How to bleed water-to-air intercooler? (1 Viewer)

This site may earn a commission from merchant affiliate
links, including eBay, Amazon, Skimlinks, and others.

Joined
Jul 24, 2018
Threads
4
Messages
72
Location
New Zealand
Hi all,

I have just finished the install of a WTA intercooler in my '92 Surf, filled with water / coolant mix as best I could but EGTs are just the same as prior and it seems no heat is making it to the heat exchanger on the front of the truck. After a bit of mucking around including setting the truck up on an angle with one front wheel raised, I eventually managed to get a good flow of coolant through the radiator filler cap that's part of the system, but no result when I took the truck for a long test run. I opened the cap back up after giving it some time to cool down and it still looked full. I'm not getting any cooling it seems, so I figure there must be an air lock. Any tips on how to bleed would be much appreciated...
 
Normally you design them with a header tank and they self-bleed.
 
Fixed for fool proof simplicity :lol:
Not helpful but a fair comment 😆

I went water to air due to space constraints and wanting to keep the A/C. I thought about a top mount ATA, but concluded it would require a fan to be worthwhile and there's definitely no space for that. Additionally I wanted the extra performance of the WTA. The pump business is a pain in the arse but I'll get there in the end so no regrets. I went with a cheaper aftermarket version of the Bosch pump and possibly the actual reason the system wasn't working was simply because the pump failed unrelated to any air in the system. I guess I'll never know. Anyway, a new genuine Bosch pump is on it's way from Australia.
 
Can you tell us a bit more about your setup? What you are using? Ideally some pictures?

The “easiest” way I’ve found to do water to air is have the water pump directly connected to a reservoir/tank… so as you fill the tank the pump can push water through the system. Even if the intercooler is higher, the air can be pushed through. Again though, this heavily depends on pump, line size, height difference, intercooler angle etc etc.

By yeah, a well designed system should self bleed automatically.
 
Can you tell us a bit more about your setup? What you are using? Ideally some pictures?

The “easiest” way I’ve found to do water to air is have the water pump directly connected to a reservoir/tank… so as you fill the tank the pump can push water through the system. Even if the intercooler is higher, the air can be pushed through. Again though, this heavily depends on pump, line size, height difference, intercooler angle etc etc.

By yeah, a well designed system should self bleed automatically.
Thanks for that Feistl. Upon further investigation is seems my problem was most likely that the pump simply failed so maybe I was just unlucky and got a faulty one. Apparently these pumps can run dry ok and while it will shorten the life, it definitely shouldn't have failed within half an hour. Possibly there was no air-lock at all and the system was purged ok.

The new pump should be here within the next couple of days. I'll drain the system and then hook the garden hose up to it and give it a good flush to make extra sure there isn't any swarf or other nasties in the system.

As I understand it, not all WTA systems have a header tank or reservoir. I'm not planning to run anything more than 12-14 PSI boost so I don't think the system will be working very hard or get very hot. Space is at an absolute premium in the engine bay so hopefully I can get a good result with the design as is. Just getting the intercooler to fit was enough work!

The coolant lines are all 3/4 inch, the intercooler is 11"x9"x3.5" and the radiator is 26"x7"x2". Pump is the regular Bosh type. System is very simple. Most of it is visible in the photo below, from here the lines simply drop down to the rad. Filler cap is the highest point in the system, slopes nicely down to the IC and from there down to the pump, then drops down to radiator. There's one section of coolant hose that kicks back upwards before joining the radiator but other than than it's all very tidy.

tempImageeFCfAp.jpg
 
Last edited:
just make sure the pump itself isnt getting air locked, thats what kills em, partially fill it from an open connection if you have to to make sure the pump is filled with coolant first, then fill the rest from the cap.
did you plug the other bolt hole on the bottom of the intercooler? they go right into the pressure side.... not that that had anything to do with the pump failure
 
just make sure the pump itself isnt getting air locked, thats what kills em, partially fill it from an open connection if you have to to make sure the pump is filled with coolant first, then fill the rest from the cap.
did you plug the other bolt hole on the bottom of the intercooler? they go right into the pressure side.... not that that had anything to do with the pump failure
Hmm... ok cheers, can you elaborate on what 'the pump getting air locked itself' means exactly? And how quickly can that kill a pump? I'm pretty confident that the coolant will reach the pump from the cap just fine. If I fill then bleed as best I can and get a nice consistent flow of coolant through the cap, does that mean I'm safe?

I might fill the system as much as possible then give the truck a bit of drive over some bumps and hills, then top up before I turn the pump on to try and reduce the risk a bit. I've read comments from other people who say they just fill as best as possible then drive a bit, top up, rinse and repeat until it won't take any more liquid and that's good as gold.

I can give this a whirl too maybe:

 
And yeah I sure did plug the other bolt hole and used a bit of plumbing tape to get a nice seal. All my work on the air side seems fine. Holds 10 PSI with no issues as far as I can tell.
 
The issue you can face without having a proper reservoir is its much harder to bleed out any air, as bubbles can be carried around the system over and over, as the buoyancy of air isn't actually that good, so a high flow pump can easily suck down the air bubbles. A reservoir allows the fluid to slow down, and bubbles to rise to the top.

The pump getting "air locked" is when it sucks and pushes through a bunch of coolant, but sucks in new air... Then it can't move any fluid, so it sits there no moving any coolant meaning it gets hot and can die.

The other potential problem I see is your pump is sucking from the heat exchanger, which isn't ideal. Pumps are really good at creating pressure, but not suction. So you're effectively trying to pull the water through the heat exchanger, whereas you normally have a reservoir/filling point connected directly to the pump, that way there is no resistance/suction required to get the fluid into the pump, and the output pressure then circulates around.

I'm not saying that won't work, just its not ideal and does make bleeding difficult.

The other major benefit of W2A compared to A2A, is it takes a lot of heat energy to raise the temperature of water, meaning even though air to air is technically more efficient at dissipating heat, W2A is better at keeping a much lower PEAK temperature. EG A2A might AVERAGE say 50 degrees celcius, but have peaks of 120 degrees. A W2A with the same size front heat exchanger might average 60 degrees, but only peak at 80. Now given its the peak temperatures that kill engines, W2A can be "better".

But the critical part is the volume of water... The more volume of water the lower the peak temperatures... My W2A setup has a system volume of 11L, meaning even a really really long up hill run in the middle of summer won't see intake temps raise very much at all, whereas a A2A would suffer heat soak and much higher peak temperatures.

So yeah, question is is there ANYWHERE you have room for some sort of tank? You could even put a tank under the car (I put one above my rock sliders) and just make sure you protect the lines.

Anyway, keep bleeding and see how it goes
 
Last edited:
The issue you can face without having a proper reservoir is its much harder to bleed out any air, as bubbles can be carried around the system over and over, as the buoyancy of air isn't actually that good, so a high flow pump can easily suck down the air bubbles. A reservoir allows the fluid to slow down, and bubbles to rise to the top.

The pump getting "air locked" is when it sucks and pushes through a bunch of coolant, but sucks in new air... Then it can't move any fluid, so it sits there no moving any coolant meaning it gets hot and can die.

The other potential problem I see is your pump is sucking from the heat exchanger, which isn't ideal. Pumps are really good at creating pressure, but not suction. So you're effectively trying to pull the water through the heat exchanger, whereas you normally have a reservoir/filling point connected directly to the pump, that way there is no resistance/suction required to get the fluid into the pump, and the output pressure then circulates around.

I'm not saying that won't work, just its not ideal and does make bleeding difficult.

The other major benefit of W2A compared to A2A, is it takes a lot of heat energy to raise the temperature of water, meaning even though air to air is technically more efficient at dissipating heat, W2A is better at keeping a much lower PEAK temperature. EG A2A might AVERAGE say 50 degrees celcius, but have peaks of 120 degrees. A W2A with the same size front heat exchanger might average 60 degrees, but only peak at 80. Now given its the peak temperatures that kill engines, W2A can be "better".

But the critical part is the volume of water... The more volume of water the lower the peak temperatures... My W2A setup has a system volume of 11L, meaning even a really really long up hill run in the middle of summer won't see intake temps raise very much at all, whereas a A2A would suffer heat soak and much higher peak temperatures.

So yeah, question is is there ANYWHERE you have room for some sort of tank? You could even put a tank under the car (I put one above my rock sliders) and just make sure you protect the lines.

Anyway, keep bleeding and see how it goes
Thanks for taking the time to write all that. Aah okay, so that does make sense and almost makes me wonder whether I could find a way to run the pump at a lower speed when bleeding to give the air more of a chance to escape. Yes, I understand your comments regards ATA vs WTA and the role the coolant volume plays. I based my design on what a number of other forum members have done in their 2LTE LJ78s. These system were sans reservoir or header tank I believe, but possibly I should go back and check that. My thinking was that given the radiator is a good size and, it would seem larger than what many people use, in addition to the fact I don't plan to raise the boost much over stock anyways, a low volume system would be sufficient. It seems my setup only takes about 2 litres, even with the more generous 3/4 inch coolant hose so that's a long way off your 11 litres. It ought to flow very well though which should help with its efficiency too.

I think your interpretation of the direction of the pump is incorrect though. It is pulling down slope from the cap through the WTA to the pump then spitting it out down to the radiator, which was how I designed it after read the same thing elsewhere. I can't quite get my head around why the directionality makes any difference though. The mass of the liquid and the friction should be the same either way.

I don't have a coolant temp sensor in the system anywhere but was planning to gauge the performance by EGTs which I am very familiar with in this vehicle. One forum member is running 18PSI in his 2LTE with a WTA setup that is almost identical yet claims he's only seen the WTA coolant temps rise to 40C. My Surf is smaller and lighter than the LJ78 and I've never actually felt the initial peak power to be quite as terrible as some people claim. However once the truck has been on boost for any duration, up a steep hill climb for instance, the loss of power and efficiency due to high intake temps becomes obvious very quickly. So my gut feeling is that the system will be sufficient for the application as is, as long as I can get on top of the bleeding issue if that is in fact what killed the first pump and that's not something I know one way or the other.

The WTA has been a big project for me, I'm pretty handy but prior to this had never really worked with metal so had to learn how to cut steel, weld and to bend aluminium with mapp gas etc. Additionally the install required quite a number of components in the engine bay to be moved with wires and hoses extended etc so that was a lot of faff too. So I'm pretty keen to get it over the line finally as is, if this is possible. From what I've read on here it seems this should be true, but I am open to being convinced otherwise.

My mechanic says he's happy to give it a nudge if I chicken out. I have the new pump here now so possibly I'll have a crack over the weekend first. Maybe I can fill it under pressure from the garden hose then drain some of the water from the system using the drain hole in the radiator while adding pure coolant from the top so I end up with the right mix and no air. I'll keep pondering...

image0.jpeg
 
Any thoughts on whether I could damage the pump by forcing water through with a garden hose while the pump is off?
 
Any thoughts on whether I could damage the pump by forcing water through with a garden hose while the pump is off?

If I remember correct that is what I had to do and it didn't hurt anything. You'll be very happy with this setup once its working.
 
The garden hose method seems to have done the trick but I think actually the issue all along was just a faulty pump. The Bosch is smooth and quiet whereas the knock-off pump was a lot louder and had a bit of a whine to it. I can hear that there is still some air in the top of the intercooler though which is a bit frustrating.

Performance wise the system is good but a little disappointing at present as max EGTs have only dropped about 10%. What is pleasantly surprising though is how much better the truck performs up hills. Next I will move the number plate which is currently blocking a lot of airflow to the IC radiator and I suspect that will help quite a bit. Unsure whether I will look at adding a reservoir or not. Thanks to all contributors
 
res helps for sure
 
The garden hose method seems to have done the trick but I think actually the issue all along was just a faulty pump. The Bosch is smooth and quiet whereas the knock-off pump was a lot louder and had a bit of a whine to it. I can hear that there is still some air in the top of the intercooler though which is a bit frustrating.

Performance wise the system is good but a little disappointing at present as max EGTs have only dropped about 10%. What is pleasantly surprising though is how much better the truck performs up hills. Next I will move the number plate which is currently blocking a lot of airflow to the IC radiator and I suspect that will help quite a bit. Unsure whether I will look at adding a reservoir or not. Thanks to all contributors

Can you show a picture of your intercooler radiator?

I think I went with one about 27"x8"x2". It gets a lot of airflow in my setup.

Its hard to remember, but I think my EGTs dropped by 250-300F compared to no intercooler.

Edit, here's my thread from way back: LJ78 Air/water intercooler working! - https://forum.ih8mud.com/threads/lj78-air-water-intercooler-working.745872/

After I did that, quite a few people followed the same setup.
 
Can you show a picture of your intercooler radiator?

I think I went with one about 27"x8"x2". It gets a lot of airflow in my setup.

Its hard to remember, but I think my EGTs dropped by 250-300F compared to no intercooler.

Edit, here's my thread from way back: LJ78 Air/water intercooler working! - https://forum.ih8mud.com/threads/lj78-air-water-intercooler-working.745872/

After I did that, quite a few people followed the same setup.
It was actually your setup that I based the design on 👍 It was a lot harder fitting it in the Surf though! From what I gather, you were already running boost above factory when you did your intercooler? I was running stock 10PSI and the max (ish) EGTs dropped from 1000f to 850F with the IC in. I've read that intercoolers only really start to make sense above 10PSI and after puzzling I figured probably at stock boost I wasn't getting much bang for buck so increased the boost to 12PSI and sure enough, EGTs dropped again down to sub-800 (probably helped by the ECU limiting fuel). I'm confident if I up boost again to 14PSI they'll drop even further and then would probably be close to what you've achieved with your setup. Just not game with a 31 year old turbo 😬 I have a new high-flow CT20 to go in very soon though. Have pondered TD04 with adapters etc, but will probably have to rule that out on cost grounds. I think peak power will be fine, it's just the slow spool that sucks but hopefully that'll improve a bit with the new turbo.

Rad photo (get's pretty good air flow through the bumper too):

IMG_6264.jpeg
 
I cleared the system of more air by unfixing the radiator and tilting it around vertical each way, then unfixed the IC and tilted that vertical, pouring coolant in etc. Made no difference to performance so probably I had it pretty good before with the hose trick. Pretty sure it's more or less completely free of air now. Refund for the dodgy pump pending. New pump good as gold 👍
 
I cleared the system of more air by unfixing the radiator and tilting it around vertical each way, then unfixed the IC and tilted that vertical, pouring coolant in etc. Made no difference to performance so probably I had it pretty good before with the hose trick. Pretty sure it's more or less completely free of air now. Refund for the dodgy pump pending. New pump good as gold 👍

Looks great!

My regret on mine was that I did not place the pump low in the system. I think then I would not have had the priming issues I had when I first setup.

Make sure the lower barb is the flow inlet on your front intercooler radiator.

Also make sure the air/water intercooler doesn't have trapped air either.

After going for a drive, put your hand on the air inlet to the air/water intercooler vs air outlet. Should be a big difference. I found I could not even touch the inlet, but outlet was maybe 40C. You can also feel the sides of the air/water intercooler to make sure the water is cooling it fine.

If you can borrow a FLIR thermal camera, you can have a look at the setup that way and learn quite a bit. You could also use a couple temperature probes to measure air temp before and after the intercooler.

Once your egts are solid, you can turn up fuel a bit at the spill control valve. This makes a huge power increase, and will help your spool a lot too.
 

Users who are viewing this thread

Back
Top Bottom