Because I designed and made my own heads with some own philosophy I wanted to do some tests. I did a test once making a vid on an idling engine but it did raise some questions. The nest test I wanted to do is using a steady heat source and a steady airflow to test on most equal conditions.

The idea is to make some slide in module to fill up a crankcase wit in the center a heater and next to it a temperature sensor to measure the heat on combustion room level and the airflow is done by a fast PC fan.


A 1st test of how hot the "engine" can become w/o any cooling, The heater is 12v rated running on 13.5v. 132 degrees Celsius is not bad.



My heatcam can not set a minimum and maximum temperature range and so the color palette adjusts itself to the automatic minimum and maximum temperature it detects. Taking pictures of cooling heads with different temps can be confusing if temperatures are different and so it color palette The minimum is just the room temp and is about 20 degrees. To fool the meter I did get a small so called PTC heater with a temperature of 110 degrees. Because IR devices have an issue with getting the right temperature on plain aluminium I did stick a small black cooling fin on it. As you can see the red cursor it detecting the hottest part and the white cursor which you can point your self is on the plain surface giving an almost 20 degrees difference. This is only a difference made by the meter, in real both parts have the same temperature. In the test there is also a plain aluminium OSMAX R2105 head used to see if it makes a difference compared with an anodized one. But the sensor inside the engine is always reading the right temperature.



The test is done in 2 steps, with the whole engine in a full airflow and with a simple body simulation where only the head is placed in the airflow. With all engines the "body" was set on the same height of 8cm.



This is all the equipment to test with, sadly the Rody R2101 head did not fit, it needed a slight smaller diameter buttonhead. Instead I used the silver looking OSMAX R2105 head. Again, keep in mind that the IR readings on this head are a bit off.....

The heads are:
- Novarossi Brave, which I see as the most standard design full size head.
- Novarossi Ielasi 9CL, which is the tapered or also called as an xmas tree shaped head.
- OSMAX R21 head, this one is from the stock 5 port onroad engine.
- OSMAX R2105, it is a less fin lighter version of the R21
- My own Tooms Tuned head
- On its way but not in yet is the Hasi Tuned head with straight up fins, I will test that one later.



The 4 colored anodized heads within a same color scheme range with a full engine cooling w/o a body simulation.



Funny to see is that again my head is full colored and the others are on top a little bit cooler as my movie did show..

Oh, the test is actually very simple, just power it all up and wait until the temperature is stabilized.

Here pics of the heads with the heatcam and the display of the internal sensor.







Funny to see the Novarossi heads are in disadvantage but because the whole engine is in the airflow it is not realistic.

So now with the body simulation:








So the temperatures within the combustion chamber are higher which is logical. With the lighter color crankcase you can see it is missing some cooling. And yes, normally the chassis plate will also provide some cooling. But good to see my head is showing the best results.

that is some awsome testing Roelof

I have been thinking. The movie I made a while ago:
The design of air cooling must have fins. Fins does enlarge the total surface that is in contact with the air and the effect of fins to transfer heat to the air is much better than a flat surface (which my head basically has). Most heads have thin fins that gives a kind of resistance for the heatflow, that is why with the normal head the outer tips and most on top stays cooler because the large distance from the heat source and ambient temperature has also more influence at the farthest spots . My head with a bit like closed in fins and the fins are much thicker and shorter to the heating source will take up much easier the heat and does not transfer it to the air that easy.

But with an airflow it will make a difference. With the normal heads the airflow can bend off from the head. The tree trunk on the middle of the head can bend off the airflow so there will be less airflow on the rear side of the head as it also gives a blind spot behind the tree. My head has actually tunnels. When the airflow comes in it is forced to go all the way through, only the small plughole and the screw holes are the only unwanted spots where the air can flow out but there is no kind of "ramp" in the head that some of the the flow is directed to those holes so the loss of airflow is low.

That theory can be proven The Windsible Is a Tiny Wind Tunnel for Your Desktop

Made the airflow visible....

So is the R2105 the most effective head? I wonder if the older Magnesium heads offer anything?

Commercial available yes. Probably because the R2105 is not anodized or different anodized. Plain aluminium should work better than colored but black surfaces seems to work a fraction better above lighter colors, not sure about a compare with non-anodized. I do not think I can measure that alhough I have my head in 3 different colors, red, blue and black.
There is also the question which types of aluminium are used. My head is made of a 2000 kind of aluminium which contains more copper giving a slight better heat transfer.

Cooling head airflow can act like an extra wing, generating a little downforce on the motor...

R2105 would definitely be the best option

Yes and no...

The R2105 head is for sure lighter than all my other heads and is giving a slight better cooling but because it it is not colored you will get completely different temperature readings when using an IR meter, it can be more than 20 degrees C lower than with a colored head. If you are a guy focussing on the temperature you know from other OS engines then you will ruin the engine.

Very cool testing.

I wonder if there is a way to CNC a head with some fins to direct the wind to dissipate heat out and some fins directed down to cool the crank case. This is super interesting stuff for sure.

With advanced software any heat transfer can be shown in a simulation based on some CAD design. Then you can also simulate different materials like magnesium and copper but aluminium is price and weight wise the best choice of material although there are many different kinds of aluminium with also different heat transfer specs. In theory the 2000 kind of aluminium is the best due a higher grade of copper.

Like my head with a so called tunneled airflow is seen more, a good example is a car radiator and the better brands processor cooling but also my HarmanKardon TC600 car amplifier has a tunnel shaped heatsink cooled by a fan blowing air through the tunnel.

Crankcase cooling is a bit done by the chassis plate, it is not the spot where the heat is generated but it does help a bit in the total cooling of the engine. The actual heat is generated within the sleeve with has the best cooling contact to the head due the 4 tight screws holding the head. I can remember the 90's with the 235mm 2wd class, back cover cooling ribs became popular but we also noticed it could give tiny issues with tuning. For a part the crankcase temperature does help vaporizing the mixture and so to cold or to variable cooled can have an influence on the tuning..

Roelof

What do you think it is that drives the design of the modern cylindrical heatsink?

Have you tried to put between the cooling head and the combustion chamber where they touch thermal compound like MX-6 or PTM7950.

they make wonders improving the thermal conductivity between the CPU’s and coolers.