MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Water pump question

The clearance between the impeller and the pump body on my waterpump (early type, three bearing engine) is about 1.7 mm. (0.066", I believe.)

Is this clearance excessive? Can it be the explanation why my MGB gets so hot in traffic, with a new and clean radiator?

I cannot find any data on this for the MGB pump, but the midget manual states this clearance should be very much smaller - .020 to .030.

Tore
Tore

I don't think that is excessive - the pump is not a positive displacement one and it really only assists the thermo convection effect.

Cooling while stationary is more a function of the engine fan and the radiator - alos the guages are far from accurate - what do you mean by overheating? The needle can rise well into the Hot sector without the water in the engine actually boiling.
Chris at Octarine Services

By overheating I mean that the needle is steadily approaching boiling point - 212F on my instrument - in 10 minutes or so after coming to a standstill. But I have always been able to find a way out of the jam before it passes 212F, and never had the nerve to find out if it will actually start boiling!
Tore

Whilst my mechanical fan roadster can show idle temperatures a bit higher than normal in very warm weather I wouldn't have thought that would be an issue in Norway in March. What year i.e. mechanical fan or electric? If mechanical (which I would assume since you imply gauge markings in degrees) it sounds more like a problem of the fan not pulling air through the radiator when stationary than a problem with the pump. However there are two types of pump for the different engines and whilst at least one incorrect combination (if not both) can be assembled it does result in less coolant flow than there should be, I understand.
Paul Hunt

64 MGB, original three-bearing 18G engine, mechanical three-bladed fan, and the original, first type of water pump with grease nipple. And it is not overheating in March - the problems i referred to was last summer. Still, this is not California - summer temperatures are usually quite modest.
Tore

Tore,

I would be inclined to test the guage by immersing the bulb in a pan of boiling water to check that 212 on the guage is boiling point.

Don't forget that the engine coolant won't biol at 212 becuase it is under pressure from the rad cap - boiling point is nearer 230 degrees F under 15 lbs pressure if I remember correctly....
Chris at Octarine Services

Might want to check the fan is spinning in the right way direction whilst you are at it?
Steve Postins

Tore

Definitely calibrate your temperature sensor. Is should be the same bulb/capillary tube type that is also used in the A. My A gauge reads about 20 deg F high when checked against an infra red temperature meter used to test it.

One other thing to check, you say the radiator is new and clean, but has the block been flushed out in any way?

HTH

Larry
58A
Larry Hallanger

Tore,
Have you checked the thermostat; 74°C/165°F Summer, 82°C/180°F Winter. Sounds as though it has a higher one fitted (88°C or 92°C maybe).
About 10°F rise on the guage is normal after switch off or in traffic as the thermostat works over a range of 20°F from starting to open to fully open and will respond to heatsoak from the engine when airflow slows or stops.
Rich.
Rich

Re impeller clearance. I have no idea what it should be but I just had to replace my water pump. I have always used the cast iron ones with a cast iron impeller (just an old fashioned engineer) and when I changed out the old pump, I was surprised to see that the impeller clearance on the old one was about 4 times the clearance on the new one (visual only). As the machined surface on the impeller on the old one was fine, the different clearance was definitely due to impeller positioning on the shaft.

A water pump is a simple centrifugal pump and the clearance between the impeller can affect flow capability. More clearance means more slip and less pumping efficiency. If the bearings are tight with no slop this clearance can be quite small.

After this experience, I will examine all the water pumps in stock at my dealer and get the one with the smallest clearance as there is obviously a variation in manufacturing.
Richard Smith 1

Tore,

For one MK1 original water pump impeller clearance may vary from 0.020" up to 0.030" as mentionned in my "Autobooks". I send you the document to your own Email adress .
Regards.
REnou

Steve;; If ever you find the fan is turning the wrong way then you will have at least four speeds in reverse and only one in forward.
I think you might mean that the curvature on the blade is installed backwards.
Sandy Sanders
conrad sanders

OK - the clearance SHOULD be 20 - 30 thou.

I have just measured a couple of new pumps and they are nearer 70 thou!!

Chris at Octarine Services

70 thou - that is almost exactly the clearance of my old pump!

I wonder - if the clearance is not important, why did the factory bother to specify that it should be from 0.020" to 0.030", as in the "Autobooks" drawing I just got from Renou?

Tore
Tore

Both early grease plug type, and later sealed types are shown in the period BMC manual. Both drawings show 20 - 30 thou clearance.

The wording which relates to this is in "Reassembling",

"(d) Ensure that a clearance of .020 to .030 in. (.51 to .76 mm ) exists between the impeller vanes and the pump body (c, Fig. C.3)."

Personally, I would read this as a guide to setting the clearance when reassembling rather than a strict tolerance. I do not think it will change your cooling at all with this type of pump.

There is usually ample cooling available in temperate climates. Ensure the correct stat is fitted, 82° max, and try a reverse flush or even a chemical flush.

Rich.
Rich

I meant to include a couple of thoughts,

The clearance may be specified to ensure the seal is correctly compressed when replacing the impeller; the drive hub goes on the spindle first and is fitted flush and the impeller goes on last. Also, it may ensure the front edge of the impeller doesn't touch the pump body, and the rear doesn't touch the block. This clearance may only apply to OE pumps, pattern types may be different.

Rich
Rich

Sounds sensible, Rich.

Just come to think of it - in oil pumps, the clearances are very small and extremely critical. But they usually work on a different principle? Are they "positive displacement" pumps?
Tore

Tore,

Yes - oil pumps are positive displacement eccentric lobe pumps.

Oil is sucked in due to the space between the lobes increasing as it passes the inlet port.

The oil is then pushed out of the outlet port as the space between the lobes decreases.

Oil pressure is created because the oil cannot flow through the bearings as fast as the pump delivers it - hence the pressure relief valve to stop excessive pressure.

On the other hand the water pump ( and the heater fan) are just centrifugal pumps that spin in the fluid and cause the fluid to move in the desired direction - if pressure builds up with these pumps then the pump stalls, just spinning to no effect!

They are very crude and are there just to augment natural forces, thermosyphon in the case of the water pump.
Chris at Octarine Services

Yes I agree with Chris, a positive displacement pump seals the fluid it's pumping, then reduces the volume so it is displaced. The resulting fluid pressure may be important; clearances will affect the seal and hence the effectiveness. Some oil pumps are piston type with sprung balls for valves but the principle is the same; the fluid is positively displaced as it cannot go anywhere else.

The water pump acts more like a fan would do in air, it moves the coolant round but the clearances are greater and there would be no hydraulic lock-up of the pump if the fluid could not move.

Rich.
Rich

I feel I am getting more knowledgeable every minute! Just one more question - on the Midget, there is a radiator shroud (the right word?) enclosing the upper part of the fan. This was not there on the first Sprites, but was introduced to improve cooling efficiency.

Is not this comparable to the waterpump clearance? A small clearance gives the water fewer escape routes, so that more water is forced in the right direction?

Tore
Tore

Yes correct in the case of the fan!
Do not read this into your water pump clearances though. If your impeller was 30 thou closer at the front the gap at the back would be 30 thou greater so no net improvement.
There's no sealing intended, remember that the impeller has to turn even when the stat is closed.
Rich.
Rich

Regarding my earlier posting about the impeller clearance, I have some news to report.

Bearing in mind that my recent water pump replacement was required because of seal failure, rather than rust-out or impellor deterioration, a week with the new pump (the one with much tighter impeller clearance) shows an improvement in performance. The main area of improvement is in water temp consistency and recovery from start and stop driving conditions.

Based on the other comments and my memory of the visual clearnace differences I observed when I replaced the pump, I would guess that the clearance on my old pump was in the .070 area and the clearance on the new one is in the .020 to .030 range. And the permance IS better than the old pump.

I hope these observations help.
Richard Smith

Richard,
By "Permance" would you mean performance?
you say " The main area of improvement is in water temp consistency and recovery from start and stop driving conditions." And I remember to have got a ride in a 65' B, I noticed this exactly like you, after a traffic light, at startup, the water temperature decreased down to 80°C more quickly than on mine car
Could the narrower impeller clearance be important to give a higher and more sustained flow ???
Renou

Thanks for catching the typo!

Although a centrifugal pump (that's what the water pump is) is not a positive displacement pump, it does have 'slip' and the 'slip' generally occurs at the gap between the impeller and case.

It may be a little easier when you understand what is going on in a centrifugal pump. Although not easy to see in our water pumps, the water enters in the center of the impeller and is 'slung' centrifugally outward by the impeller to exit under pressure out the pipe that connects the pump to the radiator. As the pump tries to build pressure at the outer edge of the impeller it is easy to see that its ability to build pressure (and flow) will be affected by the gap between the impeller and the case. The bigger the gap, the more leakage will occur. And, in the case of the water pump note that the back of the impeller is open to the water in the block all around its periphery - the leakage that occurs through the gap simply returns to the engine and never gets to the radiator.

This situation is also worsened by performance at idle - pump speed is low and pressure and flow are low as well. The 'recovery' time will be affected by the pump's ability to move water when you accelerate away and then later reach a higher speed. The more water it can flow the quicker the recovery will be.

Logic tells me that the tighter the gap, the better pump performance will be.
Richard Smith

To me, this makes a lot of sense. And I like to think that the old boys at BMC did have serious intentions when they specified that this gap should be between .020 and .030.
Tore

Richard - wrong way!

The impellor SUCKS water from the rad and throws it into the block.

The gap between the impellor and the casing is going to make a minor difference in that some water will pass through the gap but the bulk of the flow is thrown outwards and rearwards away from the gap.
Chris at Octarine Services

Yes Chris you are right, but basically I agree with Richard point of view . Lower gap in the pump produces a slightly higher builtin output flow by reducing the sleep in coolant and consequently a much more efficient cooling capacity due to the slight increase in flow speed. Of course a flow bench test would confirm all this.
Renou

Renou,

I would challenge the "much more efficient" - the prime force at work in the cooling system is the thermosyphon effect which is driven by the hot water rising in the engine and flowing into the radiator where it is cooled by the airflow and returns to the bottom of the engine to be recirculated.

The water pump only "assists" this action.

If for arguements sake we say the two forces are equal then the pump contributes half the circulation and if the increased gap in the pump contributes say a 10 % loss then the overall efficiency gain by closing the gap is only 5% - a marginal gain.

More cooling efficency would be gained by a deeper core in the radiator which would strengthen the thermo syphon effect.

However, it is possible to overcool an engine and, contrary to most people's natural reaction to want to see the temperature needle close to the N mark (or 190F) the engine is actually better off running at 212 degrees or well into the H!

The pressure in the cooling system raises the boiling point as does the addition of antifreeze. The localised boiling in the cylinder head actually carries away more heat as the latent heat of the coolant is greater than the heat required to raise the temperature before boiling.

Driving at altitude has the reverse effect - the boiling point is reduced ( to about 200 degrees at 6000 feet) and you either need to increase the pressure cap rating to compensate or run the engine cooler.

At the end of the day it is personal preference as to how hot you run your engine but I really would not worry about water pump clearances!
Chris at Octarine Services

Yes I will agree with all your args . But BMC documentations is somewhat troublesome indicating narrow gaps in all the water pumps as well fot the B, A and even for the Twin Cam MGA. And as Richard previously said, the narrower gap seemed to make temperature recovery a bit quicker. As far this thread was full of interest, a lab check would carry a little more than our speculations.
Thanks to everyone 's contribution to this post .
Renou

Chris
I don't buy your thermosyphon argument. I had several 3=6 DKW s. They were cooled by thermosyphon. The system was carefully set up for that purpose. The head water flow (no valves as it is a 2 cycle engine) slopes upward toward the radiator. Three pipes about 2" in diameter all went up and joined a large diameter pipe ascending at a 30 degree angle to the top of the radiator. The radiator is mounted at the rear of the engine as high as possible and the pipes and hoses were in excess of 2" in diameter. There were no valves or ports in the head where heat concentrated. The waterways were large and unrestricted around the cylinders. The engine put out about 1/3 the horsepower of the MGB and had a radiator twice as thick and about 2/3 the rectangle size. It worked pretty good. The major problem was carburetor icing in the winter as there was no warm radiator heated air bathing the engine as the radiator was in the rear. They couldn't restrict the flow of the water with a thermostat so you had a shade you could pull up in front of the radiator to reduce air flow.

The MG in contrast has about 6" from the top of the head to the radiator. The hoses are small and restricted with a thermostat and twists and turns. The water comes in the front of the engine and if it followed the route of least resistance would go straight up to the outlet and cool only the front side of cylinder 1. There would be virtually no cooling flow around the exhaust valve as the water would be moving so slowly it would only be steam in the vicinity.

Hot spots that get cooled by steam get hotter not cooler. Steam is 1300 times less dense than water. That means it carrys 1/1300 as much heat. Converting water to steam absorbs energy it is true, but steam will not carry much heat energy away. If your steam theory were true an engine would run cooler with no radiator cap. At non pressurized atmospheric pressures the steam would form at lower temperatures and keep the engine cooler??

One of the reasons for a hi pressure radiator cap is to keep steam from forming at hot spots in the engine. You may never run you engine over 200 f, but there are hot spots under hi load well over that temp. By having a a 15 lbs pressure cap you can raise the boiling steam pressure hi enough to stop steam pockets from forming.

One of the major advances Ford made from the model T to the model A was a water pump.

Barry
Barry Parkinson

Haynes:
"Coolant is circulated by a thermo-syphon, pump assisted system."

Factory Manual:
"The cooling system is pressurised and the water circulation is assisted by a pump attached to the front of the engine..."

However what percentage is due to what effect is not stated, but it would appear that both thermo syphon and pump are involved.

Cheers
ian F
Ian Fraser

The description is technically accurate, since in the absence of the pump turning there will still be some thermo-syphonic circulation and hence cooling. But as anyone who has lost a fan belt and not noticed the ignition warning light will know, that on its own is nowhere near adequate even for light throttle cruising in other than very cold ambients.

All very interesting, but going back to the original question either the engine temps are normal and the indication is high, the engine is generating too much heat that the cooling system cannot get rid of, or the cooling system cannot get rid of the amount of heat it is supposed to. The first can be confirmed or eliminated by checking the coolant temperature when the gauge reading shows high.
Paul Hunt

I just checked my temperature gauge, as many of you have suggested, and it turns out to be dead accurate. Surprise for me, too!

Today, I visited my faourite local MG parts dealer (OK - 110 km away, but there are not many of them in Norway!) He is an old engineer that used to work on turbines, and his opinion was that the clearance is definitely important. Anyway, he had the right MGB water pump with .020 clearance, so I bought it! When I get the engine back in the car in two or three weeks time, we shall se...

Tore
Tore

Chris,

Boy was that a goof! You are right, of course.

Sorry for the confusion...

However, the other stuff oabout centrifugal pumps is ok.
Richard Smith

Paul is right - no fan belt = overheat.
Richard Smith

Very interesting discussion. I have nothing technical to add, but due own and drive a 65 Austin Healey Sprite and a 66 MGB. The Sprite has always run at normal, even at outside temps of 85 to 90 degrees F. The B has always been a heat problem in continued slow driving at the same temps. On engine overhaul of the Sprite, the timing was advanced way too much and the engine always ran hotter and had less zip. Changed the timing to specs, runs cool again and is quicker. I have a new radiator in the B, run it with 50/50 antifreeze/ H20, and also tried using a blanking sleeve at the head......still runs hot. I put a fan on the front, helps very little, but cools quick on shutdown. So perhaps it is the nature of the beast. I have not checked the guage, perhaps that is the next step along with an engine backflush. Also as someone said in the thread, it works and there is no engine problems, but it bugs me.
James Hanks

James

I'd check timing on the B as you did with the Spridget, but possibly your B is clogged. I've run them in all temps up to our best English summer days of 90°F - don't laugh guys it was back in 75 and 76 remember :-) - anyway they never overheated, but they weren't old cars back then.

I've just put a new (not recon) rad on my present car and backflushed it. Took the stat out and disconnected the heater and ran water through every way. It now sits at whatever temp the stat is, 180°F/82°C.

Go for a backflush and maybe a chemical flush, I don't think there's a problem doing that in a B, anyone done it?

A small point is you'd get better cooling with 25% or 30% antifreeze, to some extent that 50% mix will be negating the water wetter you use.

Rich
Rich


James, Rich,
It's boring, feeling that something might go wrong.
I too believe that a clear and efficient chimical block cleaning helps to cure the "overheat" issue. I am under rebuilding a new engine for my car and I was surprised how clean the internal block walls came back from this operation, nearly as new, without speaking the cast renewed appearance to head cylinder inside. My radiator refurbisher spoke me about coolant additive used in trucks and coaches engines cleaning along any internal deposits and avoiding the "clogging up" , any thoughs about that ?
Renou

Renou,

Once cleaned out it will stay that way if you ALWAYS use antifreeze and change it annually. I have seen engines with no rust or scale in them at all - just like the day they were made - the owners had run them with antifreeze in the water from new.
Chris at Octarine Services

The block in my 67 gt is a 79. The car I pulled it from was rough and neglected. It had badly recessed valves (unleaded gas) and some flat cam lobes and well over 100k miles with no major engine work. I bought it from the 2nd owner who had neglected the car for its last 20,000 miles.
The cooling system - including radiator, was like new. There was no cylinder ridge and the bores and crank were all within stock specs.
The original owner had used synthetic oil, Mobil 1, and had used antifreeze.
Barry




Barry Parkinson

This thread was discussed between 30/03/2005 and 10/04/2005

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.