I originally typed this up for a DSM website, but due to popular demand, I have posted it here so people can link to it.
There are all kinds of misconceptions when it comes to turbo sizing, flow rates, and compressor efficiency, so I have typed this article with the intent of clearing some of them up.
I will focus this discussion on the most common DSM turbo upgrade, the 14b to 16g switch. However, due to all of the other options that people still use, I will bring up a couple others, such as the 20g, and the common Honda practice of using a huge garrett compressor wheel at very low boost pressures.
First, the common DSM upgrade beliefs. People think that if you upgrade from a 14b to a 16g, you will instantly flow more air due to greater compressor efficiency, making more power, requiring more fuel, and all this will take place at mundane 15 psi boost levels, on pump gas. Wrong!
The fact of the matter is, especially at the lower boost levels DSM’s commonly run on pump gas (not to mention the lower boost levels that are run by Hondas), compressor efficiency differences do not create a large advantage in terms of mass-flow. It is the mass flow of the turbo and engine combination that actually matters when one is trying to find out fuel needs and power output, and with some semi-simple math, one can calculate the differences between the 14b, 16g, and any other turbo you so choose.
For the following examples, we will assume that the car is question is a 2 liter DOHC import motor, thus assuming very good volumetric efficiency, on pump gas, with a 60% efficient intercooler. With a good FMIC, you will see even less gains then those below. All boost pressures are measured at the compressor outlet, and all temperatures are measured in the intake manifold. The pressure ratio (PR) is taken from the turbo inlet pipe to the intake manifold, and assumes a turbo inlet pressure of approximately 14 psi (due to flow loss from the air filter and the intake piping.) and an intercooler and IC pipe pressure drop of 2 more psi. Compressor inlet temperature is 60 degrees farenheight.
Example one: 14b, 15 psi, 315 CFM, PR = 2.26
Compressor Efficiency: 73.5%
Charge temperature: 593R (133* F)
Example two: 16g, 15 psi, 317 CFM, PR = 2.26
Compressor Efficiency: 77%
Charge Temperature: 589R (129* F)
At this point, we can already see a couple of things developing. First, there is not a huge difference in compressor efficiency between a 16g and a 14b at a PR of 2.26. Second of all, the temperature difference is very small.
If you run a tad bit more math, you can figure out the mass flow rate, either by converting from CFM, or by using the ideal gas law. I prefer to use the ideal gas law, just because that way I do not have to use the same CFM number that I used to calculate the compressor efficiency, and I can check my numbers.
For the sake of information, in order to use the ideal gas law, we assume that the displacement is the volume of the motor (2 liters) times the VE (95% at 6000 rpm) times the amount of full cycles of the motor in the chosen time interval (1 minute). This is just half the rpm (3000). Pressure is the pressure ratio, 2.26 in this case. T is the temperature of the air in the manifold in Rankin (because you must use absolute). R is a constant, found by plugging in the “base” units of all of the above (22.4L, 1.0, 1.0, 460). N is what we want to find, because we can easily convert the number of molecules to the mass flow, by knowing the average molar mass of air.
Ok, after you crunch some numbers, you arrive at the following: The 14b is flowing about 25.2 lb/min, while the 16g is flowing 25.4 lb/min. Not a big difference, if you even want to call it that. This is about 2 horsepower, requiring a negligible amount of fuel (because the change in fuel flow is one-tenth the change in airflow).
Now, a more extreme example. It is common practice in the turbo Honda world to sell kits with large Garrett turbos, yet run at boost levels around 10 psi. The 50-60 trims Garrett turbos are the most common, so for this example, I will compare the 14b (again) to a 57 trim Garrett wheel, at 12 psi of boost.
The motor in question has now become a 1.8 liter, and I will increase the VE to 110%, because the Honda guys get all uptight about their “higher flowing heads.” I could write a whole article on how they can’t be that much better than other import heads, like that of the 4G63, or the 3SGTE. All other numbers will remain the same as they did for the first example.
Example 3: 14b, 12 psi, 330 CFM, PR=2.05
Compressor Efficiency: 74.5%
Charge Temperature: 582R (122* F)
Example 4: 57 Trim, 12 psi, 330 CFM, PR=2.05
Compressor Efficiency: 76%
Charge Temperature: 581R (121*F)
This example obviously has an even smaller difference in terms of mass flow and power potential. There is no reason a 57 trim is necessary at this boost level.
So now the question is, why even go to a larger turbo? My intent was not to discourage people from upgrading, but just to clear up some beliefs about turbos that were not true. A 16g is a good upgrade over a 14b, because the 16g is capable of boost levels over 24 or 25 psi, while the 14b is stuck operating around 21 psi. This will bring very large power increases, especially on race gas, or on a well prepped car.
Also, it is important to note that there will always be some degree of VE gained whenever a switch is made to a larger turbine housing. While larger housings will create more lag and raise the boost threshhold, they also decrease the restriction seen by the exhaust gases, and thus decrease the pressure within the exhaust manifold. The larger the pressure differential between the intake and exhaust side, the better the VE gains. With a switch from a 6 cm^2 housing to a 7 cm^2 housing at boost levels around 1 bar, there will be a small but noticeable power increase due to this. Obviously, this is more effective as power and flow levels rise.
The same is true for turbine wheels, which I feel are an often overlooked part of the turbo sizing process. By making a switch to a higher flowing turbine wheel, you are decreasing the pressure ratio needed across the hot side of the turbo in order to generate a certain shaft speed, and thus a certain boost pressure. This means that, for the same post-turbo exhaust pressure, the exhaust manifold pressure (backpressure) will be decreased, and VE will be increased. Again, you can see very solid power gains, without taking the compressor wheel into account.
Kyle Tarry 2003