MrDangerUS

I have re-charged the battery and tried it again.
I tried it with the pin 1 to -12v and 4,3 to +12V - it started, but with 3 sec delay. Next, I tried it with 220 ohm resistor between pin 4 and 3- the same result. When I connected just pin1 and 4, (no PWM signal), all the same happened, except the speed dropped to ~50% after an additional 4 sec.
.

Mackhack

Yes, but try it without pulling the PWM high and see if the pump starts then.

M275FAN

Has there ever been a comprehensive list made up of pumps/ heat exchanger upgrades that are current as of 2020? Whats worked and whats hasn't? Installs ranging from direct bolt on to reverse engineering? Since this thread spans 7 years and almost 500 posts.

MrDangerUS

I tried it with the pin 1 to -12v and 4,3 to +12V - it started, but with 3 sec delay.
Next, I tried it with 220 ohm resistor between pin 4 and 3- the same result. When I connected just pin1 and 4, (no PWM signal), all the same happened, except the speed dropped to ~50% after an additional 4 sec.
I had an ammeter hooked up during the test, which showed 1.95A inrush current and 1.35A steady state. It never got above 1.95A.
I'm puzzled...

Mackhack

Strange and I never saw something like that with those pumps. I’m glad you didn’t buy that pump from me.

Welwynnick

What was different between your first test when it didn't work, and the second test when it did? It sounds like you're doing the right thing and the pump does work.

Regarding the current measurements, were they performed on the bench or when installed? I'd expect the current to be higher when installed. One the bench the load on the pump is low - it won't consume much electrical power if it's not generating much pumping power.

Nick

Welwynnick

This thread is it. Sorry, there aren't any straightforward drop-in upgrades unless you count the late model coolant reservoir.

The most important thing with the V12TT charge cooler is to make sure the system is properly bled, which is difficult.

Nick

M275FAN

Pumps show any decrease in temp or not worth it?

Welwynnick

Sure they do. The Bosch pump is OK, but upgrades are better. The Pierburg CWA-100 is everybody's favourite pump these days

M275FAN

Thanks for the help

Welwynnick

Well, I got bored with social distancing, and thought I would try and find out a bit more about what made intercoolers tick. Charge cooling is a big subject, and not many people publish useful information, but Wagner Tuning buck the trend by telling you about stock and upgraded equipment, and how they perform. So I took advantage of their website, and compared a variety of intercoolers. I started writing about this in another manufacturer forum, but they weren't very interested, so I thought I would come back here.

I took intercooler dimensions (in cm) and actual flow rate measurements and looked for trends and correlation. I calculated the frontal area, as seen by the ambient air passing through, and the cross-sectional area, as seen by the charge air passing across. I also calculated the volume of the intercooler matrix, which I think is the most important overall parameter from the point of view of cooling capacity and flow capacity (it's also consumes cost and space). I then compared the parameters against each other. Sometimes there were no significant trends, but a few combinations looked interesting.

Area/length is an arbitrary measure of the shape of the IC - whether it's short and fat, or long and thin. It's the number of sq cm total cross-sectional area seen by the charge air, divided by the cm length of the tubes the air has to travel down. It predicts that thin IC's have small area, low A/L and high restriction, and short IC's have high A/L, low restriction and low pressure loss. It's analogous to an electrical resistor, or a thermal conductor, where charge or heat flow through a material is proportional to the cross-sectional area, and inversely proportional to the length of the conductor. Do heat exchangers behave in a similar way?

Quantity: 1 = FMIC, 2= SMIC
Length = distance travelled by charge air through heat exchanger matrix
Width = width of SMIC or height of FMIC
Thick = thickness of heat exchanger matrix
Total W = total width (double for SMIC)
Area = Thick x Total W = area of matrix as seen by charge air
Volume = length x thick x total W
Inlet = inlet port diameter in mm
Outlet = outlet port diameter in mm
OE g/s = mass flow rate of stock IC
Mass flow rate = Wagner IC flow rate g/s



So is there a correlation between flow rate and the size and shape of the IC?
It's not a direct relationship, but it's fairly good.
Broadly speaking, the higher the A/L the higher the mass flow rate.
Large IC's are better than small IC's.
SMIC's have higher flow than FMIC's.
Large charge air ports flow better than small ports.
The BMW M4 is a water cooled chargecooler, and flows well for a small matrix.
Neither of the BMW 335d coolers flow well for some reason.
The EVO IX exceeds expectations due to it's large inlet / outlet.
The RS6 has the highest flow of any stock IC.
The highest flowing IC's aren't necessarily the ones with the largest matrix - they need short tubes and large ports, too.

Wagner took all these measurements at a pressure differential of 175 mbar or 2.5 psi, which is quite a significant drop.
By contrast the max pressure drop across paper air filters is usually about 5-10 inches of water, around 0.3 psi.

So what's the significance of mass flow rate? BHP roughly equals 1.1 x flow rate in g/s.
And the pressure drop increases with the square of the flow rate (or roughly the square of the power).

I also looked at temperature rise measured during a dyno run, against the intercooler matrix volume, when normalised against engine capacity, and against engine power. There are lots of different parameters that I compared, but these gave some good correlation:



As you'd expect the bottom left of the chart is dominated by the upgraded IC's - suffixed by "W". How big does the IC need to be? The simple answer is the bigger the intercooler, the lower the charge air temperatures. The middle ground seems to sit around a temp rise of 30 deg C, which corresponds to an IC matrix volume of four to five times engine capacity. For water cooled charge coolers, this means the volume of the IC and HE combined.

Notice how much better the upgraded Macan IC is - way out of proportion to volume over capacity or area over length. I think this is because the stock IC is low and thick, and this means ambient air is choosing the path of least resistance, which is to go up and over the IC and straight through the condenser and radiator instead. The Wagner IC is full height, so all the air has to go through the IC. It's a recurring pattern in Wagner tests.




I wasn't sure which parameters would show the best correlation - the capacity or the power ratio, but both ended up with a pretty good match. In the above chart, low and left is obviously good. Again, the cluster sits around a temp rise of around 30 deg C. This corresponds to about 40 bhp per litre of IC matrix volume, which seems like a good rule of thumb, and perhaps a bit less power than I expected (see the opening posts of this thread).

Nick

Khoi Bung Phe

Nick,
Thanks for all your contribution to the group. May I ask for the secondary HE, I can get any radiator as long as it's larger than the stock one? no restriction? and do you think I can run a dual BOSCH pump for my setup. Thanks again Nick

Welwynnick

Any radiator as long as it´s larger than the stock one - yes, as long as you can physically fit it in the space available, AND get the inlet and outlet hoses in place, AND you can bleed the radiator.

Dual Bosch pump - yes, but you need to consider the flow characteristics of the system and therefore whether the pumps should be in series or parallel. Series for high resistance and parallel for low resistance system.

Nick

FxFormat

Nick, i was thinking of adding a secondary HE to my 2009 SL600, i was looking at the blank space today below the front bumper crash beam, and there's a lot of space. Thinking of running the stock HE along with a CXracing one of 24x8x2.5. Would the bosch 010 pump be sufficient to push water through both HE? I would get an HE with ports on the same side, facing to the driver's side and just run a tube from the outlet of the stock HE to the CXracing HE inlet then from that go to the outlet to the intercoolers..

FxFormat

This is what i'm working on, my car has a ton of room behind the bumper this is how i will mount the secondary heat exchanger. I hope the Bosch pump alone can move all that water through both HE

FxFormat

Don't know who's still reading this thread anymore but i just did some test runs and logging in my IATs. Air temperature is 90 degrees and sunny, cruising at 70mph IATs stayed at around 115-120F, when i floor it, it'll get down to 107-110 but as soon as you cruise it creeps up to 115ish. Stop and go traffic is still a problem, at a redlight it'll climb to 150s, but as soon as you give it some throttle it immediately get down to 125-130 going at residential speed. Is the pump RPM dependent or something? When I floor it, temps would dip down until i let off, thinking about just making the pump run all the time now if that's the case. I felt my heat exchanger that i just put in, it's cool to the touch even driving 30 minutes in this hot weather.

So overall, with the extra heat exchanger my temps are 15-20 degrees lower overall than before, i'm not sure if have any air bubbles in the system or not but i'm happy with it so far. What i like about it most is how fast it recovers, temps at redlights would get really hot then as soon as you take off it gets to a temp where computer doesn't pull timing.

Mackhack

You have an electric pump, rpm doesn’t matter because the belt doesn’t drive the pump. What matters is the reduced air flow at the red light. That’s why usually the fan turns on at a stand still. As soon as you drive the air can flow again with much more effect than your fan and therefore you see an immediate drop in IAT. Hope that helps.

FxFormat

That part i understand, but let's say i'm cruising at 70mph, temps stay at 115 degrees, now if i nudge the pedal it starts to go down, and if i floor it it'll drop. As long as you're on the throttle temp will drop, but as soon as you let off, it'll slowly rise again. Speed isn't the factor as cruising at 80mph doesn't cool it down anymore than going 65mph. That's why i'm curious if throttle position has some effect on the flow of the coolant.

Welwynnick

That's the more difficult bit to answer. Your IAT behaviour is normal. Don't have time now, but from memory it's down to the position of the temp sensor, and how the pressure of the charge air at that position affects the temperature at that position. Compressed air is hot, expanded air is cold.

Nick

FxFormat

Cruising at 70mph, not on the throttle outside temps 90 and humid IAT 118F



Nudge the throttle a bit barely hitting boost, it'll drop and continue to drop until you let off IAT 108F



If you actually floor it and keep on it, I saw brief moment of under 100F IAT. It's scary though because you're doing 120 + mph in no time and that's a ticket to the American jail. I think i'm happy with the results, i'd imagine in the Fall when the ambient temps dip into the 50s-60s it'd run much cooler.

Chevota

Fx; Keep in mind that air going into the eng is a big factor. Imo the pipe between the the air filter and turbo if hot as hell because the hot rad air is blowing on it and at a stop in traffic it's much worse because there is much less air flow thru the eng bay so the air is even hotter. I mean if you wanted to heat something with rad air, the stopped idle condition and the placement of the intake tubes just about ideal.
Speaking of that, imo, it looks as if these cars barely have enough air flow exiting the eng bay under all conditions which needs to be fixed, but that's another story.
So the heavy steel tubes to the turbo are just cooking, as are the ones exiting the turbo but imo the ones before the turbo are worse. Point is you have hot pipes and very little air flow in them at idle condition so they just get hotter and hotter. The air inside the tubes, since it's moving so slow, it absorbing that heat but the flow is too little to really cool the tubes so basically all that's happening is the air getting hot as hell which explains the super hot air intake temps.
When you gas it the air is flowing much faster so less time to collect heat which gives you instant lower temps. Well, imo much faster than the RTD can read but even with that delay the temps drop fast don't they? And the pipes start to cool down thanks to two things; one is the cool air inside is sucking heat out, and the air exiting the rad is cooler since air flow is up so the tubes are being cooled from both sides.

On the freeway the eng is making much more heat than at idle but air flow in/out the eng bay substantially better so the tubes still get blasted with rad air but it's much cooler air than idle. My guess, and from what others say, the IC pump doesn't kick on until ~120F? But from what I see I it also does not kick on at idle, so I'm guessing it needs both the temp and a high enough MAP to kick on and MAP at idle is very low, as is cruising. So based on what I see you need to be both heat and higher MAP to trigger the IC pump. So on the freeway you get more air cooling the intake tubes, then add a little gas and the pump kicks on and you get even cooler. That's what I see with my car and I assume you guys see the same.

Basically I just wanted you to consider the turbo in/out tubes and the role they play in it, and the on/off of the IC pump. So ideas I've had are to make the IC pump turn on when the brake lights are on, and wrap my intake tubes with an insulator, like that DEI tape. I do not want to pull my tubes to wrap them and make it pretty, plus I don't like the thought of the glue the tape has. I worry it'll bake on like epoxy and I'll be screwed getting it off later. So I'm thinking of using silver mylar that doesn't have glue and just wrap the tubes with say a 3" strip and secure with zipties. Ideally I want to wrap it first with cloth, like cut up white under shirts into 3" strips to act as a an air space to keep the mylar off the tube. Since all the tapes and other coverings sold are expensive, I planned on cutting up an emergency blanket. One blanket should be enough for a couple wraps and they're like $2?
I was thinking of just wrapping the important tubes, which are the ones feeding the turbos, and just wrap from the top to as far as is reasonable to reach which is what gets the most rad heat, then see what it does to temps...
If the rad fan does not run full blast while stopped, and I don't believe it does, then maybe wire it to the brakes as well. I mean why wait until everything, especially the intake tubes, get super frikkin hot before turning the fan speed up? Get a head start right?


Welwynnick; I've never believed the story about coolant dwell time in a heat xfer condition, meaning the conditions of an eng radiator or water IC cooler where the part making the heat is included in the system. When I was like 19 or so I had overheating issues in my pos hotrod, like pretty much everyone else in my area did, and everyone believed the old tale about keeping the water flow slow. Even the auto shop teachers in school firmly believed it. Of course it cools the water in the rad to a lower temp, but there's also less of it coming out to do the cooling! So the eng is overheating while you wait for that trickle of cool water than cannot cool the engine and you overheat... Plus there's a side effect called hot spots in the heads so when you slow water flow you get them much more easily and you're basically screwed. So I put a restriction on the system to see if there was something happening I didn't understand. There wasn't, it immediately ran hotter plus the aforementioned hot spots appeared MUCH sooner... I had several different orifice restrictions to play with and I quickly saw the bigger the orifice the cooler it ran. So then I did what everyone said to NEVER EVER do; remove the thermostat altogether with no restriction. It's said the water will then flow so fast it'll never cool and you'l over heat. I believed this to be bs from day one hearing it, long before I started tinkering. The result? Well, let say I've never ran a T-stat or any restriction in any hot rod since. And thanks to the increased flow, hot spots only appear when I'm already overheated, which happened much less but still does because of high ambient temps and eng loads. Also to prove the flow thing a bit further; I later modified my water pump to flow better, and guess what, it ran even cooler... If you opened the rad cap at idle it was flowing like an oem car would a high rpm. Then if I revved the eng is was flowing much more than any oem system I've ever seen. I'd say ~4x more but hard to gauge, but it was like Niagra Falls in there. To compare, I once put an elect pump in once that claimed 30gpm, or maybe it was 35, but it was an absolute joke compared to the mechanical pump.
Another way to look at this heat xfer flow bs; if the water coming into a rad is say 200F, and you slow the cooling so it's 150 exiting, then the colder exit side isn't cooling as well now is it? Meaning the lower the water temp is the less heat transfer takes place. If the water is flowing much faster and the exit side is say 190, then overall there is much more heat being removed. What I saw was the eng ran cooler so the eng exit temps never got as high which can only be explained by the aforementioned increased efficiency of the rad thanks to increased flow. So despite cooler water entering the rad, which was 170-180 now, and thus less delta temp to cool it, it still worked better!!!
I know these stupid computer controlled cars don't like lower eng temps but if I could get the water flow up and eng temp down to say 180F without changing the fuel mix, I think it would be a win across the board. So, does anyone know at what temp the fuel goes into normal operating mode? Meaning no extra fuel to compensate for a cold eng? If it's already 180 or less then I may tinker with that... Cooler eng means less heat on the intake tubes, cooler combustion chamber so less ping under boost which hopefully means more boost and timing.
My point here is about flow which I believe directly applies to the IC system. The more flow the better. Of course we want to avoid cavitation which is hard to predict but will no doubt show up, and avoid so much pumping that's it adding more frictional heat than the increased flow is helping. Considering how restrictive the system is I think both of those are factors to consider. Fyi I don't believe any significant friction is generated in the lines or coolers, imo it's almost all made in the pump. This no doubt explains one poster saying adding a second pump only made things worse. That tells me the pumps make more heat that I thought so it really needs to be considered. I'm sure the pumps impeller design has a ton to do with that heat, and of course cavitation. Cavitation and overall efficiency is why I modded the mechanical pump on my hotrod, which cavitated less and flowed more in one easy mod, a mod that doesn't apply to these pumps so don't bother asking. Someone suggested slowing the water flow in the IC with a restriction, which I think is a super bad idea. If you do you'll now have more friction and cavitation, yet zero bennies. If you really want to reduce flow to see what happens you need to reduce the pumps speed. You could control that from inside the car with a wire and a rheostat or transistor(s) and pot. Cruise, idle, floor it or whatever with various voltages applied to the pump and see what the AI temps are. Of course the results will only apply to that pump in that IC system so what one guy sees with his 010 pump won't apply to other pumps, but it will show if any reduced flows help. I don't believe it will with an 010 or similar pump so I won't bother, but maybe someone has the initiative to do it. And keep in mind that if a lower voltage to the pumps shows better cooling, it is without a doubt not from reduced flow, but increased flow. Meaning it was cavitating before and now it's not so it's working better and thus flowing more. I don't believe this will happen with an oem pump like the 010, but I bet with bigger pumps using a crappy impeller it will.
One more thing, which I think you guys already know; is coolant (antifreeze) is crappy coolant. Water is coolant and the more antifreeze you add the worse the system cools. Since I live where it never freezes I only run enough antifreeze to lube the water pump and keep corrosion down. Using 50/50 is just a bad idea imo, at least for me because I basically live in the frikkin desert so i need all the cooling I can get. It's 118 outside as i write this, the dirt in my yard reads 145 with my little thermal reader so imagine how hot the asphalt is. The rad is attempting to cool the eng with that hot *** air just above asphalt on the road and people wonder why I have cooling issues. This is why I've tinkered with cooling where I'd imagine most people do not have a need.

8899

Wow...this is some thread!!! HUGE thanks to @Welwynnick for all of your research.

I recently purchased this Pierburg CWA100-3 pump from here: https://www.ebay.com/itm/7-04934-54-...72.m2749.l2649

....and would like to get my tech to install it, replacing the stock IC pump, on my 2005 S55.

I would be extremely grateful for some tips/tricks on doing this remove/replace. Specifically, will it easily mount up in the space where the existing IC pump sits? And what about the wiring/connectors, what modifications are needed to get this pump running?

Thank you!

Mackhack

I make custom made adapter harnesses for plug and play, if you’re interested PM me.

MrDangerUS

Bosch 010 pump is not that efficient and is heavy, see diagram.
I have changed to Davies Craig EBP40-9040 Pump, which weighs only 1.25 lb and flows more than Bosch xxx010
.

Krazyleg

This has been one of the most amazing reads I’ve ever encountered. Note I read a variety of books about 4 per week, on any and all subjects. This was captivating. The entire process of cooling everything Benz in the innovative fashion you have with complete disregard for intended use but rather focusing on actual capabilities and unintended compatibility is simply stellar. I too drive a W220 but of the 65 variety. No track or anything like that just enjoy experience. I don’t technically have a need for this alteration at this time but I’m contemplating the IC pump as a preventative or an upgrade. This post is fairly old, I was curious if you have any updated thoughts or recommendations about replacing the OEM pump. Currently 77,000 miles. Just did ABC pump, pulsation damper, accumulators, seat pump, trunk pump and struts, combar and some sundries.

Regards,
Frank