Small Boiler

[jdurand]

Problem with small boilers is surface to volume ration is bad. A lot of radiating surface with a small volume.

[editor@glue-it.com]

Problem with small boilers is surface to volume ration is bad. A lot of radiating surface with a small volume.

Yes, and I've made this worse by making the length versus diameter quite large - the opposite of a sphere...

However, when I "turn the wick up" on the burner it gets hot very rapidly. So, a lot more work to get the balance of this design right.

[editor@glue-it.com]

Making some big assumptions I've made a very simple mathematical model of the boiler


Then using the cooling curve data I've estimated the heat loss to be 35W at the operating pressure of 40psi (140°C).

This then allowed me to look at estimating the heat input from the first design of the methylated spirit burner. It is a bit variable, but probably expected as it uses a chicken feeder and there is some air flow.



The heat input for this first design is around 54W. This is why I was seeing a very marginal capability for this boiler.

[editor@glue-it.com]

I've now run a few experiments to understand the optimum distance between the spirit burner and the boiler. Normal practice is that the hottest point is at the tip of the inner blue flame, however, there is a depth to the boiler surfaces as shown by this diagram.



So, using the simple assumptions, a new firebox and some more measurements on the instrumented boiler I get the following curve:



This plot shows that the peak in heat output is when the burner is around 85mm from the base of the boiler. Also, it shows that this firebox and burner is producing nearly double the amount of heat that the original design generated.

I checked this calculation by looking at the boiler temperature versus time data. The original firebox and burner took just under 8 minutes to take the water from 60°C to 100°C. The new firebox and burner took just over 2 mins 30s for this same temperature change. You have to consider heat losses and heat input, but this makes sense.

[editor@glue-it.com]

I wrapped the boiler and electrically heated it so I could estimate the heat loss. This way I could then compare unclad to maximum cladding.



At the designed operating point of 40psi and 140°C bulk water temperature the heat loss unclad is ~35W and with maximum cladding it is around 12W.


The cladding is a lightweight fleece, then a aluminium coated bubblewrap.

This has made me realise that I could make a low voltage miniature immersion heater - I will post something on this.

[editor@glue-it.com]

I think I now have most of the parameters for an optimised design of the firebox for the small vertical boiler. At last I can finish it and get back to the traction engine. So, I ordered a bar of 2" mild steel hex. However, when it turned up I just couldn't turn it into swarf....



So I thought I could make the firebox from sheet metal, trouble is my sheet metal skills are not that great.



However, I'm quite happy with this, now to add a door for the meths burner and some louvres for air.

[editor@glue-it.com]

I've been struggling to get heat into the boiler. So I instrumented it to understand the heat losses and to optimise the burner location.

Then I made two types of steam raising blowers.

Left: compressed air jet Right: electric suction fan

What is interesting is that both give a similar performance. Below is my calculated heat input to the boiler versus time for the two blowers and you can see them switched on, off, on and then finally off.


The heat input to the boiler using the simple methylated spirit burner increases from ~50W to ~350W with either blower on.

[editor@glue-it.com]

A video describing the blowers and the data

[editor@glue-it.com]

Finally I'm back to the 1/20th scale Burrell traction engine. My new problem to solve and design is the cylinder block.

I have an idea for creating a plate that is silver soldered to the boiler and forms the joint. The cylinder is then mounted one side and the boiler the other.



Me offering up the Stuart 10V that is 1.5x too big at 3/4" bore and stroke. The Burrell will be 12mm bore and stroke.



This bracket would be easiest made from brass, however, it's best to avoid using brass due to dezincification.

[editor@glue-it.com]

Back onto the traction engine and whilst waiting for material for the cylinder block I roughed out a steering wheel.



This is 22mm in diameter (~7/8")