907 Overheating

Overheating Diagnosis
"Tim Engel" <tengel@isd.net>

Mark Belchamber <markb@mirkwood.clara.net> wrote:
No blockage to heater matrix, no thermostat (temporary measure) and the whole thing's been flushed. All pipes are clear and I've back and forward flushed it all.

PUT THE THERMOSTAT BACK IN !! Regardless of whatever else may be wrong, the system will not work properly without the thermostat! Really! Running without the thermostat will just compound your witch hunt.

The only suspicious thing is the fact that I can't SEE the coolant flowing through the header tank. I can't feel the water flowing from the large bottom hose to the smaller hose going to the header tank, which splits to the larger pipe going to the thermostat via that metal angled bit. It's almost as if the water is static, NOT moving.
The header tank is plumbed into the cabin heater circuit. If the cabin heater is off, there will be no flow through the smaller hoses (3/4 inch ID) attached to the header tank. Flow in that circuit starts at the rear of the cylinder head on some vintages, on the rear of the intake manifold water passage on other vintages. Either way, water leaves the rear of the engine, flows through the heater valve and heater core (matrix). Then it returns via a copper pipe routed across the face of the firewall to the left side of the engine where it turns forward. A 3/4 inch hose attaches there and runs forward along side the engine, through the header tank and down to the lower radiator return hose.

If the heater valve is off, there should be no flow through the header tank. That's normal. If the valve is on and there is no flow, the heater core is probably plugged with sediment.

I've had the pump off and it seems O.K. and the water pump belt turns fine. Should I be able to see the water flowing with the header tank cap off?
See my comments above... only if the heater valve is open and the heater core is not plugged.

If I run the pump and block off all the outlets, should I feel it trying to push air?
Doesn't seem like a meaningful test, but yes... if you drain the system, disconnect the lower radiator hose, remove the thermostat and it's housing/cover and spin the pump fast enough, you should feel a flow of air entering the inlet (where the lower radiator hose attaches to the pump and exiting the open thermostat housing. That's a lot of work for something that won't tell you much about the general status of the complete cooling system.

I've taken the water pump off again and discovered that the impeller vanes are arced in a loose 'C' shape. The direction of the belt makes the vanes stroke the water (i.e. the back of the 'C' strikes the water.) Now I would have thought that to get maximum pumping effect, the vanes should cut into the water - i.e. the open face of the 'C' strikes first.
It's a centrifugal pump. The water enters the pump via the lower radiator hose and is routed internally to the center of the front side of the impeller. Centrifugal force (which really doesn't exist, but we'll avoid the physics for now... ) throws the water radially outward along the rotor vanes until it exits at the perimeter of the rotor.

Backwards curved blades are a way of slowing the water velocity and increasing pressure. Forwarded curved blades would increase velocity and decrease pressure. In a water pump, pressure is usually more important than velocity and the blades are curved backward. In "some" air fans, high velocity is the goal and the blades are curved forward.

It may (?) be possible to press the rotor onto the shaft backwards, but then it would not fit into the scroll housing. The rotor just will not fit into the housing backwards... without some serious creativity. When the pump is removed, you should see a flat, round face of the rotor's backside. Between it and the scroll housing you should see the ends of the vanes. The combination of the housing, the rotor back plate and the vane walls should create a series of radial water passages that exit all around the perimeter of the rotor. If you actually have a full length view of all the vanes staring back at you, there is something seriously wrong with how the pump was assembled... and you've probably found your problem. But I don't think the rotor can be assembled into the housing that way.

The water does not get pumped efficiently from lower rad. hose - water pump - thermostat housing with the t'stat removed even if I block off the little pipe going up to the header tank. However when the engine is turned off, the water all runs out of the bottom hose. This indicates to me that _some_ pumping is beng attempted, but it is inadequate.
With the thermostat out, IT WON'T !!

The normal circulation enters from the lower radiator hose to the center of the front of the impeller. Radially outward through the impeller into the scroll housing. The back half of the scroll housing is cast into the front face of the engine block. On the right side (carb side) of the block/scroll housing there is a rectangular hole that leads into the block's water jacket. Coolant passes through there, into the block, up into the cylinder head and

out into the intake manifold's water passage. A rubber hose connects that passage back to the thermostat housing.

When the engine is cold and the thermostat is closed, the coolant can't leave the engine and go to the radiator. Instead it is shunted back to the front of the impeller and recirculated.

The proper thermostat has a round disk added towards the bottom. When it's installed, the disk blocks off (partitions) the internal water discharge passage in the pump. Water leaving the impeller hits the bottom of that disk and is forced to turn 90 degrees and go through the rectangular opening that leads into the block. Water returning from the intake manifold hits the top side of the disk and a portion is diverted back to the front of the impeller and is recirculated.

With the THERMOSTAT REMOVED from the housing, there is nothing to force the water from the impeller to turn and flow through the rectangular hole and into the engine. It would rather just recirculate. Flowing water does not like to turn and it will take the path of least resistance. That does not include a 90 degree turn into the engine. When the thermostat finally opens, the water has a second alternative path through the radiator and back to the pump inlet. Most of the pump discharge water will either recirculate or go directly through the radiator. Relatively little will circulate through the engine. The engine is not sufficiently cooled and quickly over-heats.

That's why it's important to never run a 907/ 910/ 912/ etc without a thermostat installed. That little diverter disk is vital. SO PUT THE THERMOSTAT BACK IN!! Several people have made the point to you that the 907 needs a thermostat, yet you are still trying to run without one. If you are not willing to install a NEW, LOW TEMPERATURE thermostat, I strongly recommend you stop trying to diagnose the problem yourself and take the car to a garage that knows Lotus engines.

1) Does anyone know which way round the vanes should be?
Curved backwards relative to the direction of rotation... just as they are.

2) Is there a way of reversing the belt direction by the addition of a small pulley near the pump (everything runs clockwise at the moment and the pump needs to be going the other way, I think)
No, it's running the correct way as it is.

3) Has the P.O. put the wrong pump on (maybe Esprit vanes point a different way but the pump is similar in every other respect)
No, all 907's are created equal in that regard. It doesn't make any difference if the engine is installed in an Elite/Eclat, Esprit, Jensen or Sunbeam. The pumps all turn clockwise.

4) Can you get replacement impellers?
I believe they only come as part of a complete pump rebuild kit.

5) Am I completely misleading myself?
Yes

- has anyone come up against this and discovered it's an entirely different problem?
I would be very concerned that IF the problem was originally minor, it may now have mushroomed into a bigger problem. From the sound of it, the engine has now over-heated several times. The aluminum head is sensitive to over-heating and can warp or blow the head gasket fairly easily. In that case, one of the symptoms would be chronic over-heating. A mechanic could perform a pressure test that would indicate if the head gasket were blown. If the engine has seriously over-heated, the valve seats can loosen and drop out of the head (voice of experience).

I've just noticed that the hole _behind_ the water pump (water is directed to it via a snail shaped funnel) doesn't have water going anywhere. It seems to be a channel > that leads to nowhere.
Rectangular hole? That should be the passage into the cylinder block. That's the main route for all the cooling water into the engine. What do you mean it doesn't go anywhere? Is it blocked? If you look into it you should see the #1 cylinder liner.

Now I also notice to the right of the upper part of the w/pump on the cylinder head a blanked off hole. This looks suspiciously like a good point for a bottom radiator hose to go to if the blanking plug (if that is what it is) were removed. That way I would have water going from the pump, round the engine, to the lower hose.
NO

At the moment water could just be swirling round the pump and going up its own backside.
No, not because of the frost plug not being a connection... Yes, similar scenario, but because the thermostat is not installed...

Water flows FROM the lower radiator hose TO the pump. Not the other way around. The blanked off hole is just a frost plug. In the turbo engines and later 912's that use the uprated turbo water pump, the frost plug is replaced with a pressed in nipple. A 3/4 inch 90 degree molded hose then connects it to a second water inlet to the turbo pump. This provides another return flow path back to the water pump in addition to the one through the intake manifold water passage and cools the head better.

Don't start thinking about putting in the turbo pump to solve your car's problem. It won't help. If everything else was fine, the turbo pump would provide a little extra cooling for a safety margin on hot days. Your car's problem isn't just a safety margin. It's not working. Once you get to the root of the problem and fix it, upgrading to a turbo water pump would be a nice optional extra.

Mark, first put a NEW thermostat in the engine.

Then have a garage pressure test the engine to determine if:

  1. The head gasket is blown
  2. The head is warped
  3. The head is cracked.

    Make any repairs indicated.

    Then pull the radiator out and take it to a shop. Flushing cleans out the loose bits. However, the radiator could still be seriously plugged by hard scale. It could appear to flow coolant, but still have it's ability to transfer heat severely reduced. Have the tanks removed and the core inspected. If partially plugged, either have it rodded out or re-cored. Once you are in that deep, re-coring with a thicker core would be the best option.