TheSamba.com :: View topic - Liquid cooling with rear mounted radiators

Mon Jan 05, 2015 12:55 pm

Ive searched all over this forum (and other sites), looking at engine conversions in Bays with engine compartment mounted radiators. Obviously I am not the only one out here who doesnt want a front or belly mounted radiator. But apparently, obviously, its a real bugger to get it to work very well.

Ive got some ideas but no actual hands on knowledge with Bays, my only hands on dealt with REAL buses. Ive owned a few, the last was an MCI MC5B over the road (OTR) Bus.

On the MCI, and other Buses too, Prevost, Eagle, GM, etc., radiators are mounted up where the vents are on a Bay, far upper back sides. Because thats a high pressure area where air would naturally want to flow into the radiators.

But without low pressure behind them you would get zero air flow. On the MCI, large engine driven centrifugal blower fans form a blower box between the two outboard radiators, and vent downward into the engine compartment.

The radiators are not that huge. Approx 24 by 30 on my MC5, but thick, about 4 or 5 cores, maybe 4 inches thick. Mc8's and '9s have same size rads but are even thicker. But make no mistake, they can shed an enormous amount of heat. Many many times anything a lil VW Bus would generate. So my feelings are that the cooling problems are two fold. Not enough radiator, and not enough air flow.

Skills mentioned the entire rear of the bus, the floor, getting very warm. That in itself indicates an airflow problem. Mild Cores Bus seems to work "okayish", but still on the warm side. Is Skills engine too hopped up? Not sure. Technically, the Bus should require X amount of HP to travel at X speed, so not entirely sure where extra heat would come from. Spinning faster might be the problem, but i really feel the root of the problem has more to do with airflow and radiator capacity than anything.

Also, I see Skills and Mild Core both run the engine outlet into the bottom of the left radiator, upper goes across to the right upper radiator and out the bottom back to the engine. In virtually all applications hot goes in the top and cold comes out the bottom to take advantage of physics. Also, not sure what difference it would make, but Buses send the hot up to a reservoir, Tee off to both radiators, and "Y" the returns together before returning to engine.

I have some other thoughts for later but hoped a thread with some discussion would be knowledgeable and entertaining. There has to be something that can be done to make it work, just need someone to find it.

Mon Jan 05, 2015 1:21 pm

Some people have mounted them up top. Probably not ideal if you have a pop top though.

Mon Jan 05, 2015 2:35 pm

Thanks for the input Artvonne, seems like you've been lurking here for a while by your comments. :D

Quote: Tee off to both radiators, and "Y" the returns together before returning to engine.

I was speaking to an engineer friend on just this subject recently and this came up. Each radiator is independent of the other if I understand this.

Mon Jan 05, 2015 3:24 pm

babysnakes wrote: Quote: Tee off to both radiators, and "Y" the returns together before returning to engine.

been there, tried that. after talking with jake, the expense was more than my pride was worth.

I forget the numbers, but with belly mount I pounded an honest 80+mph running around 200*

what jake has done is amazing, but the cost was quite a lot. he wouldn't sell it to me anyway because he wasn't up to speed on my combo. that alone commands respect for him.....not willing to sell something because he hasn't tested it.

jake was pivotal in making my underbelly system work. is it where I want it? no, but I don't off road either. and not for nothing, I can't stand my phone up under my audi, and have yet to grenade a radiator after 70k.

would I take another stab at it? perhaps, but I was going for form and function. I wanted a OE appearance I didn't want a bunch of crap that didn't 'flow' with what I was trying to do.

were I to try again, it would be on someone else's dollar. I did learn a lot, but honestly, mild core is a bit afraid to push his bus much past stock speed in fear of over heating, and only God knows what happened to jsmskater.

busaru was having issues...in less than 10 miles

I can now blow donuts, do 80+ from here to boston (did it for transporter fest) without fear.

even the TDI guy is having issues. I dunno.....wasted a lot of my life and money fooling with it

Mon Jan 05, 2015 4:51 pm

Wildthings wrote: Some people have mounted them up top. Probably not ideal if you have a pop top though.

This is what I'm doing. It shouldn't be much longer before I need to order a radiator and start mocking stuff up.

Mon Jan 05, 2015 5:24 pm

skills@eurocarsplus wrote: I forget the numbers, but with belly mount I pounded an honest 80+mph running around 200*
(...)
and only God knows what happened to jsmskater.
Belly-mounted radiator on my '75, and I've run an *indicated* 75-80mph (no idea on accuracy) with temps around 190-200* on flat highway and a bit slower in the mountains (actual West Coast mountains :P ), and run it for hours on end with no problems. Summer, winter .. whatever. It just works.

Last I heard about JM is that work was keeping him so swamped he hasn't had time to fix it after the oil pump blew, and I don't think he ever got an accurate read on actual operating temps other than Not Blowing Up Between New Mexico And Washington. No idea how Nepenthe88's setup is working out, but I think he was having trouble with heat at speed...

Mon Jan 05, 2015 5:46 pm

yep. belly is for sure the way to go for a inexpensive effective way to cool.

55superbeetle did a roof mount that was hidden in a luggage rack with great success so that is possible too.

I have made my system (other than the radiator) pretty much 'off the shelf' and at 159 for a aluminum 2 row shipped to my door, as big of a pain as it may be, a replacement is a day away.

Mon Jan 05, 2015 6:15 pm

skills@eurocarsplus wrote: 55superbeetle did a roof mount that was hidden in a luggage rack with great success so that is possible too.

Rats, been beaten to it i see. This is exactly my plan. Shouldn't be able to see it when I'm done. I don't want to lose any ground clearance.

Is there a thread on that build? Search didn't turn up anything.

Mon Jan 05, 2015 6:36 pm

can't comment on water cooled VW setup, but can on cooling water cooled SCCA race cars. Four things to consider if you are trying to max good cooling for space.

1) use as many rows and tubes as will fit, and the more the fins the merrier. Not all rows, tubes and fins are created equal so keep it in mind when shopping price before performance.

2) Make sure the inlet hose(s) is/are held open by mechanical means if you can - especially if you know they are longer than they should be. I have seen some beefy hoses collapse at speed, and as a result choke down the amount of water that can move thru them. We noticed it on the dyno first time we saw it and could not believe what we were seeing - but remedying it really made a difference in cooling at faster RPMS. Don't assume it is the workload that is causing the temperature to rise.

3) Water pump impeller cavitation can and does happen if one does not watch pump speeds

4) Run the right mixture. Too much water or too much coolant can affect the cooling.

One other thought - if you can avoid it never let the head(s) be the highest cavity in the system. If you do have issues then steam will collect there and ruin your day.

Mon Jan 05, 2015 6:39 pm

One thing not brought up is that skills has an early bus, the later bus inducts more air through the larger scoops. Air flowing through the engine bay seems to be one of the biggest issues, the mechanical driven fan could work if done correctly. It's gonna be a headache and take a lot of R&D. I don't see the outside temps to ensure the in bay set up would work in hotter climates.

Mon Jan 05, 2015 6:43 pm

Search for "Roadcow"

Mon Jan 05, 2015 6:52 pm

ah, thanks. That's huge and in the back still. My plan is using front luggage rack, planning on glassing a second one to overhang the windshield a little bit. Pipes running along the rails for coolant and a dummy pipe looping around the front to support the "scoop" and mount some offroad lights onto.

Mon Jan 05, 2015 9:11 pm

if you are trying to run pipes up to the roof, to the front of the bus I am going to bet you will need in line coolant pumps.

I think part of my problem was having the weight of the coolant trying to be pumped up/over/down/around was more than the pump could handle.

as SG points out, straight water cools the best. 50/50 mix will add about 15-20* to the mix.

I am curious to see how some of you make out. I had quite a few expensive scanners to log what my ecu was seeing in regards to coolant temps.

another issue, even running a 091 with tall tires wasn't ideal for the Subaru at 75mph+. you are turning 4-4,200 rpm's. way too much for the Subaru imho for extended trips

backing off to 65 I watched my ECT drop an honest 10-15* which told me 1 of 2 things:

pump was either starting to cavitate

or

the coolant was moving so fast through the radiator it couldn't stay put long enough to actually cool.

it's all related. high rpm's=high oil temps which is cooled by water (in the case of us swappers)

my re-geared transmission puts me at 3k at 75. I am actually thinking of dropping the tire size a hair to gain a couple hundred rpm's, but I want to work with what I have first.

as jake raby always said "it's in the combo". <<<<<<<that's real. if you think that you can get away with using stock gearing and cruise 75 good luck. I was having issues with a 091, let alone what a 002 would do for you.

the most important thing I did learn from this swap was everything has to play together. my reworked trans is just a stop gap till I can go with a 5 speed.

I will be taking donations :D

Mon Jan 05, 2015 9:31 pm

tristessa wrote: skills@eurocarsplus wrote: I forget the numbers, but with belly mount I pounded an honest 80+mph running around 200*
(...)
and only God knows what happened to jsmskater.
Belly-mounted radiator on my '75, and I've run an *indicated* 75-80mph (no idea on accuracy) with temps around 190-200* on flat highway and a bit slower in the mountains (actual West Coast mountains :P ), and run it for hours on end with no problems. Summer, winter .. whatever. It just works.

Last I heard about JM is that work was keeping him so swamped he hasn't had time to fix it after the oil pump blew, and I don't think he ever got an accurate read on actual operating temps other than Not Blowing Up Between New Mexico And Washington. No idea how Nepenthe88's setup is working out, but I think he was having trouble with heat at speed...

JSMSkaters ran fine from Las Cruses to here - no tweaking at all. I know he has some refinements but no time yet as stated due to his high demand at work. Unfortunately - Im in the same time boat most the time - ADRR. Joe did come over after xmas but all we did was rub on my heaters :wink:

Mon Jan 05, 2015 9:39 pm

Wow, lot of replies, very cool. But lets keep the discussion focused on compartment mounted radiators. Not under, up top, or out front.

Skills, and Mild Core, do either of you have any pics of how you tried running the rads parallel?

Both of you have twice the radiator you should need, but for some reason its starved for airflow.

Electric fans. First they were used to augment the engine driven fan for AC. Honda was one of the first to run solely on electrics, but the read out front getting ram air. It can cool itself without fans once its moving. But none of them really move that much air, they dont have enough power.

Side scoops. I dont think the later larger scoops are any big deal. Maybe some, but the main problem is something is blocking airflow. The aircooled motor has a very powerful centrifugal fan. With the motor blocked off around the compartment, and air coming in from the vents is "pushed" out through the engine.

My diesel pusher had a big ass radiator on the back, and the engine drove a huge pusher fan blowing rearward. Come to think of it my CASE skid loader did the same thing, and so did my Bobcat excavator. That CASE really blew some air and the fan made some noise. All three blew through the radiator, not suck.

I have seen a few try rear facing radiators but they all had electric fans. Has anyone tried a rear rad with an engine driven pusher fan?

And seriously, if you want a fan to work its best, it has to have a shroud.

Mon Jan 05, 2015 10:02 pm

Sorry to hijack, Thanks for the info. I'm hoping 5 gallons of coolant and plenty of airflow off the windshield will be on my side. A lift pump but that won't to hard to add after the fact if I have to.

Mon Jan 05, 2015 10:26 pm

No applogies necessary. But to help you out, It shouldnt be any problem pushing it up to the roof, because the other half is returning at the same time, so they balance out. Its a closed system, up, down, or sideways isnt the issue. If the motor can pump coolant to the front bumper and back, it can pump it to the roof.

But the slightest leak up there and your done.

Anyone thought of putting the filler neck up higher in one of the vent cavities, through an access door?

Tue Jan 06, 2015 1:34 am

Where everyone makes the mistake with this is using a cheap, easy to source radiator that has little thermal conductivity. The second mistake is improper fan controllers and thermostat, along with believing the suby water pump will do the job all by it's self

Tue Jan 06, 2015 4:21 am

try looking here;

http://www.rjes.com/html/radiator_guide.html

I'm looking at a subaru installation in my Doka - as I have a lot of space between the rear 'firewall' and the fuel tank i'm proposing to install my radiator where the rear firewall is with an under slung scoop and ducting to the radiator through the cargo floor. The space is huge so I should be able to use a larger rad. I've seen before an alfa romeo engine in a bay window where the battery was relocated and an under body scoop attached to ducting which went through the battery trays on both sides of the car direct to the radiator - the radiator was located directly where the engine bay opening was. He'd also ducted the lower parts of the bay window air intake vents so there were 4 ducts blasting the radiator. It was hideous but effective, i'm sure a prettier solution could be achieved
:)

Tue Jan 06, 2015 7:23 am

Here is another rear mount solution. High quality detailed build.

https://www.youtube.com/watch?v=VslAr_w736M

Seconds the buy a better radiator to start comment. Ultimately even this set up did not provide complete cooling under all conditions with a 50/50 coolant mix. To increase cooling margin, a small auxiliary under belly radiator was installed without a fan. If I understood his build correctly, this was to achieve zero fan use while moving.

He did some measurements of airflow while driving and found that the air velocity across the rear radiator was almost zero.

Think about this, if we assume the air vents (late bay) on the body are 4" x 12" that is 96 sq-in of area not subtracting the area of the grills. The radiator linked above is 36" x 8" or 288 sq-in. That is a 3 to one ratio assuming the vents are directly into the flow which, they are not. So if you are driving 75mph the airflow through a rear mounted radiator sized above will be less than 25 mph to the first order.

What I am trying to point out, is you are generating the heat load of 75mph HP and cooling with 25mph airflow or less. Adding fans will boost this to ???. Engine driven fans run all the time. Yeah, but at a cost in HP.

The moral of the story above is using the side vents severely limits the airflow you can get through a rear mounted radiator and you will have to maximize the thermal efficiency of every other component in the system to compensate. This means a high thermal efficiency radiator, understanding the coolant mix, making sure that pumping losses of the coolant are minimized, proper ducting and about 10 other things I don't know about because I'm not a cooling system engineer and have not taken the time to trail and error a solution.

Learn from those who have gone before, this a hard problem to crack and those who have done it have invested lots of time, money and thought. If it was easy, EMPI would have put a crappy kit out by now.