Car water pump barrier to flow when off?

[bluecap]

When a car is parked and not running... is the radiator water (or coolant) pump totally shut that water from the radiator side can't enter the engine or in equilibrium (what's the exact term)? The water pump is a centrifugal pump.. does this totally shut the inlet and outlet from mixing so that when you have a leak in the engine side.. would the radiator water decrease or would it not be affected because the centrifugal pump serves as barrier?

Or for bigger illustration. Supposed all the engine side coolant are gone. Can the radiator water or coolant still be full because the water pump is serving as barrier for any water flow between inlet and outlet?

 

[berkeman]

When a car is parked and not running... is the radiator water (or coolant) pump totally shut that water from the radiator side can't enter the engine or in equilibrium (what's the exact term)? The water pump is a centrifugal pump.. does this totally shut the inlet and outlet from mixing so that when you have a leak in the engine side.. would the radiator water decrease or would it not be affected because the centrifugal pump serves as barrier?

Or for bigger illustration. Supposed all the engine side coolant are gone. Can the radiator water or coolant still be full because the water pump is serving as barrier for any water flow between inlet and outlet?

Definitely not my specialty, but AFAIK, there are no one-way valves in that cooling loop...

http://denverexpresscare.com/wp-content/uploads/2014/11/sb_0007AUSMB.jpg

 

[bluecap]

Definitely not my specialty, but AFAIK, there are no one-way valves in that cooling loop...

http://denverexpresscare.com/wp-content/uploads/2014/11/sb_0007AUSMB.jpg
View attachment 220574

Thanks. What kind of water pump design (even other home water pump, etc.) has one way valves right in the design (and not a separate valve that you can add on in the water pump)?

 

[Tom.G]

As @berkeman indicated , there are no valves in an automotive centrifugal water pump. There is a thermostatic valve mounted in the engine block that controls the engine coolant temperature, but they generally don't completely seal closed.

If you are asking about pumps that prevent flow when not operating, then most Positive Displacement pumps qualify. Often their valves are reed valves or ball-and-spring check valves. They could also have cam driven valves as in a four-cycle engine. Then there are the Peristaltic pumps, and Gear pumps. For instance the oil pump in your car is a gear pump.

If you are talking specifically about a car and radiator, remember that both the engine and the radiator have both an inlet and outlet. You would need to block flow in several places to prevent a system-wide fluid loss.

 

[Ranger Mike]

typical automotive water pump is not totally sealed. Water will leak thru the impeller. These pumps have a "weep hole" that will drip coolant once the front impeller seal wears the point of failure. It is possible to loose quote a bit of coolant because of this. The weep hole is purposely drilled when manufactured to let the driver know there is a seal failure. If we had a total seal pump very little coolant would leak and the drip would stop once coolant in the pump housing had dripped dry.

 

[bluecap]

As @berkeman indicated , there are no valves in an automotive centrifugal water pump. There is a thermostatic valve mounted in the engine block that controls the engine coolant temperature, but they generally don't completely seal closed.

If you are asking about pumps that prevent flow when not operating, then most Positive Displacement pumps qualify. Often their valves are reed valves or ball-and-spring check valves. They could also have cam driven valves as in a four-cycle engine. Then there are the Peristaltic pumps, and Gear pumps. For instance the oil pump in your car is a gear pump.

If you are talking specifically about a car and radiator, remember that both the engine and the radiator have both an inlet and outlet. You would need to block flow in several places to prevent a system-wide fluid loss.

how about vane pump used in the power steering fluid? does it prevent flow or not?

 

[Ranger Mike]

Same situation, not a total seal, oil pump not total seal.

 

[bluecap]

Same situation, not a total seal, oil pump not total seal.

note oil pump is different from power steering vane pump...

 

[Ranger Mike]

be it vane, impeller or rotor...none are total seal..they all can "leak" fluid

 

[sophiecentaur]

I can't imagine what the context is for this question but a centrifugal pump will allow water to pass in either direction when it is not turning; it's little more than a 'fan' with no valve or seals to impede reverse flow. Also, water can always flow via the bottom hose, the return path from radiator and the level on both sides will soon end up the same if there is a leak on either side.
Can you reveal the reason for this question? Are you trying to explain some odd behaviour in your motor car cooling system?
If the engine is 'blowing out' coolant, you may have a leaking cylinder head gasket.

PS In the past, I have heard of a complete blockage of water flow due to the inner fabric skin of the bottom hose coming loose and filling up the entrance to the engine waterways. That could account for water in the radiator but not in the engine. I think water hoses are better these days, though. Your fault must be at the bottom of the system and not near the top. where the pump is often found.

 

[bluecap]

I can't imagine what the context is for this question but a centrifugal pump will allow water to pass in either direction when it is not turning; it's little more than a 'fan' with no valve or seals to impede reverse flow. Also, water can always flow via the bottom hose, the return path from radiator and the level on both sides will soon end up the same if there is a leak on either side.
Can you reveal the reason for this question? Are you trying to explain some odd behaviour in your motor car cooling system?
If the engine is 'blowing out' coolant, you may have a leaking cylinder head gasket.

PS In the past, I have heard of a complete blockage of water flow due to the inner fabric skin of the bottom hose coming loose and filling up the entrance to the engine waterways. That could account for water in the radiator but not in the engine. I think water hoses are better these days, though. Your fault must be at the bottom of the system and not near the top. where the pump is often found.

Well. There was a leak in my coolant. I brought it to Honda factory. They replaced the following brand new:

1. Radiator unit
2. Upper hose
3. Lower hose
4. Bypass hose
5 Rotary Air Control Valve Hose
6. Throttlebody Hose In
7. KL Sealing Cap
8. New Honda Coolant

However, it hasn't solved the leak! There is no visible leak beneath the car (it's not the water pump or other hoses). But the cooling still leaking from the radiator. I don't want Honda to guess and replace the engine without knowing what part exactly is leaking. They can't tell because they don't use UV dye, UV googles and UV flashlight. Is it safe to put UV dye in the coolant for diagnostic purposes?

What I'd like to know is this. Can you tell or any symptoms to look for if it's a cylinder head gasket leak or the engine has already microcrack right in the block? Any way to tell the difference? Just for reference. Also what are the list of gaskets inside the engine as far as the coolant path is concerned? Thanks.

 

[sophiecentaur]

Try a compression tester on all cylinders. If one is a bit low then it's likely to be the Cylinder Head Gasket. Engines with 'wet liners' or with other configurations have other places where a leak can be happening. What's their problem with a revealing dye? It could save hours of technician time. there should be someone with intimate knowledge of the engine and cooling system somewhere who would know more than PF contributors - although I have found that 'repair men' tend to be absolute rubbish at analysing faults and will not even consider discussing likely causes of problems.. The strategy is to change parts, starting with the most likely one to fail and working downwards. Needlessly expensive but it's usually the customer who is paying.

 

[bluecap]

Try a compression tester on all cylinders. If one is a bit low then it's likely to be the Cylinder Head Gasket. Engines with 'wet liners' or with other configurations have other places where a leak can be happening. What's their problem with a revealing dye? It could save hours of technician time. there should be someone with intimate knowledge of the engine and cooling system somewhere who would know more than PF contributors - although I have found that 'repair men' tend to be absolute rubbish at analysing faults and will not even consider discussing likely causes of problems.. The strategy is to change parts, starting with the most likely one to fail and working downwards. Needlessly expensive but it's usually the customer who is paying.

They don't have a dye test kit. They just look at the bottom of the car. Their pressure gauge is out of order. So before I let them replace the cylinder head gasket and spend over $200. I just need to know something. The following is picture of the old radiator cap... what is the yellowish substance? Is it the coolant corrosion inhibitor breakdown product or could it be sign oil is getting inside the radiator? But I read someone said it normally happened in radiator.. does it? Do you guys radiators have it too?

Even my new radiator has it... I removed some bits before the picture was taken..

Did you also see the yellowish substance? What is it?

 

[Tom.G]

That 'unknown substance' sure looks like someone put a dose of sealant in the system. And the second photo with the milky white color of the coolant is also characteristic of a sealant. If you bought the car used, that may be why the previous owner got rid of it.

Lacking a system pressure test (which would be my next step), or a dye test, @sophiecentaur gave the good suggestion of a compression test. Also try to find out if there is a Factory Service Bulletin for the problem. These are bulletins from the manufacturer to their dealers about common problems on a specific model that may even be covered by an extended warranty. They are usually not publicized and only covered when the customer complains. Perhaps @Ranger Mike would know where to track these down.

If you are comfortable working on the engine you could pull the spark plugs and see if any of them are discolored. If one is discolored there is a good chance there is a coolant leak into that cylinder, either from a head gasket or a cracked head or cylinder block. Put the plugs back in the same cylinders they came out of. That way the next person that inspects them won't get different results.

You can also check the oil on the dipstick to see if there is any discoloration. Coolant leaks into the oil frequently show up as a white or gray deposit floating on top of the oil, sometimes as a foam.

In any case, you need to find a different mechanic/dealer; that one took you to the cleaners and is waiting for another chance. I would probably keep ALL documentation including conversations with the dealer and documentation from all others that work on the vehicle. Then when it is fixed, raise Holy H*** with the manufacturer about dealer overcharging, selling unneeded parts, and not fixing the problem.

Please keep us updated about your progress... and Good Luck!

Cheers,
Tom

 

[bluecap]

That 'unknown substance' sure looks like someone put a dose of sealant in the system. And the second photo with the milky white color of the coolant is also characteristic of a sealant. If you bought the car used, that may be why the previous owner got rid of it.

No. The 2nd radiator in the pic is empty (no coolant inside) with the unit already taken out and in my floor. The white stuff is the surface of the inside metal part of the radiator. I was asking about the yellowish stuff. It was in my old and new radiator too. Btw I'm the original owner of the car bought brand new in 2004. It's a CRV. What can usually form it (see blue arrow):

Lacking a system pressure test (which would be my next step), or a dye test, @sophiecentaur gave the good suggestion of a compression test. Also try to find out if there is a Factory Service Bulletin for the problem. These are bulletins from the manufacturer to their dealers about common problems on a specific model that may even be covered by an extended warranty. They are usually not publicized and only covered when the customer complains. Perhaps @Ranger Mike would know where to track these down.

If you are comfortable working on the engine you could pull the spark plugs and see if any of them are discolored. If one is discolored there is a good chance there is a coolant leak into that cylinder, either from a head gasket or a cracked head or cylinder block. Put the plugs back in the same cylinders they came out of. That way the next person that inspects them won't get different results.

You can also check the oil on the dipstick to see if there is any discoloration. Coolant leaks into the oil frequently show up as a white or gray deposit floating on top of the oil, sometimes as a foam.

In any case, you need to find a different mechanic/dealer; that one took you to the cleaners and is waiting for another chance. I would probably keep ALL documentation including conversations with the dealer and documentation from all others that work on the vehicle. Then when it is fixed, raise Holy H*** with the manufacturer about dealer overcharging, selling unneeded parts, and not fixing the problem.

Please keep us updated about your progress... and Good Luck!

Cheers,
Tom

Yes. I will let them test the spark plug next week. I'm collecting information this week and observing the volume lost just to be informed customer in the event they want to try replacing other parts that are not broken. Thanks.

 

[Tom.G]

Well, in that case I can only (wildly?) speculate.

Known:

• Coolant disappearing. (how rapidly? under what conditions {driving, parked -- for hours or days, hot or cold?})
• Foreign material present. (hard like stone or concrete? crumbly? mushy? firmly attached to surfaces or easily removable?)
• Radiator, hoses, and cap recently replaced (what mileage / time frame since replacement?)

Speculation:

• The foreign material could be corrosion due to not replacing coolant often enough or using wrong coolant. Or it may be introduced foreign material. Since you have eliminated sealant in the system, I lean toward corrosion, or even deposits from low quality make-up water. In either case you may want to have the cooling system flushed. It depends on how much contamination there is, your tolerance for uncertainity, and your trust in your mechanic.
• Coolant disappearing. Obviously a leak somewhere. Many of the obvious causes have been replaced.
  • A prime suspect is the water pump. At 14 years old it would not be a surprise to have the shaft seal fail. These will often leak only with the engine running and the cooling system pressurized. These leaks can be hard to spot because they are hidden behind the water pump drive pully. If necessary, driving several miles with dye in the cooling system and then inspecting will spot them.
  • A leak at the coolant overflow tank or the hoses to it.
  • A leaky cap on the overflow tank not letting the system pressurize. (this would reduce the boiling temperature of the coolant)
  • Leaky heater core; but you would probably smell the hot coolant in the passenger cabin.
  • A corroded expansion plug (core plug) in the block. Again, not a surprise at 14 years old. (If an inaccessible one is leaking, adding sealant to the system will get you several more years of use.)
  • and we are down to the prior guesses of head, head gasket, or block.

Leaks into a combustion chamber often show white smoke out the tailpipe at startup or under heavy acceleration. I'm not talking about the usual steam cloud on startup in cold weather. Unfortunately it's a case of "If you've seen the comparison you will recognize it the next time." You may also be able to smell coolant in the exhaust at startup. Have a helper start the car while you hold your hand a few inches behind the tailpipe. Angle your hand as if it was a mirror and put your nose maybe 6 inches from your hand, then start the car and sniff.

Some (many?) auto supply stores in the U.S. sell UV flashlights in a kit with Yellow tinted googles for around USD15. The last time I checked (4yrs. ago) they wanted about USD30 for the dye. With a little effort you can extract dye from felt-tip highlighter pens available at office supply stores. Remove the back end of the pen and push the tip/cartridge out the back. Then soak the felt in a small amount of the fluid you want to mark to extract the dye. I haven't tried that approach using water so I don't know just how water soluble the dye is. I know it works for brake fluid, power steering fluid, that kind of stuff.

Well, past my bedtime. Keep us updated.

Tom

 

[bluecap]

Well, in that case I can only (wildly?) speculate.

Known:

• Coolant disappearing. (how rapidly? under what conditions {driving, parked -- for hours or days, hot or cold?})
• Foreign material present. (hard like stone or concrete? crumbly? mushy? firmly attached to surfaces or easily removable?)
• Radiator, hoses, and cap recently replaced (what mileage / time frame since replacement?)

Speculation:

• The foreign material could be corrosion due to not replacing coolant often enough or using wrong coolant. Or it may be introduced foreign material. Since you have eliminated sealant in the system, I lean toward corrosion, or even deposits from low quality make-up water. In either case you may want to have the cooling system flushed. It depends on how much contamination there is, your tolerance for uncertainity, and your trust in your mechanic.
• Coolant disappearing. Obviously a leak somewhere. Many of the obvious causes have been replaced.
  • A prime suspect is the water pump. At 14 years old it would not be a surprise to have the shaft seal fail. These will often leak only with the engine running and the cooling system pressurized. These leaks can be hard to spot because they are hidden behind the water pump drive pully. If necessary, driving several miles with dye in the cooling system and then inspecting will spot them.
  • A leak at the coolant overflow tank or the hoses to it.
  • A leaky cap on the overflow tank not letting the system pressurize. (this would reduce the boiling temperature of the coolant)
  • Leaky heater core; but you would probably smell the hot coolant in the passenger cabin.
  • A corroded expansion plug (core plug) in the block. Again, not a surprise at 14 years old. (If an inaccessible one is leaking, adding sealant to the system will get you several more years of use.)
  • and we are down to the prior guesses of head, head gasket, or block.

Leaks into a combustion chamber often show white smoke out the tailpipe at startup or under heavy acceleration. I'm not talking about the usual steam cloud on startup in cold weather. Unfortunately it's a case of "If you've seen the comparison you will recognize it the next time." You may also be able to smell coolant in the exhaust at startup. Have a helper start the car while you hold your hand a few inches behind the tailpipe. Angle your hand as if it was a mirror and put your nose maybe 6 inches from your hand, then start the car and sniff.

Some (many?) auto supply stores in the U.S. sell UV flashlights in a kit with Yellow tinted googles for around USD15. The last time I checked (4yrs. ago) they wanted about USD30 for the dye. With a little effort you can extract dye from felt-tip highlighter pens available at office supply stores. Remove the back end of the pen and push the tip/cartridge out the back. Then soak the felt in a small amount of the fluid you want to mark to extract the dye. I haven't tried that approach using water so I don't know just how water soluble the dye is. I know it works for brake fluid, power steering fluid, that kind of stuff.

Well, past my bedtime. Keep us updated.

Tom

Brilliant! Ok. I'll collect data like how many coolant volume lost when parked and when operating at full engine luminosity.

 

[Ranger Mike]

If you have over 75,000 miles on the engine chances are very good you have blown head gasket. Possible cracked block or head but very high for gasket failure.

Go to a garage that has hydro carbon gage used to test emission in exhaust tailpipe and test for fuel emission in coolant tank reservoir.
Quickly determine if you have cracked engine block or blown head gasket. Takes a few minutes and is cheap.

 

[sophiecentaur]

Another sign of a blown head gasket used to be a change of colour of the coolant. They used to have an indicator in the fluid which would show a lowering of pH due to acidic gases, leaking from a combustion chamber. Also, a bad gasket leak will reveal itself by irregular running of the engine.
I have to say, this repair company who are supposed to be sorting out the problem seem to be pretty incompetent. If they have no test facilities and rely on changing parts, one at a time then they are ripping off the OP. @bluecap should start complaining loudly about the way they are tackling the job. Threats of a bad review can sometimes encourage companies to 'improve' their attitude. Also, if this Honda dealer is an 'official franchise' you could take it higher to the Honda Company and ask them for help. You could, at least, get some good will financial help from them.

 

[Ranger Mike]

good points all...there are Mechanics and there are parts replacers. I think you are doing exactly the right thing by researching this. If you have not wrenched before , tackling a cylinder head gasket replacement is a huge step. If you have the tools, the time and a decent garage area, it can be done.

 

[Averagesupernova]

Generally a bad head gasket will cause coolant overflow due to combustion pressures getting inside the cooling system. At an idle the opposite can happen. Coolant is drawn into the cylinder, then when the throttle is opened a white cloud will come out the exhaust. Also, a head gasket can cause contamination in the oil.

 

[sophiecentaur]

I reckon the best advice we can give you would be not to use that particular car repair workshop. It looks to me that they are taking the micky by charging you money for their approach to fault finding. They are supposed to be professionals and they should be at least as good as a bunch of us amateurs, chatting on a forum.

 

[bluecap]

Hi, I brought it in again to Honda and they replaced a gasket and the coolant is no longer leaking. Good. Their reasonings are hoses are cheap so may as well replaced old hoses (that's what they initially did.. they really don't have colored dyes).

Anyway. I'm checking the coolant but I noticed something odd.

One hour after turning off the engine, I opened the radiator cap and the coolant level is full (to the base of the filler neck):

But after another hour, it settles below it (you can even see the gap in the right side where it has lost 5cc or 5ml of coolant.

What's odd is that after parking it for 2 days without using it.. the level stays that way. So the coolant is no longer leaking (the reservoir is not missing any coolant).

But what cause the initial settling down? You may say it's because hot coolant expands and cold coolant contracts. But when I turned on the engine a while ago. It fills up the neck base so suddenly. So its not getting the coolant from the reservoir. The coolant is supposed to be filling up every part of the engine and pathways. So what could explain this odd behavior. Could it be bleeding problem? But Honda said they bleed the air out already.

I need to know something about the relationship between the radiator and reservoir. I know the reservoir gets the extra coolant when pressure is so high and it triggers the radiator cap to divert the coolant to the reservoir. So when you use a car and shut down the engine say after 30 minutes of use. How fast before the coolant in the reservoir gets back to the radiator (is it in seconds, minutes or hours)? This may not be related to the odd behavior above but just want to know.

The reason it's not connected to the odd behavior is because when the level of the radiator is in the second pic where it has settled down and I add coolant to the filler neck.. it settles down again to the second picture even when the engine is not turned on.. so what could be inside the engine that's causing the changing in the level of the coolant in the radiator (in the first and second pic)? it's not a leak because it's not settling down further. Thanks.

 

[sophiecentaur]

Initial drop in level: sounds to me like there are some voids in the system that gradually filled up when the engine was first run. That would look like a loss. Now they are full of coolant and not air, there is no more apparent loss.
Happy driving and try to avoid looking at the temperature gauge every ten seconds.

 

[bluecap]

Initial drop in level: sounds to me like there are some voids in the system that gradually filled up when the engine was first run. That would look like a loss. Now they are full of coolant and not air, there is no more apparent loss.

No. The initial or any drop in level occurs whenever the engine is cold.
Again one minute or one hour after using the car. The coolant level is full to the filler neck. Then after another hour, it settles down and remain that low level without further settling down. You have to start the engine to make level rise up even after 30 seconds of using then shutting down. What could causing that behavior? Doesn't your car behave this way too?

Happy driving and try to avoid looking at the temperature gauge every ten seconds.

 

[bluecap]

Initial drop in level: sounds to me like there are some voids in the system that gradually filled up when the engine was first run. That would look like a loss. Now they are full of coolant and not air, there is no more apparent loss.
Happy driving and try to avoid looking at the temperature gauge every ten seconds.

About these voids in the system. Is it not that for the cooling system. The system should be entirely bled of air so the coolant should reach every molecule of the engine, etc. even when parked? Or are voids unavoidable? what components can contain the voids?

 

[Tom.G]

It might take 50 to 100 miles for the system to completely purge, especially if the heater in the passenger compartment if it is not being used. Try running the heater for a while.

Here are some other possibilities to check.
If the vehicle does not have a cooling system overflow tank, then this is normal behavior. (The top tank of the radiator acts as an expansion tank.)
If the vehicle does have a cooling system overflow tank and you must continually add coolant to the radiator but not to the tank:

• Check the hose between the radiator and overflow tank for leaks/cracks
• Inspect and clean the sealing surfaces of the radiator neck and the radiator cap
• Replace the radiator cap

If you don't have to continually add coolant I would consider it a personality quirk of the vehicle. I once had a vehicle that wanted the engine oil ¾ quart low. If topped up it would be back down within 100 miles. If left alone, it stayed ¾ quart low.

 

[Ketch22]

The drop that you are seeing can be a normal system. The radiator cap has two gaskets, one is at the "top" of the neck and it should seal the system. The other is at the bottom of the neck and it should seal the radiator.
When you start the car the coolant expands and as pressure develops due to expansion related to temperature the overage causes the bottom seal to open and the "excess" fluid escapes into the cap area and down the reservoir hose to the reservoir. When the engine is turned off the coolant cools and contracts. This causes reduced pressure which draws fluid from the reservoir to fill in the space. * This will only happen until the pressure equalizes with the static lift height from the reservoir. * When the engine is completely cold there is often a small gap at the bottom of the neck. Opening the lid to check will only make the challenge more apparent as it causes an atmospheric opening that reduces the low pressure.

To check if this is the condition you are experiencing start with filling your radiator to the top of the neck when fully cold. Then fill the reservoir to a significantly "overfull" condition. Then drive for a period of time after the engine is fully up to temperature. It would be expected to expected to smell coolant as you drive, this is do to the overfull condition. Then, park the car in a level site and allow to cool until all parts (radiator, block, and heater systems) are cold. Check the coolant level and the drop in the reservoir. Then repeat the drive and cool cycle. Then repeat again. One should expect consistent level in the reservoir when cool which is the actual maximum fill level. The manufacture normally adds a little safety factor to allow for temperature variations. Also after the third time one may expect that conditions are fairly normalized. The gap from the bottom of the neck to the actual water level is a reflection of the vacuum generated by the reduced volume due to cooling. This should stay relatively consistent over time.
If, after a period of time the gap changes or the reservoir level changes one can infer a leak or a defective radiator cap.

 

[Ketch22]

Tom G beat me to it. The same condition may be caused by the evacuation of air from varies isolated portions of the system. This is why you need to watch for a change over time as opposed to an initial change ( which may last for a couple hundred miles) followed by a static state, as opposed to a consistent and continuing change.

 

[Ketch22]

About these voids in the system. Is it not that for the cooling system. Or are voids unavoidable? what components can contain the voids?

One of my rides is a first generation MR2 Toyota. This is a mid engine, Front Radiator car. These are legendary for being difficult to bleed the air from. The engine heats the coolant which moves to the cab heaters which have their own thermostat. When the second thermostat opens the coolant goes to the radiator which has it's own thermostat. Air bubbles and challenges exist in several places.
Bleeding the air is very difficult at the most polite expression.