Supercharger Performance and Engine Performance Parts

Posts Tagged ‘CFM’

The Lysholm / Autorotor Supercharger is a unique and very practical chager. The unit is a great compromize between a positive displacement supercharger (that creates boost pressure by over pumping and over feeding the engine with air) and a compressor (similar to a turbocharger) that compresses the air inside the supercharger housing before sending it out to the charger piping.

The unique three-five design of the twin screw chargers relies on a 3 lobe and a 5 lobe rotor intermeshed to capture the air flowing into the supercharger for inter-screw compression. The combination of an intermeshed 3 lobe and 5 lobe rotor means that the rotors inside the housing are operating at different rpms with a ratio of 5:3 to keep the rotation of the lobes (3 lobes to 5) in synchronsim. This complex design allows the rotors to capture air (in its natural volume) from the back of the blower housing, and push it foward as the screws rotate. As the air is moved forward it is captured and compressed between the intermeshed rotors as well as being pumped (in positive displacement) from the inlet port at the back of the charger housing to the outlet port near the front.

Because of this unique design, screw style chargers are able to outperform simpler rotor based chargers in two aspects:

1- The blower is able to acheive a higher pressure ratios because the compression is combined between positive displacement (overfeeding) and between direct compression of the air (inter-screw compression).

2- Since the air is compressed inside the housing, the housing is able to ingest and move more air (higher CFM ratings) for a similarly sized roots style blower.

These kinds of superchargers boast great adiabatic efficiency of up to 68% while at the same time being able to deliver that high efficiency at high boost levels of up to 18 psi. With such a high potential peak boost level, these chargers are capable of matching the top end delivered by a typical turbo system, without the lag and throttle delay disadvantages of spooling a turbo. Because of the positive displacement nature of this charger, the charger will always be able to make boost at any rpm so long as they bypass valve is closed.

Some applications for this unit include the OEM install of 1.1Liter Lysholm / IHI hybrid in the Mazda Millenia motor.

Other kits include the BBM upgrade kit for the Volkswagen Corrado. The kit replaces the G20 centrifugal supercharger with the lysholm twin screw system. As you can see in the dyno below, the twin-screw outshines the smaller centrifugal unit both in the lower rpms and in the top end and this exactly the result of the inter-rotor compression keeping flow and efficiency up to par at higher rpms while positive displacement fills out the low rpm torque.

Here is an overview of Lysholms available chargers:

Lyshom	pressure ratio	Boost	CFM	HP	effeciency	displacement (liters)
1200 AX	2.2	18	636	424	64	1.2
1600 AX	2.2	18	848	565	66	1.6
2300 AX	2.1	16	1059	706	65	2.3
2300 R	2.2	18	989	659	68	2.3
3300 AX	2.2	18	1236	824	66	3.3
3300 R	2.2	18	1236	824	66	3.3

For more Information please visit:

Lysholm

BBM

Technorati Tags: Autorotor, boost, CFM, Displacement, Lysholm, pressure ratio, Rotor, supercharger performance, VAG

Supercharger performance is proud to present the newly updated power calculator. The only calculator built for enthusiasts by enthusiasts…

Get your copy today !

Technorati Tags: boost, CFM, compressor map, header, Intercooler, N2O, Power Calculator, RPM, water injection

The whipple supercharger is a unique and very practical chager. The unit is a great compromize between a positive displacement supercharger (that creates boost pressure by over pumping and over feeding the engine with air) and a compressor (similar to a turbocharger) that compresses the air inside the supercharger housing before sending it out to the charger piping. 

The unique three-five design of the whipple screw clearly showing how the lobe from the three lobe screw tightly fits between two lobes from the 5 lobe screw to compress the air for inter-screw compression.

The secret to this style of ‘hybrid’ blower is the two intermeshed rotors of different lobe numbers (see illustration). The combination of an intermeshed 3 lobe and 5 lobe rotor means that the rotors inside the housing are operating at different rpms with a ratio of 5:3 to keep the rotation of the lobes (3 lobes to 5) in synchronsim. This complex design allows the rotors to capture air (in its natural volume) from the back of the blower housing, and push it foward as the screws rotate. As the air is moved forward it is captured and compressed between the intermeshed rotors as well as being pumped (in positive displacement) from the inlet port at the back of the charger housing to the outlet port near the front. 

Because of this unique design, screw style chargers are able to outperform simpler rotor based chargers in two aspects:

1- The blower is able to acheive a higher pressure ratios because the compression is combined between positive displacement (overfeeding) and between direct compression of the air (inter-screw compression). 

2- Since the air is compressed inside the housing, the housing is able to ingest and move more air (higher CFM ratings) for a similarly sized roots style blower. 

So how is the whipple best used. Some people are interested in SIGNFICANTLY boosting their small displacement motor to make it not only have better low rpm torque but also unrestricted peak RPM power. In two of our articles (one, two) we have talked about how you can combine a typical roots style charger for low rpm instant boost, with a high rpm solution of turbocharger or even a centifugal supercharger that is sized proparly to elevate your motor to the required peak psi -- that above which your typical roots style supercharger may not be able to provide effeciently.

Well here’s the whipple solution. If you use a whipple charger, then you have the best of both worlds, you have a positive displacement charger that has no spool up lag, as well as internal compression allowing you to achieve high PSI levels without the need to for overspeeding your blower to do so. 

So, with the use of a whipple charger you can have a fairly flat torque curve from zero to redline giving you very predictable traction and launch control (which is why whipples and other screw type chargers are popular in drag racing or coming out of corners in road coarses). A predictable and linear torque curve also is more forgiving to overgeared cars and more forgiving with different driving styles. 

A video of a whipple-charged GT-500 mustang and dyno showing the infamous flat torque curve…

Here is an overview of whipples available chargers:

For more Information please visit:

Whipple

Technorati Tags: CFM, effeciency, Rotor, twincharging, Whipple

The LS1 powered Camaro Z28 is a great car, with plenty of power and great modification potential. One of the typical upgrade paths for Z28 owners is high flow heads, high lift moderate duration cams, and healthy doses of nitrous oxide. In unleashing this nitrous driven frenzy on the LS1 engine, the LS1 has proven to be a fairly robust contender in the face of a 200 horsepower shot of nitrous unleashed on the motor at 2500 RPMs, some people have even gone as far as spraying 300 shots on their stock LS1 with great success.

One of the great things about nitrous is it gives us some insight on the power of our engine. Since nitrous is typically delivered all at once in a single shot, then it usually produces huge torque figures at lower rpm ranges, and as the RPM’s rise, the torque increase from the nitrous shot drops down, as horsepower stays maintained.

The reason for this is the basic relationship between torque and horsepower.

Horsepower (hp) = Torque (ft.lbs) * RPM / 5252

By applying this math we find that a 200 horsepower shot sprayed at 2500 RPMs add about 420 ft.lbs of torque to the motor. And if we look at a dyno for a stock LS1 engine such as this one I found for a 2002 6speed SS, we find that the motor alone is making about 290 ft.lbs of torque at 2500 RPMs.

The motor on nitrous will be expected to make: 290 (motor) + 420 (nitrous) = 710 ft.lbs of torque. Quite impressive for our 5.7 L V8.

Now if you’re using 200 and 250 and 300 horsepower shots of nitrous, you will get bored fairly quickly of the amount of money and time that you waste refilling the nitrous bottle, and you may be willing to pay one time lump sum to have that power avaialbe to you ‘on tap’ rather than in the bottle.

Using our same formula and taking our 710 ft.lbs of torque up to the LS1’s redline of 6200 RPMs I find that I can fairly safely set my peak horsepower goal to be:

710 * 6200 / 5252 = 838 horsepower.

So let’s think about this for a minute, I can make 515 horsepower using a 200 shot worth of nitrous, or I can make 838 horsepower using a centrifugal supercharger setup all the while not exceeding my bottom ends’ well known, tried and true, withstand of 710 ft.lbs of torque. I think this is a very safe adventure into supercharger performance.

Looking at the stock LS1 dyno we can see that peak power is delivered early at 5200 RPMs through the use of conservative 202*/210* duration camshafts (intake/exhaust) in stock format. According to my calculator, that duration cam would typically put peak power around 4500 to 5000 RPMs which corresponds to our dyno shoot.

Before we find out how much boost we should expect to need to reach 838 horsepower we should factor in the fact that we plan on making 838hp at 6200 RPMs rather than 5200 RPMs through the use of a proper supercharger camshaft. Again using my power calculator I come up with an ideal cam configuration of 210*/244* (intake/exhaust) with a target LSA of 112 to 115 degrees. And based on our new peak power RPM we find that we will probably need 18psi of boost to make 838 hp @ 6200 RPMs (on the other hand we could make 838 hp @ 5200 on the stock camshaft at a much higher 24 psi… I personally would rather make the power with RPM at lower boost than use much 20+ PSI figures on a restricted motor).

Based on my calculator, the following falls into place for our 838 horsepower stock block LS1 build up:

Quite an impressive build we are about to take on. Now these are calculated figures. Not all of them will end up being exact as we need to source available parts that work for our target build.

Supercharger:

Procharger offers a supercharger kit for the LS1 based off of its P1SC2 supercharger (or the even larger D-1SC) which

P-1SC1 impeller...

are at least capable of delivering 1200 CFM @ up-to 32psi of boost if needed. Perfect for our requirements of 1200 CFM and 18psi.

Camshafts:

Looking around for a bit I found thunder racing’s cheater camshaft which is perfect for our application: 214* intake duration, 230* Exhaust duration, and 117* LSA. Notice in the description they say ‘responds very well to nitrous’, well high boost supercharger applications have similar cam requirements to nitrous oxide in that you neither want excessive overlap (to waste your nitrous or boost into the exhaust manifold, and at the same time your duration requirement on the intake cam is reduced (to reduce overlap and because of the compression of the air via nitrous or a supercharger means that you can ingest more air volume in a shorter duration of time). A great find.

 Thunder Racing Custom Camshaft
“CheaTR” - 214/230 .601/.575 117 LSA. Off Idle-6800 RPM Power Band. Broad power range. Works well with stock exhaust manifolds and catalytic converters. Stock like idle. Minor tuning required on automatic transmission cars.  Responds very well to nitrous.  Due to the fast ramp rate of this camshaft, the use of 1.8 rockers is not recommended.  Double valve springs and titanium retainers required for this cam.

The ideal figures we got for cold side and hot side piping are 190mm and 170mm respectively. These figures are calculated based on less than 1 horsepower loss in the intake system and less than 0.01 psi pressure drop. Now air is a compressible fluid. If it needs to go through a 90mm intake system rather than a 170mm intake system it can do so. However, there is a certain pressure required to force the air to flow through this bottle neck of a restriction and that shows up as a pressure drop.

As we said earlier we’re going to need 18psi of boost to reach our target hp goal on the motor giving the parameters given in the calculation. But if we use a 92mm throttle body costing us 18hp and 0.25psi, a 125mm intake hose (as we have calculated later on for our MAF housing), costing us another 6hp and 0.1 psi, as well as an undersized header with another 3 or 3 psi in back pressure from using 2” primaries rather than 2.4” primaries, then the overall pressure losses in the system can add up to about 4psi. What this means is that in the end, when all is said and done, we will probably make our target 838 horsepower somewhere between 18psi and 22psi because some of our parts were undersized.

For the supercharger cold-side piping, the supercharger has a 3.75” inlet which we will expand out to 125mm or a 5” intake using our new 125mm MAF housing and a 5” air filter.

For the supercharger hot-side piping, the supercharger has a 3” outlet which will take out to 3.5” using 3.5” piping into our procharger air to air intercooler and all the way into a 92mm (3.62”) throttle body.

This 92mm throttle body is the largest I could find for a bolt on LS1 throttle body and comes as part of a package with the Fuel Air Spark Technologies (F.A.S.T) LSX style intake manifold for the LS1.

This throttle body will cost us 18hp at our target power level, but we should more than make that up elsewhere. Why so? Because all these calculations are done based on a stock motor (stock intake, header, exhaust) figures. The only modification I have factored in here is Cams and boost pressure. If I were to factor in the fact that the factory motor is restricted by the factory intake, header, exhaust systems, and that it could potentially gain 3 or more psi of pressure losses by upgrading those items, and if you factor in that we in the process of installing our supercharger package are in fact upgrading those same items, then it’s safe to say that my 838 is a somewhat conservative estimate of what a Cammed LS1 can put down at 18psi of boost. However, I like to do my math conservatively and be pleasantly surprised by the results later on J.

Intercooler

The kit also comes with procharger’s twin high flow intercoolers which are each sized at 11 x 9 x 4.5 or a core volume of 445.5 cubic inches. In total we have 891 ci of intercooler cooling volume (between both intercoolers) compared to our original calculations of 1470 cubic inches. Instead I would call up procharger and try to get a larger core (to try to stay away from needing water injection for this application) … and choose something like their Air to Air core at 27.5” X 12” X 4.5” which they rate to 1300hp and should do very nicely on our 838 (or more) horsepower build. Furthermore this intercooler has 3.5” inlets and outlets which match our chosen throttle body for a good matched package.

Headers

Kooker race header with 2" primary and 3.5" collector.

This was actually pretty tough to find. There are probably many supercharged LS1 powered Camaro’s out there, but yet the only header that I could find that comes close to our requirement (2” primary, short 12” (or longer) runner length into a 3.5” collector) seems to be the Kooks race header with venture collectors. The kooks header comes with 2” primaries and 3.5” collector.

Exhaust

For the exhaust there are two options:

1-      To do a custom dual 3.5” exhaust with 3.5” X-Pipe and under car turn downs releasing the exhaust gases before the rear differential (see picture), or even using some custom 3.5” side pipes if you like.

True dual x-piped exhaust with under car exit before the bell housing.

2-      To make a custom 3.5” exhaust cutout section right after the 3.5” collector exit on our cook headers, and then from then on reduce the exhaust back down to a typical 3” exhaust and get an aftermarket single or dual 3” exhaust that will be used for normal street use. When full power is demanded, the 3.5” exhaust can be unleashed right at the headers at the flick of the switch by opening the e-cutouts and dumping the exhaust before the cat-back.

Fuel System:

According to our calculations, we need to be able to supply 350 liters per hour (lph) of fuel at a constant fuel rail pressure of about 40psi. Most typical single fuel pump upgrades use a Walbro 255 lph pump instead of the stock pump. Although this pump alone is good with the proper hotwire kit to 610 crank horsepower, it will not be enough for our target power figure. One option is to install a kenne-bell boost a pump which can increase the pump voltage from 14volts (stock with a hotwire kit) to 17 volts or 20 volts. Increasing the voltage from 14volts to 20 volts potentially gives us a 42% increase in flow making our pump capable of delivering 362lph of fuel which can meet our requirements.

Lonnie's performance fuel kit

I think for our application, I would rather cut it safe than cut it close. Who’s to say that we end up geared with our pulley system for exactly 18 psi? Who’s to say that we don’t end up making 850 horsepower with our setup when all is said and done? Who’s to say that the motor won’t want to run a one point richer air fuel ratio at that power level and so our fuel delivery requirements will actually be higher by about 8 to 10% to cover that extra point of air fuel ratio.

Instead a better approach for this application would be to use a complete fuel supply kit, including dual in-tank pumps, a hotwire kit to deliver full 14volts to the pumps, complete with new fuel feed and return lines (matched for 800hp) and some high flow fuel rails to make sure that every injector (weather it is right at the fuel feed or at the end of the fuel rail closer to the return) sees the same amount of fuel pressure, because the fuel rail is large enough that there is no significant pressure drop between injector #1 and injector #4.

One such kit is provided by lonniesperformance in conjunction with racetronix.

The kit includes

Double Pumper Kit & Complete Custom Fuel System
Camaro/Firebird ‘99up LS1 Double Pumper -- Twin High Output Fuel Pump Kit,  Wiring Harness & Hobbs Switch -- Fully Assembled & Tested - Requires your sending unit for modification

Includes -8 Supply & -6 Return lines, fuel rails, regulator, filter, & all fittings needed to connect to double pumper sending unit.
LS2 fuel rails optional.
LS7 fuel rails optional.

Fuel Injectors

Delphi 75 lbs/hour injectors

Again we can pick these up from lonnies performance. By our calculations we need 630cc/min injectors (or 63 lbs/hr injectors). It’s a good rule of thumb to have injectors that will deliver your fuel needs at around 85% duty cycle. Which means to deliver 63 lbs/hr of fuel per injector we need a fuel injector that can deliver 74 lbs/hr.

Driving larger injectors to 85% of their maximum capacity is ultimately safer than trying to extract 100% of the capacity out of a smaller injector. The reason is that an injector that is running at 100% duty cycle is more likely to fail, and any variation in power level or boost level means that you have no room to increase your fuel delivery because your injectors are maxed out.

A set of 75# (or 75 lbs/hour) Delphi injectors will do the trick for our build.

Tuning

One simple route for tuning is typically to use a matched pair of an upgraded MAF sensor and upgraded injectors. For example using a 100% larger MAF sensor or MAF housing recalibrates the performance of the factory computer around the use of injectors that are also 100% larger than stock.

This approach works well for low boost packages or normally aspirated buildups of the healthy 5.7L V8.

The largest MAF housing I could find was a 100mm MAF which is maxed out at a reading of 511 grams per second of air. If we do some calculations we find that 511 grams per second comes out to about 545 hp worth of air which is shy of our eight three eight horsepower goal. If we were to use this kind of setup on our car, the ECU would not know the difference between 600 hp and 800 hp because at both situations it would be reading full MAF voltage and giving the command to the injectors to go Full on at 100% duty cycle.

There are two reliable ways we can go about solving this problem:

1-      Upgraded MAF / Injector combo

850 horsepower / 545 hp = a mass ratio of 1.56 (or an increase in area of 56%).

TSP 100 mm MAF

To make our stock MAF sensor capable of reading 850 horsepower it is possible to transplant the sensor into a housing that has an area that is 56% larger. Doing the math we find that transplanting the sensor into a 125 mm housing can do the trick. The problem with continuing to transplant the stock sensor into a larger housing is two-fold:

As the housing gets larger, then at very low flows (such as at idle) the small amount of air flowing through the larger 4.7” pipe can get turbulent, and as it gets turbulent, then the sensor reading oscillate as the air spinning inside this 4.7” pipe hits the sensor in waves (rather than in nice laminar flow). What this does is it gives inconsistent readings to the ECU about the amount of air flowing, and results in a wandering air fuel ratio because the fuel supply itself can be oscillating based on the reading.

One of the things talked about in mechanical engineering is the flow profile of a fluid in a pipe. Obviously the walls of the pipe have some resistance to the air flow, and this wall resistance means that the air travelling in the center of our 4.7” pipe is at a higher velocity than the air travelling on the surface walls of our pipe (because this air has some frictional forces reducing its velocity). As our flow pipe gets larger then our sensor gets moved farther and farther away from center of the pipe, and thus our sensor’s reading is less accurate because it’s measuring the slower air closer to the pipe walls.

If you find that you have either of the two problems (a poor idle and wandering air fuel ratio, verified with a wandering OBD-Log of your MAF reading, or a sensor that seems to be under-reading the amount of air as you get into boost verified with comparing your required injector duty cycle to sustain your target air fuel ratios compared with the amount of air that your sensor is reporting) then one typical solution is to use some fine wire mesh to create your own MAF screen. The screen that we see common on factory sensors helps reduce this velocity profile for the air as it enters the sensor for metering and makes the air closer to laminar at that point, such that any sample of that air being metered is very similar to the velocity, density, and temperature of the rest of the air being unmetered by the sensor.

Matching our 125mm MAF, with our 75lph injectors we can then do the rest of our tuning using the factory ECU and a flash tuner.

2-      Speed density setup:

A speed density setup eliminates the MAF sensor altogether replacing it with a manifold pressure sensor, air temperature sensor, and uses both RPM and throttle position readings as well to approximate the car’s air flow requirements based on those inputs.

I prefer directly metering air myself rather than extrapolating it from pressure and some assumptions about volumetric efficiency. When you directly meter air, if you for example open your exhaust cutouts (as stated earlier in our exhaust section) and those cutouts allow your engine to breathe in more air, then that air will get metered and your ECU will know about it adding more fuel. If you use a speed density setup then when you get into situations where your motor is breathing in more air at the same (or even lower) boost pressures because of a change in volumetric efficiency then your ECU will be completely oblivious to the change.

Water Injection

Since we got cams for our setup (to reduce our peak boost requirements for our target horsepower) and since we were able to source a good sized intercooler for our power goals, we may not need water injection at all on this build. This is not by accident, but rather by design. In our calculations we came up with 12 gallons per hour of water injection (typically this figure comes in as 10 to 15% of the fuel delivery of our car).

If you have a 12 gph water/methanol injection setup and a healthy 2 gallon tank in your trunk, then you will run out of water/methanol mix in 10 minutes of full throttle time.

It quickly becomes clear that you would rather invest in some power upgrades (to lower your peak boost level and thus your peak inlet air temperatures) as well as invest in a healthy sized intercooler (again to lower your peak inlet temperatures) so that you can steer away from having to drag behind you a water/methanol tanker to keep up with your supplemental injection needs.

Spark plugs

Typically 1 step colder spark plugs are needed for every added 100hp as well as a 2-3* ignition timing retard for the same, at least that is a good starting point, and then hotter plugs or more ignition advance can be added when tuning.

The stock NGK spark plugs for the LS1 is a TR5 which is a 14mm 3/8” hex plug with 18mm reach and a projected tip, a tapered seat and is resistive for reduced EMI.

After looking NGK’s part code I think the best start would be BR10ECMIX iridium plug (or a similar BR10ECS copper plug which is cheaper). The spark plug has 1mm more reach, but is a non projected tip which we need for a high power application and is 5 heat ranges colder than stock as required by our additional 500hp. Furthermore the ‘CM’ designation means that this is a compact plug with a low profile or even side discharge ground strap so both the spark plug gap is reduced (for high boost) and the ground electrode is moved farther away from the center of the combustion chamber and reduced in size to prevent it from becoming a hot spot and an instigator for detonation.

Expected dyno:

Based on an expected boost curve and a stock LS1 dyno, I’ve created the following expected dyno for our car.

-          Blue lines are stock hp and torque.

-          Green Lines are Stock + 200 shot of n2o (peak torque at 700 ft.lbs and peak hp at 500hp).

-          Red Lines is our custom supercharged setup. If you look closely you see that I have not crossed the 700 ft.lbs line on this setup yet we are able to reach our peak power goals over 800 horsepower. Reliable, progressive power delivery from zero to redline.

LS1 stock, vs 200 shot vs 18psi + supporting mods centrifugal

 Some Procharged LS1’s on youtube …

Links to parts:

Thunder Racing (Split patter camshaft for 6800 rpm supercharger application)

Procharger (P-1SC2 supercharger kit and race intercooler)

Lonnie’s Performance (Complete fuel system solution up to 1000 hp)

Kooks headers (2” primary headers into 3.5” collector)

DMH (electric exhaust cutouts)

Burns Stainless (X-pipes and custom exhaust)

FAST (intake manifold and 92mm throttle body)

Texas Speed and Performance 100mm MAF

Clubplug (NGK spark plugs)

Technorati Tags: boost, Camara, camshaft, CFM, engine performance parts, Intercooler, kooks header, LS1, MAF, Procharger, supercharger performance, twin-pump fuel system

Engine Performance Parts to improve supercharger performance…

I am compiling a guide on information on how to pick the exact engine performance parts to fit your target power requirements. Basically I want to eliminate all the guess work out of tuning and save you some money from having to do things over and over again.

While I was doing research for ‘buying the right intercooler’ I got lost, honestly. There are two types of information you will find out there:

1-      One class of articles is written by engineers talking about pressure differentials, thermal efficiencies, enthalpy and multi variable equations that are very remotely related to flow, horsepower, torque, supercharger rpm or other things that we KNOW that we can use as an input to our equations. (Basically this science needs to be translated to layman’s terms)

2-      The other class is a group of random trial and error advice by enthusiasts, press releases and other materials that you find online.

Here’s what we do know:

First let’s talk about how intercoolers work. There is some debate about whether the intercooler is like a heat sink whose function is to absorb thermal energy from the incoming air to prevent the heat from reaching the engine, or whether the intercooler is like a radiator, where the air flow over the intercooler is responsible for extracting heat from the inlet air charge.

The true answer is both are correct…

The air running through the intercooler spends very little time inside the intercooler and slowing it down for more thermal exchange (like we would coolant in the radiator) would mean preventing air from reaching the engine which is a restriction on power. Because the air spends little time in the intercooler, the intercooler usually has multiple passages, internal ribs, and fins inside of it to maximize the surface area contact between the intercooler aluminum and the compressed air molecules. In this sense, the overall volume of the intercooler, and the overall surface area of its internal surfaces are like a heat sink that absorbs the heat energy out of the compressed air. In this aspect it makes sense that the larger our intercooler, the better. Furthermore it also makes sense, that the more complex and intricate the internal passages of our core, the more heat we will be able to extract out of the charge air. Of course the flipside of this is that very complex internal passages can create turbulence and restrict airflow so ultimately there is a balance in good design between internal complexity and flow capacity.

When we start out, the intercooler is cold, and with our first power run, as the hot compressed air runs through the intercooler, the heat is transferred to our heat sink (which is the intercooler) and nice cool air is left to enter the engine. After the first run, the intercooler is warm; and if we were to make a second power run back to back, the intercooler will not be able to SINK much heat because it is already somewhat heated. This is where the intercooler as a radiator comes in, the heat that was transferred from the air to the intercooler core, needs to be taken away either by cross flowing air in an air to air intercooler, or by cooling fluid in an air to water intercooler, or even by an ice-water bath for drag racing applications. Without harvesting the heat that the intercooler has absorbed out of the compressed air, the intercooler will heat up run after run until its temperature is the same as the compressed air heating it. At this point there is no temperature difference between the air and the intercooler core and we can no longer SINK any heat.

Some cars have their intercoolers located under the car’s hood (like the Mazda Sentia / 626). In this kind of installation the intercooler is mostly a heat sink and will be used for a few passes till it soaks, once it soaks it needs to be left to cool till it returns to under hood temperatures before it can be effective again as an intercooler. From this we gather, that any intercooler no matter how small, or poorly placed is better than no intercooler because at least for that first power run it will potentially increase horsepower.

Now I’d like you to keep this information in mind while we talk about intercooler dimensions…

There are three main dimensions to the intercooler, the height (H), width (W) and (D) depth and based on that there are some physical concepts that we want to think about:

Cross Sectional Area:

Height x Depth = the cross section of the intercooler and is related to how well the intercooler will flow and whether or not it poses a restriction to the intake flow. This is the area of the surface facing the compressed air as it travels through the intercooler. Just like free flowing intakes, throttle bodies, and exhausts, if this area is undersized it will pose a flow restriction and reduce performance.

Width:

Width = the length of the intercooler and if you have a same side inlet/outlet intercooler then your intercooler length is effectively 2*W. This is the distance the air has to travel through the turbulent and complex intercooler core. The longer this length is, the more pressure drop there is in the intercooler so it’s not advisable to have too wide an intercooler because we’d be waste turbocharger compression in intercooler pressure drop, neither is it advisable to have a same side inlet/outlet intercooler where the air has to travel a long distance in the core.

Frontal Area:

Width x Height = frontal area of the intercooler which faces the incoming ambient air, a good sized frontal area is required to ensure that the intercooler doesn’t heat soak and that the rushing air stream is able to cool the intercooler efficiently (like a radiator) for you to be able to make back to back power runs. As we increase this area, we expect the intercooler to have better control over its peak operating temperature and have better repeatability no matter how long we stay in boost (good for standing mile races for example or all day road racing events).

Depth:

Depth = the depth of the intercooler, usually the intercooler is front mounted in front of the radiator… if you increase the depth too much (and especially without proper air ducting to the intercooler and airfoils between the intercooler and radiator) then you may slow down the incoming ambient air enough that your radiator starts overheating. So increasing D gives us better intercooler performance and more flow capacity (H*D is the cross sectional area mentioned above) but it reduces engine cooling efficiency so it must also be controlled.

Last but not least:

Total Volume:

Height x Width x Depth = the total volume of the intercooler, which is an indirect measure of the internal surface area of the intercooler. The larger the volume, the larger the heat exchange surface area, the more heat we can sink out of the air in an extremely short period of time (the 100 milliseconds  or so that the air spends inside the core). Obviously the bigger the volume, the better the cooling and the worse for pressure drop. Again this number needs to be controlled.

How  do I know if the intercooler I have now is adequate?

Intercooler efficiency can be tested in two ways:

1-      Thermal performance

a.       Measure the temperature difference between the intercooler inlet air and intercooler outlet air and use this delta T to compare between the intercoolers you have available to you. The best intercoolers out there can drop air temperature by over 100*F and get you within 20* of ambient air temperatures. If your factory intercooler can already accomplish similar results then there may be no need to upgrade.

b.      Track the temperature of your intercooler in a prolonged power run, or on back to back power runs. The design and placement of the intercooler should be adequate that the temperature rise over time (with say 60+mph air hitting the intercooler) should be controlled, if the temperature rise is too steep then you may need a better ‘radiating’ core with more frontal area, better air guides and air foils, and better placement with high pressure air in front and low pressure air behind it… we’ll explain more about this later. 

2-      Flow performance

a.       Measure the flow through the intercooler core at 28” of water (standard for most flow meters), or measure the overall intercooler pressure drop at the flow rate required by your target horsepower. If the intercooler is on the car, measure the differential pressure across your intercooler at peak hp figures.

The best intercoolers will have less than 1psi of pressure drop (typically 0.5 to 0.9psi) at peak boost and horsepower. If your intercooler is within these power figures then there may not be any need to upgrade.

Now going back to selecting the best sized intercooler for your application, it would be very tough for me to figure out the exact math of how to optimize your intercooler size, and then I would have to translate that math to ‘car terms’ of power, inlet air temps, supercharger outlet temps, pressure ratios and boost pressures…etc

Here is another solution; one thing engineers like to do in dealing with this kind of a problem plotting statistical data on a chart and looking for some trends…

I found some 30 different intercoolers online with either flow tests (CFM), or Dyno tests (HP) or both, and since we know that it takes roughly 1.5 CFM of air to produce 1 HP (depending on density) then I combined both sets of data both for flow tested OEM intercoolers and for aftermarket ‘engineered’ intercoolers to produce the following graphs:

Flow in CFM vs. Cross Sectional Area:

Cross Sectional Area vs CFM

This is a plot of flow in CFM (vertical) vs. cross sectional area (squared inches) for the 30 cores that I had data for. As you can see there is a linear relationship between flow and area which is expected. So we can use this as a guideline to figure out (for a given depth D) of available cores, what the minimum height of our intercooler must be to get good flow performance.

One thing to note here is that these flow measurements were taken at 28” of water pressure or 1psi. As we know from supercharger theory, the more boost pressure (and the higher the pressure ratio) the more compressed the air is. Air at 15psi of boost is actually half of its volume compared to 0psi (or 1psi). So making 700hp (1050 CFM) @ 15psi (on a 3.5 liter 6 cylinder for example) may require only 42 squared inches of cross sectional area (because the air is at half its original size) whereas making 700hp (1050 CFM) @ 3psi (on a 7.0 liter 8 cylinder for example) may need a larger 91 squared inches of cross sectional area. So make sure you factor in your pressure ratio before choosing your cross sectional area.

Here’s my second chart:

Horsepower vs Core Volume (Cubic Inches)

This is a plot of horsepower (vertical) vs. total core volume (cubic inches) for the 30 cores that I had data for. As you can see there is a linear relationship between horsepower and volume which is expected. The more horsepower we want to make, the more air we need to ingest. The more air mass there is; the more energy that mass can carry (at the same temperature compared to a smaller mass) and thus the more intercooler core we need to sink that energy into our intercooler.

I think between these two charts it becomes now possible to go back to my ‘twin-charged’ Toyota Celica and say:

I wanted to make a peak of 320hp @ 20 psi. That equates to 480 CFM @ 2.36 Pressure ratio.

Starting with a standard 3” deep intercooler core, let me figure out my other 2 dimensions:

Minimum cross area = ((480/2.36) + 12.84) /11.63 = 18 square inches = D*H

Intercooler height = 18 / 3 = 6”

Total volume = (320 – 50.17)/0.533 = 506 cubic inches.

Intercooler width = 506/18 = 28”

So my ideal core size seems to be 28” X 6” X 3” which is a pretty reasonably sized front mount intercooler.  

Now 28” is a reasonable intercooler width for pressure drop. If this figure were too large I would go back and use a 3.5” deep core for example. Likewise, if my intercooler height of 6” would not fit behind my bumper I could go back and increase depth slightly and redo the calculations.

Pressure drop across the intercooler is really important to track for a supercharged car because unlike a turbocharger, we can’t just increase boost pressure with a boost controller, we are limited with superchargers to the gearing we have available in our supercharger pulley. So wasting any of this boost is really bad for performance. This is why it’s really essential to neither undersize the intercooler to choke off the engine, nor to oversize it as to create a big pressure drop.

For more intercooler information:

A great website by ARE cooling about intercooler design

Technorati Tags: CFM, How To, Intercooler, Supercharger