won't fully vaporise. This causes the fuel to remain out of suspension and some cylinders may receive more nitrous oxide and/or fuel than others (note that this is not a problem on port injected fuel injection systems). By placing the nitrous injection ports closer to the intake valves, the fuel remains in suspension. However, less cooling takes place since the nitrous oxide does not exist within the intake system for as long. All is not lost, however, since this also means that much of the nitrous oxide will enter the combustion chamber in liquid form. As hinted at earlier, liquid nitrous oxide is about 300 times more dense than vaporised nitrous oxide, which translates into much more oxygen injected into the combustion chamber. Cheaper kits introduce the nitrous oxide near the throttle body because usually only a single nozzle or injector plate is required, as opposed to the more expensive fogger systems that utilise separate nitrous oxide and fuel jets for each intake runner.

Without taking the nitrous oxide system into consideration, the engine can be designed with dependability, drivability, and fuel economy in mind. The engine will idle smoothly, get good gas mileage, and easily pass MoT emission testing. Such an engine will also be significantly cheaper than a more powerful engine built to the same level of dependability (or if that isn't possible, *two* powerful engines of greatly reduced dependability).
Expensive carburettor and head porting isn't as big of an issue as it is on a normally aspirated non-nitrous engine because the extra oxygen in the nitrous oxide is being delivered initially as a liquid.

If an engine running on nitrous oxide were to suddenly go lean, however (which is more likely on a dry system than on a wet system), look out. Without the extra fuel the engine will run dangerously lean, and will melt pistons if you are not careful. Nitrous oxide already increases the combustion chamber temperatures when used, which is potentially dangerous on an high-compression turbo engine. As mentioned above, there is strong evidence that the free nitrogen helps to keep the exhaust valves running cooler, while promoting a "smoother" burn within the combustion chamber.

Reliability-wise, the theory that the free nitrogen acts to slow down the flame front within the combustion chamber means that the internal parts of the engine are subjected to much less peak stresses. If you plan on using nitrous quite a bit, larger exhaust valves should be used in conjunction with special nitrous cams. A large bore header would be a good idea, as would be running spark plugs one (or even two) heat ranges colder. Nitrous oxide and fuel pressure gauges would be a must, as would an air/fuel gauge and an exhaust gas temperature gauge.

If you're really worried about the extra heat, water injection could be timed to operate in conjunction with the nitrous oxide.

Nitrous Oxide

* Molecular weight : 44.013 g/mol

Solid phase

* Melting point : -91 °C
* Latent heat of fusion (1,013 bar, at triple point) : 148.53 kJ/kg

Liquid phase

* Liquid density (1.013 bar at boiling point) : 1222.8 kg/m3
* Liquid/gas equivalent (1.013 bar and 15 °C (59 °F)) : 662 vol/vol
* Boiling point (1.013 bar) : -88.5 °C
* Latent heat of vaporization (1.013 bar at boiling point) : 376.14 kJ/kg
* Vapor pressure (at 20 °C or 68 °F) : 58.5 bar

Critical point

* Critical temperature : 36.4 °C
* Critical pressure : 72.45 bar

Gaseous phase

* Gas density (1.013 bar at boiling point) : 3.16 kg/m3
* Gas density (1.013 bar and 15 °C (59 °F)) : 1.872 kg/m3
* Compressibility Factor (Z) (1.013 bar and 15 °C (59 °F)) : 0.9939
* Specific gravity (air = 1) (1.013 bar and 21 °C (70 °F)) : 1.53
* Specific volume (1.013 bar and 21 °C (70 °F)) : 0.543 m3/kg
* Heat capacity at constant pressure (Cp) (1.013 bar and 15 °C (59 °F)) : 0.038 kJ/(mol.K)
* Heat capacity at constant volume (Cv) (1.013 bar and 15 °C (59 °F)) : 0.029 kJ/(mol.K)
* Ratio of specific heats (Gamma:Cp/Cv) (1.013 bar and 15 °C (59 °F)) : 1.302256
* Viscosity (1.013 bar and 0 &degC (32 °F)) : 0.000136 Poise
* Thermal conductivity (1.013 bar and 0 &degC (32 °F)) : 14.57 mW/(m.K)

Miscellaneous

* Solubility in water (1.013 bar and 5 °C (41 °F)) : 1.14 vol/vol