Turbo Fundamentals by Dennis Grant
Here is a collection of post's made to the D.S. mailing list by a Chrysler Employee about turbocharging...
How Turbos Work (or: The Closest Thing to a Free Lunch)
Before we start, we have to take a second to review a little grade 10 physics - The Ideal Gas Law. In short, gas temperature, pressure, and volume are all related. Compress a gas (reduce the volume) and pressure and temperature goes up. Let it expand, and temperature and pressure go down. Increase the temperature, and the pressure goes up (in an enclosed space) or the volume goes up (it expands). Finally, gases want to flow from a high pressure area to a low pressure area, and the greater the difference, the bigger the push. (Pop a balloon, little bang. Pop a welding O2 cylinder, big bang) OK, a 4 stroke engine produces work by expanding a gas in a confined space where the high pressures created can push against a piston. Furthermore, that gas is heated by the process of creating it (unlike a steam engine) so you get even higher pressures - and more power. Unfortunately, most of that heat (which is the same as energy) is dumped overboard in the exhaust before we get any chance to use it. It's just not in the cylinder long enough to transfer all that heat into mechanical energy, and it's not practical to make cylinders "tall" enough to extract every last bit of work from that hot expanding gas. So, what can we do about it? well, we can point the tailpipes out the back to try and get some thrust - except that aside from some very rare circumstances, the gas volume isn't high enough to get any worthwhile push. (A few older Indy Cars actually created a couple of pounds of thrust from their exhausts, but that's not enough to be really useful) OK, how about sticking some sort of auxiliary engine in the exhaust flow? Steam engines did this for years... Enter the turbocharger, a turbine fed by exhaust gasses, connected to a compressor via a shaft that compresses intake air into the engine. More air in the cylinder means more fuel can be burnt per power stroke, more burnt fuel means more hot gas, more hot gas means more power - and more boost too. This is the closest thing to a free lunch you'll find in engineering, because you're taking heat (energy) that would otherwise be wasted and getting usable work out of it, with almost no tradeoffs. You gain a little complexity, and added manufacturing costs, but there is no real performance hit from adding a turbo. "But doesn't the turbo increase exhaust backpressure?" Under boost conditions, no. Here's why: when the exhaust valve opens, the pressure inside the cylinder is much much higher than the pressure at the turbo inlet. That cylinder pressure "blows down" very quickly, but we're on the exhaust stroke - the cylinder volume is decreasing very rapidly, and from the Ideal Gas Law, that tends to keep the cylinder pressure higher than the turbo inlet pressure. Finally, when the exhaust stroke is nearly done, and the pressures are nearly equal, the intake valve opens, the intake pressure (we're under boost here!) "blows down" into the cylinder, and presto! we have a higher cylinder pressure again. (I'll discuss backpressure - I _hate_ that term, it's misleading - in greater detail in a later post) That's enough bandwidth for today.
Alright, yesterday we determined that a turbo was a device that could be used to get useful work out of otherwise wasted energy, to day we'll discuss how that happens in more detail. It is a common misconception that the exhaust turbine half of a turbo is driven purely by the kinetic energy of the exhaust smacking into it (like holding a kid's tow pinwheel behind your tailpipe) While the kinetic energy of the exhaust flow does contribute to the work performed by the turbo, the vast majority of the energy transferred comes from a different source. Keep in mind the relationship between heat, volume, and pressure when we talk about gasses. High heat, high pressure, and low volume are all high energy states, low heat, low pressure, and large volumes are low energy states. So our exhaust pulse exits the cylinder at high temperature and high pressure. It gets merged with other exhaust pulses, and enters the turbine inlet - a very small space. At this point, we have very high pressure and very high heat, so our gas has a very high energy level. As it passes through the diffuser and into the turbine housing, it moves from a small space into a large one. Accordingly, it expands, cools, slows down, and dumps all that energy - into the turbine that we've so cleverly positioned in the housing so that as the gas expands, it pushes against the turbine blades, causing it to rotate. Presto! We've just recovered some energy from the heat of the exhaust, that otherwise would have been lost. This is a measurable effect: Stick an EGT upstream and downstream of the turbo, and you see a tremendous difference in temperature. So, in real world terms, what does this tell us? All else being equal, The amount of work that can be done across an exhaust turbine is determined by the pressure differential at the inlet and outlet_ (in english, raise the turbo inlet pressure, lower the outlet pressure, or both, and you make more power) Pressure is heat, heat is pressure. Raising the inlet pressure is possible, but tough. Lowering the outlet pressure is easy - just bolt on a bigger, free flowing exhaust. I've seen a couple of posts from people who added aftermarket exhausts, who report "my turbo spools up faster now" Well, that's because by lowering the outlet pressure, you increased the pressure differential, and now the exhaust gas can expand more, and do more work. That increased work pushes harder on your turbo, and it spools up faster. You should also see less boost drop at redline, because if an exhaust system is flow-limited, once you pass the flow limit of the system, any additional gasses you try and force through it only raise the outlet pressure. Higher outlet pressure, lower pressure differential, less work, less boost. [note that the compressor side comes into play here too - that's another post DG] That covers Turbine Theory. Tomorrow - the Compressor Side.
Having covered what a turbo is, and how the exhaust turbine works, we now turn our attention to the compressor side of the turbo. (If you thought yesterday's post was a little verbose, just wait 'till you see this one :) If you can extract work from an expanding gas via a turbine, then it stands to reason that you can compress a gas by driving the turbine shaft with a power source. In other words, the compressor side is just the turbine side driven backwards. The exact same physical lays apply, just now in reverse: we take a low pressure, low temperature gas, do work on it with the compressor vanes, and get a high pressure, high temperature gas at the outlet. That temperature increase is unfortunate, and will cause us problems later on - and we'll come back to it in a bit. While the turbine and compressor sides of the turbo are essentially the same, they are _not_ mirror images of each other, and the reason why is due to the chemistry of combustion. A given volume of air will burn an exact amount of fuel, in a ratio of air:fuel about 14:1. The volume of exhaust produced is much greater than the volume of the air used to create it, and the resulting exhaust pressure is much higher than the boost pressure will ever be, so the wheel and housing designs are completely different. Which leads us to turbine/compressor design. Turbines are wonderful devices. They are light, and _very_ efficient, but they also tend to suffer from a limited RPM range. That is, a turbine/compressor is very efficient at a certain RPM/flow capacity, but if you vary the shaft RPM very much, the efficiency drops. Run too fast, and the turbine blades cavitate and (aerodynamically) stall, and flow drops. Run too slow, and the blades aren't getting enough "bite", and flow drops. Here's an example. The M1A1 Abrams tank weighs about 55 tons, most of it in armor. (Steel and depleted uranium) It has a gas turbine engine that produces 1800HP at the wheels... er, tracks, which is enough power to move that beast at about 70 MPH. The turbine is amazingly small, and while I don't remember exactly how much it weighs, it seems to me that it's on the order of 300-500lbs. Compared to the weight of the rest of the tank, the engine might as well not be there! However, the design of the turbine was optimized for WOT operation. At WOT, the turbine gets better gas mileage than an equivalent diesel at the same power point, but at idle, the turbine efficiency drops, to the point where gas mileage (per minute of operation) is **lower** at idle than it is at WOT! Turbines are fantastic powerplants for vehicles that can run at a constant RPM all day - like tanks, boats, airplanes, IndyCars, etc. For vehicles that need to be run at different engine speeds, they don't work so well. (although if somebody invents a good infinitely-variable-ratio transmission, look out!) So, getting back to turbochargers, what does this mean? Well, a turbo is really a single speed device. We're only producing enough exhaust to generate boost at WOT, and we have boost-limiting devices to keep the turbo running at a constant speed (once it gets there) so, if we know how much boost we want to produce at WOT, and we know how much air we are consuming at WOT and full boost, then we can select a turbo (really, we're selecting a compressor wheel and housing combo) to maximize the turbine efficiency at that flow point. Well what does _that_ get us? A smaller turbo. That is better, because the smaller the turbo, the less rotational inertia you have to overcome, and the faster the turbo accelerates to it's WOT speed (and the associated boost level) The time delay between opening the throttle and the production of full boost is commonly referred to as "turbo lag" and is the single most hated "feature" of turbos. Ever wonder why the turbo on the 2G is so small? It's been exactly matched to the air consumption of the engine for the driving style of Joe Public - who rarely, if ever, exceeds 4500RPM. Reducing lag has another important side effect though. If you have a datalogger, and plot the boost curve of your vehicle, the area under that curve determines your transitional power band. Do a little calculus, and you find that increasing that area - even without increasing the peak boost point - increases the torque available to accelerate the car by a large amount. One of these days, one of our tuner guys is going to get a flow bench, and a dyno, and work out the air consumption of his motor at a certain boost point, and select a compressor wheel and housing combo that maximizes efficiency at that point (describing how is beyond the scope of this post - in a nutshell, you compare pressure maps) and go really, really fast. If the tranny stays together. :) Tomorrow: Wastegates and Intercoolers and BOV's - Oh My!
The story so far: We have determined what a turbo is, how the exhaust turbine functions (and what affects its performance) what the inlet compressor is (and what affects its performance) and hinted a little at what selecting a turbo requires. Today: Intercoolers and Wastegates and BOV's - Oh My! So, yesterday we left off with high pressure air leaving the compressor outlet. Unfortunately, physics has worked against us this time, and the act of doing work to our inlet air to compress it has raised its temperature. This is bad. Not only are we reducing density, we're increasing the possibility of the great bugaboo - detonation. Remember, the onset of detonation is usually the limiting factor on the amount of power a given engine can produce, and that increased intake temperature (as measured at the intake valve) increases the chance of detonation. So we have to cool the air back down again, without losing any pressure. That's the job of the intercooler, basically a "air radiator" placed in the flow stream between the turbo compressor outlet and the intake manifold. There's really not much else to say about them, except: 1) The more you can cool the air flow, the better. This _normally_ means the bigger the intercooler, the better. (There are some smaller coolers that are better designed than the lower-end "big" coolers though, so size does not necessarily indicate effectiveness. 2) The cooler must be placed in a location where ambient air can flow through it. This means that your cooler must have an intake path and an _exhaust_ path. Mounting a cooler flush against a plate does no good! 3) There's always a pressure drop across a cooler. How much depends on the cooler design. Wastegates A turbo is a positive-feedback device. The more boost you make, the more exhaust you make, which makes more exhaust, which makes more boost... in a vicious circle. So we have to have some way of limiting boost. What we _really_ want is a way of keeping the turbine operating at a constant speed (see yesterday's post) so that we can maximize the compressor efficiency - remember that turbines like to run at a single speed. However, as measuring turbo RPM is not practical, and as boost level is directly related to turbo speed, keeping the boost constant is the wastegate's job. The wastegate is just a valve that opens when we have exceeded our desired boost level, and allows exhaust to flow around the turbine, instead of through it. This lowers the pressure differential across the turbine, less work is done, and the turbo slows down. The only "gotcha" with the wastegate is that it must be able to flow enough gas to let the turbo slow down. If it can't, then you get "boost creep" where boost levels slowly grow as the car remains under boost. Bad. BOV Everybody likes BOV's because of the nifty sneeze sound they make. However, a BOV is an evil device. It's taking your precious boost and venting it to someplace else. Bad! Unfortunately, it's a necessary evil, and we have to live with it. Here's why: You're under boost, the turbo is fully spooled, and life is good - then you shift. That means your foot comes off the gas - and the throttle plate slams shut. Suddenly, instead of flowing in a continuous stream through the engine, the intake air smacks into the closed throttle plate. The turbo, which is still spinning and producing boost because if it's rotational inertia keeps producing pressure, and the intake stream, caught in between a rock and a hard place, jumps in pressure. In fact, you get a high-pressure shockwave that travels from the throttle plate back to the compressor vanes, that once it gets there, is a little like poking a stick into the spokes of a bike wheel. The repeated shock is hard on the compressor vanes and the shaft bearings, and in any case acts like a brake, slowing the turbo, and requiring it to be spooled up again. The BOV sits in between the turbo and the throttle plate, and if it detects the shockwave created by a shift, vents it elsewhere - either to atmosphere, or back to the inlet side of the turbo. So, we lost boost pressure, but we kept the turbo spooled... tough to say without a dyno if that was a fair trade on a race vehicle. On a street vehicle, it was definitely a good idea, because we spared our expensive turbo a mechanical shock.
For those of you who may have missed a part of the "Turbo Fundamentals" series, or those who may not have wanted to sift through all my Never-ending Sentences (TM) here's a summary of the last week's posts, plus the odd important point that may have slipped through the cracks last week. 1) Turbos reclaim energy that would otherwise been dumped overboard in the form of heat, by using exhaust gasses to spin a turbine, which in turn spins a compressor, which compresses the intake air. 2) Compressed intake air makes more power, because it allows you to burn more fuel per power stroke of the engine, and because it helps scavenge the engine. (The new, compressed, intake charge "blows out" the remaining exhaust gasses) 3) The amount of work done by a turbo's turbine depends on the amount of gas flow through it, and the pressure differential across it. 4) You can improve the turbine pressure differential by installing an exhaust system with a higher flow capacity than stock. It's impossible to have "too much" exhaust flow downstream of the turbo. 5) The intake compressor works best when it has been specifically sized for an engine's flow requirements and boost levels. 6) The best way to choose a compressor wheel and housing is to call the manufacturer of the turbo, and answer all their questions. 7) Intercoolers are our friends. They reduce the temperature of the compressed intake charge after the compressor has heated it. 8) An intercooler is only as good as the air flow into it ***and out of it*** 9) Wastegates limit boost levels by acting as a "rev limiter" for the turbo. 10) "Boost creep" indicates a wastegate that is too small. 11) High boost motors require good ignition systems. Most top end stumbles and misses are ignition problems. 12) There's no substitute for a day on the engine dyno. There, that should about do it. Keep in mind that I've just summarized 80 years of turbocharger development and theory into a couple of pages over the last week. I haven't covered everything, and I've simplified a number of concepts where I could without losing the important stuff. (the physics of what goes on between the head and the turbine alone could make a book all on their own!) While my posts should help get you by, if you're really interested in this stuff, there's a lot of good books that cover this material better than I do.
We got questions....
"But if increasing the pressure of the gases at the turbo inlet (exhaust manifold) will >produce more work, then why is everyone boring out their exhaust manifolds? According to your post, would this not lower the pressure differential and therefore decrease the heat and energy transferred to the turbine wheel? Does this mean boring is counter productive and we should only be polishing it to smooth out flow? "
Well, that depends. :) If the exhaust manifold is not flow-limited, then yes, increasing the diameter of the manifold runners is counter productive. If the manifold is flow limited, then increasing the diameter of the manifold runners may do you some (or a lot) of good. The pressure may be a little lower, but now you've got more gas available at the inlet to expand and do work with. "But you didn't say anything about the amount of gas in your other post" Yeah, I know. It's tough to summarize a mildly complex topic and hit _all_ the points. :( Part of the problem is that pressures in the system are all very time dependant. That gas coming out of the turbo outlet isn't a steady source, like a hose, it's a pulse. The pressure at any point in the system varies with time, and it's difficult to "nail down" system performance in absolute terms. After copping out like that though, let's examine a flow limited manifold runner. The pressure in it is higher than in a runner that is capable of flowing exactly the amount of gas produced, but it won't be able to sustain that pressure as it gets bled off through the turbine, and not enough gas can flow up to the inlet to replace it. Pressure will drop to the sustainable point - lower than the ideal case. Enlarge the ideal case though, and you've gained nothing - unless you can increase the production of gas to the point where you're back at ideal again.
"What about people who wrap up there downpipes to keep heat in and the exhaust gases hotter after the turbo. The hotter gases travel faster right, and that decreases pressure (good)? Or, the higher retained heat keeps the gases hot and the pressure stays higher and that would lower the pressure differential (bad)?"
Heh, welcome to the real world. :) Both are legit ways of examining the problem, and either scenario is possible, depending on the flow capacity of the exhaust system. If it is capable of flowing the exhaust, then the retained heat will serve to increase the velocity, lower the pressure seen at the outlet, increase pressure differential, and increase power. If, however, the system cannot flow the gas produced, then the retained heat only serves to increase the pressure seen by the outlet, decrease the pressure differential, and reduce power. Theory is a wonderful thing. It lets you gain an understanding of what's going on, allows you to ask meaningful questions, and leads you to try things in an intelligent manner. Ultimately though, when you start talking about _specific applications_, there's no substitute for trying it out and testing it. "will boring out my exhaust manifold make more power?" Well, calculate how much exhaust gas you should be producing at redline at your desired boost level, and then put your manifold on a flow bench. Are you flowing enough? What's the current pressure at the turbine inlet? What was it when you tried a bored-out manifold? (Here's the kicker) What happened when you put the changed engine on the dyno? Crew chiefs in F1, IndyCar, NASCAR - all these guys have forgotten more theory than I'll _ever_ know, and they still end up trying stuff out on the dyno. But trying stuff _without_ knowing the theory is just blind guessing, putting monkeys on the typewriters and hoping for Shakespeare.
"Thanks Dennis for your posts. It's great to have someone who has a background in this stuff share their knowledge."
Thanks. For the record though, I don't work in engineering here, I'm an IS weenie. (An IS weenie that's been racing for quite a while, but an IS weenie nevertheless)
"I've have a question about forced induction vs. NA. With forced induction, you have to lower the compression ratio to prevent preignition. Doesn't that lead to lower thermodynamic efficiency (when not under full boost, i.e., most driving) vs. a NA engine, since the combustion temperatures are lower with the lower compression? In other words, aren't NA engines more efficient for passenger cars than forced induction engines? Or does better thermodynamic efficiency not always translate into better MPG?"
Wow, ask a simple one, why don't ya? I don't think I'm qualified to give you a _definative_ answer on this one, but I'll take a stab at it. Firstly, forced induction engines lower the mechanical compression ratio to prevent _detonation_ not preignition. No, I'm not being picky, they're 2 different things, preignition being a premature ignition of the mixture due to a hotspot (normally glowing carbon deposits on an excessively dirty combustion chamber). Preignition is fairly benign as such things go, detonation can slag your engine. But yes, that lower mechanical compression ratio does reduce the amount of power produced per cubic centimeter of displacement when not running under boost, and you're right, most daily operation is not under boost conditions. So that would mean that the NA engine should be more efficient. However... If I remember correctly, the extra compression didn't make all _that_ much difference on its own - 10-15 HP on a Pontiac 6.6l moving up a point or so. The real benefit that all those 60's muscle motors got out of the extra compression was the ability to run bigger/longer cams. (As cam duration and lift increases, you need more compression to make use of it, at least on big V8's) Secondly, the turbo motor is much more efficient under boost - and when you are under boost, you're accelerating hard and burning more fuel, so the turbo gains efficiency when it's needed most - so it may make more difference to overall efficiency. I guess the easiest way to tell is to check the EPA MPG ratings for the FWD 2.0 Turbo Talon and the FWD 2.0 NT Talon, and see how they compare.
Thanks for visiting
Donovan's Dodge Garage