115, 90 & 75mm bar/plate on left. 73,57,37mm tube/fin in right.	Our bar/plate on the left & K&J tube fin on the right.	Nissan skyline oe on top. Adrad on left & K&J tube/fins on right.
Real close up of the three different header plate designs.	Difference  in fins & tubes between K&J closer & ADRAD cores, even though both are tube/fin	The extruded tubes. Integral strength over bar/plate. We remove any swarf before welding or it's straight into your engine !	When I was in the factory, we discussed rib size, flow tested the tubes & this is the result - 1.28mm ribs instead of .80mm

                     Plate tube & fin cores are the physically weakest core in that we have allot more of them come through our workshop for repairs than any of the other three, & too reinforce that, they usually have a larger leak for the damage &/or the leak be further away from damaged area. The core has three components too each tube. Two stamped side plates with fine folded fins fused internally. I believe the plates are "stretched" too far to form the tank on each end, thinning the alloy to where it will crack with minor stress. They are clamped together (looking like a piano accordian) & furnace brazed.                       Bar/Plate & fin cores are the second weakest, even though they are also the heaviest with a couple of the thickest components. However this is the reason for their weakness. The thin plate side walls have to fuse onto the thick plate sides to form a "tube", with both of them having to fuse onto the very thick bars to form the "header plate" for the tank to be welded to. It is hard for these three differing thickness to be held at the right temperature for the cladding to melt,     ??   & fuse the surfaces together. This is where we see most of the leaks in these cores, even from new in one brand. The charge cooling fins are fused to the thin plate walls, meaning there are 5 components too each tube, a reason that these cores tend to develop a series of weeps along the "header"  seam.     

                       Folded tube & Extruded tube seem to be of similar strength although with two tube sides, fine fused internal fins & two header plate joints, folded tube should be weaker with 5 components. However, I have seen both types of cores come in for repairs quite beat up & out of shape, without leaking. Eventually the hot/cold & pressure/vacuum cycles will open up cracks.                                                                                                                                                                                      The Extruded tube core is possibly the strongest core as it only has three components in it's construction, a tube & two header plates. It is also the most consistent performing unit as the internal charge air cooling ribs are one piece with the tube walls whereas the above three cores have tubes with fins fused too them & if the specifications vary in production for any reason ( & it does occasionally), then an area of fins can exist where they are not in full contact with the tube wall, bottling up the heat in that area. The same also happens with external ambient finning, although this is common to all four cores & can be found with a simple visual/physical check.

 Header Plate/tube joints, bar/plate joints

Bar/plate top & tube fin bottom. Note how the header plate is proud approx. 7mm/side making a wider overall 'cooler.

Same pair intercoolers from opposite direction. difference is more graphic.

Folded tube/fin on top is okay, tube/fin bottom left is best & right is worst for air flow entering tube.

Bar/plate on left has intergral header plate, whereas tube fine has to be proud to weld  tank to.

The 8 components that make a complete tube in a bar/plate core

The 2 components that make a complete tube in a tube/fin core. Note flux/cladding around tube.

The 3 components that make a complete tube in a folded tube/ fin core. This has a tabbed header plate to hold tank on.

The 2 components that make a complete tube in a plate tube/fin core, including the tank - shown cut in half here.

I have not worried about fin/ribs in the above pics. as they are common too the internal structure of every tube & would unnecessarily clutter the pics. They are highlighted in their own section.

  Finning - or Ribs.

Both Adrad cores. Different fpi.'s for different applications.

Adrad top & K&J  bottom, different fin shapes & pitches.

Garrett top & ours bottom. Straighter uniform fin allows more air through.

Charge air fins, straight uniform fins allow intake charge through with less pressure drop.

More open ribs of Adrad give 17% less pressure drop. Radius on ribs adds to surface area.

Longer ribs help pull heat out of air better, but more restriction.

Cut away bar/plate core shows fins at 90 degree to each other, very efficient but high pressure drop.

This is actually an abient air fin, but very similar to charge air fin. Louvers direct air flow but also cut flow dramatically, so fpi important

The parts  of an intercooler that do all the "work" are the fins, or in the case of extruded tubes, ribs. Their purpose is the opposite when in charge air flow to ambient air flow. In charge air flow, their job is to soak up the heat in the passing air & deliver it to the tube wall. The trick is too pull the maximum heat out of the air with the least resistance to flow. The different designs are flat fin (cheapest & least effective), punched fin & closely followed by louvered fin (dearest & most effective at heat soak but with highest pressure drop). Extruded tubes are formed by forcing an aluminium block through a die resulting  in a complex one piece tube, offering 100% heat transfer between the rib & tube wall. This does not always happen with fins, because sometimes the factory does not apply the correct "crush" to the core when clamping in assembly, resulting in no bond between the tube/fin during furnace fusion. It appears to be more of a problem with internal fins than external, but luckily, is not that common. Extruded tubes have internal walls that form continuous compartments with ribs inside each compartment. These vary between manufacturers, but the same rules apply - more ribs/inch, more heat soak & pressure drop. Also the more ribs in contact with the tube wall, the more efficient. Because of the extrusion process, ribs are thicker than fins & there is always more open space in the middle of the tube. This results in a small reduction in heat transfer if the tube is too long, as the air " straight -ens out"- reduces turbulence due to the smooth compartment & rib walls. With Ambient air flow, the fins job is too radiate the heat, so the air passing through the core absorbs the heat & transports it away from the core - usually straight into the radiator finning - but that's another problem! Adrad fold both fin edges over (hem the fin) which doubles the strength fin & lessens damage. Damaged fins should always be straightened as they inhibit air flow & so cooling efficiency.  The higher the vehicle speed &/or  the more direct ducting to stop spillage of the air around the core, the closer the fin pitch can be :- an offroad  buggy or hill climb car needs less fpi. than a circuit racer. Some thought has to be given to radiator cooling or you can create a problem that can limit the engine output more than the intercooler can increase it! Intercooler position, surface area, fpi., tanks & workload, all have a bearing on engine water temperature.  

TANKS - the single most underrated part of an intercooler !!!

This tank is extremely basic & because the outlet is in line on the other tank,- even worse!

This is our tank to do the same job So much more efficient! How much more expensive?

Line up of our Hot Chilli tanks. A couple of sizes still to come. GTR type this end.

Note the over hang a tank has to have because of the construction of tube/fin

Not all cast tanks are created equal. We could get our tanks cheaper, but we pay for quality/performance.

Our cast tank walls average 3.6mm thick, negating heat soak problems assoc. with tanks that are 6 -7mm.

The areas between the lines are the "tanks" on a Denso intercooler. The large boundary layer is extremely turbulent causing a high pressure drop for no gain.

This is where allot of my knowledge comes from, hundreds of hours testing but hundreds upon hundreds more fabricating parts too test. Results do not always cross over to the real world. Alex of Bryant Engineering.

The thickness of the alloy in tanks is very important. Too thin & heat soak is negligible, but their life span maybe short due to stresses from the continual hot/cool & boost pressure cycling. too thick & they'll last forever, but suffer heat soak, which is the time it takes the tank temperature to react to ambient &/or charge temperature changes. Depending on the use of the vehicle, the effect of  heat soak varies. Constant high speed static throttle position with a front mount (road transport) is the least effected. A four wheel drive in soft sand or with a heavy boat up a ramp, fitted with a top mount, is the most effected.  We spent 14 months making sheet metal tanks for next to nothing while I looked for a good pattern maker & foundry.  It's our customers gain as our tank walls average 3.6mm thick.

Shape or Profile

Most of the charge flows through the bottom 35% of core, with the top 65% causing detrimental pressure drop  until high boost is reached. Worse again if other pipe is on the bottom too !!

The 2  corners cause large eddies & the square inlet corners small eddies which restrict flow & distribution across the core window. Worse if pipe is on bottom other end !!

Not perfect, but very good & very cost effective. Nowhere near as important to have diagonally opposite tanks like both designs on the left. Very small eddies & small  uneven distribution across core window

Square shaped/edged tanks are taboo for performance conscious enthusiasts for four reasons:-

1) They look basic, agricultural & ugly 

2) A square edge concentrates stresses along that fold (or weld), which, depending on several factors, will shorten the life of the tank to a some extent. It's surprising the number of small weeps that can go undetected for months, especially if the turbo seals are good (no oil stains), resulting in a loss of economy/performance. This usually applies more to a sheet metal tank than cast tanks, due to their thinner wall construction.

3) Most importantly is that the square edged tanks hurt air flow & increase "static" pressure drop. The intake charge does not want to make a sharp turn with the majority of the air taking a radiused path around the edge, causing some of the air to eddy back around through an elongated 360 degree tumble. This causes friction with both the wall surface & the following air, really hurting flow figures. 

4) I don't know if any manufacturers ever think of this, but when building tanks for a tube & fin core, if more air is directed to the front or back of the tubes, then that is how it stays for the entire length of that tube, as the extruded bridge walls contain that charge & it can't mix until it reaches the outlet tank, resulting in a cooling efficiency loss. With a good bar & plate core (not a cheap plain flat fin) this can't happen because the louvers or holes in the fins allow the charge to pass across the "tube" as it travels lengthways. This problem doesn't apply when the inlet pipe is inline (parallel) with the tubes & only applies a little when big boost is being used. 

 In our Hot Chilli range of tanks, the angles & radii of the triangulated roof are different for each size tank. This is not a marketing gimmick, but shows how important it is to get it right. Only three degrees sometimes made a difference of 2% of flow but 10% of window distribution.

Configuration

Two designs of tanks are very poor in distributing the intake charge evenly across the core face. Please note that 100% distribution is impossible, if the core is high enough to offer worthwhile cooling.

One design is commonly used on Rotary powered cars coming out of Sydney. The Core is 500+mm high & has the inlet/outlet pipes in the bottom of both tanks at 90 deg angle, parallel to charge flow. However the higher static pressure drop to cooling efficiency ratio, is less important to a 20+ psi. drag car (it will increase 'spool up' time! ), but decreases throttle response & a few kw. output, in a street car.  In the pictures (below) of our answer to this problem, we have designed a dual pass intercooler that splits the core area in half, so the intake charge uses the whole of the core evenly. Attention has to be paid the the sizes however, as the air speed through the core is doubled & if the charge window is too small, a pressure build up (drop) will result.                                                  

4 of  9 tank shapes built to try to maximize distribution across the core face.

From lower down to show some entry changes made. Front 2 are absolute shockers.

In each tank we tried different baffle shapes/angles & got some mixed results. Some hopeless.

In our initial testing, 80% of charge air used end 25% of core at 264 cfm @ 25" water. Just turning bottom tank around makes a big difference

If the outlet is below this inlet, this is the worst combination I can think of. Step causes reversion, smaller diameter speeds air flow & increasers syphoning !

This is the second worst combination I've tested & it's so popular with the Rotary set. ARE has never sold an intercooler like this - allot missed sales, but we now have a better setup.

This is our solution too cooling the Rotary. Maximum effeciency from this large core giving more cooling with less pressure drop. Relocate the oil cooler & we can make it even better.

Here is how our intercooler looks in the car. Still fills up the front but gives more overall cooling with a little less pressure drop. There is a limit to the cfm flow through these cores & generally need to be thicker.

The other design mostly is used in the USA., where tanks are on the top & bottom of the core, causing the charge air to make two 90 deg. bends. By using a second manometer with a long thin probe, we found that 80% of the intake charge flows through the end 25% of tubes on our first test tank, which was of reasonable length. Flow rate was good but cooling rate was terrible. When we doubled the length of the tank, 55% of the intake charge flowed through the end 25% of tubes & overall flow dropped 9%, indicating that this design has too have very long tanks, but should be able to be avoided anyway. We also discovered flow 'syphoning' for the first time during this testing. The intercooler was making a strange noise, & with the aid of the second manometer probe, & then removing the tank & using the old tissue streamer on a stick trick, found that air was actually being sucked  back up the tubes next too the inlet end, by the same principal as a spray gun operation. This means that some of the already cooled intake charge was being recirculated through the core again. This creates a double loss as not only does the extra syphoned air create more pressure drop, it has already been cooled so takes away core space needed by the hot air. I didn't waste time measuring the % recirculation as we rarely use this design. Please note that this paragraph refers to intercoolers with the inlet/outlet above each other (on the same side). It is reduced noticeably if one tank is reversed, but still should be avoided if possible. A dual pass intercooler is heaps more efficient where both pipes are on the one side, but they require more height for the core. 

A few factors should be remembered with air flow. Definitely don't polish the inside of tanks or pipes. An ultra smooth surface causes capillary tension on the surface layer of air with the wall & will increase drag so increasing pressure drop.                A street car is not in high boost that much of the time (maybe 10% max.- or no licence, no tyres!), and for possibly 50% of the time in vacuum, with maybe another 30% at less than 5psi boost, so the engine operates as a naturally aspirated engine for approximately 70% of it's life. Now, how many hundreds of hours a year do the Group A Supercar teams spend on the flow bench trying to find even a half a percent increase in flow? A square edge tank can easily cause a 20 percent drop in flow, at low boost driving !! Some people confuse the result of poor tank/pipe design with turbo lag, when it actually may have little or nothing to do with the turbo's performance. This has been reinforced with our Air/Water intercooler R&D. program, where much shorter/straighter pipe lengths are used (reduced "turbo" lag). A crisper throttle response with increased "driveability" can actually be felt driving the vehicle,  very important not only on the street, but in Rally, Hillclimb, Gymkhana, Jet - boat, type competitions. In my flow bench testing, I've found that making air flow change direction by 360 deg. causes noticeably more pressure drop than a 180 deg. change.

The importance of squared tanks exponentially decreases as boost pressure increases, this is for air flow & distribution. If you see a car at the drags that runs 35 plus psi, with square tanks on it's intercooler, don't worry about it after the car has launched & is revving out, as power loss is minimal ( maybe 3%), BUT, if the engine could not spool up quickly, or needs nitrous to spool up, then the tank shape plays a very definite part in this problem, with Rotary engines seemingly affected more than 4 stroke engines. I have seen many drag cars, not being able too spool up onto the converter for 1 to 3 seconds, usually mostly in the afternoon heat - over cool night air, resulting in up to a quarter track lead to their opponent. The importance of the inlet tank shape also exponentially increases with the thickness of the core, from when it is 25% wider than the inlet pipe,  & especially for tube/fin cores.. The outlet tank only has approx. 40% influence on this as the inlet.   I have spent hundreds of hours on the flow bench with everything from conventional basic (opposition manufactureres) tanks, to some way out designs that would be laughed at, but you can't always accurately predict the results, & it's marvellous what I occasionally learn from these experiments. I reckon 85% of my r&d. time is aimed at tough street car applications to provide the most cooling with the highest horsepower output with the crispest throttle response. Note the word economy was not used, because it goes hand in hand with crisp throttle response.

Material

Not a pretty sight. Too thin material used.

A closer pic of the thickness guage. We use 2mm for hand hammered (formed) tanks without flat surfaces & 2.5 or 3mm on everything else.

Macro picture of 3mm sheet metal tank surface. Fold made with 10mm dia. mandrel. Can see surface cracks if made with sharp edge folder.

Surface of cast tank. Both polish up similar, but cast takes longer. Can have inclusions or blow holes that will leak if not cast proper;ly.

If the tank wall is too thick is  it will become a heat soak & counteract part of the advantage your intercooler provides. Worst example is a thick cast tank fitted to a Subaru WRX top mount - or any top mount. It is getting all the under bonnet heat (around 70°c.) at low speed & then when nailed, takes a couple of seconds to drop to it's proper temperature, which isn't that cool anyway. This has much less effect on a front mount, except when reversing back after that big impressive burnout & it won't stabilize when staging either. We pay more for our cast range of tanks, but they average 3.7mm thickness & are well worth it because of the formentioned problems. 

Can an intercooler be too big?

 YES ! In street driven applications, an intercooler can definitely be too big, maybe not for 'attitude', but for maximum performance. It is not an argument to say that Joe Blow's RX2 runs 9.6 @ 132 with a monstrous intercooler because that car may run 9.4 @ 134 with an intercooler that is engineered very close to the overall requirements of that combination. The same may be said for Fred Nerks TX3 that runs 13.8 @ 100 with his headlight to headlight monster intercooler jumping out of the grille. The car may run 13 flat @ 110 with a properly engineered intercooler/pipe setup. This is because if a 13,500 cu. cm. core drops the intake temperature to 40° c. @ 1.5 psi. "static" pressure drop, a 9,550 cu. cm. core may drop the temperature to 41°c @ 0.82 psi "static" pressure drop, & I know which will produce more power ! 

The closer to ambient temperature the intake charge gets cooled to, the exponentially larger the internal surface area of the core is needed, also exponentially increasing the friction (drag) of the charge air in the larger core, causing higher Static pressure drop. Please note that the same kw. output can be generated by two engines, but require two different intercoolers. If  one engine is fitted with a smaller than optimum turbo, then it will operate out of it's efficiency range, over revving & 'beating' the air, to a higher temperature. Even in a street application 40+ c. (eg. 120 to 160° c.) is common! This requires a more efficient cooling intercooler, with either larger volume, more cooling fins or longer tube length, than the other engine that is fitted with a larger than optimum turbo, which may only heat the charge to 105° c., placing less importance on the intercooler, but more on the inlet tank "window" (area) which needs to be larger & the tube length shorter. The first engine makes a better street driver, as the turbo will be on boost at considerably lower revs.

A really important point is that the more oversize an intercooler is, the more important tank shape is, for both pressure drop & charge distribution across the core face. Make an effort to suss out the best overall intercooler & pipework for your car, usage & driving style. It will make a worthwhile gain, & possibly at less cost! 

Turbo Lag ?  or is it
The internet is a great thing, BUT, there are so many false statements that people make sound so convincing. "I put a big front mount on my VL Commodore  & there's no extra lag at all, it goes heaps better too". Treat this person with total distrust. The only accurate (note: I did not say honest, because he may be genuinely kidding himself) part is ' it goes heaps better'. It is a physical & scientific impossibility, to not have extra lag in this case. 
If it was a Skyline or similar car, treat the person with some distrust & quiz them on both setups, as the system they replaced was certainly not optimized &/or the new system must be really suited to the car. 
    --  These figures are calculated for a system with perfect air flow, without disturbance from  poor tank shape, poor core flow efficiency, gaps in pipes, no. of bends & many more factors, so the actual real world times will be slower !
    -- The actual time a molecule of air takes to travel the system during , from idle, to 8500rpm., cannot be accurately calculated as the variables are immense. These four graphs are to help you get a handle on how important it is to run a properly designed system for street throttle response & race spool up. This is part of the reason a high stall auto, turbo, 4 cyl. drag car is usually allot more consistent.
    --  Note: -- The smaller the engine capacity - or the lower the revs used - the more important the design is. 
                 -- Boost has miniscule bearing on charge air speed/time - it changes the cfm. output, making the charge denser, BUT not the air speed velocity ! 
                 -- The lights on an ANDRA Xmas start tree are 0.400 of a sec. intervals. 
    -- We have been using these figures (in house developed computer programme) for years to help us design ic. systems for customers, maybe that's why so many of our customers speak so highly of the results our products provide.
    -- Following are four PDF files to validate the above.

	2.0Litre engine putting out 668cfm. @ 8500rpm. Core size - 610x278x109mm. 2.5" & 3" pipes. Skyline GTR tanks. 26.853L. 800rpm. -1.673 secs.  8500rpm - 0.167 secs.
	Exact same engine specs as above. Core size - 500x278x109mm. 2.0" & 2.5" pipes. Evo tanks. 19.477L. 800rpm. - 1.227 secs. 8500rpm - 0.123 secs.
	Exact same engine specs as above. Core size - 150x380x109mm. 2.0" & 2.75" pipes. As our Subaru WRX top mount upgrade intercooler- not as efficient (at cooling drop) as a front mount, BUT --    800rpm. - 0.517 secs. 8500rpm - 0.052 secs.
	3.1Litre engine putting out 1193cfm @ 8500-remember boost has no bearing on air speed!   All other specs exactly as the first file above. 800rpm. - 1.079 secs.  8500rpm. - 0.108 secs.

Pipes, Couplers and Clamps

I know they're not intercoolers, but your shiny new $1000 piece of alloy lurking behind the grille isn't much use without them. Pipes also have to be engineered or they can cause sizeable pressure drops.  The worst thing you can ask the charge air too do is turn back on itself ie. make a 180 deg turn to the right & then a 180 deg. turn to the left, like an S shape. Once air is deflected it likes to keep travelling in that direction. Asking charge air to open out at 360 degrees, as in a 50mm pipe into the middle of a 90mm flat rectangular tank,  causes a big pressure drop.   Another problem is Heat soak, especially when the throttle body faces across the tappet cover on a north/south engine or at the rear of an east west layout, making for a long pipe length from the radiator support panel to the plennum. Why is this a problem? We have measured under bonnet temps. of over 80° c. & that is not down near the turbo either. On a 26° c. day, the average charge air temp. out of our front mounts is 30° c. & this low temp. makes the charge very susceptible too heat soak from the high under bonnet temp.& naturally, the longer the pipe, the higher the rate of heat absorption. In these cases, even though we work in alloy predominately, or stainless, I recommend to do the pipe work in mild steel mandrel bends & have it HPC (or similar) thermal coated, it is worth the extra time & could be cheaper. Worst case scenario we have measured between the intercooler outlet & the plennum inlet, is 21° c. (a 25° day) on a Commodore VL with the pipe behind the radiator, with a guy in Cairns telling me he measured 26° on his Skyline R32 GTST 

When it comes time to by couplers, if at all possible, pay the extra money, buy the proper high temp. silicon ones (they look good), fit them & forget them, after you retension the clamp a week later, they do shrink under the clamp compression. I know hump hoses cause a little pressure drop, but that can be better than cracking pipes through not having enough give in the system. If there is only a short distance for the pipe, & it is between a fixed point (intercooler etc.) & the engine, use one or two humps too allow for engine movement.

Clamps are important in both sealing & longtivety of the couplers.

Intercooler Maintenance

This is the worst picture since 1997 of one of our customers horror story.

Enough too make you sick !!

Japan really must have terrible air. The black buildup actually corrodes into the alloy & must be cleaned.

This is how it should be. Back to near new efficiency, & looks the part.

An intercooler that has been in service for a couple of years (longer in rural areas), will definitely provide more cooling if it is cleaned properly. That is, soaked in a degreaser bath to remove the interior oil film that has accumulated (which may be ever so thin with a new turbo), then power flushed with water. The exterior fins should sprayed with a proper aluminium etching cleaner to remove the surface oxidation & road grime that will have formed. Both the oil film & oxidation/grime buildup provide insulating coatings that inhibit the transference of charge air heat to the ambient air. 

If not sure of the proper procedure & products to use, we strongly recommend having this done by a professional business in our field. Please see the two pictures above left, which is a 9 month old intercooler that was removed off the car for another reason, & the owner decided to do the right thing & clean it to be sure of maximum performance. He rang us & I told him we use a Degreaser, Kero & Phenol solution in our bath. He already had some degreaser at home, so bought some kero, mixed them, blocked off one end & poured the mix in to let it soak. It started to get warm after a short while, then bubble gently & get hotter. By the time he realized it was going awfully wrong & tipped it out, this was the result. It was late Friday afternoon when he rang, both of us were in a  hurry,  I didn't say & he didn't ask about different sort of degreasers and he didn't read the label which said 'not for aluminium'. It was a water  degreaser he had & mixed with the kero. The caustic base of the degreaser attacked the alloy & rendered the core completely useless, a new core being the only alternative - expensive mistake for both of us, as we helped him out with the cost. I rang a few industrial chemists & the answer was the same from them all - over the left shoulder.  For internal cleaning, only use a petroleum based degreaser, which we mix with Kero & phenol. Specific acid cleaners must be used on the external cooling fins as the average thickness is only 0.08mm - that's paper thin. Any bent over fins should be straightened  before the cleaning process, so the fin is cleaned properly & full air flow is allowed through the core. 

Check all the mounting brackets, both on the cooler & car for any signs of cracks. If you have to undo or loosen any bolts, take notice that the brackets don't 'spring' or have tension on them. If so, fix the problem so the bolts go straight in easy. This will stop any stress cracks forming or spreading.

Check all the hoses for any signs of deteriation, especially next to the clamps & in the centre of any convolutions if the hose has them. These are by far the two most common problem areas. When buying hoses, do your car a favour & pay the extra for proper high temp silicon hose. Buy it once, fit it once, & forget it (after checking clamp tension after a week). Remember that hoses shrink under compression of the clamp, so when reassembling, make sure that the clamp not only goes back in the exact same position but also around the same way, as the tabs under the worm drive mechanism are asymmetrical & could allow charge air too escape. Same with radiator hoses too. When buying clamps, always make sure the band part has a rolled edge on both sides, this stops cutting the hose with time, as they must be very tight if boost is over 10 psi. If over 20 psi. I recommend using the T-bolt clamp.

One really top, time saving trick I was given by Matt @ Chiptorque (Nerang) is to plumb an air line set at a couple psi into the inlet system & listen or feel for the leak if you're having trouble tracking it down or have doubts. A small leak is next to impossible to diagnose with the engine running sometimes. 

Repairability

This oe Nissan GTR core does not leak. Because there are no sharp creases in the damage, it will probably give several years service.

This Skyline GTST is going for scrap. Tubes probably don't leak now, but tubes torn out of header plate & tank holed.

Close up of damage. If "will leak" tubes aren't leaking now, they soon will with hot/cool & pressure cycling

When considering tube construction, one thing to remember is the strength & repair ability of the cores. The weakest core by a very big margin is the plate tube & fin core. We get allot of these through our workshop (including oil coolers of the same construction), but they are relative easy to repair as they mostly crack in the centre of the end tank concertina joins.  If the crack is in the centre of the "tank" end, then it is a throw away. They have to be hydrogen brazed as the heat from TIG welding is too concentrated, usually resulting in a weep at one/both ends of the welds. Bar & plate cores are weaker than tube & fin.  The worry is that Garrett paperwork says that it is acceptable to have leaks in the core, as long as it doesn't exceed 4 psi drop from 30 psi in 15 secs !!!  We may have had a leak in a tube & fin core over the years, but no new intercooler has ever left our shop with a weep, let alone leak.  Bar & plate cores are very difficult to repair leaks anywhere on them because of the vastly differing thickness components , but tube & fin are only hard when the leak is in towards the centre of the core by more than 15mm.. The wall thickness  of an Adrad core is 1.3mm front/rear face & 0.7mm for rest of tube and K&J is 1.0mm front/ rear curve & 0.4mm sides of tube. Any competent aluminium welder will be able to weld damage to the tubes, but burning the fins away will sort out their talent !  This advice is from our workshop, that is one of the few set up too repair properly as well as fabricate new assemblies - in fact there is not another shop in Australia, set up anywhere near as comprehensive as us, for repairs !  

Chinese Intercoolers

Import or 'Grey' Intercoolers

These are a much preferred option to the Truck intercoolers, talked about below, in fact, some very good units can be bought for a value price. If the unit has not been stored properly after it was removed from the vehicle (& this means in the wrecking yard in Japan, loading onto & in the container coming over here & storage over here), then it may in fact, cost you money &/or performance. They have been brought in to us for cleaning with a line of whitish powder 1/4 to 1/2 way up the core internally, from where they have been laying in the salt water coming over. Don't touch them. Sometimes, they have looked real good, but after we have soaked them in the bath & while power flushing, the amount of pebbles, dirt, & the odd self taping screw that comes out, is a real worry. Firstly, too us that we have got all the rubbish out, & secondly, that anybody buying one & fitting it straight up, has been lucky & it is clean, or else whatever is in there, will finish up in the engine sooner than later.  Paranoid? no, seen it happen too often.

If an import intercooler is for you, then look for a unit off a car that has at least 50% more power than your engine, as oe. intercoolers are sized by accountants, not engineers. Even if the engineers do get the final say, I'm sure they know how much some of you guy's will turn up the boost & so have a small intercooler to try & "choke" your gains ! 

Truck Intercooler Cores

There are differences between a 'truck intercooler' and an intercooler fitted into a performance car application.  However, 'truck intercoolers' is a broad term and does not sufficiently categorize the product.  Just like other types of products, truck intercoolers come in different configurations -  both bar and plate and tube and fin design. I have not seen a plate tube & fin core in a truck, they would be too weak.  However, due to their specialized application, truck intercoolers have a noticeably courser internal fin &/or external fin pitch (depending on whether tube and fin or bar and plate), than their cousins fitted into normal performance car applications. This means that although a new truck core can be used in performance applications, the core surface area/volume must be larger  in proportion to that of a standard 'performance core', & then it will give good temperature drop, but it will be at the cost of a medium/ high "static" pressure drop & higher radiator water temperatures. 

A truck intercooler core is nearly always larger because even though the engines are slow revving, they have a large swept capacity requiring a high cfm. intake charge, also there average road speed is much slower & the engine is under load most of the time. Because of the slower road speed, thinner cores per area and with less fpi., are used to keep the air speed onto the water radiator as high as possible. 37mm thick is used by truck manufacturers in there lower powered applications, but you hardly ever see them in a performance application. Trucks also have a large frontal area, giving more room for the intercooler/radiator.

We cleaned out a Volvo truck intercooler (it had plastic tanks) & during the job, discovered that it had no internal fining, turbulators, nothing, inside the charge air tubes at all. When the customer came to pick it up, I asked him how it going on his car (Cordia), & he said brilliant ! He had never measured the temperature drop, & I guess he drove pretty easy, so it may do a reasonable job, but put it on a Commodore & it would be hopeless.

The larger volume of a truck intercooler has three main drawbacks:-
1)  Higher 'static' pressure drop
2)  Increased 'turbo lag'
3)  Higher engine water/under bonnet temperatures - in front mount applications, as the large truck intercooler blocks more of the air flow through to the radiator.

Therefore, if money is a major issue than new truck cores in a performance car are a viable BUDGET alternative, especially if tanks can be made & fitted at "mates rates"! 

Used Truck Intercooler Cores

It is my opinion that used truck intercooler cores should not be used in a performance car.  Using a used truck intercooler core is a huge gamble, are you prepared to risk the welfare of your engine because of a cheap intercooler.  Why put a $100 core in front of a $2000+ engine? It also is a slow job to do properly, as compressed air has to be blown through the tubes, from the opposite end to which you are working on. 

Most truck intercooler cores can fit into a performance car application, but they nearly always have to be modified.  During this modification process, whether it be when the old tanks are being cut off, or when the core is being cut down in preparation for welding tanks on, or when the tanks/core is being cut down, swarf and fillings can get lodged in the core, some of which can be up to 600mm in length and can be imposable to blow or clean out.  The result is that they dislodge days or weeks later, after working their way through the intercooler tubes and intake piping and end up passing straight into the engine.  I'm sure you can imagine the result of that !? Another point is that bar & plate and finned tube cores are much, much harder to clean any solid particles out of as they get trapped in the fins & have to "travel" through the length of the tube, interestingly, the exact same reason why per sq. cm., these cores dissipate heat better, but at a higher "static" pressure drop.

Remember that a used truck core is for sale for a reason. If you buy one, hope that the truck was hit up the arse, because if it was hit around the cab, then the core was more than likely damaged or twisted to some extent. If the damaged section was cut off, your core will still have some stress & 'set' in it & even if not leaking when purchased, the heat & cutting when modifying it to fit, may release the stress, resulting in leaks in the header plate/ tube joints when finished. We do not do this work anymore for this reason, it happens far too often !

Picture of core in test tank.

Used truck cores in a performance car are a stupid, desperate alternative, & if you haven't guessed, a pet hate of mine. Use them in your smoky turboed A12 powered rusty 120Y, but please not in a performance car. I'd much rather see you buy from an opposition shop !!

Last words.

 Intercooling a forced induction intake charge is a compromise. We direct all our energies & R&D., at minimising the negative compromises & maximising the positive compromises!

A 20% gain in air flow through a component, does not give a 20% increase in power - I wish it did ! It might only result in a 0.5% (or 5%) gain in power - it all depends on the application.

If you have read this far, I thank you kindly, because it has taken me allot of work. I've tried to be as honest & unbiased as I could, so I really hope you found it both worthwhile & especially educational. Spend your money wisely, & you'll always be ahead.

Short Glossary (for a our full edition click here)

* "Static" pressure drop - The measurement of the drop in pressure of the air travelling through the core, friction or 'parasitic drag' measured on a flow bench (of capable capacity) at ambient temperature.

* "Dynamic" pressure drop - The measurement of the drop in pressure of the air travelling through the core, friction orO 'parasitic' drag plus the drop in pressure caused by the cooling of the intake charge (closer molecular structure of the particles resulting in a denser, smaller volume exiting) due to the mechanical design of the intercooler, or in more engineering terminology, thermal matrix heat exchanger!