Alternative bleed method?

[Simon Jones]

I've got a cunning plan & hope to try it out shortly when I fit my new radiator. Two of the big problems with the bleed process are:

* Sticking the bung back in the bleed pipe without getting scalded
* Being able to bleed 1 - 2 litres of water from the pipe without letting the level drop in the expansion tank

This is my proposed solution:

1) Fit an on/off valve in the end in the bleed pipe. I have several John Guest push-fit valves & fittings as used in for plastic water pipe (central heating, beer pump lines etc). If this is fitted securely then there is no reason why it can't stay fitted all the time.
2) A length of pipe goes to the bottom of a plastic 'sight glass' which allows you to see the coolant passing thru & more importantly see when the bubbles have settled down.
3) From the top of this is a pipe going to a 5 litre plastic tank which has two pumps in the bottom. This is just a windscreen washer bottle & 12v pumps from a scrapyard. I have choosen a type with 2 pumps to give a decent flow
4) The output from the pumps goes back up to the expansion tank.
5) The pumps are controlled by a simple circuit which interfaces to the low coolant alarm.

So, open the valve & let the coolant start to flow. The bubbles will be visible in the sight glass & because the water goes from bottom to top, it effectively creates a U-bend which will prevent air being sucked back in. When the level in the expansion tank drops, the alarm will come on & the pumps will replenish the coolant. In theory, as long the rate at which the pumps refil the tank at least matches that which the coolant leaves the bleed pipe, then you could leave it running indefinately until all the air is purged.

What do you reckon - constructive comments welcomed (no big arguments please ). Any suggestions on why this may not work or ideas for improvement?

[teenmal]

Hi Simon,would this help.

Lisle LIS24610 Spill Free Radiator Funnel Kit w/Extra Adapter
Product #: LIS24610

--------------------------------------------------------------------------------

Mal..

[mikeonb4c]

I like your style Mr Jones. Definitely on the right track if you ask me. Main issue will I think (and as you have pointed out) be getting the pumps in the bucket to raise hot coolant the required height at the required rate in order to resupply the header tank before it drains fully.

I wonder if the system could be controlled by a ballcock type switch in the bucket so that pumping takes place as soon as level is made to rise by coolant entering from the bleed pipe. This would get replenishment started as soon as depletion occurred in the header tank rather than only when it had sunk to the low coolant alarm level.

[TonyBee]

Simon,
How about utilising a pump from a computer watercooling system?
I have one that is surplus to requirements if you would like to experiment.
It is 12v and can pump 1200 l/hr and has a max head of 3m. It has 12mm tube connections on it and can operate up to 60 deg C water temp.
Here it is: http://www.swiftnets.com/products/mcp650.asp

Tony

[Simon Jones]

Teenmal, that funnel-type arrangement looks quite interesting: http://www.denlorstools.com/home/dt1/pa ... _wext.html

TonyBee, the pump looks a bit more meaty than the ones I have. It might be worth a try. I think the problem is finding any pump that will work with water that is near boiling point. I'll try putting some boiling water in the washer bottle to see how it runs & what the flow is like.

[waycar8]

Teenmal, that funnel-type arrangement looks quite interesting: http://www.denlorstools.com/home/dt1/pa ... _wext.html

TonyBee, the pump looks a bit more meaty than the ones I have. It might be worth a try. I think the problem is finding any pump that will work with water that is near boiling point. I'll try putting some boiling water in the washer bottle to see how it runs & what the flow is like.

I sugested something similar a while ago in regards to pumping the coolant back into the header tank when bleeding, but by using a pump from a washing machine, maybe worth a try but the 240 volts may be a bit dangerouse around boiling coolant , To be honest i think the windscreen washing pump seals will go under the heat of the coolant as they havent been designed to run warm water let alone hot coolant.

[teenmal]

Teenmal, that funnel-type arrangement looks quite interesting: http://www.denlorstools.com/home/dt1/pa ... _wext.html

TonyBee, the pump looks a bit more meaty than the ones I have. It might be worth a try. I think the problem is finding any pump that will work with water that is near boiling point. I'll try putting some boiling water in the washer bottle to see how it runs & what the flow is like.

I sugested something similar a while ago in regards to pumping the coolant back into the header tank when bleeding, but by using a pump from a washing machine, maybe worth a try but the 240 volts may be a bit dangerouse around boiling coolant , To be honest i think the windscreen washing pump seals will go under the heat of the coolant as they havent been designed to run warm water let alone hot coolant.

Or a old central heating pump.

Mal..

[waycar8]

why are you wanting to connect it to the low coolant alarm?, if your doing it just to bleed the system, whats point?, once the alarm goes off to start with then it will start the pumps? then coolant will be pumped from the bucket into the header tank?, so then wont the coolant alarm switch off as coolant will be passing the sensor in the header tank? thus the pumps switching back off?, then they will kick back in then back off and so on, so wouldnt it be easier just to start them manually then pull the plug in the bleed hose and let it flow?

less hassle and not much chance of leccy problems with it switching on and off constantly, thus reducing any other problems, such as fuses blowing or motors burning out or the coolant alarm going tits up as well?

Just a thought, Wayne

[TonyBee]

Simon,
Just let me know if you would like me to post it to you and pm me your address.

Tony

[Simon Jones]

The pumps will be controlled by a light activated switch which will be pointed at the low coolant alarm warning LED, so no electrical connection to the alarm. Good point about the coolant flowing into the tank shorting out the low coolant alarm, but if the pipe is put fully into the expansion tank, then it shouldn't be an issue. If I could find a pump that is happy to run 'dry' then it could run all the time & the level should be maintained at the correct level a it will be pumped back in as soon as it comes out the bleed pipe.

The other thing that may work would be to extend the bleed pipe up to the expansion tank, so it would rely purely on the force of the water coming out. If this works, then it solves all the complicated stuff. Anyone tried this approach?

[teenmal]

The pumps will be controlled by a light activated switch which will be pointed at the low coolant alarm warning LED, so no electrical connection to the alarm. Good point about the coolant flowing into the tank shorting out the low coolant alarm, but if the pipe is put fully into the expansion tank, then it shouldn't be an issue. If I could find a pump that is happy to run 'dry' then it could run all the time & the level should be maintained at the correct level a it will be pumped back in as soon as it comes out the bleed pipe.

The other thing that may work would be to extend the bleed pipe up to the expansion tank, so it would rely purely on the force of the water coming out. If this works, then it solves all the complicated stuff. Anyone tried this approach?

Aye,we are forgetting that the engine has a built-in pump

Mal..

[bigdaddycain]

I've got a cunning plan & hope to try it out shortly when I fit my new radiator. Two of the big problems with the bleed process are:

* Sticking the bung back in the bleed pipe without getting scalded
* Being able to bleed 1 - 2 litres of water from the pipe without letting the level drop in the expansion tank

This is my proposed solution:

1) Fit an on/off valve in the end in the bleed pipe. I have several John Guest push-fit valves & fittings as used in for plastic water pipe (central heating, beer pump lines etc). If this is fitted securely then there is no reason why it can't stay fitted all the time.
2) A length of pipe goes to the bottom of a plastic 'sight glass' which allows you to see the coolant passing thru & more importantly see when the bubbles have settled down.
3) From the top of this is a pipe going to a 5 litre plastic tank which has two pumps in the bottom. This is just a windscreen washer bottle & 12v pumps from a scrapyard. I have choosen a type with 2 pumps to give a decent flow
4) The output from the pumps goes back up to the expansion tank.
5) The pumps are controlled by a simple circuit which interfaces to the low coolant alarm.

So, open the valve & let the coolant start to flow. The bubbles will be visible in the sight glass & because the water goes from bottom to top, it effectively creates a U-bend which will prevent air being sucked back in. When the level in the expansion tank drops, the alarm will come on & the pumps will replenish the coolant. In theory, as long the rate at which the pumps refil the tank at least matches that which the coolant leaves the bleed pipe, then you could leave it running indefinately until all the air is purged.

What do you reckon - constructive comments welcomed (no big arguments please ). Any suggestions on why this may not work or ideas for improvement?

Sounds good simon... A "suck it and see" session is imminent methinks...

[Trouble at t'Mill]

A light activated switch? My nameshh Joneshh - Shhimon Jonshh...

Sounds like it's getting a tad overcomplicated to me

However, your other suggestion sounds like a paragon of sensibility. Would it work? Could it really be this simple?!!

Drain the whole system, flush it out if required, refill using a 50:50 mix, then extend the bleeding hose out the car door and sit it pointing into the expansion tank opening. Have additional mix ready to pour from a jug with a narrow spout to pour past the extended bleed hose into the exp tank.

Run engine to fully warm, rev engine at intervals and observe what's coming out of bleed hose into the exp. tank - ejected coolant will find it's way back into the system, air will automatically escape. Keep an eye on the water level and don't allow it to fall below the exp tank hose levels hose level - keep it topped up using the jug. If it does drop low enough to allow further air into the system, then keep bleeding until the air has gone full circle and comes out again!

Could this really work? You may need additional hands - one to control the bleed tube bung in case coolant is being expelled too quickly, and another to watch the temp gauge to provide feedback and warn of 'too hot'!!

Perhaps a funnel could be popped into the expansion tank for the hose to point in to, and this would also make it easier to add more coolant - unless the coolant being ejected might splutter and splash too much?

If this idea works, Simon, you'd be elevated to, to, er, ..what's higher than a Supreme Being?

[Simon Jones]

.

The idea was to devise a one-man operation solution. It's not that I dont trust my other half with the business end of the coolant system, but I'd like to see if it can be done. The other goal to is to find a solution to the question of when the system is fully bled. If we can find a method which doesn't require continual intervention, then it could be left for an hour or so to just sort itself it.

If all goes to plan, I'll have a go at the 'long pipe' method this weekend

[francophile1947]

If the bleed hose goes up to the level of the "expansion" tank, will the coolant still come out of it? - it will be the highest point in the system