Refractory Experiment In progress

Further experiments.

Test ingredient variations. 

This is the formula by volume.  10 parts perlite  is a guess, it can be increased in a later experiment after these are evaluated.

10 : Silica Flour or Sand
  6 : Native Clay
  2 : Portland cement
  3 : Lime
  5 : Perlite

The variations from the mix are listed for each sample.  Play sand is a quartz sand.

Mixes tested.

1. ✅  the formula aka control
2. ✅  local sand instead of silica flour 
3. ✅  play sand instead of silica flour
4. ✅  silica flour without Portland cement
5. ✅  play sand without Portland cement
6. ✅  local sand and silica flour 5 each
7. ✅  local sand and silica flour no Portland cement

𝥷   Not started     ⑇    Dry mixed     ✅ Drying     🔥  Firing

Whats wrong with this picture ?  6 and 7 are identical but 7 lacks portland cement so it is lighter.  So yeah, I knew which one it was. LOL

If all goes well tomorrow I can hydrate them and move the flags to ✅
Currently thinking a day of natural drying in the house then bake them below the boiling point of water till dry.   Then fire.  This is a bit of a rushed schedule and I reserve the right to change my mind. LOL

Aside:  I knew a geneticist who said "It's all book keeping".

All were hydrated in one afternoon.  Containers were covered with snap on lids assumed to be mostly air tight.  Solo brand containers.

In addition to the normal sample and additional one was made with 1 spoon of sodium silicate.  Total of 14 sample.   I should have enough to make 1 or 2 more samples from each bottle.

It took considerable time to create these samples.  I am now thinking a week to dry.

The perlite is not of a uniform size.  I think it would be better to sieve it so a more uniform sample could be made.

 It would also work better to add the perlite to each sample as it was hydrated.  That would ensure a more even distribution.

---------------
Its the next morning about 12 hours after the tests were cast. 
Moved the labels to the lids as they were not going to make the trip on the sides.

Took the lids of the samples which we heading in the direction of dry.  So I gave each 2 teaspoons of water.  Shortly after casting I had given each sample a dribble of water.  It will be interesting to see if the absorb this water and how long it will take.

If you are thinking the water is only needed to hydrate the Portland cement you might be mistaken.  The samples with no Portland cement were just as non-wet, don't want to say dry.  It may be that water enables processes that unlike the Portland cement bonds are not destroyed by heat.  If this is or is not the case adding water will at worst waste my time.   And if you look at it this way I should have gone out and purchased chimerical refractory.  But at this juncture I  have more time than money.

The original tile/brick was mixed to a consistency of peanut butter.  I would call that ramable refractory.  These were slightly more fluid maybe more along the line of a castable.

It seems that ramming the refractory produces better mechanical interlock between the particles.   This provides additional strength in the green and fired stages.     However if you are using a commercial refractory follow the makers instructions.   This is about DIY.

Again with not so good image of my sand.

It looks like the sand particles would interlock when forced together.  A good thing.  But note that the original test tile/brick was made with silica flour instead of sand.

---------------
48 Hours after cast

IB needed water
2B and 5B were soon to need, and were watered

B is identical to A but has a small spoon of sodium silicate added.  Adding sodium silicate to this mix may flux it so maybe a bad idea.

1. A B 
2. A B
3. A B
4. A B
5. A B
6. A B

I would not read anything into this.  Too little data and not enough control over how much water was added.  Different samples had different mixes with different weights and densities.   Even if all had the same amount of water to start it would not have usable meaning.  I record it here because it is an observation and multiple observations can provide clues which can be investigated.

Why am I bothering to keep these moist when the Portland cement bonds will break down anyway?    The lime and possibly the sand may form pozzolan type bonds which need water to form and may not break down with heat.

---------------------------------

Saturday November 17th.

Interesting difference between sample 4 and 5.

1. ✅  Mix from previous video
2. ✅  local sand instead of silica flour 
3. ✅  play sand, quarts, instead of silica flour
4. ✅  silica sand without Portland cement
5. ✅  play sand without Portland cement
6. ✅  local sand and silica flour 5 each
7. ✅  local sand and silica flour no Portland cement

Both lack Portland cement. But sample 4 with silica sand is much firmer than sample 5 with play/quartz sand.   

Sample 5 show no signs of firming up, it is about the same as it was mixed.  I had been thinking it was the lime which was doing the bonding but apparently the silica sand is.  Strange because the lime only sample from a video or two back dried/set into a nice puck not unlike the refractory.   Interesting.

Thinking I should try crumbling the lime only sample.  See if there is any bonding past drying going on there.  The silica sand only sample dried to a puck but easily crumbled to dust.

 I see I should have made a sample with local sand and no Portland cement.