4 deg C difference is not 39 deg F. More like 7 degrees F. That’s the intake temperature difference.
The difference in cylinder temps will not be so great with a one-pass system: the heat exchangers have a gap underneath, and consider the intake valves opening and closing will bounce slugs of air back and forth between the intake ports. Some mixing there in that.
I wouldn’t bother turning off the pump - mind you, the factory system already has a PWM signal that speeds up and slows down the pump, but on what basis, I have no idea. I’m surprised that nobody has replaced the original pump and used the PWM signal to control the speed.
Cold water would hurt any of these pumps at all - where does it say that in the study guide? What breaks pumps is cavitation and two-phase flow. But typically, these pumps fail often, at least in other applications, it’s weird I haven’t heard of the B8 S4 applications failing - maybe because of the pump speed control.
Ok I used my conversion calculator wrong, thought that sounded extreme. ( damn metric system is foreign to me) is your scale saying our OEM IC drops the temp from 256f-140? If so than my design has to be at least 25% better… No way it drops only 7 degrees, even with reduced flow I would expect cooler temps than a 7 degree drop…
you may be right about the temp evenly distributing across cylinders but I would wanna know for sure.
Page 37 says something about extreme cold temps and high viscosity increase current draw…
No, I’m still stock. Posted thread on AZ asking why pump waits for 122 degrees to start. He replied saying he does lower the set point through ecu tuning.
Much more expensive for a variable drive pump… Most wouldn’t pay the increase costs.
I don’t think your taking the compounded temp drop from the top pass cooling down the Boost much more than the OEM top pass. Then passing through a 2nd pass of water that’s also virgin cold water. At the reduced speed your bringing in more cold water. If we increase speed than we are compounding the effects even further.
It looks like AMS increased the speed with their 2nd pump causing less temp loss of the water leaving the IC, meaning colder heat exchanging points from the OEM IC. My design with increased flow could have extreme potential, imho…
I just took another look at my AMS kit, that thing is massive. It will definitely be able to increase the volume enough to drive my system with authority. Match that with properly sized hoses, it might flow enough to warrant an even bigger pump than the AMS pump depending on tolerances.
So AMS says 5th gear upgraded system makes 120 degree IAT with a post intercooler water temp of 99 and pre intercooler temp of 67. I think my system with AMS would bring IAT down to no higher than 107f IAT with a post intercooler water temp of 85 and pre intercooler temp again at 67. Compared to 140f IAT post intercooler water temp of 122 and pre intercooler water temp of 92… That’s in 5th gear!
I’m gonna make a bold prediction that my design could produce IAT, in AMS test parameters, of at very highest 107f in 5th gear wide open throttle. If that was true could you basically run race gas style timing? Not sure if much colder air creates similar results to race gas…
So you can follow along with how I calculated this,
I used log data to come up with the airflow (convert to kg/s): call this M
I calculated the compressed temp heat of the air at 15 psig (2 atm, remember it’s absolute pressure not gauge): call this T air hot
I calculated the log mean temp for the stock configuration, call this LMTD, using:
T air hot: air temp from blower (calculated above)
T air cold: post-intercooler temp (your 140F/60C number seems optimistic, but it jives with a couple logs I’ve seen in really cold weather)
T water cold: you said 92F (33C)
T water hot: you said 122F (50C)
I then calculated the heat exchanger duty (Q) using the OEM example, using Q = MCpdelta T
Using the OEM numbers, I calculated the overall heat transfer coefficient for the exchanger U using Q = UALMTD.
I then used the same duty Q to determine the “beard” LMTD. The “beard” layout, as expected, has a greater LMTD, in this example 49C instead of the OEM 46C. This is probably expected since there is double the water flow. If there was double the water flow in the OEM example (which may not be possible/feasible due to pressure drop across the water side of the exchanger), it is worth noting you’d see a similar result.
Using the higher LMTD from the “beard” layout, I calculated the new (higher) duty Q, and plugged this back into Q = MCpdelta T the new post-intercooler temp. To solve properly, you’d have to go back and iterate several times (calculate the new LMTD, then the new Q, and so on), but this is a pretty good estimate and I’m lazy.
As a note, I used the kindest example possible for your scenario, a full countercurrent heat exchanger, to avoid having to guess the correction factor for a cross-flow heat exchanger. This would probably exaggerate the gains in your layout, if anything, compared to the cross-flow exchanger. For the OEM exchanger, I calculated the whole exchanger, like it was one whole countercurrent exchanger. For the “beard” exchanger, I calculated only 1/2 the exchanger with half the air, and the corresponding 1/2 the temp drop at the same flow as OEM, since you wish to double the water flow using a bigger pump. This indicates it would cool the air further.
I see a couple errors I did on the sheet, but they aren’t big and wouldn’t affect the result much. (e.g. just measured the outside of the blower and the HX is more like 11" x 2.5" tall)
Just curious, why do you know your design has to be 25% better? and that no way it only drops 7 degrees? without any calculation?
I’d easily admit my spreadsheet is a estimate with many assumptions, but it’s the best I could show you quickly using the tools in my head, and without cracking open a textbook like Kerns or whatever. I don’t think I could know any gains for sure without actual data - logging the temperatures for real.
A lot of performance parts are based on hunches, and come up short, only to be exposed on this very forum.
I think APR could have done a lot more homework with the size of that heat exchanger in their 1740. It baffles my mind to see it’s pretty much the same size.
Lol dude. Is this how APR engineered their 1740 kit? I mean, it HAS to be better. It’s bigger
Where do you come up with your numbers? Don’t make predictions - not with actual temperatures, lol.
And, no, a few degrees cooler air is not a substitute for race gas.
Meth/water may be a substitute for race gas, but not just because it cools the intake charge a little - it’s because it cools and ballasts the compression and combustion event INSIDE the cylinder to prevent knock and reduce peak temps.
It’s an educated guess, but certainly not fact. I’m pretty good at visualizing complicated mechanical design, but I can completely understand why your skeptical or flat out think I’m wrong. If OEM design drops temps as much as 116 degrees than my system will remove quite a bit more. My water is consistently much colder because it only gets heated 1, and starts each journey at extremely temps when using AMS data.
First, can u confirm the 256 f to 140 f temp drop of the OEM system?
If so incoming water is heated up 30 degrees. You don’t see how I came up with those numbers? I’ll explain my logic. I cut that in half because the 2nd pass of water is the same temp as the first pass, opposed to the OEM design being much warmer on the 2nd water pass. Air temps in AMS graph are consistently 20 degrees higher than the leaving temp of IC water… I did that because I honestly think it’s the least my idea is capable of, and will get some one trying to prove or disprove. I deal with this kind of theory everyday at work. A co-worker is trying to run some calculus on this design…
Look at AMS site, B8.5 data was a little more thorough. That’s how I got gpm.
Not sure if you saw this (cuz it was in the A4 section), but APR has posted a teaser about a new/better/bigger intercooler, presumably for the Magnusun Stage3 kit:
Argh, I think you are right. I was reading up on the Stage 3 kit for the S4 and I probably cross read those treads and thought it was for the Stage 3 Super Charger. sigh Never mind.
My expansion tank, filled with 12 psi of compressed air, the water feeder feeds at 12 lbs controls the pressure rise as the heat pump loop becomes hot. No one in my field really understands this stuff, my customers love me
Nice, I’ve put in a few large Evapco open-loop towers over the years at our plant as we add capacity. Worst part is trying to get contractors in to clean the fill. Gross. Usually they do it once, then refuse to come back the next year.
Finally I found a company run by a guy named Bubba that will simply replace the fill for a decent price. Bubba and his crew are patched bikers, it’s hilarious when they come in with lifted Hummers and trucks and park wherever the fuck they want around the plant. Nobody tells them to move their trucks, lol.
With your evaporative cooling experience, I expect you to have some ideas with water/meth injection
This is the equation for the temperature rise in a roots blower:
T2 = 0.286 x T1 x (P2 /(P1-1)) / vol_eff
T2 is the outlet temp (absolute units, say 298K)
T1 is the inlet temp (absolute units, say 467K)
P2 is the outlet pressure (absolute pressure, 2 atm)
P1 is the inlet pressure (absolute pressure, 1 atm)
0.286 is a constant derived from the ratio or heat capacities for air, which is 1.4, and in this equation (1.4-1)/1.4 = 0.286
vol_eff is published by eaton in their compressor map…in our case it’s around 0.67 on the map.
This is where I get 124C (256F) outlet temp.
As an aside, for fun with calculations in thermodynamics, it seems the blower is consuming around 35 bhp from the pulley at our logged 1200 kg/hr, assuming around 80% efficiency.
There’s no theory or calculus involved, beyond using the ln() function in your spreadsheet: what you’re looking for is the Log Mean Temperature Difference, LMTD.
I did a bit of a over-assumption by using a countercurrent heat exchanger in the calculations instead of a cross flow heat exchanger, and doubling the flow in your configuration and not the OEM. I realize now that all of the gains in my spreadsheet for your configuration come from doubling the flow. If you double the flow in the OEM arrangement, you would come to the same result, maybe better.
You would need to use a program like ASPEN to properly solve this - I have it at work, but it’s honestly been years (a decade? more?) since I’ve used it.
Audi did the right thing by using the lowest temperature coolant at the bottom pass of the exchanger, where the air is coolest, to maximize the LMTD, then using the partially heated fluid on the upper pass, where the air is hottest, and the LMTD difference is still great. As a result, the bulk of the duty (heat, Q, in kW) of cooling the air is done in the upper stage heat exchanger with the pre-warmed fluid, then the coolest possible fluid is used at the lower stage heat exchanger to achieve the lowest possible exit temperature. This is similar in principle to any series of staged equipment, i.e. pumps, turbines, etc.
The possible gains in your configuration come from overcoming the pressure drop limitation of the heat exchanger, without changing the internals and spending a lot of money. IF, and only if, the heat exchanger has a high pressure drop across it, then a higher flow pump would achieve better performance by making the flow between the upper and lower stage parallel instead of in series, as you suggest.
I still don’t think it’s worth it unless someone is willing to flow-test the exchanger in both configurations. And I elect you
Yes Audi used the dual pass correctly, but IAT would still be even lower if the core (read average) temp was dropped, plus the increased flow removing the the dual pass…
Oh shit it’s 1am, I need to shut my mind off… Sorry I’m obsessed…
Thanks. You may be one of the only other ones here who knows the hell of ASPEN…lol. or HYSIM/HYSYS. You have the big steel furnaces at your plant, IIRC…
Last night I went to a friends repair garage and did some hands on testing of expansion/header/surge/reservoir (all interchangeable names for that stupid tank) and proved with out a reasonable doubt that the stage 3 tank was placed too low and is the main design flow of this stage 3 supercharger. I’ve passed my findings onto magnuson, APR, and RSW. I’m convinced I’ve ruffled enough feathers that a redesign of the flawed placement is a few days away.
I will not ramble on as no one seems interested in the theory, but I’ll leave this here.
(Bottom on the page read about “surge tanks”)
Now that I single handedly resolved the air bound intercooler problems for stage 3, can we go back to speculating about how much more power this 1740 will or will not make when running properly?
To be honest, I can’t believe no one else has chimed in to say " IF the tank is lower, then BB is definitely right." Do you guys work on your own cars? The evidence is right in front of you. Read the link, look at the pics, decide for yourself. I’m not here to change your mind. I don’t care what u think. Stage 3 will be fixed. I’m looking forward to the results.
I’m right, and if you are not capable of coming to the same conclusion with the evidence already laid out then your just gonna have to wait till I’m proven right by some one you trust. So I’m gonna sit back and watch you make a fool of yourself while you try and prove me wrong.
This should be good.
If my efforts haven’t been appreciated here, then it’s time to move on.
Tough crowd
