DIY charge cooler upgrade

[Craigshag]

I am thinking of upgrading my charge cooler (rad for supercharger) there are kits out there for about 1k but, i was thinking of having making my own with an aftermarket rad.

Has anyone on here made there own kit with off the shelf parts? E.g rad, booster pump etc.

 

[mercules]

I am thinking of upgrading my charge cooler (rad for supercharger) there are kits out there for about 1k but, i was thinking of having making my own with an aftermarket rad.

Has anyone on here made there own kit with off the shelf parts? E.g rad, booster pump etc.

Just get the Merc racing one from the states mate it's good quality rather than flapping around with what fits and and materials etc. I paid around £400 all in if I remember correctly. Just get some hose ,clamps and coolant in the UK.

Sent from my SM-G950F using Tapatalk

 

[RyanJonS4]

Xpower used a rad off a Jaguar I think, sure someone will chime in and tell us which one. I’m pretty sure it was a fair bit cheaper than the merc racing one as well.

 

[mercules]

Xpower used a rad off a Jaguar I think, sure someone will chime in and tell us which one. I’m pretty sure it was a fair bit cheaper than the merc racing one as well.

Yea he did a dual stacked one so I think kept in the stock one just added the Jag. There's a couple posts if searched.

Sent from my SM-G950F using Tapatalk

 

[forty3]

Just get the Merc racing one from the states mate it's good quality rather than flapping around with what fits and and I materials etc. I paid around £400 all in if I remember correctly. Just get some hose ,clamps and coolant in the UK.

+1 for merc racing

 

[Cerbera9]

I also had the Jag dual rad set up then swapped to the Merc Racing HX as it’s a neater install. I wouldn’t say there is much performance difference between the two but they are both better than the OE set up.

 

[Adamantium]

If it just takes a larger radiator, why is there a market for these expensive kits?

 

[forty3]

If it just takes a larger radiator, why is there a market for these expensive kits?

Clever branding?

 

[jdp1962]

If it just takes a larger radiator, why is there a market for these expensive kits?

Because if there's a way to separate people from more money than is necessary, someone will find it and exploit it.

 

[Adamantium]

Are you just being cynical?

I can accept that a new header tank increases the volume of water enormously.

I also like the idea of divorcing the systems as it allows you to run different antifreeze concentrations in each. Granted I’d rather break my supercharger, but I’d rather risk that than risk my engine block.

 

[Dippy]

No. Cynical people are not allowed on the forum. Jeff is a moderator. He bans people at the rate of 2 or 3 each day.

 

[jdp1962]

Are you just being cynical?.

You think?!

 

[Adamantium]

Obviously missing a "not" in my post above - not dead keen on breaking my supercharger.

 

[jdp1962]

No. Cynical people are not allowed on the forum. Jeff is a moderator. He bans people at the rate of 2 or 3 each day.

I’ve been good lately; haven’t banned anyone for almost a week now.

As to cynicism and @Adamantium ’s question, I’m not experienced in modifying cars, but am pretty experienced in life, chiefly because of being so blimmin’ old. And I tend to find that in any venture, the scope for spending and achieving respectively ranges from the minimal to the excessive with the optimal laying somewhere in between, and there isn’t always a correlation between the two.

 

[forty3]

The header tank isn't very big with the revo and forge kits so doesn't add much volume. The Apr kit has an additional reservoir as well as the header tank which adds a fair bit more coolant but is very expensive.

I divorced the system then used the (very small) expansion bottle from a vw up as a header tank.

Did a track day on Friday and didn't notice any loss in power after driving it very hard so the merc cooler has definitely done the job for me running a stage 2 tune.

Was a nice sunny day as well so ambiant air temp was around 20 degrees

 

[Dippy]

@forty3 - have you done any VCDS logging during your time on track? It would be interesting to know exactly what the various temperatures are.

 

[youngsyp]

If it just takes a larger radiator, why is there a market for these expensive kits?

@jdp1962 covered this perfectly. It's something that happens everywhere, not just in the car modifying space, as there's always someone who will believe the hype and pay through the teeth for it.

There's a wealth of information about this subject around the web and when I looked into the same topic, the conclusion I came to was that upgrading the OE core to a Merc Racing replacement was all that was required to maximise any improvement.

In essence, there's little point in divorcing the charge cooling system from the engine coolant system as their interaction is minimal and thus any reduction in coolant temps with them divorced is equally minimal.
A higher flowing pump isn't necessary and will actually reduce the efficiency of the upgraded system (with just the uprated core fitted) due to pumping the coolant through it at too higher a rate. The coolant needs to be in the cooler for a specific length of time to take any advantage of the core's functionality. Audi who spent millions on the development of this platform, just about nailed that with the pump they fitted...

To your point about running different coolants in the divorced systems. Whilst that's a nice idea, you're in the same situation with both coolant systems so it wouldn't work in practice. In other words, both systems are constrained by the same temperature and corrosion variables and as such, would require the same minimal level of protection from a particular coolant mix.

My 2C.

Paul

 

[Adamantium]

Thanks for the input - take your point about the coolant limitations but I was trying to say I'd be more comfortable running close to the edge on the supercharger system as the casualty of getting it wrong is the supercharger body and radiator core not the engine block.

Bit confused by a few of your points though as it seems to me like you've been slightly taken in by hype. for example, the faster the flow rate, the better the cooling efficiency for the same amount of cooling, assuming you aren't getting cavitation. On top of that, the merc racing core, they make a big point of it being a quad pass core, but in the absence of an uprated pump all you've done is increase the flow resistance and path length which serves to reduces the flow rate proportionally which as stated above is a bad thing overall.

 

[forty3]

@forty3 - have you done any VCDS logging during your time on track? It would be interesting to know exactly what the various temperatures are.

No logs I'm afraid but there are tonnes on audizine for people with the merc racing cooler.

 

[youngsyp]

Thanks for the input - take your point about the coolant limitations but I was trying to say I'd be more comfortable running close to the edge on the supercharger system as the casualty of getting it wrong is the supercharger body and radiator core not the engine block.

That's still only worthwhile if you can calculate the specific heat capacity of the coolant you're mixing.

Bit confused by a few of your points though as it seems to me like you've been slightly taken in by hype. for example, the faster the flow rate, the better the cooling efficiency for the same amount of cooling, assuming you aren't getting cavitation.

That's not correct. In simple terms, it's the rad that takes the heat out of the coolant. If you're pumping that coolant through the rad too quickly, the coolant isn't exposed to the cooling effect of the rad long enough to release the latent heat from the coolant, into the air. There needs to be a balance of flow rate to suit the core. Basic physics.

​

On top of that, the merc racing core, they make a big point of it being a quad pass core, but in the absence of an uprated pump all you've done is increase the flow resistance and path length which serves to reduces the flow rate proportionally which as stated above is a bad thing overall.

The increased number of passes the rad employs plays to the previous paragraph. I.e. the more passes the coolant takes through the cooled part of the rad, the greater the amount of latent heat is released from it. The rate of flow through these passes will still dictate the core's effectiveness. The same basic physics.

Well of course the greater capacity and longer path length would increase flow resistance and reduce flow rate. However, if the stock pump has the headroom to supply the ideal pressure and capacity for a larger core, then that's not an issue. And from what I read, the stock pump is a good match for the Merc Racing rad, able to deliver the presure and capacity required.

Also, as the volume of coolant has increased, any reduction in flow rate would be offset by the increased coolant volume's greater heat capacity.

Don't forget, there's a mix of OE core configurations among the B8/ B8.5 range and I believe the same pump has been used across the board. By all accounts, it's man enough for the job with a Merc Racing sized cooler.

So who's the one that's been taken in by the hype that these full kits are required...

As mentioned, the research/ findings is out there on the web if you care to look for it.

Paul

 

[Adamantium]

Sorry to say your basic physics is way off.

If you increase the flow rate, yes there's less time for each unit volume of water to heat up, but then as soon as it flows away it is replaced by another fresh unit volume of cooler water that has a lower starting temp as it hasn't just been heated partially by the supercharger. Since heat flow is driven by the temperature delta between hot and cold, a plentiful supply of colder water is better. Think of it like having an infinite reservoir of cold water.

Regarding number of passes, the only thing that determines the amount of time the water stays in the radiator core is the flow rate which is defined by the pump. If it flows 1 litre per second, it doesn't matter how many passes there are, the water will flow through the rad at 1 litre per second. You can have one pass at low speed or four passes at high speed but one litre flows through the rad every second - except that the more passes, the greater the flow restriction and the less efficient the pump and so actually you end up at 0.95 litres per second.

As for the mixing of coolant and specific heat capacity, I can tell you that water has a better specific heat capacity than coolant, and so the closer you are to 100% water the higher the specific heat capacity. I have a feeling it's actually a linear scale proportional to concentration.

I'm not trying to antagonize, I'm just using my brain and my degree in physics to question what I have seen written. In my experience, those who excel are those who question the perceived wisdom.

I actually haven't seen anything that resembles empirical evidence of the superiority of any one system over another, including the merc racing core.

 

[youngsyp]

Sorry to say your basic physics is way off.

OK, lets park that for now.

If you increase the flow rate, yes there's less time for each unit volume of water to heat up, but then as soon as it flows away it is replaced by another fresh unit volume of cooler water that has a lower starting temp as it hasn't just been heated partially by the supercharger. Since heat flow is driven by the temperature delta between hot and cold, a plentiful supply of colder water is better. Think of it like having an infinite reservoir of cold water.

All true of course. However, you've overlooked a key factor - if the rad hasn't been able to cool the coolant effectively enough, because the coolant's been pumped too quickly through it, that "fresh volume of cooler water", won't be very cool at all. It could be of a very similar temperature to the coolant that exits the supercharge, in fact. In this example, the efficiency of the charge cooling system will be reduced and boost will be bled off and timing pulled.
But I tried to explain that in my 2 previous replies...

Regarding number of passes, the only thing that determines the amount of time the water stays in the radiator core is the flow rate which is defined by the pump. If it flows 1 litre per second, it doesn't matter how many passes there are, the water will flow through the rad at 1 litre per second. You can have one pass at low speed or four passes at high speed but one litre flows through the rad every second - except that the more passes, the greater the flow restriction and the less efficient the pump and so actually you end up at 0.95 litres per second.

Yes, mostly, except to be pedantic, flow rate is dictated by the pump. And the capacity of the core also affects the length of time a measure of coolant will stay within the core. Again though, you've missed the salient point. That being - the more passes a measure of coolant has through the core, the greater the cooling effect will be applied to it, all other variables being equal. Ergo, for a given measure of coolant, at a set flow rate, a specific starting temperature and given volume of air flowing through the core, the coolant that exits the 4 pass cooler will be at a lower temperature than that which exits the single pass cooler.

The main reason to add core passes is to increase the core's efficiency and effectiveness, as we know. Although, the more passes, the greater the volume of coolant required to fill the system, which has the benefit of giving the system a higher specific heat capacity. And this gives the benefit of a lower flow rate through the core being required, for the same level of effectiveness as that of a system with a lower volume of coolant. All other variables being equal. But again, I tried to explain that in the reply above...

As for the mixing of coolant and specific heat capacity, I can tell you that water has a better specific heat capacity than coolant, and so the closer you are to 100% water the higher the specific heat capacity. I have a feeling it's actually a linear scale proportional to concentration.

Well (fresh) water has the highest specific heat capacity of any liquid, Ammonia aside. But you'd not want to put Ammonia in your coolant system, would you?!

I'm not trying to antagonize, I'm just using my brain and my degree in physics to question what I have seen written. In my experience, those who excel are those who question the perceived wisdom.

Antagonistic no. Condescending yes!
I don't have a degree in physics but, I can honestly say I've not found that a hindrance to my life up until this point and I don't think it's particularly necessary to understand these simple concepts either.

I actually haven't seen anything that resembles empirical evidence of the superiority of any one system over another, including the merc racing core.

That's probably because it doesn't exist. However, there is plenty of well thought through testing and reported results out there for the various aspects of an upgraded charge cooling system for this platform. And as mentioned above, it's not a complicated concept, as such a bit of intelligent thought will lead you to which system should work best.

Paul

 

[Adamantium]

Didn’t mean to be condescending so I apologise for that but I think we have to agree to disagree because i just can’t follow your logic.

You’ve said a number of things that either make no sense or mean nothing.

More passes does not alone mean more efficiency not does it mean greater volume which in turn does not give the system a higher specific heat capacity.

Greater volume is better because more water takes longer to heat up and so the system has greater cooling capacity. Think of it like having a much bigger air conditioning unit, it can keep room cool more easily than a small unit because it pumps out more cold air.

Separately from that, assuming your pump doesn’t slow down, you want the water moving as fast as possible because fast moving water is better at carrying heat away.

The only reason to have a greater number of passes is because faster flow is better than slower, as I said earlier but you disagreed with. Thing is, without significant headroom in the pump’s flow capacity this potential advantage is outweighed by reduced flow rate due to increased flow resistance.

I’m convinced the merc racing core performs well compared with stock because of its size but in spite of its four passes.

There are other factors affecting the efficiency of a system or core design but I’ve not mentioned those in this thread, only flow rate and system volume.

I didn’t realise that the forge system didn’t add much volume, shame as that was the one I was leaning toward.

 

[youngsyp]

I think I've twigged why we're still having this conversation - we're coming at it from different sides.

OK, so moving the heat away from the supercharger (the source) as quickly as possible is the ideal from that point of view, for obvious reasons (the perspective you're coming from?!).
However, moving that heated coolant through the cooling core as quickly as possible would not be ideal, if you're purely concerned with the efficiency of removing that heat from the cooling system as a whole, for reasons explained above (the perspective I'm coming from).
This is where the balance comes and why using more cores will improve the overall efficiency. In essence you can move the heated coolant away from the supercharger at the ideal flow rate and use the extra passes through the cooling core to increase the core's efficiency, and heat transfer out of the system, with same the higher flow rate.

Are we on the same page now, with the holistic view?

Paul

 

[Adamantium]

No we’re not, sadly.

But it’s fun to try.

Just sticking to the facts, forgetting anything to do with cores.

Physics + well established design principles of heat exchangers says:

Faster water flow through cores transfers heat better - that means from supercharger to water and from water to air.

More volume of water in the system means the larger amount of water will not get as hot as the standard volume. This means it will consistently keep intake temperature lower than it would otherwise. (If in doubt, take it to an extreme where you have no recycling of water just an ininifte source of freezing cold water).

So from the above we know, bigger system is better, faster water flow is better.

Now consider the multi pass core. Theoretically, from above, faster water flow is better so multipass core should be better ONLY because multiple passes speed up flow speed they DO NOT increase flow rate.

I wonder if this is the point where we are coming unstuck.

The pump flows say 1 litre/sec (arbitrary number for ease of maths).

Assuming the core is one litre, then every second a litre of water goes through the core.

If it’s a single pass or a quad pass, it doesn’t matter, a litre goes through the core every second.

In the quad pass though water will travel along each pass, which contains only 250ml every 1/4 second.

So for the quad pass to flow the same 1 litre per second, the water through it is travelling four times faster than in the single pass.

Because of the statement above about greater speed being better, the quad pass forces the water to travel four times faster, faster = better, therefore quad pass should theoretically be better at removing heat than single pass.

But, in a quad pass, all the water is effectively being pushed through a narrower core with a quarter of the area over four times the distance, this creates MUCH more work for that pump to do and suddenly it no longer flows 1 litre per second, it now flows 0.75 litres per second.

So the question left is, is the gain in speed having a more significant effect than the loss in flow rate through the whole system?

Remember that the supercharger is unchanged so now it’s part of the system is operating at lower efficiency with no upside since its core is only getting 75% of the fresh cold water that it used to.

The point is, the above trade off becomes irrelevant if the stock pump has enough performance in reserve to be unaffected by the increased restriction of the quad core.

Just read this back - so sorry for the long and again condescending post - not sure how else to write it without sounding arrogant and superior.

Genuinely apologise for that!

 

[youngsyp]

In the spirit of reducing bandwidth usage, I've quoted against the statements I didn't agree with

Faster water flow through cores transfers heat better - that means from supercharger to water and from water to air.

OK so now I understand what you're trying to convey with that comment.
I thought I'd do some reading as the wording you've used around this didn't line up with logic, to me (clearly!). The key word missing is turbulence. So it's not as simple as stating faster flow = better heat transfer as that's only the case if that increase in flow also increases turbulence in the flowing liquid. It's this increase in turbulence that increases the interaction between the liquid and the heat exchanger and results in an increasing reduction in liquid temperature.

So from the above we know, bigger system is better, faster water flow is better.

Careful, you're making assumptions that 'we' understand the context in which your stating this. It may seem obvious to you but specifically, a bigger system is only better if it contains a greater volume of coolant. And as we've now established, faster flow is only better if it results in greater turbulence of the flowing coolant.

Now consider the multi pass core. Theoretically, from above, faster water flow is better so multipass core should be better ONLY because multiple passes speed up flow speed they DO NOT increase flow rate.

This is assuming the multi pass core has narrower coolant channels than the single pass core? If the multi pass core has the same diameter coolant channels as the single pass core, it won't increase flow velocity over the single pass core.

The pump flows say 1 litre/sec (arbitrary number for ease of maths).

Assuming the core is one litre, then every second a litre of water goes through the core.

If it’s a single pass or a quad pass, it doesn’t matter, a litre goes through the core every second.

In the quad pass though water will travel along each pass, which contains only 250ml every 1/4 second.

So for the quad pass to flow the same 1 litre per second, the water through it is travelling four times faster than in the single pass.

Because of the statement above about greater speed being better, the quad pass forces the water to travel four times faster, faster = better, therefore quad pass should theoretically be better at removing heat than single pass.

Agreed given the highlighted assumption.

But, in a quad pass, all the water is effectively being pushed through a narrower core with a quarter of the area over four times the distance, this creates MUCH more work for that pump to do and suddenly it no longer flows 1 litre per second, it now flows 0.75 litres per second.

Agreed.

So the question left is, is the gain in speed having a more significant effect than the loss in flow rate through the whole system?

Remember that the supercharger is unchanged so now it’s part of the system is operating at lower efficiency with no upside since its core is only getting 75% of the fresh cold water that it used to.

The point is, the above trade off becomes irrelevant if the stock pump has enough performance in reserve to be unaffected by the increased restriction of the quad core.

Agreed. And this seems to be the case with the stock pump and Merc Racing core, as previously mentioned.

Just read this back - so sorry for the long and again condescending post - not sure how else to write it without sounding arrogant and superior./QUOTE]It didn't, you're good!

I think we're there?!

Paul

 

[Cerbera9]

Interesting thread

I run a Merc Racing HX, stock pump mod, divorced system & reservoir and the difference from the stock set up is noticeable esp on 2nd, 3rd, 4th pulls etc manly due to the systems ability to recover and deal with heat soak more effectively.

 

[Adamantium]

Ok we are getting there. Mostly in agreement.

Yes you are correct re turbulence, but in practice it’s very difficult to not generate turbulence, you’d have to go out of your way to design in laminar flow promotion so in reality turbulence is an automatic byproduct of faster flow.

But I think you are missing the central main point. The quad pass can have exactly the same diameter of coolant channels and will still nigh on quadruple the speed (not the flow rate).

If the two cores are the same size, they have the same cross sectional area.

As before if they both have volume of 1 litre and the flow rate is the same of one litre per second in both systems, then each second the entire core will have its water replaced.

If in one second a litre flows through both cores, but in the quad core it has to make 4 passes through the core in the same time, then it has to go 4 times as fast to get through.

Single core 1 x 1 litre pass in one second

Quad core 4 x 250ml pass in 1/4 second each.

But the length of the pass is the same ie the width of the core. So in the single pass core the width is travelled in one second, in the quad core the width is travelled in 1/4 second. Ie 4 times the speed but quarter the volume - constant flow rate.

It’s not possible any other way.

 

[forty3]

Physics and fluid dynamics aside, temperature logs taken from various members on audizine show a decrease in temperature over standard with any of the larger heat exchangers. This proves that they work as expected.

Divorcing the system also seems to help although not as much.

Running the supercharger pump at 100% all the time also helps. This is done by cutting one of the wires to the pump. Not sure what effect it will have on pump life tho.

 

[youngsyp]

Yes you are correct re turbulence, but in practice it’s very difficult to not generate turbulence, you’d have to go out of your way to design in laminar flow promotion so in reality turbulence is an automatic byproduct of faster flow.

So it's not difficult to generate turbulence in this scenario then, as it's a byproduct of the process...

The quad pass can have exactly the same diameter of coolant channels and will still nigh on quadruple the speed (not the flow rate).

Maybe using diameter is confusing. Let's try volume.

So lets be clear, each single channel, in either core, shares the same volume (from being the same diameter).
A quad pass core (with 4 channels) will contain 4 x the coolant volume as the single pass core (with 1 channel).
As such, if you're using the 4 pass core, 4 x volume of coolant needs to be pumped over 4 x the distance, to get from input to output, as the single pass core with the same channel volume. In this instance, you'd need a 4 litre per second pump to get the cores volume of coolant from input to output, in the same time as for a 1 litre core using a 1 litre per second pump. Basic stuff.

As before if they both have volume of 1 litre and the flow rate is the same of one litre per second in both systems, then each second the entire core will have its water replaced.

As above, they can't both have the same volume if the coolant channels are the same diameter and the four pass coolant path is exactly 4 x that of the one pass core.

If they share the same volume, that will dictate that the 4 pass core will have smaller diameter coolant channels, than the single pass core, will it not?

If in one second a litre flows through both cores, but in the quad core it has to make 4 passes through the core in the same time, then it has to go 4 times as fast to get through.

Single core 1 x 1 litre pass in one second

Quad core 4 x 250ml pass in 1/4 second each.

But the length of the pass is the same ie the width of the core. So in the single pass core the width is travelled in one second, in the quad core the width is travelled in 1/4 second. Ie 4 times the speed but quarter the volume - constant flow rate.

I don't disagree with this as it makes sense.
However, as before, this theory dictates the coolant channels, in the 4 pass core, are 1/4 of the volume (or diameter) as the coolant channels in the single pass core.
What I was disagreeing with is that a core with 4 times the volume and 4 x the path length will not exhibit a faster flow velocity than a core with 1/4 it's volume and 1/4 it's path length. In this example, to move the same volume of coolant in the same time frame, the 4 pass core will need a pump exhibiting 4 x the flow rate.

Paul

 

[youngsyp]

Physics and fluid dynamics aside, temperature logs taken from various members on audizine show a decrease in temperature over standard with any of the larger heat exchangers. This proves that they work as expected.

Exactly, which is where this all started.
You can over think and theorise until you're blue in the face but the numbers are out there, from those who've already experimented and published the results.

Paul

 

[Adamantium]

Ok, so we’re on the same page.

Not sure why you would theorise a core that was four times the size.

As I said, everything else being equal quad pass means four times the fluid speed.

Rather than the individual passages being smaller, the quad pass would just use a quarter of the number of channels. Changing channel size would not mean all else being equal.

I’m not doubting that every after market solution is better, Im sure they all are but that primarily because they have both much more water and much bigger cores.

Whole reason I posted was for comparative info. I’d like to choose the best option. If saying don’t overthink it because they are all better than stock then I might as well just go for the cheapest - which wouldn’t be foolish I suppose.

Thing is, it’s nice to make an informed decision rather than just picking at random.

I think fewer passes, maybe two, and a better or additional pump with much higher capacity is the best choice. Which would that be?

 

[Cerbera9]

All the HX’s I’ve seen have the same 17mm inlet and outlet as stock so surely they are all limited by this factor? They are all going to do a better job than the standard item and be more efficient so you pay your money and take your pick

 

[mercules]

Ams has the pump its near 2k for the kit.

Sent from my SM-G950F using Tapatalk

 

[youngsyp]

Not sure why you would theorise a core that was four times the size.

I didn't, you did, or at least that's how it read from you responses!

But then I did try to point that out...

Anyway have you made a decision on what kit your going to buy?

Sorry, couldn't resist.

Paul

 

[Adamantium]

No I ****** haven't! Help me!

and I never did - all my theoreticals had two identical cores one with one pass the other four passes.

 

[Adamantium]

I have a contact at AMS, might see if I can get one for trade. Have heard bad reviews about quality though.

Big fan of the AMS guys, met them a few times at Sema when I was heavily into GT-R tuning.

 

[RyanJonS4]

Ams has the pump its near 2k for the kit.

Sent from my SM-G950F using Tapatalk

That’s the 1 I have as I’m an add mans dream