water cooled head that will fit...

[Frankenstein]

darn tootin it was a pain in the arse!

nah, it just needed to make it to MKII, or version 1.0.2 or something...

basically i didnt give much thought to the idea, just made it... now, if i had sat down and seriously contemplated how to get the thermosiphon to work properly, and had oriented the inlet/outlet and radiator to do so...it would have worked

one, inlet as low as possible. 2, outlet as high as possible. 3, head has to be basically level with center of radiator, and as short and straight a run with hoses as possible. no sharp bends and no loops etc.

"polarised" isnt really the word, but the head simply needs to be plumbed so the water can flow the correct way. ^^^up^^^, as hot (less dense) water rises. no pockets, no air/water traps, just as smooth and unrestricted a flow as possible.

completely pointless, but it would have worked!

the 120mm PC radiator was adequate.


being unpressurised does limit heat transfer to an extent, and there has to be allowance for water expansion upon heating as well.

coincidentally, only just the other night i suddenly realised why the pump i had made didnt work! should have made it from plastic or something nonconductive rather than metal/alloy. eddy currents! so bleedingly obvious now...


only advantage of water cooling is when you are stuck at traffic lights on a hot day with an engine prone to overheating... and you can make multi cylinder engines smaller because you dont need fins in between. oh, and they tend to run that tiny little bit quieter. slightly more efficient with multiple cylinders as they can all have the correct mixture rather than running the center pair slightly rich...


anyone interested in buying build 1.0.2 if i make it? will include radiator and all your choice of combustion chamber volume/design, whatever!

Here's that sketch, it follows your idea of thermosyphoning but has a small reservoir and would have had a hand made radiator.

The motor doesn't have a problem as is except possibly when it is very hot out, I can feel the loss in power, and with a large 2 gallon tank the engine almost definitely overheated.

Originally the cooling effect would only be there if the engine and radiator temperature difference would be large enough. Since it will still heat water and the extra copper it does add higher thermal capacitance, and if (when) the coolant gets hot enough it will simply circulate since physics say so. It is vented at the 'recovery tank' and the head would have coolant pipes interference fitted on both sides. I hoped to run 2 lines to the radiator, one from each side of the head, this would self regulate the amount of coolant to each side in the loop since hotter temperatures heat water faster so it should flow quicker as a result.

I want to use 3/8 outer diameter copper tube interference fitted to holes drilled in the cnc cut aluminum head. The unique 3-tube design is simply to get more water to copper surface contact at the head for the most heat removal, at the exiting junction the temp will be basically uniform as it heads up to the radiator.

The reservoir would have very small tubing to act as an air escape at the top of the radiator and a second one to act as backup, while both will help to create a down force to help circulation behave a bit more stable. I would add an inch or 2 of tube at the very bottom of the radiator with a plugged end to trap particulate at the lowest point and will also serve as a drainage point.

Of course buying the head would be easier than making one. Though having both in and outlets on the top of the head doesn't seem ideal, I think heat and vibration could create trapped air/steam pockets somehow. Or just fight the general flow of water since purely thermosyphoning tends to be touchy of bends. It also requires cool water to up before comming back down to the surface to be cooled, which can further create a thermal lock (cool water falling on hot water wants to mix the 2 creating turbulence instead of cool water slipping under the warm or pushing it up and out.) That also makes one area disproportionately cooler while the outlet is going to be the hottest. If the 2 "lets" are next to each other then the hot water trying to leave cools down before leaving, and the cool water comming in heats up with heat that we are trying to remove... Placing in and out on opposite sides of the heating area works better, and putting the in and out at opposite sides along the direction of heat flow (generally upwards, and definitely upwards in our case) makes it more efficient.

Your head with forced cooling would work better since it doesn't rely on the magic of gravity to force it up then down and around and up again.

Edited to include sketch*

 

[gary55]

It would be nice if they just made one with cooling fins on it as well. Then if you lose your coolant oh well no biggy.

 

[Frankenstein]

It would be nice if they just made one with cooling fins on it as well. Then if you lose your coolant oh well no biggy.

That was also part if the effect of my head idea... With the fins still in good shape, along with the extra heat sinkage of copper, and depending on how I connect it, will act as a larger sink to trap heat, even if the coolant leaks out entirely the coppers mass pulls and traps the heat out away from the piston better, while the aluminum absorbs heat from the copper and quickly dissipates it to the moving air.

If there is direct copper solder joints from the heat collector to the radiator then the heat will naturally move into the connected work, and helping to cool the engine via air cooling on copper fins. This is the major advantage of a copper tube being interference fitted, the strong contact surface naturally wicks away heat from the head, the larger or greater numbers of connection the better the heat removal.

Most people who tackle performance desktop computers will begin to see more elaborate copper heat transfer designs to cool a cpu. Some are simple whereas others can be very complex aluminum and copper sheets and tubing like materials that jut off in dozens of places to maximize heat distribution to the cooling fins. In part this is like that design but intelligently incorporates a natural phenomenon of physics, and uses rising and lowering densities of a fluid to exchange heat more rapidly between the 2 heat sinks (head and radiator.)

It will also look motherf**king cool.

 

[Frankenstein]

Sorry I forgot to include my photograph of the sketch, please look at this:

 

[Steve Best]

You will need larger pipes. 1/2" or 5/8" to flow. One large pipe flows better than 2 smaller ones. The mass of aluminum will keep chamber temperature even. You might want to insulate your "up" pipe.

I like the sketch. I would think only one in and out port would be needed.

 

[Frankenstein]

You will need larger pipes. 1/2" or 5/8" to flow. One large pipe flows better than 2 smaller ones. The mass of aluminum will keep chamber temperature even. You might want to insulate your "up" pipe.

I was thinking of using a rubber or clear tube to reduce heat lost on the up cycle.

Half inch seems a bit large, I might Consider it but not convinced, I still remember playing with putt putt boats, a coil of copper tube over a simple tea candle, back then 1/8th inner diameter tube seemed plenty for strong water flow to make them work. 3/8 took cross a large pond like a tiny motorboat with fair speed with a large picnic burner.

At the moment I feel this is the appropriate gauge to work with, I'm dealing with a very small motor which already has most of the cooling surface worked out for it. The cooling system rarely needs to run at its max flow to keep the motor temp down. At its hottest points the airflow over the radiator should create a large enough difference that the water will flow with greater force throughout a smaller tube. Larger tubes have less surface contact per volume and and flow slower (normal hydraulics) and the larger tube further ruins the effect by storing and transferring more heat. A smaller tube makes the heat transfer more dependant on the liquid, and so greater temperature differences are seen. This further helps things move along.

I will further check out thermal capacity of volumes of water and change tubing if determined it would work better, though the temp that would be reached by these engines is pretty small, and so less loss to tubing as possible while also best heat transfer and "urge" to circulate with smaller tubing by making the engine heat up a small amount of water as opposed to what it can cool. If you pick a bigger pipe the more water must be heated before it starts to circulate, if you can't heat it fast enough (big volumes of water with relatively low surface contact cause this) then it won't be doing anything and it becomes pointless untill you are really cooking the engine. Some older water cooled motors use valves, and you turn off the water flow so you can bring the engine temp up to spec faster.

Come to think of it my truck uses a sensor to turn on the flow to the coolant, like most vehicles. Since I don't want a sensor I want to make it small enough to work almost as soon as the motor does, but not so big I need a sensor or valve to control how much heat I give away (cold motors ate no good.)

 

[Steve Best]

Most automobiles and trucks these days still use a wax thermostat valve to shut off the water flow during warm up and cold running because the pump is engine driven. Any of the sensors are for the engine computer to determine conditions for running (fueling). Thermosiphon engines typically don't use thermostats because their flow and performance is driven by temperature differential. The hotter the engine gets, the better they work. Pretty simple system but needs to be sized for demand. Tube sizes and rad capacity will probably require some experimentation to determine stasis.

The rubber or plastic hose on the "up" line would likely work well, but I think your ideas on tube sizing is a bit awry. Try it. See how it works.

 

[Frankenstein]

Most automobiles and trucks these days still use a wax thermostat valve to shut off the water flow during warm up and cold running because the pump is engine driven. Any of the sensors are for the engine computer to determine conditions for running (fueling). Thermosiphon engines typically don't use thermostats because their flow and performance is driven by temperature differential. The hotter the engine gets, the better they work. Pretty simple system but needs to be sized for demand. Tube sizes and rad capacity will probably require some experimentation to determine stasis.

The rubber or plastic hose on the "up" line would likely work well, but I think your ideas on tube sizing is a bit awry. Try it. See how it works.

I will ask one of the engineers who specializes in thermodynamics for their opinion, it's a lot of work to just screw up with the wrong size tube, but I have a feeling this would be an appropriate size..

 

[gary55]

thermodynamics, variations in conductive and radiant heat dissipation, thermal capacity of varying tube volumes, I remember when it was questionable weather or not to just leave the thermostat out in phoenix. Theory was the water past through to quickly to cool the engine. Didn't make sense to me. Hot and cold is hot and cold, and one chases the other. Damn sure left it in in Ohio. With heaters and defrosters the quicker the better. Are we still talking about china girls? I mighta had a beer or two, carry on. I'm still trying to grow a pair and jb my reed housing.

 

[HeadSmess]

when i did my little setup, i contemplated the idea of using a (rather large) fuel pump for water circulation. never actually tried it.

and of course, it would be nice to trim most of the cylinder fins away and wrap the whole lot in one big cylinder, weld it up and have the cylinder water cooled as well. seems a bit silly to only have the head water cooled.

for the amount of heat, i dare say thermosiphon is adequate, main issue is conveniently sized (and very effective may i add) radiators only have 8mm fittings, so thats the size pipe i used. for good thermosiphoning, bigger is better. you only need a very slight temperature difference to start the flow, but you have to minimise turbulence as much as possible. worth reading up on some technical stuff from 1900's era when the water pump was only just starting to be considered. it made the patternmakers job a lot easier, not having to consider gravity and thermal "traps". but, as long as cold water can be "sucked up" through the heating element (head) and "squirt out" at the top, it works fine.


meh, no idea what happened to all the files i had for the one i made, but i would have to draw it all out again anyway to suit the 80's stud pattern. biggest challenge was drawing it so a suitable size end mill would fit! and working with the lump of ally i had handy... finish that attachment i made for the lathe for cutting the hemispherical combustion chamber. another challenge, having an ideal hemisphere volume, and just the right length for the spark plug thread!


copper may sound like a good choice, but try and find the price of a 3 inch bar of it...and then try machining the crap! best to stick to straight ally. and two piece, dual metal heads are not ideal, theres a huge resistance built up between the joins of any two metals, regardless of how good they are at conducting, and how well they contact each other.


heat loss through the pipes is irrelevant also.

awry?

i havent heard that word since i used it in an english essay (for which i got 100%) and the girl that just had to get the top score in everything said it wasnt a real word...boy, did she kick up a stink.... and then shut the hell up when i handed her a dictionary

whats "AWE-REE"? so much for getting top marks, didnt even have a decent vocabulary!

the teacher of course, kept her mouth shut through all of this, just sat there smirking...