"Split-Flow" Cooling Explained w/ Pics

[Cowracer]

Guys,

This was a mod that I have been wanting to try ever since I read about the hummer guys putting fittings in the rear-head block off plates. They use those fittings to take water out of the heads and dump it back to the radiator return in an attempt to increase flow and help shed heat.

To me that approach is completley ass-backwards. Taking water out of the heads will only make matters worse. What needs to be done is to ADD cool water to the back of the heads.

So what I did was to tap and drill holes in those block-off plates for 1/4" pipe fitting with hose barbs for 3/8" heater hose. Likewise, I drilled and tapped holes the the plate where the water pump mounts at the front of the engine.

Normally, water leaving the pump must pass through the block, absorbing heat, on the way to the heads. These new holes tap into the pressurized water coming directly out of the water pump. This water is as cool as any in the system, as it just came from the radiator.

So far, my engine is still in the break-in phase, so I haven't pushed it hard at all, but in daily driving, there appears to be fucntioning well. Once I get 1000 miles or so on the engine, I will do a towing run. I have figured out a way to measure rear of head, and front of head temperatures, and I can easily block off the bypass lines for comparison.

But for now, here are the pictures

http://www.gatewaydata.com/truck/cooling1.JPG

This shows the fitting installed on the water pump mounting plate.




http://www.gatewaydata.com/truck/cooling2.JPG
This picture show the arraingment of the hoses at the back of the motor.



http://www.gatewaydata.com/truck/cooling3.JPG
Front view of the motor, showing the hoses on both sides. The right hose clears the alternator bracket when its installed.

NOTE: I am not sure if the left side fitting will clear the vacuum pump if used.


http://www.gatewaydata.com/truck/cooling5.JPG
A slightly blurry shot of the block off plate detail


Tim

[CareyWeber]

Tim,

That is the first time I have thought that this idea has made any sense at all.

Now I may have to try this MOD before I just thought the idea was well kinda dumb, but adding cool water to the rear of the heads is a good idea. :cool:


I'm eager to hear how it works for you.

Carey

[TJ Moose]

Having done the Heath balance flow conversion (which IS a bitch with the engine in place) it has some similarities to your setup, although I know you're saying you put water in at the back of the engine instead of taking it out. For the sake of thinking about what we're doing - probably need to think heat flow more than coolant flow - all the coolant is - is a medium for taking heat from one place and moving it to another. So as long as we're pushing heat from the block, and moving it to the radiator where we can get rid of it, and keeping the block temp relatively uniform throughout - its a good thing. Admittedly, this is more high concept than the exactitude of how we accomplish the task at hand - but as long as we're moving heat and keeping the block temp uniform - there probably isn't a right or wrong way to do it. I know I'm liking how my Heath setup is performing - and the install was doable (once I got my head wrapped around laying on top of my engine for a number of days) with the engine in place. If this works well for you, that's good stuff too.

[JohnC]

I'm really skeptical about any of these ideas. If you take water out of the back of the heads, you reduce the flow through the heads and likely raise the temperature farther forward. If, on the other hand, you introduce water into the back of the head, you're reducing the flow through the engine to the head, which may cause trouble elsewhere...

Too much for my brain to process!

[Billman]

I'm not seeing how you will get flow through those hoses?

[gmctd]

FYI - points to ponder -

The single stat 6.5 coolant system is constant recirculative until the t-stat opens, at which point a separate valve plate on the t-stat closes the recirculate port, and all flow goes to radiator.

The '97-up single and dual stat system is constant recirculative until the t-stat(s) open, at which time coolant flow is ported to the radiator, without blocking the 1" recirculation port to the engine.

This was done to relieve the sometimes wild heat-load swings, an attempt to maintain loaded\unloaded thermal stability in the block and heads.

[Kennedy]

I'd also be concerned about life of the hoses and protecting them from chafing on the lower intake etc.

[moondoggie]

I am REALLY interested in following what you're doing here. The way our engines

[Cowracer]

TJMoose and JohnC:

The idea is to rid as much heat from the heads as possible. Heat transfer is always from hot to cold, and the greater the differntial between the two, the greater the transfer.

All other things being the same, water entering at the back of the head at 180* will remove more heat from the heads than water at 200*

Taking water OUT of the system at the back of the head will help cool the WATER but it does nothing (or at best very damn little) to cool the head.

What I was trying to accomplish was to augment the water at the back of the heads. NOT replace it, but just inject some cooler water in order to drop the overall temp of the water going into the heads. The risk was starving the block for water, but I figured that two 3/8" hoses would not affect flow through the block that much.

Billman:

Water will always flow from a higher head (or pressure) to a lower one. The highest head in the coolant circuit is at the outlet of the pump, the lowest is at the inlet of the pump. The total flow of all the water out of the engine is through the upper radiator hose, which is at a lower head than the engine itself. It does not matter if the water goes through the block, or through hoses to get to the back of the cyl head. Once at the rear of the cyl head, the head presssure is higher than at the front, where the cross-over and radiator hose is.

Where there is a difference in head, flow is guaranteed.


GMCTD
Never thought about that, then again, there is probably much I haven't considered. Thanks!

JK

You bet. I plan on watching them closely for signs of distress. I bought good quality heater hose, and I routed them as best I could away from anything with sharp edges. I may add a protective loom in the future.

Moondoggie

Caution is the watchword. The nice thing is that I can revert back to stock in about an hour with 4 pipe plugs, if it dont work out

I am NOT advocating ANYONE trying this yet. I am still in the initial tryout stages of an uproven modification. I just wanted to let you know what I am doing.


I love thinking outside the box, sometimes it works, sometimes (well, most of the times) it don't. If it does, call me hero, if it dont, call me fool.


Tim

[turbovair]

How about a couple check valves to ensure coolant can ONLY flow from the pump to the heads and not the other way around?

[grape]

considering the remarkable track record of these engines, GM really had the engineering down pat........I can't imagine doing anything different from the way GM did them :rolleyes:

the other gm head I have experience with cost $2k per set bare. Then with another $6K worth of rockers, valves, seats, guides and cnc portwork you get to read the little peice of paper from GM that says "for best performance use 1 quart of sodium silicate per radiator fill up for competition use." GM finally got their heads to not leak with the SB2.2 :rolleyes:

some of you guys must mistake engineers at GM for careing about performance instead of profit........ain't gonna happen in a street vehicle. But for you diehards that think GM can do no wrong, take a look around at how much power you can pick up with a stock LS1 with a cam change........then tell me gm cares about performance in street cars.

[john8662]

Originally posted by grape:
But for you diehards that think GM can do no wrong, take a look around at how much power you can pick up with a stock LS1 with a cam change........then tell me gm cares about performance in street cars. Just to prove the obvious point, there are MORE variables to consider than pure profit verses performance.

emissions, this is for real for production cars, at the track no one is sniffing exhaust pipes for what they emit, nor looking for kitty cats.

Driveability, how well all around does an LS1 with a hot cam run? Great at WOT, but not too great for sitting at a light (being shook). How about starting performance, hot weather, cold weather. General Mess has to consider how this product is going to perform in a variety of conditions. Also, what is SAFE for the consumer to drive?

Maintenance, adjusting rockers due to high lift on a race engine is another consideration not usually done by the normal driver. Most of GM's LS1 engines run the 100k mark being relatively untouched.

But I'll give you the point that GM's engineers are probably looking at what it's going to cost in the long run. Think compact designed block used in the 6.2/6.5 and block cracking.


Cowracer,

I'm actually believing that this system will work, since you're after saving the heads by reducing temperature. You might also consider running a temp probe in a freeze plug to see what the temp is around the cylinders in the block. My reasoning behind this is because you will be taking flow from going to the block (some) and it will be hard to see the affect of block temperatures since you're going to possibly see lower temps in the heads, which is where the water temp sender is. So you could be seeing artificially cooler temps than whats occuring in the block.

[JohnC]

Originally posted by Cowracer:
...What I was trying to accomplish was to augment the water at the back of the heads. NOT replace it, but just inject some cooler water in order to drop the overall temp of the water going into the heads. The risk was starving the block for water, but I figured that two 3/8" hoses would not affect flow through the block that much. ...Yeah, my point is that the water you're adding at the back of the heads detracts from the water flowing to the heads through the block, and also raises the pressure in the heads, further detering flow through the block. It would follow that the water that does flow from the block to the heads would then be hotter than in the absense of these hoses.

The more I think about it the more I think the solution is simply more flow through the whole engine, but that's what the later cooling mods are all about...

[CareyWeber]

Originally posted by john8662:I'm actually believing that this system will work, since you're after saving the heads by reducing temperature. You might also consider running a temp probe in a freeze plug to see what the temp is around the cylinders in the block. My reasoning behind this is because you will be taking flow from going to the block (some) and it will be hard to see the affect of block temperatures since you're going to possibly see lower temps in the heads, which is where the water temp sender is. So you could be seeing artificially cooler temps than whats occuring in the block. [/QB]Maybe a restrictor in the two added lines taking coolant to the rear of the heads. I would think if you had the 97 and 130 GPM water pump it would help ensure that the block still got enough coolant.

Carey

[moondoggie]

Good Day!

It's great everyone that can is helping Cowracer with his practical issues. I had just hoped someone might be able to address the theoretical ideas I presented... :(

Blessings! :D

[GMC Hauler]

So,

After reading what GMCTD wrote about the different type of thermostat, is it advisable to change all 96 type thermostats to 97 and newer types?

What about blocking off the heater hoses for summer towing? This should help increase some flow through the block.

Anyone ever thought about a larger water pump pulley?

[More Power]

A smaller WP pulley will increase its speed of rotation, and may increase GPM flow. Cavitation might eventually come into play if pushed too hard.

Having an engine equipped with a variety of temperature sensors before/after the mods is the only real way to determine the effects of changing the coolant flow paths. Tests would need to include a variety of engine loads (heat production) and rpm ranges (WP flow rates).

Frankly, I'm skepticle of any engine cooling improvements (lower head/block temps) using mods beyond a stock coolant flow strategy (i.e. 1997 cooling mods), but I'm always interested in data.... smile.gif

Jim

[gmctd]

Heater hose block-off valve will increase recirculating coolant temps.
Flow is out of t-stat crossover manifold, thru heater, into radiator cool-side tank, then back thru waterpump.
Provides a mix of cooler water with the recirculated water.

For those who have not - '97 cooling upgrades have been advocated by this forum since ca'97, with little or no ill effects or serious complaints.
Really helps loaded, when towing, or hot-rodding.
Just do it! ! ! !

Moondoggie - several marques have used the constant recirculating coolant method thru the years in their respective gasoline engines, even Chevrolet.
Gasser engines are notoriously easy to cool, tho, even with the smaller and smaller radiators, and proliferation of electric coolant fans.

GM\Detroit Diesel apparently considered it absolutely mandatory, over several design changes, to deal with the massive thermal output of the indirect injected 6.5 heads, in turbo application.

Again, I would direct any gentle reader to the Member's Area, 6.2\6.5 head repair article, excellent cut-away head pic, to get a sense of the problem.
Should help, imo, with any pre-\mis-conceptions most folks have concerning gasser head or direct injection Diesel head-cooling technology.

Imo, of course, as usual.......

[Quack_Addict]

Originally posted by grape:
some of you guys must mistake engineers at GM for careing about performance instead of profit........ain't gonna happen in a street vehicle. But for you diehards that think GM can do no wrong, take a look around at how much power you can pick up with a stock LS1 with a cam change........then tell me gm cares about performance in street cars. High performance and engine/powertrain durability normally don't go hand-in-hand. There's a lot to consider... emissions, warranty, fuel economy (ever hear of CAFE?), costs, etc.


I don't really think the 3/8" hose mod is going to help in cooling the heads a noticeable amount. Keep in mind, the fittings are 1/4" NPT, no? Pressure = Force / Area... and a 1/4" I.D. doesn't have much flow area... and this isn't taking flow restriction into account due to the length of hose and bends in it.

[BobND]

There are different cooling issues with spark ignition engines than with our diesels.

Supposedly, reverse-flow cooling systems are designed to cool troublesome "hot spots" that cause engine-destroying pre-ignition.

Pre-ignition is not a problem here, and (especially on turbo engines without piston oil spray jets) it seems keeping the block as cool as possible helps prevent piston burning or cylinder scoring.

So, changing the coolant flow could POSSIBLY trade one problem for the other.

[More Power]

Cowracer, Don't let anything said here dissuade you from experimenting. Motivation and creativity combined with unanswered questions move the technology forward.

The idea of altering coolant flow paths comes up about once a year here in this BB, and the question of its usefulness remains unresolved - or at least proven and documented. smile.gif

The engineers out there may be able to help devise a plan to monitor coolant and engine temperatures at various important points to check for variations in temps, then judge the value of of those changes.

Jim

[Cowracer]

Originally posted by More Power:
Cowracer, Don't let anything said here dissuade you from experimenting. Motivation and creativity combined with unanswered questions move the technology forward.

The idea of altering coolant flow paths comes up about once a year here in this BB, and the question of its usefulness remains unresolved - or at least proven and documented. smile.gif

The engineers out there may be able to help devise a plan to monitor coolant and engine temperatures at various important points to check for variations in temps, then judge the value of of those changes.

Jim Jim, I started to reply to your earlier post (and I actually thought I did, but it never showed up, so I guess I did something wrong).

The comments here will not dissuade me at all! I value contrary opinions, even when I am right and they are wrong! ;) :D


Data Collection is the name of the game...

I have access to all kinds of nifty insturmentation here at work. I could monitor this thing a dozen different ways, but there isn't and abundace of ports to the water jackets in the heads on a 6.5, and I am a tad leary of drilling holes in the heads just to put in some thermocouples.

I could monitor surface temps of the heads and maybe even get thermal images, but I dont know how benificial that would be.

After getting a thousand miles or so to get the new engine well broken-in, my testing plan is to tow an ~8000# load with split-flow cooling operational a 35 mile circuit on at least an 85* day. I will be recording engine coolant temps at 3 locations, rear of head, front of head, thermostat housing probe on a graphing datalogger.

The route I am thinking of will include interstate, state hiway and secondary county roads.

I will then let the truck high idle for 20 minutes to stabilize coolant temps, and repeat the circuit with the split-flow system disabled, again recording the temps in those three locations. That should prove if split-flow cooling has provieds any benefits worth the effort.


Tim

[ 07-14-2005, 11:21 AM: Message edited by: Cowracer ]

[gmctd]

FYI - Each coolant passage between each pair of valves has an exit just below each intake port on each head.

Conveniently cap-plugged, those passage ends may be tapped for an insertable probe, whereby individual cylinder temperatures may be monitored during experimentive testing.

Iirc, Bill Heath's system was derived and proven, using the multiple-probe method

p.s. - neglected to mention it, but I wholeheartedly second MP's presentiments on this

[ 07-14-2005, 01:40 PM: Message edited by: gmctd ]

[GMC Hauler]

You should ask Bill how he tested and what he found.

I am running one of those systems. A few things I have noticed and have been wanting to say, but havent:

1. It has no problems bleeding. I refilled the system normally, and ran it. It filled up with liquid. How do I know, you ask?? It got real hot, engine coolant hot. Coolant level in the reservoir stopped dropping after a week and has stabilized. The balance flow tank still gets hot. I have been running for 3 months now.

2. I am not seeing increased temperatures at the front of the heads, even while towing. Remember, you have two temperature probes already installed, one in the crossover for the PCM, and the other in the front of the drivers side head forward. These temperatures havent changed since install.

So I guess all that needs to be checked is rear temperatures coming out of the two rear outlets. I don't have the time now, but after I get to CT, I could modify each one to take a temperature or install some temporary connections. This way, we could look at temperatures coming out of all 4 areas. If none are hotter than the others, then most likely we have a winner.

I am a skeptic sometimes too, and wouldn't mind proving this to myself.

By the way, does that water wetter stuff actually work???

Still, is it advisable to change the thermostat type from the 96 type to the 97? would this reduce flow through the radiator to a point where the system is efficent? Has anyone tried it???

Edited to ask more questions.

[ 07-15-2005, 07:25 AM: Message edited by: GMC Hauler ]

[moondoggie]

Good Day!

[i]

[TurboDiverArt]

[quote]Originally posted by moondoggie:
[b]
My comments about reverse cooling were theoretical

[DA BIG ONE]

I just had an idea of using the washer type (cyl head temp)heat sensors at each cylinder for data collection. So, would any of you know if this is a solid idea, and what bolts should be used intake, exhaust, glow plugs, or?

[moondoggie]

Good Day!

TurboDiverArt: Thanks so much for the excellent reply - I'm gonna read all that stuff. I knew SOMEONE on the Page would know about this. I REALLY appreciate great technical information without a (probably deserved) "you moron" attached. :D

Sorry I can't help Cowracer with his situation - sure would if I could.

(on edit 19 Jul) Read both last night - excellent. Thanks again.

Blessings!

[ 07-19-2005, 07:30 AM: Message edited by: moondoggie ]

[tom.mcinerney]

This is a nice discussion. I think extra baseline coolant flow to head should decrease cracking. Art's point about air is a potential issue.

I've been meaning to raise a related question:
The marine engines employ water-cooled exhaust manifolds , and water-cooled turbos(i think).
Many of us have noted the psgr side head has an excess of cracks over drvr head {at least in pre '97 coolant upgrade rigs}, presumably a result of the additional thermal load imposed by the utube and turbo on psgr side.

Matt at Peninsular should be able to confirm that the heads in marine applications have less cracking than dry automotive, and less excess of psgr over drvr side. If so, then overboring the small coolant-metering holes in the head gaskets in the vicinity of the turbo(upstream) might equalise temps between both heads.

If the heads and gaskets are interchangeable, there is nothing obviously compensating for the heat load of the turbo side.

More Power to you, Cowracer! Echoing an earlier post...Ah Vants one too!