Boiler Insulation

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

glue-itcom

Well-Known Member
Joined
Apr 10, 2013
Messages
216
Reaction score
298
With the tiny vertical boiler I made I managed to estimate the thermal conductivity of the multi-layer insulation:
  1. Teflon sheet – a thin baking sheet wrap
  2. Medium weight paper wrap
  3. Polythene – this was a thick clear polythene bag
  4. Medium weight paper wrap
  5. Air gap – created using thin wood veneer strips
  6. Bog oak outer layer
This gave me an estimated thermal conductivity of 0.06W/mK

small-boiler-cladding.jpg


This insulation had a reasonable thickness of around 4mm, hence the overall insulation was rather good.

With the miniature traction engine I'm making I need to shrink this insulation down to ~1.5mm in thickness. The multi-layer insulators that were originally design to keep liquid hydrogen cold are interesting as the thermal conductivity is several orders of magnitude better than this. However, you still need enough space to isolate the layers.

So I have made a new boiler thermal test object to try some different approaches.


I have just tested:
  1. Teflon
  2. paper
  3. aluminium foil
  4. paper
  5. aluminium foil
  6. paper
  7. Brass outer wrap
This looks rather good, just need to run some calculations and I will post the results.

I'm thinking of trying perforated paper and hence add air pockets and reflective foil. Also, I need to try a thin sheet of silicon.

However, I would be interested in ideas for testing.
 
You might want to consider a material called "FibreFrax". It is a direct replacement for asbestos, and comes in thin sheets, and also in pads of different thicknesses, for wrapping around a boiler.
Although the pads look very puffy, when you wrap a small locomotive boiler with them, and then put the outer brass shell over them, the material squishes right down and becomes a very thin layer.
 
You might want to consider a material called "FibreFrax". It is a direct replacement for asbestos, and comes in thin sheets, and also in pads of different thicknesses, for wrapping around a boiler.
Although the pads look very puffy, when you wrap a small locomotive boiler with them, and then put the outer brass shell over them, the material squishes right down and becomes a very thin layer.
Hi William, I have some ceramic blanket material for my kiln, will have a go. Thanks, Nigel
 
I've done some initial testing and repeat testing to see if the cooling curve is repeatable.

unclad-test-repeatability.jpg

and it's looking rather good. I can work with this.

As you can see, each test is around 2 hours by the time I've set it up, heated it up and then it's cooled down to below 60°C - so all of the tests might take some time.....
 
In the meantime and whilst I'm waiting for things to cool down ;) I have been looking at the modelling of this lumped boiler test object. OK, it's a very simple model, but it's still worth listing out the assumptions:
  • Lumped point mass.
  • Single temperature measurement represents the temperature of the object.
  • Heat capacity does not change with temperature.
  • Local air temperature represents ambient conditions.
  • Local ambient conditions are uniform throughout testing.
  • Heat is only considered to be lost through ends of the object and copper surface.
  • Heat lost through the end is just a function of object temperature and not test insulation.
  • Mass and heat capacity of the object is just the mild steel core and copper shell.
 
The unclad versus maximum (is it maximum?) insulation has now been tested and the data shows a significant difference. I think enough to bookend the problem. Quite frankly my test object would be better if it was about 300mm long to make the losses at the ends less significant. However, for now I can test a range of insulation between these bounds.

unclad-vs-max-insulation-heat-output.jpg
 
Kaowool is good and often used on models, comes in 1.5mm sheets.

I'd be worried about your paper taking on moisture or even water and becoming a soggy mess
 
Kaowool is good and often used on models, comes in 1.5mm sheets.

I'd be worried about your paper taking on moisture or even water and becoming a soggy mess
I'm testing some ceramic fibre matting, will plot this against the other things I've tested. The issue is that any ceramic based material tends to break down and go to powder over time. I've been using a teflon sheet as a first wrap. I will try silicon. I assume I can buy kaowool from any good model engineering supplier?
 

Latest posts

Back
Top