Yes, another cooling system issue

[FFRWRX]

I filled my cooling system from the front (the rad) and the rear (the coolant tank). I ran the car until the gauge cluster indicated normal temp. The side pipes and rad were cold, so obviously the coolant was not flowing. I let it cool down and it didn't suck anything from the overflow bottle. I've read most of the threads on issues people have had with getting all the air out of the cooling system, so here is what I did.

Removed the thermostat and added a few holes in it to let the coolant flow through it somewhat.

I got out the vacuum coolant filling gizmo:

149187

I couldn't pull a vacuum with it. I went to the back and managed to hook up my brake bleeder to pressurize the system with air into the coolant tank. It held at 10 psi for at least 5 minutes. The rad hose at the front was obviously pressurized.

149188

So, no leaks then.

I went to the front and tried again with the vacuum thing in the rad (and the hose at the rear coolant tank that I used to pressurize the system put back on) and still can't pull a vacuum. It will go to about 5 on the gauge and no higher; supposed to go to around 26 which it usually does when I use it. As soon as I release the air from it, it immediately goes to zero.

I did that at the rad since the rear coolant bottle has coolant in it and trying to pull a vacuum with coolant in the system makes a horrible mess as it shoots out the outlet of the vacuum gizmo.

So this seems very strange. Doesn't seem like the system has any leaks, but can't pull a vacuum. Just for fun I googled my situation and found people on other car forums asking about using a vacuum coolant filling device. It's funny the reactions, which are mostly "why would you try to pull a vacuum on a system that is supposed to have pressure, it will never work". Maybe there is something to that if there is a seal (water pump) that will hold pressure but not a vacuum?

I'm reluctant to fill the system again since if I have to take anything apart again I'll get even more coolant on the floor, my arms, etc.

Any suggestions would be welcome.

Rick

[Ajzride]

This is normal. You can't pull a full vacuum until all the coolant is out. If you try to pull a vacuum with coolant in the system, the coolant boils at ambient temps and the vacuum will hold steady there at the coolant vapor pressure until it all boils off, then the system will continue on with the vacuum until it reaches 26.

What modifications do you have to try and let air escape?

[FFRWRX]

I haven't done any mods for air escape. I suppose I should do the Wayne mod and T it into the line that right now goes from the coolant tank at the rear to the rad (with a non-pressure release cap on it now). Some people seem to have got the air out without any mods and I was hoping to be one of those people. :)

I know what you are saying about pulling a vacuum with coolant still in the system, but I think it still should have worked. I don't think the vacuum gizmo is pulling enough of a vacuum to boil water out of the system. It should have been able to pull a vacuum of something like 10-20, no? But you could be right since something strange is happening.

[Ajzride]

I would expect it to behave exactly as you described with water in the system. If it isn’t get enough vacuum to boil it then it is at least getting enough to start sucking it out slowly. When I vacuum filled my system I left a shop vac hooked up to it sucking it dry for about an hour, then put a heat gun in the neck of the radiator for another hour.

[aquillen]

I think water boils around 160F at 20 inches vacuum, unless I just read an engineering chart incorrectly. So it would vaporize rather slowly at ambient, and that wouldn't explain the process not working for you. If you have a loop in the system with air in it and water on both sides of the loop, you have to have enough vacuum, and as well the fluid must also be able to move in such a way that the water "lifts" out of the way and lets the air pass. Such a loop can typically be in the radiator outlet humped up over the steering rack and back down - a great place for the air trap. Complicated way of saying you may have to lift the front of the car up and also add a fast squeeze-release on the rubber hoses while you do the vacuum tricks to help gurgle-work that air to where it can get out of the loop and then out of the system. (the latter process seemed to be my solution).

Of course this only thinks about the front of the car.

[SnailFoot]

Shouldn't the vacuum at the radiator try to suck the coolant into the system from the coolant tank back at the engine? or has that been sealed off?

[FFRWRX]

I did a small experiment. I modified the top of a mason jar so I could seal the vacuum coolant bleeder onto it. Put some water in the bottom of the jar and pulled about a 22" vacuum easily. Nothing happens to the water. Poor picture, but...

149221

With a very strong vacuum the water would boil out, but that would take about 29" of vacuum and you could only get that with a proper vacuum pump. If you want to have some fun, make a setup like this and put a marshmallow in the jar and then pull a vacuum.

If I used the vacuum bleeder at the rad with the coolant tank cap off, it would try and move some coolant to the rad, but I think air pockets would still prevent that from happening. With the system sealed up, it must be pulling in air from somewhere as it should have pulled a decent vacuum with the system going for almost a minute and not going above 5" of vacuum. But it holds pressure fine, so it is like a 1-way leak.

Many people seem to say the Wayne mod helped them, so I don't see any reason not to do that; other than what I have to remove to get access to drill and tap.

I could raise the front of the car 1' or so and see if that helps with the air pockets.

Thank you very much for helping me with this.

Rick

[J R Jones]

Rick, I have been watching this drama and thought "this is not as tough as it looks". But, I have not bled my 818, it is still under construction.
I am concerned enough that I may fabricate features to make bleeding more efficient.
As I see it we would be best off filling coolant from the bottom and bleeding air out the top(s).
I will install a fill port where convenient at a low point in a radiator tube. Fill with a funnel/stand pipe.
I will install air bleeds at the high points, like the radiator if there is no cap.
I have a Toyota MR2 SW20 and that bleed has gone routinely, no drama. Note:
https://www.youtube.com/watch?v=BNUSxZksFLA

The only odd part is squeezing the coolant hose near the thermostat to encourage the air pockets to push-out.
jim

[aquillen]

And - don't recall if you said above somewhere that you sealed the overflow reservoir connection back to where it hooks into the system. It is likely a one way valve so when system is cooling it can pull liquid back into the coolant system from the overflow reservoir. That same connection if not sealed would pull the fluid out and then air once all drawn out. Easy one to check.

[FFRWRX]

Rick, I have been watching this drama and thought "this is not as tough as it looks". But, I have not bled my 818, it is still under construction...

I would suggest that you do any mods you think you might need before the build is finished. I originally thought the same as you...how hard could it be to get air out of the system? I just did the Wayne mod (added a bleed fitting to the system) and it is not easy to do with the engine in place. Also, the pipe you tap into has a pretty thin wall and at best you are getting 2 threads. I used J-B Weld which is currently curing.

So it may be overkill to do what you are suggesting, but if you need it later, do it now.

[FFRWRX]

And - don't recall if you said above somewhere that you sealed the overflow reservoir connection back to where it hooks into the system. It is likely a one way valve so when system is cooling it can pull liquid back into the coolant system from the overflow reservoir. That same connection if not sealed would pull the fluid out and then air once all drawn out. Easy one to check.

Could it be that simple????? No, I didn't think to plug the connection from the coolant tank at the back to the overflow bottle. But you are right, a vacuum in the system should pull coolant/air from that bottle....continuously. I'll look into that when I go back out. Maybe it is not mentioned in instructions for using a vacuum bleeder on a cooling system since normally you would do it at the rad, which automatically takes that side feed to an overflow bottle out of the system. Interesting....thank you.

[fletch]

I also recommend a small bleeder hose from the top of the radiator to the degas tank (the one with the pressure relief cap). I used an 1/8” hose from the vent port of the radiator cap fitting and defeated the pressure relief function of the cap installed there. Similarly, I used another small diameter bleeder hose from the heater core back to the degas tank. I also perforated the outer ring of the thermostat as you did. These helped in getting the system to self bleed during several hot cold cycles. And finally, after each top-up during the filling process I used a pressure test kit to squeeze the coolant past some stuck air pockets and repeated squished the soft hose fittings to get the air to move. This approach has worked well at least 3 times for me.

It’s a pain. Just be sure to keep your eye on engine temp and have patience.

[FFRWRX]

I also recommend a small bleeder hose from the top of the radiator to the degas tank (the one with the pressure relief cap).

Yes, I have that. My radiator has a port in the filler neck that normally goes to an overflow tank. Since I have a non-pressure relief cap on it, that port is not hooked up. There is also a dedicated port on the top of the rad that goes to the degas tank at the rear.

[FFRWRX]

Problem solved. No big deal. Don't know what all the fuss is about with getting the air out of the system. :)

When Art suggested checking if the rear overflow tank was letting coolant/air in under vacuum, I thought to myself "but then I would see that the coolant level went down in the bottle". So I checked the bottle and there was no coolant in it (and I started off with it 3/4 full).

At that point I had already put in the Wayne mod. As I said, this is not particularly easy with the engine installed, but is doable. I didn't have a proper fitting, and getting them these days is a bit of a pain. We are still in lock-down in Ontario so you can't just go in a store and browse around. You order stuff online and then go and pick it up when the order is ready; usually a day or so. So what I had on hand were these:

149237

A fitting from air hoses, and a barbed one with the threaded part a little larger than I liked. So I cut off the "barb" from the air fitting, bored out the other one a bit, and soldered them together. I also machined off the hex and the threads from the larger one; didn't have to but machining brass is easy and fun.....makes chips that look like gold according to my wife.

149238

The problem then was that this was not a standard NPT fitting. It measured about 18 TPI and matched up pretty close to an M14x1.5 thread. So I drilled and tapped for that, and used J-B Weld as well since there isn't much thread in the aluminum part on the engine. As I said, not easy to get in there to drill and tap, but possible.

I T'd this into the line that goes into the degas tank from the top of the rad at the front.

Then I did remember to plug the hose in the overflow tank:

149239

Went back to the radiator and put in the vacuum bleed thing. Pulled a vacuum right away. I could only get it to just over 20", but no idea how accurate the gauge is. Held that for a bit and then opened the valve to let the coolant in and 4L, give or take, went right in.

I have not run the car yet (I'll do that tomorrow). I don't know that I've gotten all the air out, but assuming/hoping I got out enough to let the coolant circulate and fill up any air pockets with heat/cool cycles. Maybe I didn't need the Wayne mod after all, but no harm in having it. I have a Vintage Air heater unit which is not hooked up electrically so don't know if the heater valve was open or not. If it was closed then there will be some air in the heater core, but hopefully that will take care of itself when the valve is open and goes through some heat/cool cycles. Oh, and the holes in the thermostat should help as well to let coolant circulate before it opens.

149240

So in the end it turned out to be a simple fix.....assuming the coolant does circulate when I run it. A good lesson learned here when using a vacuum coolant fill system. I was thinking Zebras instead of horses. :)

I'll post again when I've run it.

Rick

[Ajzride]

You will most likely need to rev the engine up close to 1500 or 2000 rpms to get the heater core burped. Mine would not flow enough at idle. The stock loop is about 18” and the 818 loop is closer to 18 foot (round trip)

[J R Jones]

Years ago engines (Ford?) had a thermostat by-pass hose so there was limited system flow during warm-up. One concern was not getting a slug of cold thermal shock when the thermostat opened.
More recently I have seen (production) bleed holes in the thermostat for warm-up circulation. I suppose there are thermostats with no bleed. It was not much of a hole, less than 0.1 in.
Likely part of the intent was to improve filling and purging air through the bleed hole. You guys are scaring me.

[fletch]

You will most likely need to rev the engine up close to 1500 or 2000 rpms to get the heater core burped. Mine would not flow enough at idle. The stock loop is about 18” and the 818 loop is closer to 18 foot (round trip)
Never had an issue with flow but had to put the pedal down a bit to generate enough heat to open the thermostat for sure.

[RM1SepEx]

I also lift up the front of my car a couple feet on my 4 post lift. It takes 3 solid burp cycles to get all the air out. Mine's been on the road for over 10,000 miles over 5 years. I also have the cheater hose that everyone call's Wayne's mod. My car is serial # 17, I did that mod b4 Wayne published it. Mid engine cars are always a problem to bleed

[FFRWRX]

I just ran the car and it seems successful. Some observations:

The passenger side coolant pipe got warm within a minute, so I knew the coolant was circulating; previously it stayed cold. The top of the rad started getting warm and after a few minutes, just before the temperature gauge got to "normal", the rad fans kicked on...excellent! I haven't adjusted the temperature setting on them yet, so may bump it up just a bit. It is fairly warm out, 84F, so things did heat up fairly quickly. I took a video of the T fitting where the Wayne mod hose connects to the degas tank. The hose coming in from the right is the new air bleed one, the bottom hose comes from the top of the rad, and the top one goes to the tank. Not super exciting, but here for your dining and dancing pleasure:


https://www.youtube.com/watch?v=CspCgB8B1Zs

You can see it doing it's job. I'll replace the nylon T with a metal one when I get a chance as things get pretty hot and I don't really trust it.

After running for a few minutes the engine went up to around 2800 RPM and the throttle would not respond. This happened the other day too. Turning it off and back on solved it. I don't know if this is because I disconnect the battery each time I leave the car (since I'm still working on wiring and stuff) and maybe something to do with the ECU learning. Don't really think so, but hope so.

When the engine was heated up there was smoke coming from the exhaust. This happened the other day as well. No smoke until it was up to temp. Smelled a bit like coolant burning, but hard to tell since spilled coolant was burning off parts of the engine. Not a lot of smoke, but enough to concern me. Not blue oil smoke from what I could tell, more greyish.

So generally a success, but a couple things to watch out for.

Thanks again for all the help and suggestions.

Rick

[FFRWRX]

My car is serial # 17, I did that mod b4 Wayne published it.

So it's really the RM1SepEx mod? Doesn't have the same ring to it. :)

[Rob T]

I've been scared a few times with a burning smell after doing work on my car. In all cases (so far) it has been stuff that soaked into the wrap I have on the headers and exhaust. It takes a while to burn off, especially if you are just running it with no load.