Two Guys Rally

Combustion Engine Theory: Intro

I was thinking the other day (rare, I know) it’s about time that I started up another interesting series. How about something that is at the heart and soul of almost every motorsport: The engine!

This is just the introduction to engines, I’ll cover the common terminology and give everyone a good starting point for understanding the finer design choices that I will cover later on.

Most automotive racing is powered by internal combustion engines. There are some variations within there to make things interesting. There are usually either two stroke or four stroke engines. I’m gonna focus on the latter, because two stroke engines are basically non-existent in rally. In four stroke engines, there are… wait for it… four strokes!

• Intake: The piston is moving downward, away from Top Dead Center (TDC) with the intake valve(s) open and the exhaust valve(s) closed. This draws in fresh air/fuel mixture to burn later.
• Compression: Intake valve closes, usually just after Bottom Dead Center (BDC), and the piston moves up to compress the air/fuel mix. Just before TDC (usually 10º-40º of crankshaft rotation), the combustion begins, either by igniting the fuel with a spark (gasoline) or injecting the fuel into the compressed air (diesel)
• Expansion: This is the stroke where the power comes from, the burning air/fuel mix generates a lot of heat and pressure that pushes the piston down, generating torque.
• Exhaust: Just before BDC, the exhaust valve opens up, and the piston moves back up to push all of the burnt gasses out of the engine. Once it reaches TDC, things start over again with the intake stroke.

The next choice is the type of fuel. Diesel engines run by compressing air a lot to generate very high temperatures, then inject the fuel, which combusts to generate pressure and heat that drives the engine. Spark ignition engines run a bunch of different fuels, usually gasoline, but can also include mixes of gasoline, ethanol, methanol, propane, compressed natural gas, and a few others. The air fuel mixture is ignited with a spark instead of relying on the sheer amount of heat in the diesel cycle.

There are also variations of the typical reciprocating piston engine, the most common being a rotary, or Wankel engine. The piston is replaced by a rotor with three faces housed inside of an oval-like housing that is technically known as an epitrochoid. Rotary engines have a very high power/displacement ratio because there are three power strokes for every revolution of the rotor, compared to the one power stroke every two revolutions of the crankshaft of a normal four stroke piston engine. However, sealing and lubricating the piston is a significant issue that hampers the reliability of this design

So now that this is out of the way, it’s time to get into the common terms used when describing or talking about engines. Let’s start at the macro level, things that everyone should be familiar with and move our way in.

Displacement: The amount of volume that all of the pistons displace in one stoke. This is dependent upon the number of cylinders, the bore and stroke.

Bore: The diameter of the cylinder when viewed from the top. The piston diameter is a small amount less than this, with rings to provide a good seal

Stroke: The distance the piston moves from TDC to BDC. This is dependent upon the crankshaft dimensions.

Compression Ratio: The ratio between the displaced volume (Vd) and the volume in the top of the cylinder (Vc) when the piston is at TDC. To calculate, CR = (Vd + Vc) / Vc.

Air/Fuel Ratio: Also known as AFR, or it’s reciprocal, FAR. Gasoline likes to burn within a specific range of ratios between the mass of air present and the mass of fuel present, typically between 8:1 to 20:1. the combustion can be considered the most “complete” when the AFR is stoichiometric (the wiki article does a better job explaining the chemistry than I ever could), 14.7:1 for pure gasoline, or ~14.2:1 for the 10% ethanol blend that almost all pump gas is now. This means that for every 14.7 kilograms of air that flows through the engine, the engine will try to supply 1 kg of gasoline. Ratios that are lower than stoich are called “rich”, and higher is “lean”. Given a constant set of parameters and optimized ignition advance, AFRs around 12.5-13 for gasoline give the most torque, because the fuel burns the fastest then.

Ignition Advance: Measured in the number of crankshaft degrees before the piston reaches TDC. Typically spark will be tuned to create maximum cylinder pressure around 14º after TDC. More advance is needed when the engine spins faster, because the burn speed of gasoline does not increase with the engine speed. However, the burn speed does increase with air density, and with AFR, with a maximum burn speed for gasoline being around 12.5-13. As such, timing is typically less advanced with more open throttle or higher boost pressure, but more advanced at higher engine speeds. Many design factors play a role in optimal ignition timing.

Volumetric Efficiency: This is essentially a measure of the amount of air that goes into a cylinder compared to how much a piston displaces. Since air is compressible, meaning the density changes with pressure, it makes more sense to think of it in terms of mass. A volumetric efficiency of 100% would imply that the mass of air that is in a piston is the same as the mass of whatever the displacement would weigh in the surrounding air. So taking the 100% efficiency example further, if it was a 4 cylinder 2.0L engine running at 100% efficiency at STP, the mass of air inside one cylinder would be equal to the density of air (1.184 kg/m³) multiplied by the volume (0.5L), the result is 0.592 grams of air. Doesn’t sound like a lot, but air is pretty light, and when you’re turning the engine at 6000 rpm, the engine is moving about 7 kg/min of air.

Mean Piston Speed: The average speed of the piston as it moves through a cycle. This is dependent upon the RPM (referred to as N in calculations) that the engine is running at and the stroke. To calculate, Sp = 2 * N * Stroke. Due to material strength and fatigue limitations, it is uncommon to see the mean piston speed exceed 25m/s or so, except in extremely high performance racing engines, like F1.

Brake Mean Effective Pressure: Commonly BMEP (or MEP when not measured at peak torque or power), this a way to measure how effective an engine is at making power in relation to it’s displacement and rpm. As a general rule of thumb, the more power you make per amount of displacement and the less rotational speed at that power level, the higher the BMEP is. Alternately, for those of you who know how torque, power and rpm relate to each other, the peak BMEP of the engine is at the peak torque of the engine. To calculate MEP, you need to know either the power and RPM, or torque, displacement, and number of strokes (2 or 4)

Calculating with power and rpm:

MEP = (P * Nr * C) / (Vd * N)

P is power, in HP or kW
Nr is the number of revolutions per power stroke, 1 for 2-stroke, or 2 for 4-stroke
C is a constant, use 396,000 for imperial (hp & ft-lbs) or 10³ for SI (kW & N-m)
Vd is displacement, cubic inches for imperial or liters for SI (61.02 CI per L if you need to convert)
N is the engine speed in RPM

Or with torque:

MEP = (T * C) / (Vd)

T is torque, ft-lbs or N-M
C is a constant, 75.4 for imperial, 6.28 for SI

Well, that’s it for the intro. I’m sure some of you have specific things that you would like me to go into detail on in this series, feel free to ask, and I’ll try my best to cover it!

Turbochargers! - Part 5

Welcome back to the turbocharger series. Today’s lesson should be short and sweet; It is going to address some misconceptions I hear about size. And no, bigger isn’t better. I know I have covered turbo sizing to an extent in the second part, but I feel that there are some things that I have seen recently that I must comment on.

The first of those is about something called “Trim”. I have seen this used time and time again as a size descriptor for a certain turbo. It is not! A single turbo has two trims, one for the compressor wheel and one for the turbine wheel. Usually it is used in reference to the compressor.

So what is trim? Well, if you have read part two, you may recall that it is a ratio between the inducer diameter and the exducer diameter, Inducer²/Exducer² to be specific. Now, take note of that. It is just a ratio, nothing more. Yes, it does change the flow characteristics of the turbo a bit, but it has no bearing on the overall flow capabilities of the turbo, nor it’s size.

With that off of my chest, there is one other insane issue that I see crop up from time to time. Compressor wheel upgrades. It is often viewed as a cost effective upgarde to rebuild an engine’s stock turbo with a larger compressor wheel, without changing the turbine side at all. In some cases, this is actually a good idea.

However, in the vast majority, the stock turbo has a smallish turbine side to produce boost lower in the RPM band. When an even larger compressor is hooked up to that turbine, some not so good things can happen. First and foremost is compressor surge. The turbine has the potential to spin the compressor too fast and generate more boost than the compressor is capable of handling at that airflow, which leads to surge, which is extremely bad for the turbine.

The second is something that I mentioned in part four: Exhaust backpressure. A smaller turbine will result in higher backpressure, reducing the overall efficiency of the engine, and increasing exhaust gas temperatures. In other words, the engine’s power is reduced somewhat because it has more trouble flowing air, and is more susceptible to damage due to the higher temperatures.

Moral of the story: Before upgrading a turbo, make sure that both sides complement each other well. Flow capabilities should be similar on both sides.

The Volvo Chronicles: New Parts!

As Charles mentioned the other day, the rallycross was pretty tough on my suspension. There were a few dips in which my suspension bottomed out going over it, and I’m sure the OEM dampers were having a tough time dissipating all the heat that the roughness was generating. Somewhere in all the chaos, the dampers broke. The rear dampers are not damping suspension travel at all, one of the front struts is making an awesome grinding noise, and the other is also not damping.

Tough, but it’s a reality we all face in rally; off-road surfaces are hard on cars, and you need durable parts engineered to take the abuse if you want to keep stuff around for more than a few races. In my case, it’s not really a big deal. I was aware that the dampers were degrading, and I have suspected them to be the same parts that were on the car when it rolled out of the factory in Sweden. Both rear dampers were leaking oil, and I’m sure the fronts were close to being in similar condition.

So I ordered some replacement parts, and decided to go more for better road handling than off-road handling, seeing as it’s a Volvo that really belongs on the highway at speed, not pitching around slow hairpins in the wet grass. I’m sure I’m not the only one, but I get really excited when big packages arrive in the mail! Looks like my kitty also gets really excited about packages too…

(click for larger picture)

H&R springs, should stiffen up the car a bit around turns and over bumps. There are also some Koni Sport dampers still in the mail that will hopefully be here tomorrow. I’m really looking forward to those, since the damping rate is adjustable, I can play around with it to see what firmness level gives the best handling. With 205-55-16r tires on the car, I’m suspecting I can get away with going pretty stiff before it becomes unreasonable.

However, over rough surfaces, stiffer is not better, since it reduces the amount of time the tires are in contact with the road when the surface suddenly changes height. The stiffer dampers will slow down the speed of the wheel significantly, but on the flip side, the stiffer springs will push harder against that damper to make the tire move faster.

There are a lot of things to consider when getting springs and dampers. Without delving too much into the mathematics behind it, a car can be modeled as a mass+spring+damper system mathematicaly. Solve a few second order equations and you can calculate things like the oscillation period, transient response, and all sorts of other neat things to try to match a spring rate and damping rate to your particular vehicle and preferred handling.

Really, there are three scenarios that occur, overdamping, underdamping, or critically damped. The latter is really hard to achieve, but it’s not difficult to get close. Here’s a relatively complicated picture for those uninitiated with system dynamics, stare at it for a while and try to make sense of it all, I’ll do my best to explain each one.

Underdamping is when the spring rate is too high in comparison to the damping rate. Most cars are underdamped, as this provides a more comfortable ride, and good traction in most conditions, despite the poor response time However, when the car is too underdamped, it will become uncomfortable and uncontrollable, as the car will be bouncing for a long time after hitting a bump. Think of your Grandfather’s old old Caddilac here.

Overdamping is the opposite, the damping rate is too high in comparison to the spring rate. This is bad because it puts a lot more strain on the suspension mounting hardware, as the bumps are barely absorbed by the suspension and is translated into chassis movement instead. This means that a moderate bump can cause your tires to be airborne for a moment. Obviously not good unless you’re racing on a really really smooth surface.

Critical damping is when the movement stops in the shortest time possible, technically the ideal balance between the spring and damping rates.

So which one is best? It really depends on a lot of things. Smoother surfaces can use higher damping rates to trade a little bit of traction for better response, while rough surfaces can do the opposite. the amount of suspension travel you want, the geometry of the suspension, the weight balance of the car, and even the driver’s preference all matter towards making the optimal setup for performance, or in some people’s case, comfort.

Personally, I am based towards critical damping, but only from a theoretical standpoint. More experience with suspension setup may change my opinion. Until then, I’ll just have to play with what I’ve got and make the most of it.

Suspension Setup Basics

I’ve heard that a few of our readers would like to know a little more about things like camber and toe, and the effects the basic suspension settings have on vehicle stability and control. Before reading this, keep in mind that the optimal setup for any combination of car and road can vary a lot. This is just a guide to help understand what three settings do:

1. Camber
2. Toe
3. Caster

There are many more variables in the suspension setup, but these three seem to be the most easily changed, and have the largest effect when tuning the car.

Camber is the angle of the wheels from vertical when viewed from the front. Negative camber means the top of the wheels is closer to the center of the car than the bottom. Positive is the opposite, with the top of the wheel further away than the bottom. The measurement is degrees off from vertical.

Usually the suspension in a car is designed to decrease camber as the suspension compresses. This way, when the body rolls as it goes through a hard corner, the outside suspension compresses and pulls the top of the wheel in, the inside decompresses and pushes the top of the wheel out, counteracting the roll from the body, keeping the tire closer to perpendicular with the road.

The main idea behind changing the camber angle is to maximize the tire’s contact patch for when you need it most. Typically it is set slightly negative to maximize traction during hard cornering. The downside is less traction when traveling in a straight line.

Positive camber causes more wear on the outside edge of the tire, while negative camber causes more wear on the inside edge of the tire.

Toe is the angle between the wheels and the car’s centerline when viewed from above or below. Toe-in means the tires point inwards, ie front of the tires are closer to the car’s centerline than the rear of the tires. Toe-out is opposite, with the front of the tires out and the rear in. The measurement is degrees off from parallel with the car’s centerline.

Toe mostly affects straight line stability and turn-in response. Toe-in improves straight line stability, negating the effects of things like surface irregularity, bumps, crosswind, and generally makes the car want to travel in a straight line.

The downside of this is that the turn-in response is reduced. Consider that the inside tires must travel through a smaller radius when turning than the outer tires. When turning with toe-in, the inside front tire will have a smaller angle of turn than the outside tire, meaning that it wants to go through a larger radius, and is fighting against the outside tire during a turn. As the weight is transferred to the outside tire, the effects of the inside is reduced.

Conversely, with toe-out, the car will be unstable at high speeds, anything that transfers weight to one side of the car will make the car want to turn in that direction because the tire is pointed outward. Keeping this in mind, it seems a contradiction that toe-out improves steering response. Remember what I mentioned before about the inner and outer tire’s turning radii. With toe-out, the inside tire tries to turn a tighter turn than the outside tire, which is exactly what we want. This way, the tires aren’t fighting against each other until the weight transfers to one side.

However, just like camber, any toe away from 0º increases wear on the tires; Toe-in causes more wear on the outside edge of the tire and toe-out causes more wear on the inside edge of the tire.

Caster is slightly more difficult conceptually, and it only applies to the steering wheels. The angle between the axis upon which the wheel turns and vertical is caster. The best example I can think of is a bicycle. The front wheel rotates about an axis that is not vertical, but is angled so that the axis of rotation is in front of the contact patch. When viewed from the side, positive caster means this axis of rotation is tilted backwards, the top is towards the rear of the car and the bottom is forward. Negative camber is when this axis is tilted forward.

What does this do? Well, when the contact patch is behind the steering axis (Positive caster), the wheels want to travel in a straight line, and will have a tendency to center when turning. As you would expect, the opposite is true with a negative caster, the wheels want to turn away from going straight and more in the direction that they are currently turning.

Negative caster was used a lot back in the 70’s and earlier to make the feel of the steering lighter, since less force is needed to turn if the wheels want to go in that direction. The problem there is that negative caster gives some instability when going in a straight line.

Almost all modern cars have positive caster to improve stability and ease of driving at speed. Although the steering wheel will be more difficult to turn, power steering helps that.

Aerodynamics: Drag

Aerodynamics is quite an interesting subject, and also one of the more complex. Seeing as I’m still learning this stuff myself, this series will be an introductory lesson on aerodynamics, I’ll just cover the basic concepts that are a good framework to understanding a lot of other important things.

An important thing to keep in mind is that aerodynamics is more or less the study of how fluids move (aka: fluid mechanics), with the fluid in this case being air, and a car’s body pushing the air out of the way. As long as you think of it as air being pushed around, the rest of the concepts are pretty straight forward.

First up is drag. Drag is a force opposing motion. In the case of aerodynamic drag, it’s the force applied against the car as it moves through the air. There are a few variables that affect the aero drag. The faster you go (velocity, or V), the more drag there will be. Also, more total surface area and frontal area increases drag. The frontal area (Af) can be thought of how much area the car takes up when viewed from the front. Or if the car is moving sideways, the side area would be used, or some combination thereof. There is a coefficient of drag (referred to as Cd) that is a function of the body shape. The final important factor is fluid density. The density of air varies with altitude, temperature and humidity, as I have pointed out in the past.

There is an equation that puts all of these things together to find the aerodynamic drag that will be seen:

Drag = (Density / 2) * Cd * Af * V²

Just make sure the measurement system is the same for all and the answer will be a force. As an example, let’s look at how much drag there is on a VW Golf GTI from the late 80’s going 80mph. I have a book here, Theory of Ground Vehicles by J. Y. Wong that has a list of different cars and their Cd and Af. The GTI has a Cd of 0.35-0.36, and Af of 1.91 m². 80mph is 35.76 m/s. I’ll assume standard temperature and pressure, so the density of air is 1.292 kg/m³.

Drag = (1.292 kg/m³ / 2) * 0.35 * 1.91 m² * (35.76 m/s)² = 552.2 kg-m/s² = 552.2 Newtons

Now for some fun with math to see what this means. Let’s convert the force and speed into power.

552.2 Newtons is the same as 124.1 pounds of force. 80mph is 117.3 ft/s, multiply the two together to get 14,556.9 lb-ft/s. There are 550 lb-ft/s in a horsepower, so this hypothetical GTI needs 26.5 hp to overcome aerodynamic drag at 80 mph. If we increased the speed to 100 mph, that number changes to 51.7 hp! Note that this is power at the wheels, and is neglecting any incline, rolling resistance, drive train resistance, and so forth that increase the power requirements at the crank.

I’m going to cover lift and downforce in a later article, but while it may seem obvious, one major cause of drag is fins and spoilers that create downforce while the car is moving. The extra turbulence and changes in airflow usually turn up as an increase in the Cd. Why is this? Well, the fins are designed to push air up as their way to get downforce. When a car is going down the road pushing lots of air upwards, there will be similar amount of drag. Let’s look at F1 cars, since they make good examples. The car is basically covered in wings that make enough downforce to allow the car to drive upside down at speeds over something like 100mph. They have to make a tradeoff when setting up for every race to balance between downforce and drag, which effectively means they have to choose whether the car can corner faster, or have a higher top speed. Rally faces a similar dilemma, although in rally, there is a lot less space for wings, and there are many slow hairpin turns where wings don’t do a lot, so the emphasis on wing setup is diminished.

So hopefully now it is obvious why when driving down the highway, you hit a speed where the gas mileage suddenly drops off really fast: The power required to overcome aero drag increases with the cube of velocity! Stay tuned for the next part of aerodynamics, and feel free to suggest topics that you would like to hear from me on!

Turbochargers! Part 4

Last time, I left off talking about how intake and exhaust restrictions should be minimized with a turbocharged setup to increase efficiency and overall power at the same boost level. I meant to cover how the engine must be changed to accommodate a turbocharger, but discussing efficiency ended up to be more involved than I thought it would be.

At any rate, there are quite a few things that must be considered, mostly to prevent the engine from going KABOOM! The things I will cover this time around are:

1. Ignition timing
2. Air/Fuel Ratio
3. Compression ratio
4. Boost control

Now, when I said KABOOM, I meant it quite literally. The primary concern with increasing the boost is knocking or detonation, which is when the air/fuel mixture explodes, instead of burning outward from the spark plug like the engine is designed for. The explosion flame front travels a lot faster than combustion, over 300 m/s compared to around 30 m/s for combustion. This causes a rapid spike in pressure before top dead center (TDC), which can be pretty damaging to the engine. Here is a handy diagram from Volvo that shows what happens when knocking occurs:

As you can see, the expanding circles represent areas that are burning. When knocking occurs, there is detonation instead of, or in addition to the normal combustion.

First, we must understand why knocking occurs. It really boils down to one thing: Excessive temperature. This high temperature can be caused by a few things, like intake temperature being too high, compression ratio too high, excessively hot cylinders and pistons. Something as unavoidable as the increase in temperature and pressure from the normal combustion can cause detonation in another part of the cylinder. Lower octane fuel also burns more readily, contributing to knocking. As quick side note, this is why most turbocharged cars recommend premium fuel.

Increasing the boost pressure increases the final pressure and temperature within the cylinder significantly, which as we now know, greatly increases the chances of knocking. In order to increase the boost a lot to make more power, we must try to prevent knock.

The first way to prevent knocking is to retard the ignition timing. The ECU usually does this on the fly based on signals from the knock sensors mounted to the cylinder block. Retarding the timing may seem counterintuitive, since waiting longer to ignite the mixture means the temperature and pressure is higher, because the mixture is still being compressed. Ignition is almost always before TDC, and ignition timing is measured in degrees before TDC, a negative value indicating that the timing has been retarded to after TDC. Once again, a nice little diagram from Volvo for visual reference:

The curve indicates pressure within the cylinder, a spike occuring after ignition, and a sudden drop-off when the exhaust valve opens at the very end of the cycle.

However, later ignition means that the hot gas from combustion stays in the cylinder for less time, reducing temperature, preventing knock. It also means that if knocking persists, the pressure build-up occurs later, while the piston is traveling downward, decreasing the intensity of the pressure spike. However, the negative side to retarding timing is reduced power output, so we want to run it as close to the optimal timing for the RPM as possible without causing knock. This will typically range from 40º to 30º before TDC, depending on the geometry of the cylinder head and piston.

Another easy way to prevent knocking is to richen the Air/Fuel Ratio (AFR). When there is more fuel present, the final exhaust temperature is ultimately higher, but the extra fuel acts as a thermal damper of sorts, since it takes more energy to heat up more fuel during compression. The temperature during compression is what is ultimately what determines if there will be knock. Making the AFR too rich will also reduce power, and cause the engine to consume a lot more fuel, a doubly bad thing to do. Despite this, many ‘high performance’ tunes will richen the mixture significantly to allow much higher boost levels to be run without knocking.

Getting into the engine itself, there is a very straightforward way to allow for a lot more boost without causing knock. Just reduce the compression ratio, it reduces the final pressure within the cylinder, preventing knock. This is why most turbocharged engines have pretty low compression ratios compared to their normally aspirated (NA) counterparts. For example, my S70 has a compression ratio of 8.5:1, while the NA version of my engine has a compression ratio of 10.3:1. Charles’ WRX has a compression ratio similar to mine, at 8.2:1, and the NA Impreza of the same vintage has a compression ratio of 10:1.

The reduced compression ratio compensates for how there is almost twice as much air in the cylinder as there would be at WOT without the turbocharger. The nice thing is that the final pressure within the cylinder is higher with a turbocharger than is possible with an NA engine, because of how the air is cooled down in the intercooler between the two compression stages. Without getting into the specifics of the math, a higher pressure usually yields a higher efficiency, meaning the engine extracts more power out of a certain amount of fuel. In theory, a turbocharger can be used to increase the fuel economy when trying to reach a specific horsepower target. In reality, turbocharged cars often get worse gas mileage due to the lower compression ratio, and tuning of the ECU for extra power over efficiency.

The other important factor that has not been discussed in detail yet is controlling the amount of pressure that the compressor makes. As I explained in part 1, the compressor and turbine wheel are attached by a shaft. To control the amount of boost the compressor makes, the speed at which it is spinning must be controlled somehow. This is usually done by letting exhaust around the turbine wheel, through something called the wastegate, instead of forcing it through the wheel. The external view of a wastegate that is integrated into the turbine housing looks like this:

(click for larger picture)

The wastegate itself is a vent hole right before the turbine wheel that allows exhaust flow into the exhaust pipe with a valve that is pulled closed by an actuator. In the above picture, you can see a rod come out of the right side of the picture, and end at a small arm. That rod and arm are connected to the wastegate and the actuator. The actuator is vacuum driven in this case, boost pressure is supplied to a solenoid that is controlled by the ECU, with two output ports, one to the unpressurized portion of the intake, and the other to the actuator. The solenoid bleeds off pressure as needed so that the actuator can be controlled by the ECU as the boost level changes.

How the wastegate is controlled changes some important factors, such as how quickly the boost pressure ramps up, if there is “overshoot”, where the turbo temporarily exceeds the target boost level, and so forth. In most stock turbo setups, the wastegate starts opening at a pressure significantly below the target boost level, causing a slower increase up to the maximum boost. This gives the least amount of overshoot, which is good for safety reasons, but bad for performance. One option is to increase the pressure required to start opening the wastegate, which will decrease spool time, but potentially creating overboost situations that may damage the engine if you’re running close to the limit of it’s capabilities.

Another choice that must be made is the wastegate type. There are some turbos that do not have an internal wastegate, and require an external one, either from something like the 5 bolt Garrett flange that has a port for an external wastegate, or by using an exhaust manifold with a tube coming off to go to a wastegate. Typically, the more air the wastegate can flow, the better control over boost pressure there is, to a point. Once there is too much flow, it is hard to have fine control over the boost levels. If there is not enough flow, boost will creep above the target level, which is not good.

So not much math this time around, although if anyone wants me to, I’d be happy to review my thermodynamics notes and explain the Otto cycle and why higher combustion temperature and pressure is better. There is always more stuff to cover on turbochargers, so stay tuned for part 5, coming soon!

Turbochargers! - Part 3

Welcome back to the Turbochargers! series, where I get to have fun rambling on about one of the most effective ways to create a lot more power from an internal combustion engine. If you haven’t done so already, I recommend reading part one and part two and grabbing a snack before continuing on with this one.

There are several things I left open ended in part two that I would like to cover this time around. First is compressor efficiency, and other things, like how having a turbocharger affects the engine itself, will follow.

Simply put, the efficiency of a turbocharger is how much work is put into compressing the air compared to how much work would have been done in an ideal world. What is different about the ideal world? Well, things like turbulence, heat transfer between the blades and the air, the effects of sound, air’s high and low pressure points within the compressor wheel, and so forth. Nothing can ever be 100% efficient, so we just try to get as close as possible. Newer turbos are generally more efficient than older ones, thanks to improvements in modeling technology, more experience in design, improved bearing tech, and stronger materials to name a few.

The efficiency of the turbo really affects two things:

1. Exhaust pressure right before the turbine inlet
2. Intake air temperature after the compressor wheel

Both of these are very important things to keep as low as possible. I’ll touch on exhaust pressure (commonly referred to as back pressure) more later, since it ties in with a few other important things regarding turbo selection and engine design. The intake air temperature is pretty obvious, the lower the temperature at a given pressure, the more dense the air is, which means more air can get into the cylinders per stroke, mass-wise. This ultimately means more power, if it is not immediately obvious why, I have gone over the effects of temperature on engine performance in more detail before.

I won’t get into detail on the theory behind the calculations involved with efficiency and intake air temperature, but if you really need to know this stuff for some bizarre reason, go do some research on adiabatic compression. For the calculation, you need to know 4 different things to find the compressor outlet temperature, which I will designate as To for temperature outlet:

1. Ti: Compressor inlet absolute temperature (ie: Kelvin or Renkin, add 293.15 or 457.69 to Celcius and Farenheit, respectively)
2. Pi: Compressor inlet absolute pressure (ie: psia)
3. Po: Compressor outlet gauge pressure (absolute works too, but you will have to modify the equation)
4. n: Compressor efficiency (ranges from 1 to 0, typically around 0.7 to 0.6)

This may seem a little messy, but it is straightforward. Plug the values into the upcoming equation and you have the outlet temperature.

We can find the pressure ratio to help us simplify the final equation, and help us relate to the compressor maps, since they are given in terms of airflow and pressure ratio between the inlet and outlet pressures:

Pressure ratio (Pr) = (Po + Pi) / Pi

The equation used for finding the compressor outlet temperature:

To = ((Ti*Pr)^0.283)-Ti)/n+Ti

So if we have a car that is running with an 80ºF inlet temperature, 14.2 psia inlet pressure, 10.0 psig outlet pressure and 70% efficiency…

Pr = (10 + 14.2) / 14.2 = 1.704

To = (((80ºF + 459.67) * 1.704^0.283) - (80ºF + 459.67)) / 0.7 + (80ºF + 459.67)

Do the math and you get To to be 665.2ºR. The units are significant here, since we did all the temperatures in absolute value due to the ratios involved, the result is an absolute value. To convert, just simply subtract the number needed to convert it back to relative, 459.67 for imperial units (Fahrenheit and Rankine) and 273.15 for metric (Celsius and Kelvin)

So the outlet temperature is 206ºF, I usually round to the nearest integer, since these calculations are hardly accurate due to the complexities involved. Either way, that is pretty warm, eh? It gets much hotter with more boost and less efficient compressors. This is what we use intercoolers for.

Many intercoolers are rated up to a certain horsepower, but I find this a rather silly notion. The calculations involved with the temperature drop across the intercooler are quite complex due to the nature of the geometry of the intercooler, and I will omit them simply because we usually don’t know things like the fin height, depth, thermal resistance between the plate and fin, and so forth. It is possible to calculate the outlet temperature based on an airflow speed through the intercooler, speed of the intercooler through the air, and a lot of geometry, but it’s still an estimation at best.

So when picking an intercooler, my advice is to use as big of an intercooler as will fit in the area you’re working with, since bigger intercoolers can remove more heat and usually have a smaller pressure drop across them, which means your turbo can do less work to get the same pressure at the intake manifold. Just remember that the more volume it has, the more air must be put into it when the boost pressure increases (read: throttle response time increases)

In a similar vein, be careful of how much tubing is used to install the intercooler. The bigger the diameter, the less restriction, which is always good, but there is more volume. To avoid excess restriction, try to use as few bends in the intake path as possible, and when you need them, use a bend with as large of a radius as will fit, since that will give the least restriction to the airflow. The whole idea with the intake is to allow it to flow as freely as possible without increasing the volume, thus lag, too much. This is something that you will have to figure out on your own, or talk to other people who have done similar modifications on the same car as yours to find their opinion on how to set things up.

The same thing applies to the exhaust side of the engine too. There are two evils with exhaust restrictions, reduced power and increased exhaust gas temperatures.

I see people say things like “This engine needs a little bit of back pressure to perform properly”, and then I end up laughing a lot. The camshaft profile was designed to create optimal torque with some specified amount of back pressure. Reducing the pressure may reduce torque, but only because that is how the cam profile is set up. Change the profile some and you will end up with more power with less pressure. I’m not going to get into cam profiles yet, since it is an area that is beyond my understanding for now. With turbocharged engines, this is not a concern at all, since the turbine creates an enormous amount of back pressure.

This pressure is created by the work needed to spin the compressor wheel, and the geometry of the turbine wheel and housing. The smaller the overall turbine assembly is, the more pressure it generates at a given airflow. This is why larger turbos tend to generate more power at the same boost level as a smaller turbo. However, as we went over in part two, a larger turbo almost always spools later in the RPM band. This means that while the peak power will be higher, the total amount of energy the engine is capable of putting down to the road is lower.

Getting back to what I was talking about before with back pressure, with a turbocharged car, it is best to keep the back pressure as low as possible, since the turbine generates a substantial amount of pressure for the engine to deal with. This pressure is not constant either. Increasing the boost increases the back pressure even more, since neither the turbine wheel or the compressor wheel are 100% efficient.

In addition to this, the turbine creates energy through the difference in pressure between the inlet and outlet of the turbine wheel. Once again, due to the nature and inefficiencies of the turbine, every small increase in pressure after the turbine wheel creates a larger increase in pressure before the turbine wheel.

Why is this so bad? Well, as I pointed out before, you can make more power with less backpressure. You may have to modify the cam profile to make full use of it, but the net result is more power, which is our goal. The other is equally important. Higher exhaust pressures increase the exhaust gas temperature (EGT) with everything else being held constant. When pushing an engine close to it’s limit, a close eye needs to be kepts on the EGTs to make sure that ridiculous things like melting a piston or warping the manifold don’t happen. Plus, lower EGT’s typically mean a longer engine lifespan, since there is less thermal stress on the parts.

So how to reduce exhaust pressure? Quite simple really, use the biggest diameter exhaust pipe you can fit into the car, straight-through mufflers are a huge plus, use as few bends as possible, and possibly most importantly, the part known as the downpipe must be capable of supporting the airflow.

The downpipe is often the most restrictive part in the exhaust after the turbo (known as the turbo-back, all the parts after the turbine housing) aside from the mufflers, because the exhaust is the hottest in that part. Hot air means low density, which means a high volume for the same mass. This low density creates a high airflow velocity, and drag increases exponentially with velocity. Just like the rest of the air stream, try to ensure that the downpipe has a large diameter, smooth bends, a smooth interior surface (roughness causes more turbulence, which almost always increases the resistance to flow), and the turbo will thank you.

Well, I think that’s enough for this time around. I didn’t cover quite as much as I wanted, but the topics I did cover were gone into a lot of detail, which is good. On the plus side, I already have a few ideas for part four. I always welcome comments, questions or suggestions, so feel free to ask and I’ll do my best to help you out.

So what is everyone’s personal motto for the next month? Less restriction is better!

Electromotive Sequential Transmissions

The standard transmission, a.k.a. manual, is found in most cars with an H-Pattern mechanical gear selector. But there are faster and easier ways to select gears. What I would like to focus on in this post is the Electromotive Sequential Transmissions.

Sequential Transmissions

First I need to explain what the difference between a standard transmission and a sequential transmission is. In the standard transmission it is possible to select any gear at any time (given enough force). However in a sequential transmission you must either select the next or previous gear from the one the transmission is using. For example, if I were in 3rd gear I could either choose 2nd or 4th. It would be impossible to skip to 5th or 1st without going through the next in the sequence, hence their name ’sequential’. I am told that this can make the mechanical workings of the transmission simpler (see motorcycle transmissions) but I wont even attempt to explain it (ask Mark).

Motorcycles are the vehicle most commonly associated with sequential transmissions. Their gear selector has 3 positions: Up Shift, Down Shift and No Change. the ‘No Change’ position is the default position. If we transition the sequential transmission into a Car nothing changes. There will be 3 positions for the gear selector, but we may make the transmission a little more complex in order to reduce weight.

Electronic Gear Selectors

Mechanical gear selectors are heavy, and depending on how they are constructed may make it harder to select gears, so we can make the gear selector electronic. That will save on weight (what is used to shift is a few thin copper wires) and allow us to place the selector anywhere in the cockpit. It also has the added benefit of reducing driver fatigue as it will be physically easier to select gears. It is not just as simple as saying “Let’s make it electronic” and then we’re done: we have to get the force to select a gear from somewhere. That somewhere can be a few places. We can use compressed air to operate a pneumatic arm that controls a short mechanical linkage on the transmission. Perhaps, if we have enough electrical current coming from the alternator we can use that to power a solenoid to operate a short mechanical shifter (on the transmission).

Both Electro-Pneumatic (Electro because the gear selector is electronic) and Electro-Mechanical Sequential Transmissions have their downsides. The Pneumatic variety require compressed air to be stored in the car, but this could be lighter than the alternate Electronic Solenoid approach. However refilling or punctures could make a less competitive race car or even end a race day. While the Electro-Mechanical types have much more weight associated with the system and are slower to respond (shift time matters!!!).

WRC vs Rally-America

Electro-whathaveyou Sequential Transmissions are a common sight in the WRC, but the H pattern is the only transmission allowed in the rules of Rally America. And that brings me to this: Why does Rally-America disallow Electromotive Sequential Transmissions? Is it a strategy to make Rally Racing cheaper and more accesible? Is it to keep the cars closer to stock so fans can say: I DRIVE THAT! ? I’d like to hear your opinions on this and I would like to go into further detail about Electro-Pneumatic Transmissions in the future.

Engine Control Units (ECUs)

ECUs are black boxes that make your fuel injected engine run. They are commonly microcontrollers that take some inputs in, compare those to a table and spit out an output or two.

What are those inputs? Basic ECUs look at only a couple things, including: Air available (MAF/MAP) and Engine RPM. The Air Available is related to the throttle position. The Engine RPM equates to how long the valves are open for and combined with the air available the computer can figure out how much air is entering the cylinder. With that key piece of information the ECU can then figure out its output.

What is the output? We are still considering the basic ECU here. It has one output, and that is the amount of fuel it should spit into the cylinder/manifold. Really, it is even simpler than that. The output is just how long the fuel injector fires for.

But of course we can make the ECU do way more than manage a single fuel injector. Now they commonly control 4+ fuel injectors and at least the spark plugs. We could also slap on a boost control unit for any turbocharger that is in the car. While we are at it let us add variable valve timing to the mix and control that with the ECU. We could even make it more accurate by adding a bunch of inputs…like:

• Oxygen sensors on the exhaust
• Many temperature sensors (intake, post turbocharger, engine block, oil, exhaust)
• Throttle Position sensors
• Boost Pressure sensors
• Fuel Pressure sensors
• Valve sensors
• etc…

All of this in a pursuit to more accurately control the amount of fuel we inject into each cylinder at a given time. Some of the inputs determine how much air is available, while others are guessing at how much power the driver really wants. And one important input is watching to see how wrong the ECU’s previous decision was, so it can self correct (hint: we breath what it looks for).

However, what I find funny is that all the ECUs I have ever seen achieve this by looking up fuel injector times in a table. What takes a human mind ages to do succesfully a computer can do in practically an instant. But with every input we add a dimension to the tables. This makes me wonder what the future of the ECU is. Will we start making simple mathematical formulas to take these inputs and quickly produce an output? Maybe we will augment the result from the table with some formulas, but I don’t know yet.

What I do know is that, as computers get faster and more robust, so do automotive microcontrollers. This will allow us to continually add to the giant black box we call the Engine Control Unit. The more we add to it, the better off cars will be. We can give even more control to the driver (not likely) and allow them to choose economy or performance, or maybe even torque curves and rpm limits (Variable Valve Timing).

Hopefully I was able to get you somewhat interested in ECUs. Maybe I’ll be able to start a series on ECUs much like Mark’s Turbocharger series. Perhaps even one day, work on my own custom ECU for a rally car.

Turbochargers! - Part 2

In part one, I went over what a turbocharger does in relatively basic terms. This article will expand on the concepts I explained there, so if something does not make sense, read of it again! There is a lot to understand when it comes to designing a turbocharged engine, or selecting a specific turbo for an engine that you already have. The latter is undoubtedly more important to us, since we don’t have the resources to design an engine from the ground up.

So what things do we need to keep track of? Well, there are a few important things that take precedence over all others, and can depend on the goals of your design:

1. Air flow supported by turbocharger at different boost levels
2. Compressor trim
3. Turbine A/R

I’ll go over trim and A/R before airflow, since the pieces will fit together in a more sensible manner that way.

To understand trim, we need to know what the inducer and exducer are on the compressor wheel. Both are pretty straight forward. The inducer size is the diameter of the compressor blades at the inlet. The exducer is, as you guessed, the diameter of the compressor blades at the outlet. The opposite is true for the turbine wheel, since the exhaust gas flows in at the edge and out through the center.

We can calculate trim with a simple equation: (Inducer)² / (Exducer)² * 100

So for an Inducer of 61mm and an Exducer of 82mm, the trim is 55. As a general rule of thumb, the larger the trim, the more air a turbo can flow, while smaller builds boost earlier, assuming that all other things are held equal. That said, it’s easy to change things to make a smaller trim turbo flow more than a larger trim, since the trim is just a ratio, not an actual size descriptor.

The Turbine A/R is equally straight forward, it’s the ratio between the Area of the turbine housing inlet to the Radius of the turbine housing.

Pretty simple, just divide the area by the radius to get A/R. A turbocharger with a smaller A/R will generally spool sooner and faster than one with a larger A/R, with everything else held constant. Again, like with the trim, it’s a ratio, so a larger turbo with a small A/R may spool slower than a small turbo with a larger A/R.

Now that we have those two out of the way, the air flow is pretty straight forward. Most turbocharger manufacturers have “flow map” that show how much airflow the turbo can support at various boost levels, along with the approximate wheel rpm and overall efficiency. For example, the relatively small Garrett GT2860R’s flow map looks like this:

It may be a lot to look at for the first time, but it’s actually pretty straight forward. Along the x-axis is the mass air flow, and going up in the y-axis is the amount of boost that the turbo can make at that airflow. The lines that start out horizontal and curve down to the right is the rpm the turbine and compressor wheels would be spinning at to make that amount of boost at that air flow rate.

Go past the right edge of the map and you get into a pretty inefficient region of the turbo, meaning that the compressor wheel will be heating up the air a lot when compressing it. The left edge is called the “surge line”, and you want to avoid running the turbo in this region at all costs, as it is very bad for the turbo and will lead to a premature failure. More on that later.

The y-axis’s label is exactly what it means. The pressure ratio is how much higher the output air pressure is than the inlet pressure. If the inlet of the compressor sees 14psia, at a pressure ratio of 2, the outlet will be 28psia, which is 13.3psig of boost at sea level. This is where it is important to keep the difference between relative and absolute pressure in mind. Relative pressure is the pressure above the ambient pressure, which is usually around 14.7psia (sea level). Absolute pressure is exactly that, the actual pressure created. Absolute pressure cannot be below 0, ever.

The airflow can be calculated pretty easily, though there are a few assumptions we’ll make to simplify the process.

1. “Standard” atmospheric conditions, meaning air density is 0.0749lb/ft³, or 0.002645lb/L
2. Engine volumetric efficiency is constant at 95%, meaning the throttle valve is fully open.
3. Pressure losses in intercooler, air hoses and air filter are negligible. (They aren’t in reality)

Now all we have to do is multiply a couple things:

Mass Air Flow = Air Density * Engine Displacement * (RPM/2) * Volumetric Efficiency * Pressure Ratio

For example, a 2.3L engine at 6000RPM with 10psig of boost:

Pressure ratio = (10psig + 14.7psi) / 14.7psia = 1.68

0.002645lb/L * 2.3L * (6000RPM / 2) * 0.95 * 1.68 = 29.12lb/min

So how does this help us interpret the flow map? Well, here is where 29.12lb/min and 1.68 pressure ratio lies on the map:

As you can see, it’s outside of the efficiency range of the turbo. While it does appear that it can flow that much air at that pressure, it will be heating it up excessively in the process, putting a large demand on the intercooler, and increasing the backpressure in the engine to excessive levels. A bigger turbo would be preferred, but we have to be careful how big we go. Two obvious things, and one not so obvious. First, the bigger the turbo, the more air it can flow. Second, bigger turbos take more time and airflow to spool up. Third, and possibly the most important, but often overlooked, is that we need to make sure we don’t go past the surge line of the turbo. We can prevent this when the throttle is suddenly closed by using a compressor bypass valve or blowoff valve, but we must make sure that the turbo isn’t so large that the engine causes the compressor wheel to hit the surge line under moderate load (low air flow) with high boost pressure.

We know now that the GT2860R is too small for our example engine, so let’s move a step up in the Garrett turbo family to the GT3071:

This might actually be a little bit too much turbo, but leaves the option for us to increase the boost a whole bunch without running out of breath. We’ll just have to be careful not to hit the surge limit at lower rpm if we do so. Careful boost control with the wastegate and electronic boost controller will work well.

Speaking of wastegates, there are two very important things used to control the pressure ratio and airflow of a turbo to prevent overboosting and crossing over into the surge area. The wastegate is one of them, the compressor bypass valve (CBV) and blowoff valve (BOV) is the other.

What the wastegate does is allows exhaust gas to flow around the turbine wheel, instead of through it. This allows the engine to flow a lot of air without forcing the turbo wheels to spin at full speed all the time. Typically what happens is the wastegate is closed when below a certain boost level, and slowly opens past that level to control the amount of pressure the compressor wheel generates. This also reduces backpressure on the engine a bit, which is always a good thing.

The CBV and BOV are on the intake side of the engine, and are meant to prevent the compressor from surging. They do the exact same thing, with one minor difference. The BOV vents air to the atmosphere, while the CBV puts the air back into the intake, after the mass air flow sensor and before the compressor. What happens is when the pressure gets above a preset value, the valve opens to allow the compressor to maintain the air flow it was doing before. Typically what happens is the throttle valve will close suddenly while turbo is making a lot of boost with lots of airflow (ie: upshifting during hard acceleration). The closure of the throttle basically drops the air flow to 0lb/min. Look at the compressor map, and note how this is left of the surge line, definitely a bad thing. The CBV and BOV allows air to keep flowing during this time to prevent surge. Then the boost will decrease or engine airflow will increase again, at which point the valve will shut. There is a very important thing I would like to note about CBVs and BOVs. Never use a BOV on an engine that uses a mass air flow measurement (MAF) system. When you vent air to the atmosphere at the BOV, it has already gone past the MAF sensor, and the ECU thinks that this air will be going into the engine. As a result, there will be fuel injected for that amount of air, when it is really less. This leads to an excessively rich condition in the engine. This definitely not an optimal solution, for a whole bunch of reasons, from how excess fuel decreases performance to the damaging effects on the catalytic converter when the engine runs too rich. Since the engine only runs rich while the BOV is venting, it’s not the end of the world.

Now you know a little more about turbochargers, and what you need to do in order to select the right turbo for your applications. There is still more to cover though, such as more in depth calculations, what to do with the AFR and ignition timing when boost increases, why compression ratio must be decreased to allow for more boost, and so forth, so stick around for part 3.