This recipe was originally released on /r/Floathouse. Many thanks to Michael Eliot and Andy Thomas for releasing it. Now, on to the instructable!

What is Geopolymer Concrete?

The term 'geopolymer' can be confusing because when we hear the word we are used to thinking in terms of plastic. But 'geo' refers to rocks, as in 'geology,' so what's actually being referred to is the polymerization of rock-based materials, which is a very weird concept.

What happens when you mix a batch of geopolymer cement is an alkali activator literally breaks down the chemicals of an alumino-silicate flyash material then rebuilds it in long polymer chains, basically stone polymer. And when it sets it's as hard and strong as a good concrete, if not harder, and much more flexible than most concrete, by several times. This makes it crack resistant as well.

The alkali activator is liquid lye prepared with water. Lye is often used in making soap, or pretzels. It's considered a dangerous, corrosive material, but handled right it's about as dangerous as making soap, which anyone can do. It's quite cheap too, I was able to buy 10 pounds of pure lye for ~$30. I didn't notice any noxious fumes coming off it, but best to mix this stuff in a well ventilated area as well.

With the lye solution we add a chemical called waterglass, which can actually be made from lye if you're willing by heating it considerably.

If you don't make it yourself it can be a bit expensive in small quantities and is probably the most expensive component of geopolymer concrete. But it's chemically better to make your own fresh waterglass from lye, it results in better geopolymer cement.

After this the geopolymer needs to be heated for the next many hours. About 24 hours at 85° up to 4 hours at 200°. During this time it doesn't need to be kept wet, like normal concrete, and is in no danger of curing too quickly and cracking. It will not off-gas water either, it actually incorporates water into its chemical matrix after splitting it into oxygen and hydrogen.

This is the way to make type-F geopolymer concrete, which is low-calcium, and low-calcium is the key to seawater resistance.

There's also a type-C geopolymer formula useful for landed applications. When water touches calcium compounds the result gives off heat. This is known as the heat of hydration in cement, and is what cures regular concretes.

Well people like type-C geopolymer concrete because it's quite similar to Portland, it doesn't need heat to cure--it generates its own heat.

What's more it cures quick rapidly, but doesn't begin curing until you give it the alkali activator.

So what some have done is mix up a great deal of wet and proportioned fly ash and aggregate in a cement truck, drive to the pouring site, mix in the final alkali activators, let them mix a few minutes, then pour like any concrete.

Geopolymer pours fairly loose typically, and conforms well to molds and shapes.

Why aren't geopolymers being used more widely right now?

One of the biggest reasons is the innate conservatism of engineers. We have a lot of experience with concrete, geopolymers are fairly new. The ASTM standard on pure geopolymer concrete only came out about a year ago.

Many people will need to do small projects with the material to gain experience and wisdom, supply chains will need to be built, etc., before we see the next freeway overpass being poured in geopolymer cement.

More information on geopolymers at the opensource ecology wiki, or at wikipedia.

Step 1: Ingredients, and Notes

I have not yet perfected the geopolymer formula, though I have learned a good bit about what to do and what not to do. I plan to put these into a short monogram and release it for everyone to try.

It was very difficult for us to discover the formula but I'm quite willing to share :)

Let me dig out my notes here...

These are the proportions by weight for our geopolymer concrete that tested out at ~5,000+ PSI. These proportions are for a 6,000 grams batch.

This is the only dangerous step in making geopolymer concrete, and it's about as dangerous as making soap, which also uses lye.

101.8 grams of 14-molarity solution Lye (sodium-hydroxide).

(This means 41g of lye and 60.7 grams of water). Be careful when mixing this together. Start with a plastic cup of water, 60.7g of it, and then add about half the lye. It will heat the water almost to the boiling point. If you see bubbles forming that's okay, just stir and let it cool. Once it has cooled a good bit, say 5 minutes or so, add the rest of the lye and stir until it dissolves as well. If you dump in all the lye at once it can boil and sputter and send caustic lye back at you, and it will burn you. If it burns you, wash the spot with water for 10 min. And be careful, because lye can burn your skin in such a way that it will do damage long before you feel any pain, so be careful.

255.7 grams of Waterglass (sodium-silicate).
15.15 grams of superplasticizer.

(Geopolymer concrete turned out to be plastic enough on its own that we omitted this from future batches as unnecessary. It's generally fairly loose. This is one of its problem! Makes it hard to prepare for spraying and plastering, but perhaps with the addition of nylon fibers it can be made thicker.)

1848 gram of mixed aggregate (sand and 7mm gravel).

One point on this, we began omitting the rock and using pure sand and still obtained a high strength value, but I suggest you play around with the ration of rock to sand and try to find a good medium point. We cut back on aggregate compared to the first pour because the first pour was extremely rocky and wouldn't even fill the mold we had. The first pour had 1715g of rock and 734.3g of sand. This mix with all sand and no rock came out very beautiful and strong, but it could be made stronger with some rock most likely. This would be a good thing to try out. Also, this rock and sand should be measured out at its wet-weight, not dry weight. So make sure it always has some water in the bag to keep it hydrated. Otherwise dry aggregate will suck water out of the alkali-activator and possibly cause a failed pour when you begin to mix them together. One more note, do not use beach sand, you want some kind of granite-sand or mason-sand. Don't use beach sand, it results in significant strength loss.

1013g of type-F, low-calcium flyash.
41g of water.

One thing we learned was to not play around with the water ratio. You can't make geopolymer thicker or thinner by adding or taking away water like you can with normal concrete. Instead this will cause the chemistry to fail. The chemical ratios have to be kept fairly consistent. That's why I say try nylon fibers as a thickener rather than trying to play with water ratios. We did a lot of playing with water ratios and had a lot of failed pours that failed to set-up.

Pam Cooking Spray (or similar)

Step 2: Mixing Process

Measure out and combine the damp aggregate (sand, rock) into a plastic bucket (do not use metal bucket).
Measure 41g of water add it in. Mix the sand and rock for several minutes until everything is well uniformly wet and mixed using a mechanical stirrer of some sort.
Measure 60.7g of water, put into a plastic container.
Measure 41g of solid lye pellets. Don't leave these standing in the air too long because they will absorb moisture from the air and become gummy.
Pour about half of the lye into the water and mix with a wooden stirrer. Allow the lye to cool down as you mix, then add more lye until it absorbs. Be careful not to add so quickly that it begins to first bubble and then boil. You should be able to feel the heat on the outside of the container and can use that to judge. If mixing large batches of lye solution you will need to mix these the day before and allow them to come down to room temperature before continuing. Cover the lye solution and continue.
Measure out 255.7g of liquid waterglass (36.5% sodium-silicate, 62.5% water). Immediately add it to the cooled lye-solution and stir together.
Pour the solution into the aggregate and mix for several minutes with a mechanical mixing paddle. We used an aluminum-tipped mortar mixing paddle on the end of a drill. The lye will off-gas hydrogen if it comes into contact with just about any metal, but we felt that once it was mixed in with the flyash and aggregate that it wouldn't be as active against the metal. The alternative was to try to coat the paddle somehow, and that wasn't a good option as we thought it would surely wear off into the mix. A tough and strong plastic-coated paddle would be idea.
Spray the molds with Pam cooking spray as the mold release (or use any similar mold release, but don't use petroleum jelly, it's been known to interfere chemically with geopolymer).
Let it sit for a few minutes, then pour the mix into a mold. I suggest wooden or silicone molds that can survive the heat of curing. We used 2.5" cube molds made of wood and previously coated in silicone caulk. Note: ideally you would de-gas the mix in a vacuum chamber to get rid of any entrained air before pouring.
Cure the geopolymer in a pre-heated oven at no more than 200° Fahrenheit. Any hotter and it will negatively affect the strength. At 200°F it cures in 4 hours. At 85°F it will cure in 24 hours. Any analogous range and length between works too (ie: you could try 120° for 12 hours). It does not need to be covered or kept wet while curing.
Remove from heat when the time is up and remove from the mold (further heat will not hurt or help it). It is now cured and has about 90% of its final strength. Within 3 days it will have 95% of its full strength, and 99% within a month.

A note about flyash: You can order a flyash type-F sample from Boral free of charge. However if you're ever in doubt there's a simply test you can perform. If the flyash is high calcium, it will heat up when mixed with a little bit of water. Calcium compounds in both concrete and type-C high-calcium flyash are what cause both concrete and type-C flyash to cure themselves by generating their own heat, what's known as the heat of hydration.

If you add a bit of water to a good amount of flyash (say the size of a cup) and it stays completely cool, then you have a low-calcium type-F flyash that is possibly a good fit for this recipe.

If you have a choice, the lower the calcium content the better. 2% calcium flyash is about as good as can be hoped for. I performed this recipe with 5% flyash that was available to me.

Good luck!

Step 3: Licensing Information

Just so there's no confusion, I am releasing this info under the MIT license:

The MIT License (MIT)

Copyright (c) <2014> <Michael Eliot, Andy Thomas>

Permission is hereby granted, free of charge, to any person obtaining a copy of this document, to deal in the document without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the document, and to permit persons to whom the document is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE DOCUMENT IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE DOCUMENT OR THE USE OR OTHER DEALINGS IN THE DOCUMENT.