Skip to main content

I thought I would start a topic for us to discuss this point.  Most of us who grew up with VW or early Porsche know the Air Cooled issues but now that we are starting to see the Subie platform being used... being a boxer motor with water jackets this bring the biggest start of the season issue to assure the cooling system is up to snuff and well sealed and free of air pockets to assure trouble free first drives. 

Original Post

Replies sorted oldest to newest

Porsche factory service manual on the Boxster instructs that the car be run up to full temp, then jack up the nose of the car using only 2 posts of a 4 post lift, essentially as high as you can get it before the exhaust tips touch the ground.  Let it sit and cool and it will purge the air to the high point, untrapping it from the engine/crossover/previous high points.  We use this same trick on Subaru powered Becks and it works very well.

We also use a vacuum bleeder for our initial bleed and that is really handy in removing all the air.  Pull the system down to 30" of vacuum, let it sit for 10 minutes for a leak check and then let the vacuum pull the coolant into the system.  Generally speaking it will keep 95+% of the air out of the system and make for an easy initial purge.  This doe snot work on a system with coolant traces, only a new bleed/dry system.

I personally do not like purge valves or petcocks in the system, as they are potential failure/leak points, so I run a seamless system from stem to stern, but I read that guys have good luck using some sort of purge on the high points also... just not for me.

OK, back to work for me... :-) 

Thanks Carey,  I appreciate your feedback and experience with this as a builder.

If you already have fluid  in the system I wonder if some sort of vacuum pump could still be used? Probably not 

A good sealed system is what we all are really trying to achieve and some are better at it than others at achieving this.  Maybe using AN fittings everywhere would help

 

Well that depends    

Bleeding the system is only needed if there is some area where the system is not completely sealed.  

The cycle usually goes like this.  Air can get sucked in and can cause an air bubble on the high spot, sometimes the coolant expands and it out flows on your garage floor, and then the water temp can climb especially at idle which is not something you need or want.  

Hence the story from Ed and others that they have a bleeding routine yearly.  

Bleeding the coolant system for most Subie I think, except those who have been able to achieve a completely sealed system seems the caution tale for saving your vehicle especially due to the length and complexity of our cars plumbing.  

Update, I believe that this is the achilles of subaru installs and your mileage can vary. 

Last edited by IaM-Ray

All our Subi cooling systems may be slightly different, but some general rules apply.

Don't overfill your expansion tank.  If/when you do, you will get the coolant spillage mentioned by Ray above.  That's why the FULL mark is about 1/2 way up the side of the tank. 

The easiest way I have found to purge a Subi rad is to use a funnel that effectively raises inlet of the rad.  This seems to work the same as elevating the front of the vehicle.  Try Amazon or your local parts store.  They are plastic, cost about $25 and have a variety of fittings for different applications.

If you google "bleeding Subaru radiators" or similar, you'll find that there are dozens of you tube instructions, interpretations, texts, regimens, and strict orders that CANNOT be deviated from, etc.  After all, we're car guys, right?

When I read Chines reply above, I realize that there may be similarities between boxer-style engines and boxster engines, but the query was regarding Subaru engines specifically, at least in the heading.  The Subaru platform does have a few unique issues.  I know that the Hines family and SE knows this, but others may be confused.

The similarities between a Boxster cooling system and our (Beck) cooling systems was the reason for my comparison.  Both have front mounted radiators that are essentially the low point of the system and like to trap air.  Having worked on several Boxsters and referencing the factory service manual, we figured we'd try the same technique on our Suby powered cars and it works well.  

@Lane Anderson @Tom Blankinship No, there is no need to bleed your coolant system unless it has picked up air.  On rare occasions we'll get a stubborn bubble in a new system and it will purge out during a drive, in which case you simply top off your rad and/or overflow.  Once that is done there is no need to mess with it.

@chines1 I do believe your tips apply to all watercooled speedsters due the basic layout of the car.  While I have two radiators the same issue occurs, the length of the pipes and the location of the rads.

Have you found that the auxiliary pump oozing at all?  I have seen some oozing sometimes at the O-Ring seam (pump is made in two parts) and was wondering if it could allow air to be sucked into the system.  I have been using Evans coolant as a precaution and it may be the reason for the oozing it seems a bit slipery compared to normal prestone. 

I have a Mezière in line pump which seems to work well for boosting at idle. 

Ray

Last edited by IaM-Ray

Good for you!  I think your piping is straighter and less convoluted and thus less restrictive than my setup... 

Mine goes up from the back, up the left side then below the rocker panel to the front up the wheelwell and then down to the rads across the front to the next rad then back over the wheel well, passenger side, down the pass, right side rocker and back up to the engine.  The pump is in front between the rads. 

Just writing it is convoluted.  

Last edited by IaM-Ray

I'm learning !  Thank you !  Just a comment on the temp. warning buzzer. You need a temp. switch not a sending unit for a gauge. Temp switches can be ordered with a specific temp. range that they must reach before they come on. Usually they will start to tickle the buzzer as they approach the bottom end of it's temp. range. It's a good audible feature to slap you in the face to tell you to look at your temp gauge. I had both oil pressure and temp. audible warning in my Speedster.

I really like the AN fittings concept for the cooling system. I think Greg at VMC uses them..............Bruce

I just went through bleeding the coolant system on my Special Edition converted Fiberfab. I was changing out a bad thermostat and couldn't get the system to bleed or even start coolant flowing to and through the radiator. I realized the thermostat I removed had a 6mm bypass hole drilled in the lip by Special Edition. They have found that this helps get the system started and bled. I checked with Carey and he told me the hole wasn't mandatory and many don't use it but they found it makes it faster to bleed out the air. After a day of trying the bleed the system I broke down and removed the thermostat and drilled the hole. I got all the air out of the system in less than an hour. Lesson learned, pay attention to the part you remove compared to the new one going in. I also had a heater core to deal with that sets higher than the radiator. I installed a Prestone fill/flush valve on in line to the heater core. This allowed me to get the heater core filled with coolant. The entire process was a big learning experience.

aircooled posted:

I'm learning !  Thank you !  Just a comment on the temp. warning buzzer. You need a temp. switch not a sending unit for a gauge. Temp switches can be ordered with a specific temp. range that they must reach before they come on. Usually they will start to tickle the buzzer as they approach the bottom end of it's temp. range. It's a good audible feature to slap you in the face to tell you to look at your temp gauge. I had both oil pressure and temp. audible warning in my Speedster.

I really like the AN fittings concept for the cooling system. I think Greg at VMC uses them..............Bruce

Bruce, I too am learning.  Do you have some suggestions on doing this.  My goal is to see what the max temp should be and then have it activate the buzzer at that point.  It makes perfect sense to have this as a safety precaution in a watercooled car. 

Ray

Hi Ray...... The last temp. switch I used was set to close between 195F and 210F. It would start clicking the buzzer right at 195F and be fully closed at 200 so it was pretty accurate. I can't remember which company made it but it came from either Kaysor or Cadillac (not GM). They were used to trigger the radiator shutters on large 3 axle diesel truck/tractors.  The buzzer was one used in an air brake equipped truck for the low air pressure warning. I think it was made by Cole Heresy. However, I think any 12 v warning device would work. In my speedster I used electronic sirens from Radio Shack

I did have a shut-off switch under the dash. This was so I could listen to the engine while starting and in the event that I wanted the ignition on while I was doing something that required me to do so.  That shut-off switch was a double pole-double-throw. There other half of the switch was to turn on one of the extra warning lights in the instrument cluster. When the light was ON, it told me that the audible warning alarms were turned OFF. This way I was reminded to "activate" my audible alarm system. It was a nice theft alarm of sorts too ! If someone hot-wired the Ign sw. the low oil pressure siren would activate.    Simple but useful

Using "sirens" turned out to be a good choice . First, I can't hear so well. Second, it's hard to hear anything with the top down ay 70mph !...........Bruce

so

Jimmy V. posted:

I just went through bleeding the coolant system on my Special Edition converted Fiberfab. I was changing out a bad thermostat and couldn't get the system to bleed or even start coolant flowing to and through the radiator. I realized the thermostat I removed had a 6mm bypass hole drilled in the lip by Special Edition. They have found that this helps get the system started and bled. I checked with Carey and he told me the hole wasn't mandatory and many don't use it but they found it makes it faster to bleed out the air. After a day of trying the bleed the system I broke down and removed the thermostat and drilled the hole. I got all the air out of the system in less than an hour. Lesson learned, pay attention to the part you remove compared to the new one going in. I also had a heater core to deal with that sets higher than the radiator. I installed a Prestone fill/flush valve on in line to the heater core. This allowed me to get the heater core filled with coolant. The entire process was a big learning experience.

Jimmy...if you could do up a simple sketch of your cooling system showing the elevations of the various parts and post it here I think that would help a few viewers. 

Here's one that looks simple yet workable. Not mine. 

Attachments

Images (1)
  • Roadster Cooling diagram
Last edited by David Stroud IM Roadster D
David Stroud posted:
Jimmy V. posted:

I just went through bleeding the coolant system on my Special Edition converted Fiberfab. I was changing out a bad thermostat and couldn't get the system to bleed or even start coolant flowing to and through the radiator. I realized the thermostat I removed had a 6mm bypass hole drilled in the lip by Special Edition. They have found that this helps get the system started and bled. I checked with Carey and he told me the hole wasn't mandatory and many don't use it but they found it makes it faster to bleed out the air. After a day of trying the bleed the system I broke down and removed the thermostat and drilled the hole. I got all the air out of the system in less than an hour. Lesson learned, pay attention to the part you remove compared to the new one going in. I also had a heater core to deal with that sets higher than the radiator. I installed a Prestone fill/flush valve on in line to the heater core. This allowed me to get the heater core filled with coolant. The entire process was a big learning experience.

Jimmy...if you could do up a simple sketch of your cooling system showing the elevations of the various parts and post it here I think that would help a few viewers. 

Here's one that looks simple yet workable. Not mine. 

Yes, they are both helpful.  The photo also.

Todd M posted:
David Stroud posted:
Jimmy V. posted:

I just went through bleeding the coolant system on my Special Edition converted Fiberfab. I was changing out a bad thermostat and couldn't get the system to bleed or even start coolant flowing to and through the radiator. I realized the thermostat I removed had a 6mm bypass hole drilled in the lip by Special Edition. They have found that this helps get the system started and bled. I checked with Carey and he told me the hole wasn't mandatory and many don't use it but they found it makes it faster to bleed out the air. After a day of trying the bleed the system I broke down and removed the thermostat and drilled the hole. I got all the air out of the system in less than an hour. Lesson learned, pay attention to the part you remove compared to the new one going in. I also had a heater core to deal with that sets higher than the radiator. I installed a Prestone fill/flush valve on in line to the heater core. This allowed me to get the heater core filled with coolant. The entire process was a big learning experience.

Jimmy...if you could do up a simple sketch of your cooling system showing the elevations of the various parts and post it here I think that would help a few viewers. 

Here's one that looks simple yet workable. Not mine. 

Yes, they are both helpful.  The photo also.

Here's another system that will work. Proven in service with many miles and the rad is the "almost horizontal" configuration. 

Attachments

Images (1)
  • Roadster new cooling diagram proposed Feb 2019.

I just had a second tank done for my car cooling my tech at LP says that the tanks should be at the same level so that the fluid can overflow into the second tank and be sucked back in when the car starts up again and the negative pressure starts up. 

They made it out of stainless steel and all shined up.  Location:  It was placed right beside the overflow tank at the same height. 

Help with Valve......So to restart this thread I had my cooling system bled and I changed one of these valves on the top of my aluminum it seems pipes that takes the coolant to the front and back.   We have been using a wrench to get at these and it is not easy.... and was wondering if anyone of you guys had seen a brass fitting that might be able to have a bolt head rather than a pepcock so I could use a racheting tools to get there. 

 

 

Attachments

Images (1)
  • mceclip0

This I can help with. There are myriad air vent types typically used in hot water heating systems, including automatic vents (Maid-o-Mist is a inexpensive and common brand). Otherwise there’s something called a coin air-vent, sold for about $2 at every HVAC wholesale house in North America.

They’re typically 1/8 MPT, but a bushing gets you to whatever you might need. 

aircooled posted:

Stan....What does this valve look like ?............Bruce

 I'll put up a variety of options in 3 posts.

Maid-o-Mist #7 Auto-Vent

Maid o Mist

IaM-Ray posted:

I take it that the valve is good for so many pounds of pressure does anyone know off hand how much pressure is in a sealed coolant system>

The linked auto-vent is good up to 75 psi. Your radiator system runs at 15 psi or less.

Attachments

Images (1)
  • Maid o Mist
Last edited by Stan Galat

A totally manual version of radiator vent is what tradesmen generally call a "coin vent". Plastic containers full of these (and the keys) sit on the counters at HVAC parts houses starting in about September, through about now.Coin Vent

The thing about a "coin vent" is that it can have a knurled top, so that you can open and bleed it with your fingers or a small pliers. They come in many flavors, but here is one example:

knurled coin vent

I'm not linking you to anything, because these are available everywhere. If you want something specific, a "Coin Vent" or "Radiator Vent" or "Radiator Auto-Vent" search will turn up a lot of options. The coin vents typically cost less than 5 bucks each.

For the automatic stuff, I'd stick with a name brand- Maid-o-Mist is big here, but Bell & Gossett, Hoffmann, etc. are all safe bets. 

Attachments

Images (2)
  • Coin Vent
  • knurled coin vent
IaM-Ray posted:

Those are usually installed at the high point.  Is there a high point in our system?  Maybe near the engine but the flow of the coolant at that area might affect it.  What say you kind Sir,  Stan ?

These would be completely unaffected by flow, as a hyronic heating system has at least as much flow as your engine.

A vent is always in the high point of a system, or at a localized high-point where air can be trapped. It's a good idea to put a vent at every one of these points in a hydronic heating system, and I'd assume if the cooling system on a Subaru is convoluted with any traps, vents in those places would be a good idea. If the heads are where coolant is trapped, I wonder about drilling and tapping the heads for coin vents. 1/8" NPT is pretty stinking little.

Stan Galat posted:
IaM-Ray posted:

Those are usually installed at the high point.  Is there a high point in our system?  Maybe near the engine but the flow of the coolant at that area might affect it.  What say you kind Sir,  Stan ?

These would be completely unaffected by flow, as a hyronic heating system has at least as much flow as your engine.

A vent is always in the high point of a system, or at a localized high-point where air can be trapped. It's a good idea to put a vent at every one of these points in a hydronic heating system, and I'd assume if the cooling system on a Subaru is convoluted with any traps, vents in those places would be a good idea. If the heads are where coolant is trapped, I wonder about drilling and tapping the heads for coin vents. 1/8" NPT is pretty stinking little.

The heads are not where the air gets trapped in a Soob conversion....in my case at least ...it was always in the hoses to the rad and back or whatever. Or in my case one high spot was the top of the coolant manifold. Other builders won't have this high spot I mention as they have a filler neck with a rad cap on top of it welded onto the top of the coolant manifold. 

I did look as much as I could at a water cooled IM at Carlilse years ago but you cannot see much. The coolant hoses they use are Marine style wire reinforced rubber coolant hoses, very heavy duty and they go in a somewhat convoluted manner from back to front and back again, up and down several times. 

Because I'm not smart enough to emulate that system I just dumbed my system down to something I was capable of dealing with. 

Those are nice little bleeders Stan showed. Not knowing any better, I used several like the ones Ray shows which are just basic radiator drain valves. 

Hope you can find the high spots in the hoses, Ray. You'll need to find every high spot, not just the highest. 

Last edited by David Stroud IM Roadster D

Your right David the hose is a monster wire reinforced marine hose and the pipes going to the front are absolute artwork.  If we find ourselves in Carlisle and were not sitting in the rain, here's hoping we have some sun, I can show you the IM plumbing... The single rad set up is fairly straight forward the dual is a bit more challenging. 

Stan's valves merits a look for sure.

IaM-Ray posted:

@Stan Galat FWIW, I have 4 high spots on the metal tubes.  One per wheelwell that has that same valve that I showed you but I am not sure of the NPT off hand.  Not sure how high the valve would be in the wheelwell, but I can try to measure how much space I have and see if the valve could be accomodated. 

I just found a spare I had for the Speedster and I'm surprised to see it's 1/4" npt. Install depth requirement is 1".  CTC part 61104, , Ray.

After Carlisle I will look into it more fully thanks for all the help you guys.  

BTW I am running Evans coolant which has no water and is a bit heavier than normal coolant.  Have you in your HVAC experience used that stuff in doing radiant heat Stan?   It is a mixture of mostly propylene glycol and some ethylene glycol.  

I did it for the -40 to 350 protection and it being a lifetime coolant. 

David thanks for the info on the part too. 

Last edited by IaM-Ray

I'm familar with Evans coolant, and I'd probably run it in a specialty car-- but almost all hydronic systems use water with an anit-corrosion additive. Some in-floor systems will use a polypropylene glycol/water mix for freeze protection, but there is an efficiency loss with it.

It's a valid point to consider, but with a disc type auto-vent, I doubt it'd make a difference.

There are lots of positive elements when considering the switch to Evans coolant, but a few negatives as well.  The overview is that no substance is as efficient as plain water in heat transfer, with a heat transfer coefficient of 1.0. By contrast, Evans comes in at between .6 and .7, depending on the testing source.  We can get water almost anywhere, but a prudent driver would keep at least a small container of Evans with him in the vehicle in case of a cooling system leak.

Admittedly, most of us don't or can't use plain water, since the climates in which we use the car mandate some kind of protection against freezing.  Additionally, our water pumps, radiators, and heat exchangers need lubrication and rust prevention for longevity, so some kind of additive is a good idea.  Adding antifreeze, either ethylene or propylene glycol, in recommended percentages, also decreases the heat transfer coefficient of plain water, not quite as much as Evans, but the decrease can be substantial.

I had a long-term problem with overheating, as many on here will remember.  What that usually means in a Subaru engine is that, when coolant temps rise, small steam pockets form near the cylinder liners due to either insufficient heat transfer at the radiator, insufficient water flow through the system, or a combination of both.  Those small spots of boiling coolant can eventually lead to head gasket failure or worse, cracked heads. 

The use of Evans wouldn't fix the problem for me, it would just mask the symptoms, by not allowing the coolant to boil until it reached 375 degrees F.  I wanted to know why I had those hot spots and prevent them from forming.  I finally figured out my cooling problems,  and presently use distilled water with an additive that lubes the water pump and helps break down surface tension, to improve flow.

I'm not suggesting that Ray's use of Evans is ill-advised by any means.  It's working for him and that's what's important.  I went a different route and my route is working for me (knock on wood, mate).  The more we all know, the better choices we can make.

Jim Kelly posted:

There are lots of positive elements when considering the switch to Evans coolant, but a few negatives as well.  .  I went a different route and my route is working for me (knock on wood, mate).  The more we all know, the better choices we can make.

Exactly . If you could spell out that different route Jim, I think that would be a very nice contribution to the  list and wannabe liquid converters.  Thanks. 

FYI I had a Vanagon Westfalia. Watercooled boxer in back with LONG lines to a front-mounted radiator. The bleeding procedure involved opening a small bypass hose in the back by the header tank, raising the front 12-14 inches, then holding revs around 3000 and bleeding the radiator in front with a small petcock. You were done when coolant pissed out with no more bubbles.

I used one of the spill less funnels when purging the air out of my Special Edition Built system along with having the front of the car raised. The cool spill less funnel (bucket) thing made all the difference keeping the mess that comes with trapped air pushing large amounts of coolant up and out under control. The guy that designed this thing has my appreciation. If you haven't seen one of these, it is basically a bucket that screws on to the top of your radiator fill.

Spill less funnel

Attachments

Images (1)
  • Spill less funnel

David,

In my case, it was a series of problems, which started with a poor design by Steve Lawing of SAS.  His aerodynamic chart and calculations for the cooling system looked like something you would expect from a bright high school freshman at the local science fair, not from an experienced builder/engineer/savant, etc.

The radiator was in a bad place originally and couldn't get enough air through it to make it work.  Lawing had riveted a 3" x 5" bent piece of aluminum to a frame member to deflect air to  the rad.  The stock 2008 Legacy rad was completely open with 2 puller fans and no shroud.

I reconfigured the entire cooling system on my mid-engine application.  I moved and redesigned the radiator, fan, and shroud which now sits just underneath the traditional deck lid on our replicas, with twin puller fans and a shroud that forces all the air through the radiator.

I changed the intercooler from air-air to air-water, moved it to the frunk with its own pump and twin heat exchangers.  Not coincidentally, a tuner discovered a clever bit of sabotage.  Someone had placed a brass nipple reducer in a water line: 1/4" npt to 1/8" barb fitting, with just enough room in the hose for some water to bypass.  I took someone's name in vain when I found out about that.

I try to look back on that time as a learning exercise, like making lemonade out of lemons, eh?  I did learn a lot about air and water movement, radiator and fan design, etc.  If that's as bad as my troubles get, I'm a lucky guy.

Last edited by Jim Kelly
Jim Kelly posted:

There are lots of positive elements when considering the switch to Evans coolant, but a few negatives as well.  The overview is that no substance is as efficient as plain water in heat transfer, with a heat transfer coefficient of 1.0. By contrast, Evans comes in at between .6 and .7, depending on the testing source.  We can get water almost anywhere, but a prudent driver would keep at least a small container of Evans with him in the vehicle in case of a cooling system leak.

Admittedly, most of us don't or can't use plain water, since the climates in which we use the car mandate some kind of protection against freezing.  Additionally, our water pumps, radiators, and heat exchangers need lubrication and rust prevention for longevity, so some kind of additive is a good idea.  Adding antifreeze, either ethylene or propylene glycol, in recommended percentages, also decreases the heat transfer coefficient of plain water, not quite as much as Evans, but the decrease can be substantial.

 

Bingo.

I did so much research and work on my overheating that I have forgotten some of the process.  I also added an oil cooler and fan, which REALLY helped lower coolant temps.  Sometimes you solve the problem in small increments.  By that, I mean you force all the air through the rad and that lowers coolant temp by 3-4 degrees.  An aluminum dual-pass rad gains 2 degrees.  You install larger fan motors and that drops temp by 2 degrees.  You add intercoolers, another 5 degrees gained, oil cooler gains 6 degrees.

Incremental gains all add up.  For most of the year, ambient air temps may not concern us, depending on our location.  Summer temps in Santa Cruz, CA differ greatly than temps in Prescott, AZ or Redding, CA.  I want to be able to drive in any location at any time of year, so I was OCD about a fix that was bullet-proof.  For my application, the car ran cool until air temps reached the mid-80's, then it would overheat.  I now have that fixed, touch wood.

Vacant real estate is scarce on our cars, so placement of cooling system parts requires experimentation.  Every cooling component manufacturer states that HIS part, whether rad, condenser, heat exchanger, fan, etc.  should be in the maximum stream of cool air.  Easy in theory, hard in execution.

Yes, the EJ253 has a built-in oil cooler that mounts between the block and the oil filter.  The problem is that oil temp is changed by the engine coolant only, not moving air.  That means the cooler actually functions better as an oil heater when the engine is warming up, than a cooler when the t-stat is open and coolant is flowing through the whole system.

An external filter with electric fan and oil t-stat has proven much more effective in keeping oil, and thus coolant, temps down.  Proof is that many of the EJ253 engine owners have removed the stock oil cooler and replaced them with external coolers with fans. 

In addition, a car with marginal overheating will operate at higher cooling temps with the internal cooler, since oil temps are normally 20-30 degrees F more than the coolant temp.  Result: instead of cooling oil, you are only raising coolant temps.  Problem eliminated with quality, external oil cooler.  I installed a Setrab 1-series, 19 row with fan and shroud, the largest I had room for.

Post Content
×
×
×
×
Link copied to your clipboard.
×
×