Turbocharger Compressor Flow Maps
for 3000GT and Stealth Owners

by Jeff Lucius

Centrifugal Compressor Performance
Compressor Flow Maps
Summary and Conclusions


This web page features the available compressor flow maps for the typical turbochargers that can be used on the 2972 cc V6 twin-turbo engine found in the 1991-1999 Mitsubishi 3000GT VR4 and 1991-1996 Dodge Stealth R/T Twin Turbo. I present the "raw" maps as supplied by the manufacturers, the maps re-scaled for cubic feet per minute (cfm) air flow rates, and the maps with engine air-demand lines superimposed on them. I explain how to read and interpret these maps.

In addition to compressor performance, other items should be considered when selecting a turbocharger for the 3000GT/Stealth, especially if the turbo is a hybrid or not built by the manufacturer of the turbo components. These considerations include:
For my complete turbocharger upgrade guide for the 3000GT/Stealth please go to 2-turboguide.htm. My Pressurization Primer presents a quick review of the processes and controls involved in pressurizing the intake manifold in the 3000GT/Stealth. For discussions of turbochargers in general, I recommend the following books: Turbocharging and Supercharging by Alan Allard (1982), Turbochargers by Hugh MacInnes (1984), Maximum Boost by Corky Bell (1997), and Motorcyle Turbocharging, Supercharging, and Nitrous Oxide by Joe Haile (1997). The magazine Sport Compact Car ran an excellent series of articles about turbos in the following issues: July 2001, August 2001, September 2001, October 2001, and July 2002.


The graphic below (adapted from Garrett Turbochargers) identifies the operational components and the air flow path of the compressor section of a turbocharger. The turbocharger's turbine and compressor wheels are located at each end of the same shaft, which is supported in the middle by a bearing housing. When the exhaust gas turns the turbine wheel, called the rotor, the compressor wheel rotates at the same speed. When the compressor wheel, called the impeller, rotates fast enough, it draws air from the atmosphere in through the inlet. As this air proceeds into the impeller it is accelerated outward (radially) away from the center of the wheel. When this accelerated air enters the vaneless diffuser part of the compressor housing, the velocity decreases and the static pressure of the air increases. Here is the technical version: using centrifugal force, the impeller imparts a high kinetic energy to the air, which is subsequently decelerated in the diffuser with a consequent recovery of static pressure from the kinetic energy. This process increases both the air temperature and pressure according to the rules of adiabatic compression and the limitations of the compressor's design. The compressed air is discharged into the intercooler system.

Compressor section
The performance of an impeller in a particular housing is quantified using a "compressor flow map", an example of which is shown below. Hugh MacInnes (Hugh MacInnes, 1984, Turbochargers: HPBooks, 159 p.) briefly describes how compressor performance is measured: 3-howcompmapsmade.htm. On the compressor flow map, the horizontal axis represents the amount of uncompressed air entering one turbo, either as volume flow or as mass flow. The vertical axis represents the amount of air compression that occurs inside the compressor; it is the ratio of the air pressure at the discharge opening (P2C) to the air pressure at the inlet (P1C). The curved lines with labels at one end such as 106,000 are the rotational speed (in rpm) of the compressor wheel. The elliptical curves with labels such as 60% represent the efficiency of the compressor, or how well the compressor achieves pure adiabatic heating of the air (higher numbers are better and mean less extra heating of the air).

Compressor flow map

Air Flow

The amount of air entering the turbo is usually measured in cubic meters per second (m3/s), in pounds per minute (lb/min), or in kilograms per second (kg/s). I personally like the m3/s that Mitsubishi Heavy Industries (MHI) uses on most of their compressor flow maps because there is no ambiguity in converting to cubic feet per minute (cfm) a rate more familiar to American hotrodders. Multiply every 0.10 m3/s by 211.8882 to get cfm.

Garrett Turbochargers (GT) uses lb/min for the air flow rate. To convert this mass flow to volume flow, the temperature and pressure of the air (that is, the density) must be known. This information is on the flow map. The number that T1C, the inlet air temperature, is divided by is GT's "standard" temperature in degrees Rankine. On the flow map above this temperature is 545R, which is equivalent to 85.31F or 29.6167C. The number that P1C is divided by is GT's "standard" pressure in inches of Mercury (in. Hg). On the flow map above this pressure is 28.4 in. Hg, which is equivalent to 13.9487 psi (pounds per square inch). You can use the first calculator on my web page 2-air-fuel-flow.htm to find that the density of air at 85.31F and 13.9487 psi is 0.0691 lb/cf. Knowing the density, the lb/min mass air flow units are converted to volume air flow units. For GT compressor flow maps using these conditions, every 10 lb/min is equal to 144.7178 cfm.

MHI will use on occasion kg/s for the air flow. On the flow map for the TD05H-16G small wheel compressor, the "standard" temperature is 298 Kelvin (equivalent to 76.73F or 24.85C) and "standard" pressure is 750 millimeters Hg (equivalent to 14.5025 psi). For MHI compressor flow maps using these values, 0.10 kg/s = 181.2415 cfm.

Note that absolute pressure (any scale) and absolute temperature (the scale is either Kelvin, K, or Rankine, R) are always used on the flow maps. One degree on the Kelvin scale equals one degree on the Centigrade scale; and one degree on the Rankine scale equals one degree on the Fahrenheit scale. K = C + 273.15. R = F + 459.69.

When volume air flow is used, the flow shown on the map is the amount of air volume entering the turbo. The volume of the air leaving the turbo is inversely proportional to the pressure (volume decreases with pressure) and is directly proportional to the temperature (volume increases with temperature).
V = n x R x T/P,
where V=volume, T=temperature, P=pressure, and n x R represents the mass. When mass air flow is used, the flow shown on the map is representative of both the amount of air mass entering as well as leaving the turbo.

Please note that horsepower is calculated using mass flow in lb/hr. For example, 30 lb/min is 1800 lb/hr of air flow. At a 12:1 mixing ratio, 150 lb/hr of fuel would be needed. Using a brake specific fuel consumption (BSFC) of 0.5, that 150 lb/hr of fuel might produce 300 crank horsepower. For an engine able to use this much flow from two turbos, 600 bhp could be developed.

Pressure Ratio

The pressure ratio, or PR, is the ratio of the air pressure at the discharge opening divided by the ambient air pressure at the inlet, P2C/P1C. A pressure ratio of 1.5 means that the air has one and a half times as much pressure leaving the turbo as when it entered. At sea level, where atmospheric air pressure is about 14.7 psi, the air leaving the turbo at a 1.5 PR would be at 22.05 psia (psi absolute) or 7.35 psig (psi gauge or boost pressure; 22.05 minus 14.7). However, in Denver, Colorado, where typical atmospheric air pressure is about 12 psi, air would leave the turbo at 18 psia (6 psig) if the PR is 1.5. Because the temperature is always higher for the compressed air leaving the turbocharger, the density does not increase as much as the pressure does. More about this below.

Wheel Speed

The wheel rotational speed, in revolutions per minute (rpm), at various values is shown on the flow map as a function of air flow rate and pressure ratio. When air flow is held constant (a vertical line on the flow map), faster rotation means a higher pressure ratio. When the PR is constant (a horizontal line on the flow map), faster rotation generally means more air flow. However, air flow does not appreciably increase after the outer tips of the compressor wheel are moving faster than the speed of sound. When the air flow reaches sonic speeds, the diffuser becomes choked and only very small increases in flow rate are possible even with large increases in wheel speed. Larger compressor wheels have maximum rotational speeds less than smaller wheels because of this limitation.

On the flow map above, the air flow regime to the right of the dotted line marking maximum wheel speed is called the choke area. The choke area is almost never noted on a flow map. To determine the choke area, you can drop a vertical line from about where the fastest wheel speed curve ends on the right side of the map. This vertical line is the approximate maximum air flow the compressor is capable of, regardless of efficiency or pressure ratio.

As can be seen near the lower left corner of the flow map, there is no significant air compression at wheel speeds below a certain rpm. A properly designed and matched turbine section is required to keep the compressor wheel spinning between its minimum and maximum speeds for the engine the turbocharger is used on and the desired boost levels. The turbine wheel is driven by the heat energy and velocity of the exhaust stream. Rotor speed and efficiency are higher when the pressure gradient across the turbine section is higher. How fast the rotor changes speeds is determined by the inertia of the turbine and compressor wheel assembly and the A/R ratio of the turbine housing. Either higher inertia (larger or heavier wheels) or higher A/R (the ratio of the turbine inlet at its narrowest point to the distance from this point to the shaft) will increase the time it takes the turbine and compressor wheels to "spool up", or increase in speed. Compressor output can be limited to a certain pressure level by controlling the speed of the turbine wheel with a wastegate, which diverts exhaust gas around the rotor.

Adiabatic Pumping Efficiency

Turbochargers compress the air by increasing the velocity of the air molecules on the compressor wheel and then decreasing the velocity in the diffuser section. When the air molecules do not move in a direction toward the discharge opening, they serve only to heat the air. The more the air is heated above that predicted by adiabatic compression, then the less efficient the compressor is. How well the compressor achieves pure adiabatic compression is called the adiabatic pumping efficiency.

For adiabatic compression alone, the air temperature, TA, is calculated using the equation:
TA = T1C x (P2C/P1C)0.283.
However, because compressors normally have efficiencies in the range of 50% to 80%, the actual discharge temperature, T2C, is calculated using this equation:
T2C = T1C + (TA-T1C)/CE,
where CE is the compressor efficiency (for example, use 0.50 for 50% efficiency) and TA is the predicted temperature due only to adiabatic compression. A 50% efficiency means that the increase in discharge temperature (not the actual temperature) is twice that expected from just adiabatic compression. A 75% efficiency heats the air 33.33% more than adiabatic compression alone. You can use my client-side JavaScript calculator to interactively determine turbo outlet temperature (as well as the pressure and density ratios). As in the flow maps, absolute temperature (Rankine or Kelvin scale) and absolute pressure must be used in these equations. You can use the calculators on my web page 2-converters.htm to convert between units.

On the flow map, the compressor efficiency decreases rapidly toward the choke area and the surge limit. The best efficiency is usually found near the middle of the map area and is often referred to as the "peak efficiency island".

Because the discharge temperature is always greater than the inlet temperature, the density ratio (DR = D2C/D1C) is always less than the pressure ratio.
D2C/D1C = T1C/T2C x P2C/P1C.
The chart below shows the relationship between PR and DR for various compressor efficiencies. The intercooler cools this hot air without decreasing the pressure very much, thus effectively moving the curves in the chart upward toward the 100% density recovery line.
Compressor PR vs DR

Surge Limit

To the left of the surge limit line on the flow map is the surge area where compressor operation can be unstable. Typically, surge occurs after the throttle plate is closed while the turbocharger is spinning rapidly and the by-pass valve does not release the sudden increase in pressure due to the backed-up air. During surge, the back-pressure build-up at the discharge opening of the compressor reduces the air flow. If the air flow falls below a certain point, the compressor wheel (the impeller) will loose its "grip" on the air. Consequently, the air in the compressor stops being propelled forward by the impeller and is simply spinning around with the wheel, which is still being rotated by the exhaust gas passing through the turbine section. When this happens, the pressure build-up at the discharge opening forces air back through the impeller causing a reversal of air flow through the compressor. As the back pressure eventually decreases, the impeller again begins to function properly and air flows out of the compressor in the correct direction. This sudden air-flow reversal in the compressor can occur several times and may be heard as a repetitive "WHEw Whew whew" noise if the surge is mild (such as when the by-pass valve is set a little too tight) to a loud banging noise when surge is severe. Surge should be prevented at all costs because it not only slows the turbocharger wheels so that they must be spooled back up again but because it can be very damaging to the bushings or bearings and seals in the center section.

Below is another example of a flow map. This one is for a MHI turbo. Even though the surge line is not labeled, you can assume it is the leftmost line on the map. The wheel speed lines terminate against the surge line. During spool up, the surge line represents the combinations of flow and wheel speed that effectively start to compress the air. If the turbo is already spooled up, the surge line represents the stall point for the impeller blades, that is, the minimum flow that can be maintained at a certain impeller speed and pressure difference across the compressor. To help explain this, please consider the analogy of an airplane wing. As an airplane is taking off, it does not lift off the ground until a certain speed is exceeded. This is like the impeller spooling up. Once the air plane is flying, if it goes too slow it stalls and starts falling. This is like surge.

Compressor flow map 2

Flow Rating

Turbochargers are often described as having a certain flow rating. As you can see, this is an extreme simplication of the compressor's performance. The flow rating is simply the volume air flow (or mass air flow) at a pressure ratio of 2.0 and some "reasonable" efficiency at a wheel speed below the choke speed. MHI typically selects the flow at the 60% efficiency level. In the flow maps presented here (other than the "raw" maps) I use a circle to note the flow rating value.


Manufacturer's "Raw" Maps

The flow maps for the TD04-09B, TD04-13G, TD05H-14B, and TD05H-16G (small wheel) were supplied by Joe Gonsowski and Garry McKissick with Mitsubishi Motor's permission. Mikael Kenson supplied the TD04H-15G and TD04H-18T flow maps. The other MHI flow maps were found on the internet without credits. Garrett Turbochargers has released flow maps for some of their turbos and these can be found at several web sites including Turbonetics, Garrett Performance Products, and Ray Hall Turbocharging. IHI Turbo America presents charts that show only the flow limits of their turbos. The flow limits are the surge limit on the left, the maximum wheel speed on the top, and the minimum practical efficiency on the right and bottom sides. Click on the link to view the flow map in another window.
"Raw" Flow Maps
   MHI TD04-09B       MHI TD05H-14B       GT T3 50-Trim       GT GT28RS 62-Trim   
   MHI TE04-13C       MHI TD05H-14G       GT T3 60-Trim       GT GT35 48-Trim 71-mm   
   MHI TD04-13G       MHI TD05H-16G small       GT TO4B S-Trim       GT GT35 52-Trim 71&76-mm   
   MHI TD04H-15G       MHI TD05H-16G large       GT T4 46-Trim       GT GT42 56&53-Trim   
   MHI TD04H-16T       MHI TD06H-16G large   
   MHI TD04H-18T       MHI TD05HR-16G6   
   MHI TD04H-19T       MHI TD06H-17C   
     MHI TD06H-20G   

Flow Maps in CFM

I "cleaned up" the manufacturer's maps and re-scaled the air flow to be cubic feet per minute (cfm). Volume air flow is the preferred horizontal axis measurement not only because our engine efficiency is measured in percentage volumetric filling of the cylinder but also because compression is constant for a given air volume flow and impeller speed, not for air mass flow. Click on the link to view the flow map in another window. For easy chart comparison, sequentially click on the links and then use the "back" and "forward" buttons in your browser.
Flow Maps in CFM
   MHI TD04-09B       MHI TD05H-14B       GT T3 50-Trim       GT GT28RS 62-Trim   
   MHI TD04-13G       MHI TD05H-14G       GT T3 60-Trim       GT GT35 48-Trim 71-mm   
   MHI TD04H-15G       MHI TD05H-16G small       GT TO4B S-Trim       GT GT35 52-Trim 76-mm   
   MHI TD04H-16T       MHI TD05H-16G large       GT T4 46-Trim       GT GT42 56-Trim   
   MHI TD04H-18T       MHI TD05HR-16G6       IHI RHF55       GT GT42 53-Trim   
   MHI TD04H-19T       MHI TD05H-18G         
      MHI TD06H-20G         

See my web page 2-3s-cfm-maps.htm for the above charts in cfm on one web page.


The main point of using a compressor flow map is to determine if the compressor part of a turbocharger is sized properly for your engine. In order to do this, you need to know how much air the engine flows. The volume air flow (VAF), measured here in cubic feet per minute (cfm), is a linear function of engine displacement (here in CI), compression cycles per minute (RPM/2), and volumetric efficiency (VE), as shown in the following equation.
VAF = (CI/1728) x (RPM/2) x VE

The stock V6 engine in the 3000GT/Stealth displaces 2972 cc or 181.36 CI. Redline is at 7000 RPM with ECM-induced fuel cut normally at 7300 RPM. Using 100% VE (that is, complete filling of the swept volume in the cylinder with fresh air-fuel charge), the maximum amount of air flow can be calculated (2-air-fuel-flow.htm). At 7000 RPM and 100% VE, our engines flow 367 cfm. Unfortunately, the design of the exhaust and intake systems and backpressure from the turbocharger turbine section limit VE to something less than 100 percent. Using the torque curve published by Mitsubishi for the stock turbocharged engine, I estimate these volumetric efficiencies for our engine at wide open throttle or full load as 91% at 2000 rpm, 95% at 3000 rpm, 93% at 4000 rpm, 90% at 5000 rpm, 81% at 6000 rpm, and 75% at 7000 rpm.

I know what you are thinking. If 367 cfm is all our engines can flow, why do we need two 400-cfm turbochargers? The reason is the turbochargers compress the air (increase the density) so that it fits into the cylinders. In order to double the air mass in the cylinders, and therefore double the fuel mass combusted and the power produced, twice as much air must be drawn through the compressor. I call this amount of uncompressed air the effective volume air flow of the engine. The effective VAF is what is shown on the compressor flow maps above and in the charts below. An engine demand line shows the effective VAF at a particular engine speed, volumetric efficiency, and pressure ratio, which is related to boost pressure. The volume of air that actually enters the cylinders at a given RPM and VE is the same regardless of boost level, only the air density is different.

Before I discuss how to interpret the flow maps using engine demand lines, I want to briefly discuss compressor efficiency. I think that more importance has been attached to compressor efficiency than should be. While I do not want to dismiss the importance of compressor efficiency (higher efficiency is always better), an efficient intercooler will compensate for most differences between compressors. To illustrate this, I drafted the figure below. The chart on the left uses the efficiencies on the compressor flow maps for the TD04-09B, TD04-13G, and TD04H-15G turbos to calculate the density ratios at the selected turbocharger pressure ratios. Because of the high discharge temperatures, the density ratios are much less than the corresponding pressure ratios. [It is interesting to note that the TD04-13G has higher efficiency than the TD04H-15G up to 230 cfm @ 1.6 PR and up to 330 cfm @ 2.6 PR.] The chart on the right shows the density ratios (density after intercooler to density at turbo inlet) after an intercooler has cooled the air under the stated conditions (as a simplification, I assume no pressure loss through the intercooler). Eighty percent efficiency may be a bit optimistic for stock 3000GT/Stealth intercoolers, but aftermarket intercoolers should achieve this level of efficiency or higher. If the ambient air temperature is much lower than the intake air temperature, the final density ratio can be higher than the turbocharger's pressure ratio.

Comparison of density ratio before and after intercooler

The charts below show effective engine air flow (for the 3000GT/Stealth) as a function of compressor pressure ratio at selected engine speeds and volumetric efficienies, assuming 14.5 psi atmospheric pressure. These engine demand lines were calculated using half the displacement so they are are applicable for one turbo. Credit for the original presentation format and the spreadsheet design goes entirely to my friend Joe Gonsowski.

The engine boost level (shown as circles on the demand lines) is the gauge air pressure in the intake manifold and is not scaled to the chart's vertical axis. For example, the 15 psi boost levels (yellow circles) at the different engine speeds lie above the turbo 2.0 PR line. This is because there are pressure losses both before and after the turbo that vary with engine speed and total air flow and so the turbo must produce more than 15 psi of extra pressure to make 15 psi of boost in the manifold. As a simplication, I assume the intake manifold air temperature is the same as the turbo inlet air temperature. Because of this, the engine demand lines may show slightly more air flow than actually occurs, depending on ambient conditions and intercooler efficiency. I do estimate pressure changes throughout the system.

The volumetric efficiencies I use for the stock engine are derived from a torque and horsepower chart published by Mitsubishi for the 1991 3000GT turbocharged engine. The VEs for a modified engine are my best guess. However, these efficiencies match well with reported max boost at the various rpm for different turbos. Now with that said, let me mention that what I call volumetric efficiency is really the engine's "natural capacity", that is, the percentage of swept volume (cylinder displacement) that is refilled with fresh air-fuel mixture (or intake charge). Volumetric efficiency is actually the ratio of the volume of air sucked through the air filter(s) divided by the displacement (total swept volume) of the engine, and will easily exceed 100% for forced induction engines.

The chart on the left shows the estimated engine demand for a stock engine. The chart on the right is my guess at the engine demand for a modified engine. The stock TD04-09B flow map is shown in both charts. These two charts illustrate a very important point. As the volumetric efficiency of an engine improves, the boost level decreases for the same air flow at a given RPM. In other words, a modified engine can flow the same amount of air as a stock engine at either (or both) a lower RPM or lower boost level. For example, look at the white circle representing the rated flow value for the TD04-09B turbo, 275 cfm @ 2.0 PR. The stock engine cannot flow this much air at this turbo PR. To flow 275 cfm (in each turbo) the stock engine must run at ~7000 RPM with the turbo at 2.2 PR (~14 psi boost). On the other hand, the modified engine can flow 275 cfm with the turbo at 2.0 PR at only 12 psi boost and 6000 RPM.

Comparison of stock and modified engine demand lines - TD04-09B

The figure above shows the situation where the engine has been improved while retaining the stock turbos. In the figure below, the same stock and modified engine demand lines are shown (with the exception of a different RPM range for the modified engine), but with the TD04H-15G compressor flow map. The TD04H-15G compressor can flow much more air than the TD04-09B and is capable of higher pressure ratios. While the stock engine can make good use of the larger turbo, the modified engine can take full advantage of the higher flow capabilities of the TD04H-15G. My modified Stealth TT engine achieves about 15 psi of boost by 3000 RPM in a roll-on from 2000 RPM in second or third gear (TD04-15G turbos installed). Maximum flow with the 15G turbo and stock motor would be about 360 cfm (each turbo) at about 23 psi boost at 7000 RPM. Maximum flow with the modified engine would be about 410 cfm at about 21.5 psi boost at 7000 RPM. Neither the stock or modified engine can flow the 428 cfm at 2.0 PR that the TD04H-15G turbo is rated at. As mentioned above, even at 100% VE the 2972 cc engine can only flow 367 cfm through each turbo at 2.0 PR and 7000 RPM (367 cfm for both turbos at 1.0 PR).

Comparison of stock and modified engine demand lines - TD04-15G

Ideally, you want to select a compressor that has a flow map that covers all the engine demand lines up to the boost levels you are prepared to use. If significant portions of the flow map lie outside of the engine demand lines, the turbo may not be appropriate for the engine.

Flow Maps with Engine Air Demand Lines

The three charts below show the standard turbo upgrade choices for stock engines. Click on the link to view the flow map in another window. For easy chart comparison, sequentially click on the links and then use the "back" and "forward" buttons in your browser.
Flow Maps with Stock Engine Demand Lines
   MHI TD04-09B       MHI TD04-13G       MHI TD04H-15G   

For highly modified engines, I made up charts for all the turbos that I have flow maps for. As you look at these, compare how well each flow map covers the engine demand lines.
Flow Maps with Modified Engine Demand Lines
   MHI TD04-09B       MHI TD05H-14B       GT T3 50-Trim       GT TO4B S-Trim   
   MHI TD04-13G       MHI TD05H-14G       GT T3 60-Trim       GT GT28RS 62-Trim   
   MHI TD04H-15G       MHI TD05H-16G small       GT T4 46-Trim       GT GT35 48-Trim 71-mm   
   MHI TD04H-16T       MHI TD05H-16G large       IHI RHF55       GT GT35 52-Trim 76-mm   
   MHI TD04H-18T       MHI TD05HR-16G6       
   MHI TD04H-19T       MHI TD05H-18G       
     MHI TD06H-20G       

I made three charts that show engine demand lines for a non-existent (but highly desirable) 3000GT/Stealth engine of stock displacement that flows perfectly, that is, it has 100% volumetric efficiency at all engine speeds (admittedly, this ignores combustion chamber volume). To get any more air flow the cylinders would have to be bored. However, even a 93.65-mm bore (0.100" overbore), which increases displacement to 3141 cc, only increases the air flow by 5.67 percent. Assuming a stroker kit could also be used to bring displacement to 3500 cc, air flow would increase by 17.8 percent.
  Flow Maps with 100% VE Engine Demand Lines  
   MHI TD05H-14B       MHI TD04H-19T       MHI TD05H-16G large   

If you cannot be satisfied with "only" 25 to 30 psi of boost, then you need a single turbo setup. A few of these have been made. If you want to go this route, the two charts below give you some idea of the boost a large single turbo can provide. Can you say 45 psi? Yes, I thought you could.
Flow Maps with Modified Engine Demand Lines For a Single Turbo Setup
   GT GT42 56-Trim       GT GT42 53-Trim   

See my web page 2-3s-demand-line-maps.htm for all of the above charts with engine demand lines on one web page.

Brief Critique of Each Compressor

TD04-09B. The stock turbo is sized nicely for the stock motor and the stock fuel system. It is too small for a modified engine.

TD04-13G. This turbo is sold by MHI as the Mitsubishi Sport Turbo Upgrade. This is an excellent turbo for a stock engine. For a modified engine, this turbo will run out of flow at higher RPM - about 17-18 psi boost max at 7000 RPM.

TD04H-15G. This turbo is as large as you want to go with a stock engine. The engine really should be modified to use this turbo. It has good coverage of demand lines (modified engine) but, for the more-demanding driver, it will fall short at the highest flow levels at the highest engine speeds - about 21-23 psi boost max at 7000 RPM. There is little advantage to increasing engine redline to 8000 rpm or above with this turbo because there is only marginally more flow at those engine speeds.

TD04H-16T. This compressor is used in the TD04HL turbo and can be mated to a TD04 turbine housing. It has good coverage of demand lines, almost as good as the 15G. It should be capable of slightly higher pressure ratios at mid RPM and a few more cfm at high rpm. There is little advantage to increasing engine redline to 8000 rpm or above with this turbo because there is only marginally more flow at those engine speeds.

TD04H-18T. This compressor from a Volvo TD04HL turbo is mated to a TD04 turbine. Inappropriate for a stock engine, it has decent demand line coverage for a modified engine. There could be some spool up or surge problems for high boost at low RPM. There is little advantage to increasing engine redline to 8000 rpm or above with this turbo because there is only marginally more flow at those engine speeds.

TD04H-19T. This compressor is used in the TD04HL turbo and can be mated to a TD04 turbine housing. Demand line coverage is excellent for a modified engine. Coverage is better than the TD04H-18T, however, efficiency is generally not as good. This compressor could potentialy allow 24-25 psi boost from 3000 to 8000 RPM with the right exhaust housing and engine modifications. At 7000 RPM, about 450 cfm should be available at 25 psi boost, and about 475 cfm at 8000 RPM at about 24 psi boost. This compressor would be an excellent choice for the 3000GT/Stealth 3.0 L twin-turbo modified engines.

TD05H-14B. This stock turbo from 1st generation 2.0 L single-turbo DSM cars has excellent coverage of demand lines for the 3000GT/Stealth 3.0 L twin-turbo modified engines. At 7000 RPM ~400 cfm effective VAF should be attainable at 20 psi boost and ~450 cfm at 25 psi boost. At 8000 RPM, almost 500 cfm should be available at about 25 psi of boost. The compressor efficiency is considerably less than some of the other turbos that flow similarly at in the upper RPM ranges. A different exhaust manifold is required for the TD05H turbine.

TD05H-14G. The coverage of engine demand lines is similar to that of the TD04-15G. A different exhaust manifold is required for the TD05H turbine. There is little advantage to increasing engine redline to 8000 rpm or above with this turbo because there is only marginally more flow at those engine speeds.

TD05H-16G small wheel. This turbo is a common upgrade for the DSM engine. Demand-line coverage is adequate but shows no advantage over the TD04-15G. A different exhaust manifold is required for the TD05H turbine.

TD05H-16G large wheel. This is another common upgrade choice for the DSM cars. The larger 16G wheel is also found in the TD06H housing. In the TD05H housing, efficiency decreases slightly but flow increases favorably for very high boost levels in our engine. At 5000 to 8000 RPM and above, this turbo offers the highest boost levels and so highest flow levels. Real 500 cfm effective VAF should be attainable with this compressor. Only the TD05H-14B comes close to this performance. A different exhaust manifold is required for the TD05H turbine.

TD05HR-16G6. This is a new upgrade choice for the DSM cars. This turbo is used in the Mitsubishi Lancer Evolution IV to VIII. I think this same wheel is used in the Evo III (but cast in mirror image?), which uses a standard TD05H-7cm2 turbine housing. The TD05HR turbine rotates reverse (the "R" in the designation) of the standard TD05H and has a twin-scroll design. The compressor inducer is a little larger (0.01") than the "big" 16G. So is this the "biggest" 16G? Max flow is better than the 16G "large wheel". Efficiency is much better than the 16G "large wheel". At 5000 to 8000 RPM and above, this turbo offers very high boost levels and so very high flow levels. Real 530 cfm effective VAF should be attainable with this compressor. A different exhaust manifold is required for the TD05H turbine.

TD05H-18G. This flow map is my speculation based on horizontally squeezing and vertically stretching a MHI 20G map. The 16G-large, 18G, and 20G compressor wheels all share the same 2.680" exducer diameter and differ in trim (and perhaps blade design): 50 trim for the 16G-large, 55 trim for the 18G and 60 trim for the 20G. In wheel "families" like this (same exducer size) higher trim usually means more flow and often lower maximum pressure ratio. Sometimes the higher trim wheel may have a bit less maximum efficiency than the lower trim wheel, such as seen in the Garrett T3 series (compare 50-trim and 60-trim wheels). So to make this map I reduced the 20G flow some and increased the maximum PR some. All "G" maps are somewhat similar in appearance so I think this speculative map may be a reasonable guess as to the MHI 18G performance. Coverage of demand lines is not as good as the 14B and 16G-large, but better than with the 20G. However, efficiency is probably better than either of the other 3 wheels. The intersection of the demand lines with the surge line indicates that 20 psi boost may not be reached till about 5500 rpm. This compressor would be good for applications where the engine is in kept in the upper rpm ranges, especially if redline is increased to 8000 rpm or more. A different exhaust manifold is required for the TD05H turbine.

TD06H-20G. This flow map is for the 20G wheel in the TD06H housing. As such it is sized too large for our engines. Flow map coverage of demand lines is poor. Perhaps when installed in the TD05H housing flow map coverage improves. Compare the 16G-large flow maps in the "Raw" section above for the TD05H and TD06H housings. Notice that the 16G-large in the TD05H housing has lost some efficiency. However, the character of the pattern has changed to improve demand line coverage and to produce higher maximum pressure ratios. A different exhaust manifold is required for the TD05H turbine.

T3 60-Trim. This Garrett compressor and housing is mated to a TD04 exhaust housing in the GT368SX turbo (sold by GT PRO). The GT368SX may also use the T3 Super 60-Trim (the flow maps are very similar). Coverage of demand lines is good overall with very good coverage at the highest RPM, even up to 8000+. Real 450-500 cfm effective VAF should be attainable with this turbo. There could be some spool up or surge problems for high boost at low RPM.

GT TO4B S-Trim. I threw this one into the mix even though I have never heard of it used on our engines. It is sold by Extreme Turbo/DSM Performance for the DSM cars. For that 2.0 L engine this turbo probably works great. For our engines (which are essentially 1.5 L for each turbo) it is sized too large and is inapproriate except for the racer that will keep the engine in the 6000 to 8000+ RPM range exclusively. A different exhaust manifold is required for the TD05H turbine.

IHI RHF55. This is a ball bearing turbo made by IHI and used on Subaru rally cars. Coverage of demand lines is excellent at lower RPM, but the flow limits suggest it may be lacking at higher RPM. Not many of these are used on our cars, and actual performance may be better than the flow limits suggest at higher RPM. A different exhaust manifold is required for the IHI turbine.

GT28RS 62-Trim. This is Garrett's "Disco Potato". I had originally thought this turbo's nickname came from the shape of its compressor flow map. However, Steven Pagano wrote me in May 2005 to tell me the real source of its name. It turns out the turbo is named after a Sport Compact Car magazine project car, a Nissan Sentra. You can read about it at this link http://www.sportcompactcarweb.com/projectcars/0306scc_projsilvia/. Now that manifolds are being made to accept Garrett turbo flanges, the new ball bearing GT series turbos from Garrett can be considered as viable upgrade options for our engines. And the "Disco Potato" would make a good choice. While it cannot reach the lofty pressure ratios of the GT35 48-trim, it should hold 25 psi boost from 5500 to 8000 rpm and has better demand line coverage than the GT35 48-trim. This compressor wheel is similar in size to the T3 60-trim compressor wheel.

GT35 48-Trim 71-mm. According to my best information, this Garrett GT35 48-trim compressor wheel in a Garrett T04B housing (with a ball bearing CHRA) is mated to a MHI TD05 7cm2 exhaust housing in the "GT30R" turbo sold by AAM. Coverage of demand lines is good from 5000 RPM and above. Efficiency is excellent where demand lines fall on the flow map. This turbo has the highest maximum PR of any shown here (GT42s excluded), capable of 3.4 PR (about 34 psi boost before pressure losses are accounted for). Over 500 cfm effective VAF should be attainable with this turbo at 8000+ engine RPM. Spool up could be a little slow (5 psi boost at 3000 RPM and 10 psi boost at 4000 RPM). However, 20 psi boost should be available at 5000 RPM, and almost 30 psi boost at 6000 RPM. This may be an excellent drag and high-performance street turbo.

GT35 52-Trim 76-mm. The GT35 48-trim 71-mm is a better choice because it has better demand line coverage and efficiency than the 52-trim wheel. Spool up should be faster with the 48-trim. The GT35 52-trim 76-mm, like the MHI 20G, is an example of a compressor than can work for our engines, and sometimes quite well under drag-race conditions, but is sized too large for everyday driving and boost levels. There are other turbos that have similar maximum flow in our engines (500-550 cfm each turbo) that have better demand line coverage and so are better suited for everyday driving where flow levels rarely exceed 400-450 cfm.

T4 46-Trim. This Garrett compressor is better suited for 2.0-2.5 L engines. It might be OK for drag race and 8000+ engine RPM use. The GT35 48 Trim, GT28RS 62 Trim, or even the TD05-14B would be better choices.

Below are two charts I made to compare how the different exhaust-side housings compare in 1/4 mile performance. The charts below do not include all the turbos mentioned on this web page, and do include some newer turbos that are not mentioned above. The data were taken from Team3S's Fastest 1/4 Mile Times and Import Power Online's Quick List & Highest Dyno Runs web pages, and from the 3SI.ORG message board. I think the comparisons and the trends are useful. It will be interesting to add more data from TD05H-14B, TD05-16G, and TD04-13T turbos. "HT12" represents the DR650 turbo (any version). The "T3-50" is the GT357 Magnum. The "T3-60" is the GT368SX (any version). The "DR1000" is the GT35 52-Trim 76-mm. Turbos are roughly ordered by compressor size.

For determining power, the 1/4-mile terminal speed is a fair judge. There are too many driver and drivetrain factors involved to use elapsed time as a comparision for engine power. The trend as seen in the MPH chart is that turbos with TD05 exhaust housings tend to produce more power than those with TD04 exhaust housings. And the Garrett exhaust housings produce even more power. Note, though, that the highest speed was with the TD06H-20G compressor in a TD05H compressor housing with a TD05H exhaust housing. If a person is staying with the factory TD04 exhaust housing, then the 15G, DR650, 17G, and GT368 would all make good upgrades and similar power, with the 15G probably producing a little less power than the others for the average owner.
Turbos 1/4-mile MPH comparo
Turbos 1/4-mile ET comparo


This web page focuses almost exclusively on compressor flow maps. As mentioned in the Introduction, there are many other factors to consider when selecting a turbocharger. Still, I would say a compressor flow map is one of the more important. Without a flow map, you have only a very poor idea of how the compressor may perform and if it is appropriate for your engine and its intended use. The flow rating value, usually supplied for all turbos, tells you almost nothing about compressor performance, except that turbos with ratings more than 525 cfm at 2.0 PR may be too large for our engine.

This bears repeating one more time. Just because a turbo is rated at 500 cfm @ 2.0 PR does not mean that the turbo flows that amount of air in our 3L V6 engine at 15 psi boost pressure. The engine mass air flow is determined by the displacement, the RPM, the volumetric efficiency, and the air density (or plenum air pressure and temperature). At a given RPM and at the same plenum air pressure and temperature, the same amount of air flows regardless of which turbo is used.

The key to being able to effectively use turbos larger than the TD04-13G is to improve the engine volumetric efficiency at higher engine speeds. This is accomplished by careful upgrading of the exhaust system (port-matched exhaust manifolds, improved exhaust manifolds, precat removal, better downpipe, free-flowing main cat, better cat-back pipes and mufflers) and the intake system (ARC2 with larger MAS and better filter or a VPC setup, better intercoolers and piping, port-matched intake manifold, maybe ExtrudeHoned plenum and intake manifold, maybe head flow work, maybe multi-angle valve seats). Also, some racers have had success with larger valves, adjustable cam gears, and different cams.

The compressor flow map does not tell the whole story (for example, turbine performance and lag need to be considered), but it is the basis for making an informed turbo upgrade selection. Ideally, all portions of the engine demand lines in the boost range you plan to use should lie on the flow map. If significant portions of the flow map lie outside of the engine demand lines, the turbo may not be appropriate for your engine.



Back Home Forward

Except for the small gif and jpg images, the content, images, photographs, text, and multimedia displayed are Copyright ©2000-2011 by Jeff Lucius and K2 Software. All rights reserved. No part, section, image, photo, article, or whole of this site may be reposted or redisplayed without permission of the author.
Page last updated February 8, 2011.