Jump to content
Sign in to follow this  

Radiator help please!!!!

Recommended Posts

Hey guys,

I've got a 300 engine going in a kxf, I'm struggling with the radiators. It's ALOT more convenient with respect to routing and rad hose points to run the rads in series, rather than parallel. It's my understanding that this will cool equally well, with alot more heat in the first rad, then the second is cooler. i think it should cool better at low rpm than parallel, but it will put more strain on the water pump.

Is there any reason I should be concerned about the strain on the water pump? It's a plastic impeller but that shouldn't matter I'd think. The bike has loads of power so I'm not concerned about the power draw.

Do I have anything to be concerned about or should I just go for it?

Share this post


Link to post
Share on other sites

running them in series will create allot more back pressure and the pump may not be able to supply the necessary pressure. i would probably just spend the extra time not to get them in parallel like the whole system is intended to be run and not have to deal with any possible headaches later. if the water pump cant supply the pressure needed to maintain the flow rate then the engine will run the risk of overheating. especially being that you are already running a bigger displacement then the radiators were originally designed for.

Share this post


Link to post
Share on other sites

actually now that i think back to heat transfer, the rate of change in temperature is dependent on the difference between the hot temperature and the cold temperature. having the second radiator already cooler the change in T is less therefore creating less of a change in temperature in that radiator. so i believe if i am remembering correctly you will not get as good of cooling with them in series.

Share this post


Link to post
Share on other sites
actually now that i think back to heat transfer, the rate of change in temperature is dependent on the difference between the hot temperature and the cold temperature. having the second radiator already cooler the change in T is less therefore creating less of a change in temperature in that radiator. so i believe if i am remembering correctly you will not get as good of cooling with them in series.

Correct. However, flow rate in the radiators is reduced when they are in parallel which reduces the heat transfer rate...but I agree that parallel is better than series due to the maximization of delta T. Also, the increased pressure as seen with the radiators in series will result in further losses in the system...which also reduces the efficiency.

Share this post


Link to post
Share on other sites

Just so we're not being redundant, here's another thread on this, but I still want to hear peoples thoughts on my specific application.

It seems to me that the bearings in the water pump are strong enough to support any load the pump could create, and the power draw of the increased load is no issue to me, the motor has loads of power, but there's a reason it was done in parallel in OE form.

Some dirtbikes are done in series, my 2002 KDX220 is in series.

I would also like to run one computer fan, and I feel like the series thing would be the most conducive to that fan being effective. If the fan is run at the bottom of the left rad, where all the heat is coming in, it will also dissipate the most heat.

http://www.thumpertalk.com/forum/showthread.php?t=858653

Share this post


Link to post
Share on other sites

To properly connect the radiators in series, water has to be taken from the head to the top of the inlet radiator, then out the bottom of that to the top of the secondary radiator, then out the bottom of that to the water pump. Somehow I don't see how that simplifies the plumbing. I suppose it could be done by filling the primary rad from the bottom and having it spill over into the secondary, but either way, the system will need to always have more water in it than is required to fill the primary, or it will cease to work.

OTOH, with parallel radiators, the system will maintain some functionality until there is so little coolant that it can't reach the radiators at all.

Share this post


Link to post
Share on other sites

I'm still torn about this guys, I talked to some very knowledgeable people I know and I've got new concerns.

I would be reducing the cross sectional area that the pump is pushing through by half, and the worst thing that can happen is if the pressure in the radiators is increased enough to reduce the pressure in the head. That would be bad. I'd be using radiators that were designed to cool the heat of 30hp to cool the heat of 50hp, but the advantage I would have with series is that I could put the single fan at the hottest point, so the cooling of the fluid itself should be fine, but now i'm mostly concerned about possibly reducing the pressure in the head.

The other side of this coin is that radiators on models with both a 250f and 450f version are very similar, the radiators on my KDX that's in series is very similar to all the other radiators I'm looking at, and I found out they make DRZ400 models with one in series and one in parallel, S and E I think. So am I overthinking this?

In order to run them in parallel I have to cut a passageway in the frame and run the other line behind the cylinder. It would be awkward to say the least.

Share this post


Link to post
Share on other sites
To properly connect the radiators in series, water has to be taken from the head to the top of the inlet radiator, then out the bottom of that to the top of the secondary radiator, then out the bottom of that to the water pump. Somehow I don't see how that simplifies the plumbing. I suppose it could be done by filling the primary rad from the bottom and having it spill over into the secondary, but either way, the system will need to always have more water in it than is required to fill the primary, or it will cease to work.

OTOH, with parallel radiators, the system will maintain some functionality until there is so little coolant that it can't reach the radiators at all.

I thought that was how it's usually done, but my KDX in series feeds all the heat into the bottom of the left rad, then spills over into the right side.

The left rad will always be pumped completely full, spilling into the right, so it wouldn't stop working unless I had the system less than half full, right? I'm not too concerned about that I guess....

Share this post


Link to post
Share on other sites

I haven't ever seen ("that I am aware of", which I say because I've seen a KDX but didn't notice what you describe) any system plumbed the way you describe. All the water cooled bikes I've owned have run the outlet hose from the top of the engine to the top of one radiator, at which point it flows down through both a hose leading back to the water pump. The radiators are connected at the top tanks, and the bottom tanks are either connected and share a single outlet, or each have an outlet that T's into the water pump inlet.

With regard to the load on the pump, there is no loading applied to the pump from pushing water through the radiators because even half of one radiator has a greater cross section than the hose running into it, even taking into account the increased frictional contact against the increased surface area of several small tubes as opposed to a larger one. The load on the pump comes from the weight of the column of water it is lifting. Now, in considering this, you have to be careful, because it's easy to first see this as a comparison of the column of water in the engine outlet hose running to the top of the radiator to the water contained within a hose running to the bottom of one radiator plus the water in that radiator. That would be incorrect, however, because fluids seek their own level, and the pump only has to generate enough pressure to lift water from what ever level it sits at statically to the point where it can spill over to the other side of the system.

Note that in a healthy radiator setup with top filled parallel radiators, water flow though the radiators is accomplished by gravity, while the pump simply moves water to the top tank where it can fall through. In your KDX, the flow through the secondary radiator happens that same way, obviously, but what's less obvious is that the same is true of the primary radiator, as the flow through it is driven at least in part by the fluid's natural tendency to seek it's own level, while the pump is creating a condition where the two sides are out of balance by pumping water that last inch over into the secondary.

Share this post


Link to post
Share on other sites

grayracer,

thanks that def gives me some more things to think about. The only thing I suspect to be wrong is I wonder about the radiator not being the most restrictive point. I was talking to an uncle who's been heavily involved in engine design for many years and he told me that the radiator is supposed to be the most restrictive point, not the hoses or outlets. I didn't realize that either, so I took my radiator, plugged up all the outlets but one, and blew through it. I could feel noticeably more resistance than blowing through a straight section of tubing the size of the outlet the hose clamps too.

Share this post


Link to post
Share on other sites
Note that in a healthy radiator setup with top filled parallel radiators, water flow though the radiators is accomplished by gravity, while the pump simply moves water to the top tank where it can fall through. In your KDX, the flow through the secondary radiator happens that same way, obviously, but what's less obvious is that the same is true of the primary radiator, as the flow through it is driven at least in part by the fluid's natural tendency to seek it's own level, while the pump is creating a condition where the two sides are out of balance by pumping water that last inch over into the secondary.

I don't think this is right. In a pumping enclosed system, cooling fluid is being both pushed and pulled. One side of the water pump has positive pressure, the other side has negative pressure. Gravity would have an effect but, it would be very small. Coolant is both pushed into the radiators and pulled out.

Share this post


Link to post
Share on other sites
I don't think this is right. In a pumping enclosed system, cooling fluid is being both pushed and pulled. One side of the water pump has positive pressure, the other side has negative pressure. Gravity would have an effect but, it would be very small. Coolant is both pushed into the radiators and pulled out.

You're correct, but the effect is not that small. The point is that rather than there being a head of pressure in the top tank forcing water down through the radiator, it simply falls through as water is pulled out of the bottom of it to be pumped through the cooling system. As proof, you already know that water will circulate with the cap off the top tank.

The only water the pump must actually lift is the delta between the water level in the radiators and the height of the point where water enters the radiator (or the spillover point if plumbed like Jeezo's KDX). If the water level is lowered by an inch, the pump must lift water another inch higher than that level to circulate. If the system is totally filled, there is no lifting, only circulating, and the load on the pump is not much different than it would be if the entire cooling system were laid out flat on the same level. The radiator core itself is a "wide spot" in the stream, where the speed of the moving water slows.

Share this post


Link to post
Share on other sites
You're correct, but the effect is not that small. The point is that rather than there being a head of pressure in the top tank forcing water down through the radiator, it simply falls through as water is pulled out of the bottom of it to be pumped through the cooling system. As proof, you already know that water will circulate with the cap off the top tank.

The only water the pump must actually lift is the delta between the water level in the radiators and the height of the point where water enters the radiator (or the spillover point if plumbed like Jeezo's KDX). If the water level is lowered by an inch, the pump must lift water another inch higher than that level to circulate. If the system is totally filled, there is no lifting, only circulating, and the load on the pump is not much different than it would be if the entire cooling system were laid out flat on the same level. The radiator core itself is a "wide spot" in the stream, where the speed of the moving water slows.

Sorry but there IS pressure on the top of the radiators, even if the coolant level is low. That pressure comes from the pump. Gravity does help provide some of the force that causes the coolant to flow through the radiators but, pump pressure has a larger influence. Water is pushed to the top of the radiators and then through by the rush of incoming coolant AND it is pulled out the bottom of the radiators by the pumps' appetite for coolant on its' intake side. Even if the radiator cap is removed, the coolant would still be pulled from the bottom of the radiators by the pump. In other words, the pump not only pumps, it sucks. :thumbsup:

Share this post


Link to post
Share on other sites

And yet, water does not gush up and out of a down flow radiator when run with the cap removed.

As far as the pump sucking, the basic truth is that nothing really does that. Coolant is moved toward the pump by three forces in an open system; nature's abhorrence of vacuum, which impels coolant immediately above that taken by the pump to fill the void, gravity, and atmospheric pressure, although this last only comes into play to the extent that no new coolant is delivered to the top tank. If the system is closed, then the weight of the water delivered to the top replaces AP as the driving force. Coolant will move only to the extent that the pump displaces it from one side to the other. If the radiator were entirely closed with no air or water entering to replace the water drawn toward the pump, it would move nothing.

There is a force generated by the pump, but the pump is not required to lift a head of water the full distance from the pump to the top tank, nor to force water through the radiator. If you modeled a similar system that had a centrifugal pump at ground level to pump water from the bottom of a 6 foot tall tank filled with water back to the top of it, you'd be able to measure pressure on both sides of the pump. With the pump off, both sides would show pressure equal to the weight of the 6 foot head of water within the cross section of the tubing where the pressure was read. With the pump on, there would be a differential displayed, but it would not be remarkable unless the speed of the moving water was quite high. Water in the system will be lifted clear to the discharge point only by the weight of the water on the tank side at at a rate defined by the output volume of the pump, which has only to circulate water, not lift it.

But if you reduce the water level to 3 feet with the same plumbing, the pump will have to raise a 3 foot head of water plus how ever much the level in the tank was reduced by filling the discharge tube to the top, and the pressures will be very different across the pump.

My point, remember, is targeted at the remark made by Jeezo's old mechanic friend who informed him that the radiator was intended to slow the flow by impeding it, which is quite the opposite of what actually happens. The radiator does slow the flow of coolant, but does so by reducing the impedance to flow. It's analogous to a one lane road flowing into a 10 lane road for a time then back into a single lane again, with the requirement that all lanes be filled bumper to bumper. The speed in and out remains the same, but the speed of the individual car in the ten lane section is dramatically reduced.

Share this post


Link to post
Share on other sites

I'm pretty sure the radiator core itself IS the most restrictive part of the system. Does anyone know how much pressure the pump creates? Most of the pressure i thought was created my the expansion of the hot coolant, not the pump. The purpose of the pump is to create a higher pressure in the head than in the radiators, right?

Share this post


Link to post
Share on other sites

The pump's job is to displace water from the radiators into the engine cylinder. Water entering the bottom displaces water hear the top displacing water in the head to the radiators. The only water the pump can physically move is that which sits between the fins of the impeller, which it expels outward by centrifugal force as it spins. In order for this to occur, however, water must flow into the center of the impeller to replace that which is being expelled or shearing and cavitation occur. Is a pressure differential across the pump, obviously, but under circulation in a filled system, it isn't very much.

Whether the radiator is the most restrictive segment of the system or not depends on the size of it relative to the pump and the cross section of the connecting plumbing. I can think of examples in automotive cross flows that might be, but it isn't the norm. As an example in specifics (even though the measurements are rough), the '06 YZ450 has radiators consisting of a total of 38 .750 x .0625" tubes, for a cross section of 1.78 sq. inch. The 5/8' hoses feeding and drawing from it have a cross section of .307 sq. inch. As you can see, that's quite a difference. There will be an increase in viscous drag in the radiator because of the increased contact between fluid and conduit, but unless there is something wrong with the radiator, that won't make up for the nearly 8x different in cross section.

Share this post


Link to post
Share on other sites

thanks for the help guys, keep it coming, I'm still not sure what I should be doing here...

grayracer, if that's the case, what does it mean for me? In either case the pump has to either displace 2x water to the high point or displace x water at 2x speed.

The guys I've talked to that I trust the most so far on the subject say that the most important factor is that the highest pressure point is in the cylinder, and that switching a pump designed for parallel to series risks lowering the pressure in the cylinder to much, by raising the pressure in the radiators. Does that sound right?

Share this post


Link to post
Share on other sites
grayracer, if that's the case, what does it mean for me? In either case the pump has to either displace 2x water to the high point or displace x water at 2x speed.
Not sure how you arrived at either conclusion.

Remember that if the system is kept filled, all the pump has to do is move water from one side of the housing to the other. It only has to lift water the last inch or so, and the pressure needed is determined by how fast the water has to move.

The guys I've talked to that I trust the most so far on the subject say that the most important factor is that the highest pressure point is in the cylinder, and that switching a pump designed for parallel to series risks lowering the pressure in the cylinder to much, by raising the pressure in the radiators. Does that sound right?

No, it doesn't. As a simple matter of moving fluid, what would the difference be in a parallel or series radiator set? Using the YZ450 radiators above as an example, each half of the set has an equivalent total cross section of .89 sq. in. The radiators are fed at some rate by the .307 sq. in. area of the hose. If the rate is 5 gpm, that means the velocity of the water in the hose would be about 63 fpm. This will slow to 22 fpm as it is pumped through one of the radiators if they are plumbed in series. If the two are parallel, the average rate of flow within the radiator set drops to 11 fpm.

The principal difference is that the series radiator set has more viscous drag due to its smaller cross section, and the water flow is faster, which exaggerates the difference in drag. The larger cross section and slower flow of the parallel set offers less resistance. We've already seen that the parallel set would be very unlikely to create more drag on the pump than the hoses do, but by cutting the cross section in half, and doubling the flow rate, the series set might become a factor. Either way, the parallel set has a lower resistance to throughput.

Now, I think you should focus on what the goal of the system is. The idea is to get excess heat from the engine, transport it to the radiator, and get rid of it quickly enough that the temperature of the engine can be kept below dangerous levels. Since the transfer of heat from the source to a cooling medium is fastest when the temperature difference is greatest, we need to keep the water moving through the system rapidly enough to extract unused heat as fast as it is created. But when it gets to the radiator, it needs to be slowed down so that as much heat as possible can be extracted. The lower speed of water in the parallel set, then, is preferable.

The efficiency of the system depends on volume, not on pressure, and on the thermal transfer efficiency of the radiator. Pressure is necessary only as the driving force that moves water across the pump, and there is no benefit to having more or less if the volume of water pumped through the system is adequate.

In the end, whichever way you go with it, the question is whether the bike will stay cool or not. If it does, your system is doing enough.

Share this post


Link to post
Share on other sites

You may want to consider figuring out some way of monitoring the coolant temp, either with a gauge or with some of that stick-on tape temp stuff. I suppose that it would be best to look at the coolant coming out of the engine and heading for the radiators to see the max temp. I'm not sure what an "acceptable range" would be though.

Share this post


Link to post
Share on other sites

The acceptable range is to be above 140 (preferably), but 5 degrees below whatever the boil point is for the particular coolant mix and system pressure is. For 50/50 EG/water at 16 PSI, that's roughly 260 as a max. Running a tape at the water out of the engine and at the water leaving the radiator is helpful, also.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Reply with:

Sign in to follow this  

×