Expert advice required re coolant measurement

I know you can get engine savers that measure low coolant levels and wonder if when you suffer a loss of coolant does the resultant drop in pressure draw coolant back from the expansion tank. If so there are some ultrasonic liquid level detection sensors that are quite cheap and could trigger an alarm if water level drops in the expansion tank. Supposed to be able to read through 10mm plastic. The other option would be to place the sensor against the top radiator hose however attachment might be an issue.
Anyone have ideas on this?

Sorry not an expert lol, but I can think of a couple of potential problems with using this on the expansion tank.

If there were a sudden loss of pressure in the cooling system the valve in radiator cap might close (as no pressure to keep it open), it might not draw in the coolant from this reservoir at all. At least initially, and by the time heat/pressure opens it from the engine overheating, it isn't going to draw anything in (usually steam pouring out). Also if there were a slow leakage say overnight, when you start up the next day the expansion tank could be full, but the cooling system might be low/empty and by the time you realise could be too late.

Only a couple of thoughts (could be wrong) but I think this would only be a reliable way to monitor coolant levels under normal operation, so you would be able to top up coolant levels when necessary.

Surely the most reliable way would be to monitor around the highest point of the cooling system proper. On the top radiator hose would be better, but as you say, could be problems attaching it. Great idea to monitor the coolant levels though, and vehicle manufacturers should make it standard fare. The old gauge on it's own is a bit useless.

I'm no expert either, but Horatius makes some good sense there.

Part of the problem with the coolant system is that to detect moisture levels usually requires some form of current to be passed through the liquid, detecting it. Modern coolants have anti-conductor additives to reduce damage to the cooling system (both the water jacket of the motor and the radiator) through corrosion - because corrosion is caused by oxidisation, which required the flow of electrons.

That doesn't mean it's impossible. The ultrasonic sensor might have some merit if you could rely on the system to always draw from the expansion tank first. If there's a coolant leak, though, that may not happen.

What would be more reliable is some means of measuring the height of water in the radiator itself. Imagine adding a 90 degree threaded elbow to the top and bottom tanks of the radiator and putting a clear hose between them - you have a sight glass. Use the same technique to put a float level inside - you solve the problem with non-conductive fluid and detection of level dropping.

Can that be done reliably? That's the key question. Has any manufacturer developed a method of determining whether or not the coolant level in the radiator is dropping?

Old.Tony said:

I'm no expert either, but Horatius makes some good sense there.

Part of the problem with the coolant system is that to detect moisture levels usually requires some form of current to be passed through the liquid, detecting it. Modern coolants have anti-conductor additives to reduce damage to the cooling system (both the water jacket of the motor and the radiator) through corrosion - because corrosion is caused by oxidisation, which required the flow of electrons.

That doesn't mean it's impossible. The ultrasonic sensor might have some merit if you could rely on the system to always draw from the expansion tank first. If there's a coolant leak, though, that may not happen.

What would be more reliable is some means of measuring the height of water in the radiator itself. Imagine adding a 90 degree threaded elbow to the top and bottom tanks of the radiator and putting a clear hose between them - you have a sight glass. Use the same technique to put a float level inside - you solve the problem with non-conductive fluid and detection of level dropping.

Can that be done reliably? That's the key question. Has any manufacturer developed a method of determining whether or not the coolant level in the radiator is dropping?

Click to expand...

Very simple and effective, and has been around for a long time. Surprised no one has done it yet!

Wonder how one of these would go in the top radiator tank (if there were somewhere to fit it of course)? Not sure exactly how they operate internally re introducing any currents or voltages that could cause corrosion, but they are advertised as an "ultrasonic" detection sensor. Somehow it senses (ultrasonically) the presence/absence of fluid in the small gap at the end of the sensor. They are even advertised as being ideal for use in "off road vehicles".

http://www.gemssensors.com/level/single-point-level-switches/ultrasonic-level-sensor

There's not a lot of distance from the top of the top tank to the top of the cores, and without actually seeing the sensor in the flesh it's hard to say if it'd fit in the upper space without manufacturing some sort of adapter - and hoping that the constructed component doesn't cop a beating from the bonnet, and allows enough space for the sensor to take a measurement.

I doubt that it would induce current in the fluid, so if you can overcome the physical hurdle, it looks the goods. Then all you need is a circuit to manage it - it would be enough to simply light a warning lamp. It might be perfect!

There are several types of sensors, some rely on conductance and I agree the electric current could interfere with the various metallic components leading to corrosion or electrolysis. There do seem to be some sensors that use ultrasound. Most of the industrial flow process ones are too big. Will keep looking and hopefully get something to try.
I agree that measurement in either the header tank or to radiator hose is the preferred position for early warning.
Thanks. I will repost when I find something more.

I can only assume you are after something similar to http://www.enginesaver.com.au/ but on a possibly cheaper scale. I haven't used one of those personally, but have seen them mentioned on the patrol forum before...

Sent from my SM-G900I using Tapatalk

Old.Tony said:

What would be more reliable is some means of measuring the height of water in the radiator itself. Imagine adding a 90 degree threaded elbow to the top and bottom tanks of the radiator and putting a clear hose between them - you have a sight glass. Use the same technique to put a float level inside - you solve the problem with non-conductive fluid and detection of level dropping.

Click to expand...

the trouble with sight glass is they degrade with heat and will often be the cause of the leak.

a sensor in the expansion tank will not work. not quick enough reaction. the engine will over heat before fluid level drops and in the case of blown head the expansion tank level goes up as the radiator level goes down.

much better to use a sensor in the tank/hose etc, many manufactures do have them as OEM.

Only feasible way to install the ultrasonic sensor seems to be in the top radiator hose. A straight through coupling out of stainless with appropriate size/thread nut welded on to accept the sensor would probably do it.

Similar idea to the systems already available that use a small voltage/current to sense coolant. For the money it probably wouldn't be worth making one, only for the problem of possible corrosion with these systems.

There is some talk that using a small alternating voltage won't cause hydrolysis. Not being a chemist, surely once you start knocking electrons out of their orbit there will be some corresponding chemical reaction of some type, whether it is ac or dc?

Here's a spiel from the "engine saver" manufacturer about it, who claims his 2v/45 microamp system has no corrosive effects.

http://www.enginesaver.com.au/EngineSaver_AC_DC_Page.html

Though I was once shown a radiator (out of a Prado from memory) that failed from corrosion within 2 yrs from brand new, the radiator bloke got his meter out and demonstrated the problem, explaining that voltage above a certain range (was in millivolts - can't remember the figure) was the cause. Would be good to get an idea from people who have used them for a length of time, to get their opinion on it.