Thermostat opens, engine gets hotter! What's going on?

[scudderfish]

I'm seeing odd behaviour in my datalogs from my Megasquirt. This is from my car in the garage tonight. Engine starts and runs and gets warmer over time. From the log, it looks like when the thermostat opens, the engine heats up rapidly.

Red line is RPM, engine idling circa 900rpm with the occasional blip. White is temperature, green is the fan switching on and off.

Why would the temperature suddenly increase like that? I'm running an Offenhauser 360 manifold. The front end is plumbed in like this


The bike valve flows coolant when I press it so I don't believe there is any air left in the system. The car is currently jacked up at the front to aid this.
The general overview looks like this

Close up of the expansion tank

And the back end of the manifold. I use the take off for the dash temp gauge sender


I can't think of the physics of what's going on to cause this. Have I mis-plumbed this? Or am I just worrying over normal behaviour?

Regards,
Dave

[Rossco]

It's only when the stat has opened that you're getting full flow past the sensor and therefore getting a true measurement of temp.

[DEVONMAN]

The thing that jumps out at me is the bottom hose. It appears to rise and then fall towards the water pump suction inlet,
Is it possible that there is potential for an air lock in this hose??

[DaveEFI]

Better (or worse) coolant flow over the sensor? Water is actually a fairly poor conductor of heat, so you can have quite a differential at different parts of the engine.

[DEVONMAN]

Better (or worse) coolant flow over the sensor? Water is actually a fairly poor conductor of heat, so you can have quite a differential at different parts of the engine.

The fact that you have the rear outlet of the inlet manifold closed is not helping the coolant flow. Coolant in the rear of the manifold is semi static.
The flow over the coolant temp sensor is from the offside cylinder head outlet and may not be getting a good flow until the thermostat opens.
There is a simply modification to help this problem as follows.
With the thermostat removed two 4mm holes are drilled into the water port which rise from the heads at the front of the manifold. This allows a better flow of coolant past the temp sensor and around the back of the thermostat.
There has been some details of this mod on this forum previously so have a search.

Regards Denis

[Stu E]

I had my water system piped with the hose from the thermostat bypass connected to the back of the water pump as in your pics. When driving at motorway speeds I found it was pumping cooled water directly onto the back of the stat making it close. The temp gauge was reading 120 but the fans wouldn't come on as the stat was shut and the fan sensor is in the top hose.

HTH Stu

[scudderfish]

I had my water system piped with the hose from the thermostat bypass connected to the back of the water pump as in your pics. When driving at motorway speeds I found it was pumping cooled water directly onto the back of the stat making it close. The temp gauge was reading 120 but the fans wouldn't come on as the stat was shut and the fan sensor is in the top hose.

HTH Stu

Ooo, that's something I'll need to look out for. My ECU controls the fan so the symptoms for me would be the fan running, but it having no effect. I wonder if I should connect the back of the water pump to the back of the manifold and block off the thermo bypass? This should avoid the problem you had and reduce stagnation at the back of the manifold.

[JP.]

Imo the back of the intake manifold is part of the by-pass cooling system ( when the stat is closed ). Normally its hooked up to the heater which is integrated in the by-pass system.
This system makes also sure the block heats up & cools parallel at the rear cylinders.

You can run without a heater but it still needs to be re-routed & connected back to the water pump ( the one you blocked off ).

Next the temp sender sits wrong. It should be on front on the intake near the thermostat for correct readings.

That coolant down pipe which makes a crazy U-turn over the shock tower could benefit also a bike valve.

[scudderfish]

Thanks for all the comments, informative and educational The picture of the return pipe is a little misleading, I can easily flex it lower than the inlet to the pump and flush any air out of it.

I have a blocked off input to the pump, so at first I thought I'd take the sensor off the back of the manifold and plumb in a line. I'd either need a restriction in it, or do what is suggested in this document http://www.v8register.net/FilesV8WN/V8N ... 9.1.08.pdf which is to connect it to the top hose.

[DEVONMAN]

If you run an 8mm bore pipe from the rear of the inlet manifold and tee it into the expansion bottle pipe at the top of the radiator then you will improve flow distribution in the inlet manifold prior to and after the thermostat opens. This should moderate the reaction of the ECU coolant sensor to the thermostat movement. Naturally, you will need to incorporate your dash gauge sensor.
In winter I run the 8mm pipe through the heater also and there is adequate flow to keep the heater working.

The two holes drilled in the manifold I mentioned earlier will further smooth things.

I use this method myself but run this 8mm pipe through my water cooler turbos. This gave a vast improvement over feeding hot coolant back into the engine via the normal bypass connections and tends to back the basis of the document you posted earlier.

Denis

[scudderfish]

Following up from this I now have a 8mm feed from the back of the manifold into the pump* :-



The waterpump is an SD1 one from the 3.5 I originally had in the car. Before the thermostat opens, the pipe from the radiator gets quite hot, the pipe to the radiator stays as cool as I expect, as if the pump is running backwards. Once the stat opens, the top pipe gets hot as expected.

The first time the stat opens, the fan kicks in and temperature drops nicely, however on subsequent cycles, the temp rises higher (I peaked at 101 yesterday) and takes longer to fall back to high 80s/low 90s. It all seems a bit odd to me.


* Re-reading the previous posts, should this not be going to the pump and instead run to the feed from the rad to the expansion bottle?

[DEVONMAN]

You could tee the 8mm pipe from the rear of the manifold into the pipe that runs between the expansion bottle and the rad. This will however slow down the warming up of the engine. It will reduce the amount of hot coolant returning uncooled into the block and lower the running temp.

With regards to the gradual increase in the running temp, Maybe the fan is not up to the job.

The water pump looks like a P5/P6 type to me and not an SD1 item.

With your current arrangement the bottom hose will get hot before the stat opens as hot coolant flowing in the 8mm pipe will rise into it given the hump in it. I think you need to loose the hump.

Regards Denis