Prototype idea..... A sand + water glass riser with a 55 gal barrel.

Idea for a riser for the rocket stove.

I have been thinking of how one could build a rocket stove riser outside as cheap as possible, and I had an idea I wanted to just throw out there and see what kind of reaction it will get.


Idea:

1) take a 55 gal barrel with the bottom cut off and open top, you now have a tube.

2) Place a 6 inch PVC tube down the center of the 55 gal barrel now pack with sand to top.

3) Now this is the tricky part,  remove the PVC and now dig under the barrel meeting up to the hole in the sand ( might bury another PVC 6 inch under before ) packing sand in barrel.

4)  The thought is now we have a J Tube created out of sand,   remove the PVC pipes,     Next coat the inside of the sand with a spray bottle with water glass so that it keeps its form..

5)  Light and see if it holds together.


I have also been thinking about doing this with a paper inside, putting clay around the paper tube then sand around the outside of that....


Just wild ideas but I figured that it should be very cheap to test.

My gut feeling is that this idea will introduce too much thermal mass into the riser, and will suck too much heat out of the flame.

Paul Wheaton uses ceramic fibre in his risers because is has a very low thermal mass and high heat tolerance.

Mike Lang wrote:My gut feeling is that this idea will introduce too much thermal mass into the riser, and will suck too much heat out of the flame.

Paul Wheaton uses ceramic fibre in his risers because is has a very low thermal mass and high heat tolerance.

Thanks for your input.

I would agree if sand  served as thermal mass.      Sand does serve as thermal mass when wet, but when dry it serves as an insulator.

I also use ceramic fibre in my aircrete stove, the results are impressive.    

Ok, How about making a dakota fire hole in a 55 gal barrel?  


This would be like a pocket rocket, only filling the inside with dirt....

If your riser is the size and weight of a 55-gallon drum, filled almost entirely with sand, how big would your radiator and manifold need to be?  If I'm understanding your premise correctly, you would be building on a much larger scale, but still only have a 6" diameter system.  

Jason Broom wrote:If your riser is the size and weight of a 55-gallon drum, filled almost entirely with sand, how big would your radiator and manifold need to be?  If I'm understanding your premise correctly, you would be building on a much larger scale, but still only have a 6" diameter system.  

Right now I am just looking at alternative building materials for the Rocket stove, and brain storming.      


The Dakota inground is a proven design, I was just thinking if we made a pocket rocket with this design with sand or clay  or cob, or for that matter perlite what would be the advantages / disadvantages.


I am considering mixing a batch of cement with sodium slicate and cast myself a  chimney to test with, probably best to make a small test sample to see if my idea has any merit.

I like the idea of putting soil in a container to form a rocket type stove.
It lends itself to use with a  sacrificial form.
The soil could be protected from moisture, becoming dry,  insulative, maybe even fired,  if enough clay is present.
That Dakota design is more of a U than a J,  and he  doesnt even try to light the fire at the bottom and have it burn sideways.
The combustion air still feeds through the coal bed, and the earth can be insulative.
I wonder if you could build a deep U in a barrel and somehow keep the soil in one side wet,  and the other, dry.
The wet side could be a feed tube,  kept cool by evaporation, and the dry side the insulated riser.
If the cooling worked, we could feed long branches without the draft running the wrong way.
I have built a J with a water cooled feed, and it kind worked,  but the water jacket was way too small.

It occurred to me to look at salt glazing as a way to solidify the insides of an earthen rocket.
Turns out that salt glazing is waterglass!
High silica clay, high heat and salt water are the simplified ingredients.
Soda (sodium carbonate) works as well as salt and is less toxic.
I am seriously wondering if we could glaze the inside of a rocket just by adding soda  to the feed tube/riser at the right time, or impregnating the inner walls with it.

--William BronsonI like the idea of putting soil in a container to form a rocket type stove.
--It lends itself to use with a  sacrificial form.

The more I think about it, it does not have to be a 55 gal drum one could use a circle of rocks like a raised bed to make the dakota stove off the ground.


--The soil could be protected from moisture, becoming dry,  insulative, maybe even fired,  if enough clay is present.
--That Dakota design is more of a Unicorn than a J,  and he  doesnt even try to light the fire at the bottom and have it burn sideways.

I agree, more like a TLUD than a J tube.

Perhaps a blower with varable speed could be used.

I also thought about putting a chimney on the Dakota.


--The combustion air still feeds through the coal bed, and the earth can be insulative.

Ashes also insulate as well, as I learned from Erinie Wisner


--I wonder if you could build a deep Unicorn in a barrel and somehow keep the soil in one side wet,  and the other, dry.

Last night I was watching a video about heat pipes, and I was thinking they could be useful in keeping the feed tube cool, or...    creating more pressure in the chimney by creating a differential between the heat and the cold, thus causing the air to move more quickly.


--The wet side could be a feed tube,  kept cool by evaporation, and the dry side the insulated riser.

Yes, I saw someone make an evaporator out of air crete   using water mist and a blower......


--If the cooling worked, we could feed long branches without the draft running the wrong way.
--I have built a J with a water cooled feed, and it kind worked,  but the water jacket was way too small.

Interesting, I would like to see pictures of that.

--It occurred to me to look at salt glazing as a way to solidify the insides of an earthen rocket.
--Turns out that salt glazing is waterglass!

I did not know that :-)



--High silica clay, high heat and salt water are the simplified ingredients.
--Soda (sodium carbonate) works as well as salt and is less toxic.
--I am seriously wondering if we could glaze the inside of a rocket just by adding soda  to the feed tube/riser at the right time, or impregnating the inner walls with it.

Yep we have yet to hit the surface of the billions of possibilities of a rocket stove.

Billions of possibilities... and most of them will work worse than the standard designs. Not to discourage experimentation, but I think it would be productive to understand some of the fundamental parameters for design and materials, and not violate those, as doing so would definitely not be an improvement. Things like high mass risers fall under that category. We know that, all else constant, a low mass, insulative riser will work much better than a high mass, not very insulative riser. (Sand is too insulating to make a good heat exchange mass, but too dense and conductive to make a good insulator.)

I think it is a fundamental practice in experimenting to fully understand the characteristics of the standard before trying variations. Have you built a dead standard J-tube core for a baseline?

Exploring possibilities on paper or pixels seems like a 100% good to me.
I have built a TLUD, bucket rocket, pocket rocket, batch rocket, and a J, none of them to specifications, all of them as a riff on the "right " way to build.
I read and asked questions about the device involved, and made the adjustments that made sense to me.
Stainless steel stock pot for the body of the TLUD.
Calcium aluminate cement for the casting of the bucket rocket.
Slab built the walls and floor of the batch box.
Water jacket for the feed of the J.
No regrets.
Works for me.

I made two pound cakes recipe for the first time  the other night.
Baked them in two different ovens, and i substituted  1/4 of the flower by weight for cocoa.
They were a little dry buyut still tasty.
I had never made this recipe before.
I will never make it that same way again.
Bog standard design does not move me.
No regrets here.

I not sure about if this is what Mart intended , but for me this thread was an invitation to riff on cheaper, lower tech ways to achieve decent results.
I am glad others before us strayed from the beaten path, or we would have no established designs to stray from.
The very idea that a J is place to start from is doubtful to me.
A batch box is more stable at smaller sizes.
A bell offers lower resistance to flow than a bench.
45 minute burn times makes it easier to tend.
8 and 10 inch batch rockets  burn longer pieces of wood than the equivalent sized J's, with no chance of reversing draft.
Shorter risers make for lower cooking surfaces.

But advocates for the J point out how cheap and simple building one is.
No need for a door, or carefully measured burn chambers.
Simpler to operate.
Just bricks, barrels, mud , sand straw, barrel , and ducts.

Why don't they all use ceramic fiber insulation?
It is clearly superior to any other matirials, unless you consider other things,  like, health risks , availability or costs.

The J-tube has been measured by those with the expensive professional equipment as being as or more efficient than a batch box, and is easy to build, even without exotic materials. Insulating the core very well, even if the hot face is hard firebrick, is paramount.

Anything that does that job is fair game for experimenting, and different locations will find different materials cheaper or more available, if that is a primary criterion.

The heat exchange mechanism is entirely separate from the combustion core, so ducts vs. bells does not factor into the core experiment comparison. The various considerations all have their places for different use cases. The point about starting with a standard design is that you can't know whether an experiment is better or worse without something to compare it to.  Ordinary people don't have the equipment to make accurate measurements of efficiency and pollution, though if you have the experience of a standard design, you may be able to detect if something gives better or worse results than the standard method.

I used an experimental method for laying out the casting sections for my J-tube core, while making the interior shape exactly to specs. One variable at a time.

Glenn Herbert wrote:Billions of possibilities... and most of them will work worse than the standard designs. Not to discourage experimentation, but I think it would be productive to understand some of the fundamental parameters for design and materials, and not violate those, as doing so would definitely not be an improvement. Things like high mass risers fall under that category. We know that, all else constant, a low mass, insulative riser will work much better than a high mass, not very insulative riser. (Sand is too insulating to make a good heat exchange mass, but too dense and conductive to make a good insulator.)

I think it is a fundamental practice in experimenting to fully understand the characteristics of the standard before trying variations. Have you built a dead standard J-tube core for a baseline?

Hi Glen,  

Thanks for your input.      

As what link would give me a "Standard" J tube?

I have just finished reading the book "Rocket Mass Heaters Super Efficient Wood Stoves You Can Build"  by Ianto Evans and I have seen a gazillion videos on youtube about building rocket stoves.
I have the first 5 DVD set of building rocket stoves from permies,  but these do not have plans, but give great theory.

About a week ago, I purchased the Ebook  "The Rocket Mass Heater Builder's Guide",  I have glance thru but I have not read.

So I am posting here not so much to build an "exact" rocket stove, but to understand what is the reason people have chosen  lets say cob,  verses fire brick,  verses  ceramic fiber board, verses other materials.

So yes, I am looking at multiple standards of building a rocket stove, from Permies, and Honey Do Carpenter, and Several others from youtube land.


I am starting to appreciate the J tube more as I am starting to understand the "why" they build it the way they do.     I am starting to question is it really more efficient than say an L design?    

I think it would be a great idea to test a J tube verses an L design and see is there  more heat with either design?     Or say Peter's batch box, is it more or less efficient  than a J tube?

I just have questions like any 2 year old, I ask why a whole lot

Jason Broom wrote:When I was about 8 years old, I am told I went to my mother and said to her, "There's two ways to learn something; by making your own mistakes or by learning from the mistakes other people make.  I'm going to try to learn from other peoples' mistakes."

The last book you purchased, and flipped through, is the one to read.  Ernie has made ALL of the mistakes, and learned the most from them, in my estimation.  Erica articulates the lessons learned from those mistakes in a very effective manner.  Their approach is logical, pragmatic and easily understood, making their book the de facto standard, at the moment.  Others have made huge contributions; Donkey from proboards, Broaudio on Youtube, and Peter Vandenberg, at many different levels.  These folks have collaborated and collated their knowledge at Paul Wheaton's gatherings.

All of this to say:  The knowledge of how best to build a rocket stove (for cooking) or a rocket mass heater (for heating) is already well-established.  Any prototype you are likely to undertake has already been done, found to be lacking in some way, and/or improved to the point where it is now part of the commonly accepted design concepts around these devices.  It is frequently suggested that folks who are new to these concepts build a rocket stove or RMH "by the book" before venturing into new ideas.  That is so you better understand how and why these devices work by learning from others.

Or, you can revel in the fact that we live in a free country, with an embarrassment of riches, where you are at liberty to make all of the mistakes yourself, and learn from them to  your own edification.  

Hi Jason,

Thank you for your input.

William Bronson wrote:Exploring possibilities on paper or pixels seems like a 100% good to me.
I have built a TLUD, bucket rocket, pocket rocket, batch rocket, and a J, none of them to specifications, all of them as a riff on the "right " way to build.
I read and asked questions about the device involved, and made the adjustments that made sense to me.
Stainless steel stock pot for the body of the TLUD.
Calcium aluminate cement for the casting of the bucket rocket.
Slab built the walls and floor of the batch box.
Water jacket for the feed of the J.
No regrets.
Works for me.

I made two pound cakes recipe for the first time  the other night.
Baked them in two different ovens, and i substituted  1/4 of the flower by weight for cocoa.
They were a little dry buyut still tasty.
I had never made this recipe before.
I will never make it that same way again.
Bog standard design does not move me.
No regrets here.

I not sure about if this is what Mart intended , but for me this thread was an invitation to riff on cheaper, lower tech ways to achieve decent results.
I am glad others before us strayed from the beaten path, or we would have no established designs to stray from.
The very idea that a J is place to start from is doubtful to me.
A batch box is more stable at smaller sizes.
A bell offers lower resistance to flow than a bench.
45 minute burn times makes it easier to tend.
8 and 10 inch batch rockets  burn longer pieces of wood than the equivalent sized J's, with no chance of reversing draft.
Shorter risers make for lower cooking surfaces.

But advocates for the J point out how cheap and simple building one is.
No need for a door, or carefully measured burn chambers.
Simpler to operate.
Just bricks, barrels, mud , sand straw, barrel , and ducts.

Why don't they all use ceramic fiber insulation?
It is clearly superior to any other matirials, unless you consider other things,  like, health risks , availability or costs.

Yep,  I am starting to wonder if this thread really belongs in an "Experimenters"  section.

Indeed you don't want to confuse newbies with all our our wild ideas,  most just want to build a rocket stove to spec, and I can see the need to keep our thread away from the newbies.

Likely it won't work if your stove is producing good heat.  The heat riser in a properly operating stove experiences near the maximum theoretical heat for a wood fire.  The first problem is melting points.  I will hit google lightly here and not dig.

First how hot is a wood bonfire?  Notice the just over 2000 degree F temperature listed.  That is still significantly below the theoretical maximum for an air/wood fire.(people have listed it in other rocket mass heater stories and I know from having tried to google it that I am not asking the right question  That said for some reason 2480 sticks in my head but don't quote me on and if it wrong someone please correct me.)  

Then look up the melting point of water glass.  There you get begining melt at 1490 degrees F.  Notice you are 500 degrees lower than what can be done in an open bonfire.  Will act like a bonding agent when it is 500+ degrees past its melting point for even a wood bonfire?  Unlikely.

But lets pretend it will hold together

Type of sand matters too.  Some sands will melt below this too.  Many common sands have melting points in the 1800 to 2500 degree range depending on their chemistry. Notice that if you view a bunch of videos on riser autopsies that the ones reinforced with fiber glass the fiber glass actually melted in some of them.  On the other hand if you were say talking aluminum oxide sand blast sand it has some of the highest melting point known to man.  Type of sand really matters.

That said you are still missing the point.  Say everything held together.  You are still violating heat riser design.  You are thinking of sand as an insulator and in a sense it is.  But for these purpose it is a very poor insulator.  It is way to much mass and not enough insulation.  A heat riser needs to be low mass so it heats up quickly and it needs to be high refractory so it reflects most of the heat hitting it back into the riser.  This is why all the recipes for building heat risers include vermiculite and it is also why many of them include things that will burn or melt out leaving tiny air pockets behind.  Go look at the insulation bricks for an electric kiln.  Notice they are incredibly light.  This is more what you are going for.  The problem with them is they are soft and not very durable and a riser is basically a sand blaster nozzle.  So it needs to be a bit tougher stuff.  And this is what everyone struggles with.  Incredible heat limiting material choices, need for low mass, need for highly refractive yet tough and wear resistant is a tough combination to find.   Then it still has to hold together thru repeated thermal cycling over a wide range of temperatures with uneven and fast heating.

As for casting the burn tube and riser in one operation simply remember that you see major temperature differences in operation between various parts.  The reason someone isn't already doing it commercially is these different expansion contraction ratios will crack things.  By having joints you have places for things to make small movements.  But monolithic casting don't have that give and crack over time.

Keep researching.  This info is repeated dozens possibly hundreds of times as people get thru the learning curve.   And this is a subject that virtually nobody understands quickly.  And the people who really know it simply get tired of boosting everyone thru the same mistakes when the information is out there if you read long and deep.

Personally I think I have given up most of my early silly ideas and slowly grown to a better understanding.  The exception is everyone saying vortex chambers at the bottom of the heat riser don't work.  I still want to try building a thien baffle into the base of one thinking it might work.(the goal being to keep all dust out of the horizontal chimney section)  But I recognize odds are they are smarter thru experience than I am and it is likely going to be an expensive failure.  From my understanding of burn tube / heat riser physics I still don't understand why it won't work.  But that probably simply means I have more I need to learn.  I encourage you to keep learning.

Well,  I guess it would be better if we didn't speculate here, since the things we are talking about violate the principles set before us by the experts.

A few reasons to use a J-tube instead of an L-tube for an indoor RMH:
The extra 90 degree bend in the J introduces turbulence which helps mixing.
The vertical J feed is self-feeding as the fuel burns at the bottom, while the L needs to be fed manually as the wood burns.
The J has been measured to typically have more efficient combustion than the L.
The J is more resistant to sparks flying out and possibly landing on the floor.
The J is easy to cover with a couple of bricks to fine-tune the air supply.
The J is more convenient to feed and inspect, if the openings of both are at the same low-ish height off the floor.

On the other hand, the L can have a longer feed chamber and burn more wood at once.
You can even feed very long logs or sticks into the L, which is not generally safe to do with a J. These characteristics are conducive to use in high heat quantity applications like kilns or maple syrup evaporators.

On another point, the heat riser does not actually get any amount of abrasion, and soft materials like insulating firebrick or ceramic fiber blanket have held up well, and given the very highest performance. The burn tunnel does get some abrasion, and the feed tube can get a lot of abrasion depending on use, so strong materials there are important.

Glenn Herbert wrote:A few reasons to use a J-tube instead of an L-tube for an indoor RMH:
The extra 90 degree bend in the J introduces turbulence which helps mixing.
The vertical J feed is self-feeding as the fuel burns at the bottom, while the L needs to be fed manually as the wood burns.
The J has been measured to typically have more efficient combustion than the L.
The J is more resistant to sparks flying out and possibly landing on the floor.
The J is easy to cover with a couple of bricks to fine-tune the air supply.
The J is more convenient to feed and inspect, if the openings of both are at the same low-ish height off the floor.

On the other hand, the L can have a longer feed chamber and burn more wood at once.
You can even feed very long logs or sticks into the L, which is not generally safe to do with a J. These characteristics are conducive to use in high heat quantity applications like kilns or maple syrup evaporators.

On another point, the heat riser does not actually get any amount of abrasion, and soft materials like insulating firebrick or ceramic fiber blanket have held up well, and given the very highest performance. The burn tunnel does get some abrasion, and the feed tube can get a lot of abrasion depending on use, so strong materials there are important.

Thanks Glen,

I have made adaptations to my L design by breaking up ceramic tile and put the small chunks on top of the bottom to protect the base,   it has been working for me well so far, and gives me a nice smooth surface to put the wood on.

I have found on my rocket oven that small wood to get the fire started then two large chuncks gives me a long burn and right at the right temp for a single load.       It does seem to be a bit of art form to learn the rocket oven, but I do love how little wood it burns

I have put a chunck of railroad tie inside the oven directly over the place the flames hit the side of the barrel.    This has given me a much more even temp for baking inside.      It is a wonderful device once you get the hang of it with the fast heat up.