Adventures in Intercooling

Status
Not open for further replies.

Catmonkey

I Void Warranties!
Established Member
Premium Member
Joined
May 20, 2011
Messages
3,854
Location
Louisiana
After Tob began his thread about installing the 13-14 Pierburg intercooler pump in an earlier model in this thread, I’ve been slowing working toward making this one of the mods for my 2012. This write-up is a compilation what I did, my findings regarding the factory intercooler system and what we can do to improve it on a budget. As far as I know, everything discussed here with respect to the intercooler system will be the same for 2007 to 2012. Much of it will also apply to the 2013 – 2014 as the intake manifold and intercooler connectors haven’t changed over the life of GT500 program, with the exception of an upgraded intercooler and the Pierburg pump. This will be a long read, but there’s a lot of things I encountered I’ve not read anywhere else. I’m going to ask the mods to lock this write up in the How-To section and start another thread in the forum for comments, questions, etc.

Back when I was installing a built 5.8 aluminum block in my car in late 2013, I upgraded my intercooler to the 13-14 unit while the engine was apart. There is not a lot of visual difference between it and my OEM unit, other than it’s supposed to flow more air. Since it does not have a larger surface area, the flow increase is accomplished through a lower fin count which likely decreases cooling surfaces. Note, there are no claims by Ford that it decreases charge temperatures. While taking the manifold apart I was somewhat baffled why Ford plumbed the intercooler for ¾” hoses and then reduced the internal diameter of the crossover tubes between the intercooler and the intake manifold hose adapter to something just over ½” (.55”). Seems to me the smallest internal diameter of tubing in the system will become the biggest restriction to coolant flow and a source for higher pressure within the system. Based on the Ford part number for the tubes, since I ordered new ones to modify, the part number indicated it’s from a model year prior to 2007. With a little more research it was apparent they are a carryover from the 03-04 Terminator which were used for the same purpose. Ford saw fit to use the same tubes in a new supercharged model with a bigger engine and more horsepower. But then Ford didn’t design the system for more than the OEM ratings anyway and initially for only 9 lbs. of boost.

Most of the data I’ve seen on these smaller OEM pumps is that they flow as advertised with little or no pressure, but as pressure increases (i.e. smaller connectors, coolant volume, bends, etc.) the pump can’t maintain the same amount of flow and drops off in efficiency as pressure increases. This is a pretty interesting discussion on that topic as it pertains to supercharged Chevys, but it also comes from a highly credible source. Note that two of the pumps in that discussion are the same and similar to the pumps that Ford has used in the GT500. The CTS-V pump is the same Bosch pump you’ll find on the 07-12 models. The ZL1 pump, while not the same as the Peirburg, has very similar flow characteristics from what I can tell. The Peirburg pump is used in other applications and if you look up Pierburg CWA 50, you’ll see it’s the same pump. Here’s the graph of pumps from that discussion.


Lingenfelter pump.jpg


In the Lingenfelter graph, note the impact of higher pressure downstream of the pump. The CWA 50 also appears to do a much better job at encountering pressure and doesn’t drop off in efficiency as the Bosch pump does with as little as 5 psi of pressure. Now, what would you think if I told you that the Pierburg pump in the GT500 is wired to run at less than 50% of its capacity in the 13-14? At that rate it’s likely to come in a lot closer to the Bosch pump in our intercooler system by virtue of increases in pressure, although it has the ability to overcome the effects of pressure (restrictions) in the intercooler system more efficiently than the Bosch pump. More on that later.

With knowledge of the small cross over tube diameters in the intake manifold, I set out to determine what the minimal diameter the tubing should be for the crossover tubes based on the design parameters of the intercooler system. In doing so, I found an even larger restriction in another OEM component and in an aftermarket part many of us are using. I started measuring the inside diameter (“i.d.”) of some of the major components that make up the remainder of intercooler system. It stands to reason that to use ¾” i.d. rubber hose, the outside diameter (“o.d.”) of the connector would have to be the same diameter or only slightly oversized. In effect, a component with a ¾” o.d. would have to have a smaller passage for the corresponding i.d. to have sufficient wall thickness to support the component. Even the i.d. of the inlet/outlet on both intercooler pumps are in the range of 5/8” (.625”), as are the inlet and outlet for the heat exchanger. It seemed to me, my objective was to make every i.d. in the system a minimum of 5/8”.

So where are these mystery tubes you ask? If you’ve changed out an intercooler, you know already know the answer to that. If you look at the front of the intake manifold, you’ll notice the adapter in the front of the manifold with 5 bolts holding it to the intake. The adapter has an inlet that is connected to the heat exchanger and an outlet that is connected to the degas tank. Many of you have replaced three of these bolts to mount an auxiliary idler for your supercharger. With all bolts removed, the adapter can be removed from the manifold. Here’s an exploded diagram of the intake manifold. Items 28 are the tubes I’m referring to.

CAC Coolant Tube Assembly.jpg


This is the crossover tube. You have two in the manifold.

crossover tube.jpg


If you take this on, I would do what you can to make sure as much of the coolant is out of the intercooler before removal of the adapter, because you may end up with a little coolant in the intake manifold. It can be suctioned out from the vacuum tube that connects to the downward facing vacuum line at the back of the supercharger. The other end terminates at the bottom of the intake manifold. Both ends of each tube have an o-ring groove cut into the tube and uses a rubber o-ring to seal off coolant from the intake or air in the intercooler under boost. Removal of the 5 bolts allows access to everything I will discuss below. I would suggest replacing the o-rings, if you attempt to do this. The Ford part no. is N802927-S and you’ll need 4. I’ve since seen posts about better o-rings made of viton material, but I have not experimented with that yet. Here is a another pic of the tubes with the o-rings. You’ll notice the o-rings are no longer rounded which is why I think they should be changed.

Intercooler tubes.jpg


Let me say this because I’m sure the question will come up, while alternatives exist to go with larger tubes in the manifold by way of boring out the intercooler and using a fabricated manifold adapter, unless you intended to plumb the entire system with a larger pump, hose diameters, and comparable inlets and outlets on all the other components in the intercooler system, I’m not sure what you would accomplish. No doubt it would create better cooling than the smaller diameter stock tubes, but it seems to me you would just move the restriction downstream to the next smallest component in the system unless the whole system is upgraded to similar size. The process of installing larger tubes also requires removal of the supercharger, the intake manifold and the intercooler so it can be sent off and machined for the larger diameter tubes. Unless you’ve taken the intake manifold off your GT500, you truly don’t realize how big a job it is. Even if you were able to open all the passages to 7/8”, I’ve yet to see anyone test just how much flow the intercooler itself is able to handle before it might become the primary restriction. I see upgrading hose diameters, inlets and outlets along with a larger pump as a max effort, and it would ultimately come at a fairly high price tag to do it right. With an ice tank in the trunk, it would no doubt be the ultimate standing mile setup. Here is the OEM tube next to the larger tube.

Tubes.jpg


A system sized to match would certainly flow some coolant.

I briefly looked into the possibility of having a machine shop drill out or open up the i.d. of the tubes to 5/8”, or mill new ones out of suitable tube stock, but I decided to give it a shot by opening them up myself. Based on measurements of the tubes, it seemed to me that they could be opened up safely to .63” and still have sufficient wall strength. While that’s only removing .040” of wall material to open up the i.d. by .080”, it’s very slow going. I initially attempt to use various flex-hone tools to open up the diameter. That proved to be a bust, as the tubes are likely stainless steel and the metal is extremely hard. I ended up using a combination of the flex-hone tools and a die grinder to open the i.d. to .63”. The o-ring grooves in the tubes are .675”, so opening them up more than that may be risky. While it doesn’t seem like much, going from .55” to .63” opens up the area by 31%. There’s also a notable difference in the weight of the OEM tube and a tube that’s been opened up. Hopefully, it relates to an increase in flow by a similar amount. Increasing the passage size also has more to do with reducing pressure, than increasing flow.

Here’s the flex hone tool, I referred to. You just chuck it in your drill and use a bit of oil with it.

Fflex hone tool.jpg


There are a lot of hours opening the tubes up in the manner I used. Part of the problem is that the tubes get too hot to handle to handle, even with gloves. My solution was to wrap the tube in electrical tape and use pliers to hold them while grinding or honing. After a while the tape adhesion turns to a gooey mess, but it cleans up easy enough. I dropped the tubes into a small container of water before attempting to handle them. Using the flex hone tool between cuts helps keep the bore straight and you can measure your progress using calipers. These pics are not that great, so it’s hard to see the difference. Notice the bevel on the inside which further reduces i.d. in the tubes at the top of the pic. That bevel is essentially gone with the opened up i.d.

Tubes 3.jpg


The next thing I needed to do was open up the diameter and surrounding wall of the manifold adapter where the tubes terminate. This was easily done with a die grinder as well. I used electrical tape on the outer walls to prevent nicking them. A nick with your grinding tool in this area may result in a leak when the system is pressurized. Here is my stock adapter next to a bored out adapter. The new adapter is a little too shiny to see the difference, but these now match the tube diameter and the transition to the sand cast lines were also opened up.

Manifold adapter.jpg


What remains unknown is how large the sand casting passages are between these holes in the adapter and the inlet and outlet tubes. Judging by the hole beyond the inlet and outlet, the passage may not be a full 5/8, because material needed have a slightly smaller i.d. to machine and provide support for the inlet and outlet connectors pressed into the adapter. But you can get your flex hone tool in there up to the bend. Extrude honing may be the only way to open these up. I then started measuring the barbs on all of the connectors in the intercooler system. I ended up grinding the tips of the hose barbs on the manifold adapter back about ¼” to get the inner diameter at the tip of the barb to be equal to the i.d. of the body of the individual connectors. At least for the inlet and outlet on the adapter, they had been formed to a smaller inner diameters at the tip (i.e. too pointed) than the straight section of the main body of the coolant fitting. Grinding the tip back will open up the i.d. and remove the restriction.

I was again surprised to see how small the inlet and outlet of the plastic 3x degas tank has because of an aluminum insert that necks the i.d. down to .48”. Having read a few complaints about coolant frothing with the 13-14 pump on an otherwise stock 07-12 intercooler system, I have to think that this reduction in the inlet may have a lot to do with that. Since my intent was to run a by-pass hose around the degas tank, I did not do anything about this restriction. I located my original degas tank and it is constructed similarly except the inlet and outlet diameter measured at .45”, which is even smaller. I assume the inserts used in both are to keep the plastic barb from collapsing or breaking over time and under the compression of hose clamps. Unless you incorporate a by-pass around the degas tank, something probably needs to be done with these inserts. If you do nothing at all with the tubes, I’d at least open the inlet and outlet of the degas tank to .55” or thereabouts to match the intercooler tube diameters. That’s a 30% increase in area right there. It’s doubtful that aluminum tanks will have this issue, but I’d measure the tip of the barb to make sure the i.d. is no smaller than it should be. Using a by-pass was my solution and might be the only solution around the inlet/outlet on the degas tank.

Since we’re talking about such small diameters, some may assume this is hardly worthwhile. I did a scaled graphic to show what I’m talking about. If you don’t think these smaller diameters create restrictions and resultant higher pressures, this may put it into perspective.


Diameter analysis.jpg


Since I have provision for another analog device on my Aeroforce gauges, I decided to add a coolant sending unit downstream of the heat exchanger. I initially got a ¾” gauge adapter, but realized that the sensor for the sending unit creates a big restriction in the line. I ended up getting a larger diameter adapter to remediate the restriction the sensor created and used silicone reducers to go from the 1.25” o.d. for the adapter to the .75” o.d. of the inline connector. I don’t know that I’ve ever heard anyone comment on the correlation between IAT2s and intercooler coolant temperatures. Here are some shots of the adapter. Sticking a pencil sized sending unit in the 5/8” internal passage of the ¾” gauge adapter was sure to become the restriction in the smaller diameter housing.

Gauge adapter.jpg
 
Last edited:

Catmonkey

I Void Warranties!
Established Member
Premium Member
Joined
May 20, 2011
Messages
3,854
Location
Louisiana
Next decision was where to mount the pump and how to mount it. One alternative is to use the stock location used the 13-14. This solves a lot of issues because you can use OEM lines from the intercooler to the degas tank. In my thinking, because of the amount of heat the pump has to endure mounted on the engine, I decided to mount it forward of the radiator support. I’ve seen a few installs using Adel clamps, but I was not that confident in that method from a vibration and noise standpoint. There’s a post by 54First and I liked what he did to mount his pump, so I did something similar with slight variation. I mounted the OEM 13-14 bracket reversed (or upside down) from what he did. I also used riv nuts and made a bracket to lower one of the mounting tabs. Here is a photo of the final install.

Pump mount 1.jpg


Pump mount 2.jpg


Pump 2.jpg

Pump 4.jpg


I made the mistake of mounting the pump before I attempted to run the hoses, but in hindsight it worked out fine. There’s just not a lot of residual real estate for a mounting location on the driver-side of the car, north of the radiator support. It did necessitate the reversing the horns in their OEM bracket, but there was adequate clearance in the frame to accomplish that without any issue. I didn’t want to locate the pump in the air stream for the radiator either. The downside to mounting the pump higher than the bottom of the inlet hose was that in filling an empty system, it would likely create an air pocket at the pump impeller that might make the pump cavitate and prevent circulation of coolant. Removing the pump from the bracket and letting it hang by the hoses allowed the pump inlet low enough to expel any air in the impeller area to move out of the pump and up into the hose after I filled the system. The only time this would be a concern is initially filling the system. I also jumped the relay to cycle the pump while the pump was in the same position to make sure it expelled all air below the highest coolant level. Here’s what that looked like.

Pump bleed.jpg


The VMP triple pass heat exchanger has the outlet mounted at its highest point, some air is going to be trapped in the upper area of the heat exchanger. By simply moving the clamp down the hose, and off of the barb, and barely pulling the hose off the barb allowed air to escape from the top of the heat exchanger. As soon as a nothing but coolant started coming out of the small opening, I slid the hose back on the outlet and clamped it. Here is a shot of that location of the outlet. It’s the highest point above the lowest points down stream of the intercooler and degas tank where air can get trapped.

Upper HE outlet.jpg


Next I needed to route my hoses and make a by-pass for the degas tank. There’s discussion in Tob’s original thread that indicated a restrictor was in the line to the degas tank in the 13-14 hose routing. Obviously this restriction in the inlet hose to the degas tank was intended to reduce the amount of flow to the degas tank so it could both act as a reservoir and a fill point for coolant and divert most of the coolant flow to the pump inlet, and not subject the tank to a lot of turbulence. I located a plastic spacer that had an o.d. of ¾” and an i.d. of 3/8” on McMaster Carr to use as my restrictor. After spending the better part of an afternoon routing hoses, I was not happy with my first attempt at running the hoses from the intercooler to the pump. My hose routes were exposing the coolant hoses to superheated air from the radiator’s fan wash. While I was contemplating other ideas on how to best improve that run of hose, lexustech48 posted how he routed his hoses in this thread. While this is what I envisioned at the outset, I didn’t think it was possible to accomplish, so I set out to attempt to duplicate what he had done.

For some reason I had a lot less clearance than he had. I realized the problem on my car is that I’m running a C&R radiator, which is much thicker than the OEM radiator. Since the degas tank bolts to the fan shroud that mounts to the back side of the radiator, it put my degas tank at least a half inch to an inch closer to the engine. In order to install a plastic T between the intercooler outlet and the degas tank like he had done, the inlet on the degas tank would need to be cut back because the tee and inlet overlapped one another. I trimmed about 5/16” off the end of the inlet to get the parts to even line up. The end result was there was no room for the restrictor. While I probably could have left it like that because of the small i.d. of the inlet, I tapped the outlet portion of the tee fitting with a 3/8” NPT tap and installed a left over plastic pipe fitting from a water fountain pump that had about the right i.d. I just cut off the barb on the fitting flush with barb for the tee once the fitting was tightened. I also tapered the inside of the tee to make a better sharp turn radius from the intercooler outlet to the tee, as well as trimmed the barb down to get it as close to the bend in the hose as I could. I cut the barb as pictured below. This comes handy in a few other areas.

Gates T2.jpg


Next step was to move down to the outlet portion of the degas tank to figure out how to route it. On the original first attempt, another issue I had was the hose ran too close to the air flaps on the radiator shroud and they could not completely open. To remedy that, the hose needed to drop down along the edge of the radiator’s electric fan shroud. Since flow to the degas tank could now be a straight shot downward, I decided to use a 90° hose fitting off the outlet and connect it with a T fitting. To cut down on the distance between the two fittings, I cut the barbs similar to what I did in the previous photo on both barbs. I then used Gate’s Power Grip heat shrink fittings to connect everything together that I have no intention of taking apart. The Power Grip size I used was 15/16” – 1 1/16”. It makes for a much cleaner install. Since I don’t really like the look of standard worm clamps in the engine bay, I found some spring-band clamps from McMaster Carr that are very similar to some of the clamps Ford used on our cars. The part no. is 7329K16 and they come 10 to a pack. Jeg’s has all the plastic Gate’s connectors, Power Grip Clamps and the 90° hoses. I ended up using (3) 90° hoses, (2) T’s and one 90° connector in my system. Here’s a few shots of the degas tank and the hose plumbing with fittings.

Bypass.jpg


Bypass T.jpg


Bypass T2.jpg
 
Last edited:

Catmonkey

I Void Warranties!
Established Member
Premium Member
Joined
May 20, 2011
Messages
3,854
Location
Louisiana
I ran into the pic limit in the previous post. There are the remainder of the pics for the section above.

Bypass lower.jpg


degas 1.jpg


Degas 2.jpg


Now I needed to connect the outlet from the T on the degas tank to the intercooler pump inlet. This might have been a location for a few 90° hose ends, but would have required connectors along with it, and I wanted to use a straight length of ¾” hose. After trying the run with a few lengths of standard rubber ¾” heater hose, I didn’t like the fact that the bend necessary to route the hose was partially collapsing the hose at the bend. I figured silicone hose might be a better material to prevent that from happening. After acquiring a length of it, it worked out fine. It was a pretty easy run from the pump to the heat exchanger, but that location is also made easy the VMP triple pass inlet. You will have to figure out another solution to your heat exchanger inlet, if you using another heat exchanger because inlets and outlets are not always in the same place. You may need to modify the heat exchanger outlet side of the hose to the intercooler as well.

Lower hose.jpg


Bypass lower 2.jpg


Bypass lower 3.jpg


Bypass installed 2.jpg


Coming off the heat exchanger is where I mounted my contraption for the temperature sending unit. This is where the other 90° hose comes into play on my particular set-up.

Guage adapter mounted.jpg
 
Last edited:

Catmonkey

I Void Warranties!
Established Member
Premium Member
Joined
May 20, 2011
Messages
3,854
Location
Louisiana
Next came the wiring for the pump. There are only two blades in the Bosch pump and 4 blades in the Peirburg pump. In Tob’s thread, there’s discussion of a harness that seems to be used by the 13-14 that can plug into a length of harness to extend the wires to the pump from the OEM connector for the Bosch pump. After reading as much information as I could on the Pierburg CW50 pump, I came across several sources that indicate that running power to pin 4 and grounding pin 1 as wired by Ford is NOT running the Pierburg pump at maximum capacity. Wiring the pump in this fashion I’m of the impression that it will only run the pump at something less than 50% of its capacity. No doubt Ford engineers are well aware of this fact, but it makes you wonder why that is the case and the pump it not wired to run full tilt. The white paper I found on the pump that goes into the pump wiring and other observations on not using pins 2 and 3, can be downloaded below.

I am by no means an electrical engineer, but what I’m deciphering from this document is that the pump runs at high speed for one-half second and low speed for one and a one-quarter seconds without any signal to the PWM pin. Based on the author’s measurements, he’s coming up with a minimum of 13% of the rated capacity as low speed and 85% of capacity as high speed. Based on my calculation, that timing interval equates to 42% of the rated capacity by constantly cycling through that routine. I did not attempt to independent verify the decrease in flow. According to the document, if you connect full power (12v) to the PWM pin, in addition to the battery pin (pins 3 & 4), the pump will run at maximum capacity. This is not possible with the two wire Ford harness, part no. DR3Z-14A411-C. I know that for a fact, because I did test the connector. I installed this harness in my car, and upon removal I was able to test the pins. Pins 2 and 3 have no terminals in the connector so there is no way to make contact with the blades in the pump motor. The PWM pin can also be used with a controller to change the pump speed and turn the pump on at a certain temperature. Lingenfelter promotes one of his controllers as working with the Pierburg CW100 and CW50.


I also noted in Departments of Boost’s directions for using the Pierburg pump with their intercooler system for the GT450 supercharger system, it also calls for a different pigtail and wiring pins 3 and 4 to power to get full power to the pump. Ford makes what is claimed to be a “repair” harness and it’s the same connector as the previous part number, but has a wire for each pin (all green) and the wires are 18 ga. instead of 20 ga. wire. The part number for that pigtail is CU2Z-14S411-AYA and it was mentioned early on in Tob’s original thread. Obviously back then no one knew there was any difference.

Here is that Ford connector wired into the pump wiring. Pins 3 and 4 are connected to 12 volt power in my harness.

pump connector.jpg


It still begs the question as to why Ford is using a lower speed on the 13-14, or if the DR3Z-14A411-C harness is even used on the 13-14 GT500. Even if it’s not, I had a member send me the wiring schematic for the CAC system on the 2013 GT500 and it only shows pins 1 and 4 as being used and 2 and 3 as not used. No doubt someone will integrate that 4 wire connector into a later model and see what ultimately happens in an otherwise stock intercooler system.

I’m not sure what will happen with these mods in my car, but I have run the pump with a jumper for about 5 minutes to see if there were any perceptible issues after the install. I’ve since had the car idling for extended periods while working on other issues with IAT2s in the 125* range. Still no foaming in the degas tank. One of the issues I was working out was no signal from the fluid temp sensor in the intercooler system. I’ve since got the temperature sensor working, but I’m in the process of swapping out the blower and that’s the point of where I am as I’m posting this. I’ll post comments in the general section once I have some time on the system.
 

Attachments

  • EVTV Pierburg CWA-50 Coolant Pump.pdf
    187.9 KB · Views: 1,471
Last edited:
Status
Not open for further replies.

Users who are viewing this thread



Top