Post by mintcake on Apr 4, 2014 12:58:14 GMT -8
Hi all..
I'm part way though testing some different mixtures. Your local clay will of course be different...
My ingredients are:
1. Local clay: crumbly-damp.
2. Perlite (random sizes) 0-5mm diameter, I'd guess.
3. Crushed charcoal
4. Waterglass from the local building supplies shop.
Preparation of samples:
I mixed the "dry" ingredients dry in the ratios below (clay was NOT hard-packed but about 20% air), stirred water into the waterglass, mixed quickly and pushed them into a metal channel.
The perlite tended to grab the water quite quickly, and after only a few minutes they became somewhat crumbly, which is why I chose this method. I guess I could have added more water, but I preferred the idea of "as-dry-as-possible" to reduce shrinkage.
I then tried to pack the samples down to be as rectangular as I could.
After a week of air-drying I stuck them into our oven at about 90C for half a day to drive off any remaining water. I then got bored and increased the temperature to 110C.
Once dry I put them along the walls of my ytong batch-box, and they got roughly one hour of full burn. The parts nearest the wood were glowing nicely by the end of the firing, but I'm sure they didn't get fired all the way through.
Volumes/masses:
My standard measure was a child's play cup (85cm3).
After I'd done everything, a similar sample of clay weighed in at 100g, charcoal 15g, perlite 10g. But... >:(I'm very confused, since the final masses don't agree with this at all. They make it seem like the perlite and charcoal weigh 40g/measure, and the clay 30g. which just can't be right, so I guess I've got massive variablilty in the clay amounts, even though I tried to pack the cups consistently. I know, I should have weighed things as I put them in. My excuse is that the kitchen scales were in use then...
The waterglass is measured in tea-spoons. I didn't vary it much yet.
Testing methods:
1.Penetration test... I started aiming for a brinell -like test, but since I've no idea what the numbers mean and I didn't find a ball-bearing the right size, I put a weight (a lump of marble, which happens to weigh 1.83kg) on a 7mm ball-bearing, which was resting on a flat part of the sample. My samples have quite a lot of voids and surface imperfections in them, so I avoided them, and measured the diameter of the impression, as well as I could with a pair of vernier callipers. Force/area gives a pressure, and I'm calling this local compressive strength. I'd frankly be very surprised if they could hold me, so I doubt this compressive strength number is a bulk value. I also tested some ytong and a (probably 100 year old) hand-made brick that I had to hand. Comparing the ytong and brick numbers to numbers on the internet the compressive strengths look in the right ball-park. For the brick and ytong I smoothed an area first using a lump of iron, to give a better surface.
2. For bending strength, I supported samples roughly 15x15mm cross section on two supports 35mm apart, and increased the load on them slowly (adding water to a 1l bucket). The load was applied at the centre of the sample, using a metal rod.
3. Heat conductivity: sample about 3cmx5cmx7mm was used as the roof of an insulated (ytong) flame tunnel, 2.5cm wide, with a blow-torch supplying the heat. The end of the tunnel was partially open to provide an exit route for exhaust gas. A 3.3cm long section of 1cm (ext.) aluminium U profile (from a sliding door) was used to contain 5 drips of water from a syringe. Evaporation times were measured, repeated six times. Early results were discarded if there was evidence that equilibrium had not been reached. The evaporation times gave a figure for power transferred, and an estimated temperature of 950C was guessed based on the orange/yellow flame colour.
For reference, the perlite data sheet seems to indicates that pure perlite has a lambda of about 0.2-0.4 for this sort of temperature, and I have a datasheet for 0.48 g/cm3 insulating firebrick which gives a lambda of 0.12 at 400C and 0.19 at 1000C, measured according to ASTM C-182, which I haven't read. The figures are probably not directly comparable.
Recipe code:
The recipe code is 4 numbers respectively: Perlite Clay charcoal Waterglass
Results
Brick: 1.14mm / 17.5N/mm2,
Dent measurements +/- .1mm, except for ytong where it was more like +/-.2mm
The "fired" Ytong was a piece that used to be next to the batchbox port. I measured 3 breaking strains from this sample, with significant variation (~20%). I've given the average, which happens to also be one of the strains.
Heat conduction: I've calculated these based on guess that the red-orange heat was around 950C. I am suspicious about sample 2's measurements, they showed significant variability.
Observations:
In testing compressive strength, once, the ball bearing went straight into sample 3 and needed digging out. It's by far the weakest sample. I don't know if this is because I was scared to compress it well enough, it was too wet, too dry, or just because there's too much charcoal in it. I might try another test of that mixture.
Sample 1 initially resisted the maximum load I could put in my bucket so far. (a lump of iron, a part-brick and water). On re-testing it (probably at a different loading point) it failed under loads it had previously survived.
Samples 2 and 3, despite being closer to the port than sample 1, have not bisqued or lost their charcoal more than about half a cm in. This probably says something good about their thermal properties, but means that I don't think I should test them yet.
I'm part way though testing some different mixtures. Your local clay will of course be different...
My ingredients are:
1. Local clay: crumbly-damp.
2. Perlite (random sizes) 0-5mm diameter, I'd guess.
3. Crushed charcoal
4. Waterglass from the local building supplies shop.
Preparation of samples:
I mixed the "dry" ingredients dry in the ratios below (clay was NOT hard-packed but about 20% air), stirred water into the waterglass, mixed quickly and pushed them into a metal channel.
The perlite tended to grab the water quite quickly, and after only a few minutes they became somewhat crumbly, which is why I chose this method. I guess I could have added more water, but I preferred the idea of "as-dry-as-possible" to reduce shrinkage.
I then tried to pack the samples down to be as rectangular as I could.
After a week of air-drying I stuck them into our oven at about 90C for half a day to drive off any remaining water. I then got bored and increased the temperature to 110C.
Once dry I put them along the walls of my ytong batch-box, and they got roughly one hour of full burn. The parts nearest the wood were glowing nicely by the end of the firing, but I'm sure they didn't get fired all the way through.
Volumes/masses:
My standard measure was a child's play cup (85cm3).
After I'd done everything, a similar sample of clay weighed in at 100g, charcoal 15g, perlite 10g. But... >:(I'm very confused, since the final masses don't agree with this at all. They make it seem like the perlite and charcoal weigh 40g/measure, and the clay 30g. which just can't be right, so I guess I've got massive variablilty in the clay amounts, even though I tried to pack the cups consistently. I know, I should have weighed things as I put them in. My excuse is that the kitchen scales were in use then...
The waterglass is measured in tea-spoons. I didn't vary it much yet.
Testing methods:
1.Penetration test... I started aiming for a brinell -like test, but since I've no idea what the numbers mean and I didn't find a ball-bearing the right size, I put a weight (a lump of marble, which happens to weigh 1.83kg) on a 7mm ball-bearing, which was resting on a flat part of the sample. My samples have quite a lot of voids and surface imperfections in them, so I avoided them, and measured the diameter of the impression, as well as I could with a pair of vernier callipers. Force/area gives a pressure, and I'm calling this local compressive strength. I'd frankly be very surprised if they could hold me, so I doubt this compressive strength number is a bulk value. I also tested some ytong and a (probably 100 year old) hand-made brick that I had to hand. Comparing the ytong and brick numbers to numbers on the internet the compressive strengths look in the right ball-park. For the brick and ytong I smoothed an area first using a lump of iron, to give a better surface.
2. For bending strength, I supported samples roughly 15x15mm cross section on two supports 35mm apart, and increased the load on them slowly (adding water to a 1l bucket). The load was applied at the centre of the sample, using a metal rod.
3. Heat conductivity: sample about 3cmx5cmx7mm was used as the roof of an insulated (ytong) flame tunnel, 2.5cm wide, with a blow-torch supplying the heat. The end of the tunnel was partially open to provide an exit route for exhaust gas. A 3.3cm long section of 1cm (ext.) aluminium U profile (from a sliding door) was used to contain 5 drips of water from a syringe. Evaporation times were measured, repeated six times. Early results were discarded if there was evidence that equilibrium had not been reached. The evaporation times gave a figure for power transferred, and an estimated temperature of 950C was guessed based on the orange/yellow flame colour.
For reference, the perlite data sheet seems to indicates that pure perlite has a lambda of about 0.2-0.4 for this sort of temperature, and I have a datasheet for 0.48 g/cm3 insulating firebrick which gives a lambda of 0.12 at 400C and 0.19 at 1000C, measured according to ASTM C-182, which I haven't read. The figures are probably not directly comparable.
Recipe code:
The recipe code is 4 numbers respectively: Perlite Clay charcoal Waterglass
Results
| Recipie | Dry mass(g) | Fired mass(g) | fired density(g/cm3) | dent diameter(mm) / local compressive strength (N/mm2) | Bend / flexural strength (N/mm2) | Heat conductivity (lambda) at red heat (w/(mK)) | |
| 1 | 2203 | 141 | 139 | 0.87 | 2.85 / 2.8 | Trial1 > 0.5 (did not break) Same sample, retested: 0.38 | 0.21 |
| 2 | 2213 | 181 | 162 | 0.66 | 4.1 / 1.4 | 0.17 | |
| 3 | 2223 | 223 | 192 | 0.63 | 6.1 / 0.6 | 0.18 | |
| 4 | 2102 | 113 | 107 | 0.60 | 4.1 / 1.4 | 0.14 | 0.16 |
| 5 | 2113 | 142 | 118 | 0.56 | 4.49 / 1.1 | 0.12 | 0.17 |
| Ytong | 2.2 / 4.7 | New: ? Fired: 0.29 | 0.19 |
Dent measurements +/- .1mm, except for ytong where it was more like +/-.2mm
The "fired" Ytong was a piece that used to be next to the batchbox port. I measured 3 breaking strains from this sample, with significant variation (~20%). I've given the average, which happens to also be one of the strains.
Heat conduction: I've calculated these based on guess that the red-orange heat was around 950C. I am suspicious about sample 2's measurements, they showed significant variability.
Observations:
In testing compressive strength, once, the ball bearing went straight into sample 3 and needed digging out. It's by far the weakest sample. I don't know if this is because I was scared to compress it well enough, it was too wet, too dry, or just because there's too much charcoal in it. I might try another test of that mixture.
Sample 1 initially resisted the maximum load I could put in my bucket so far. (a lump of iron, a part-brick and water). On re-testing it (probably at a different loading point) it failed under loads it had previously survived.
Samples 2 and 3, despite being closer to the port than sample 1, have not bisqued or lost their charcoal more than about half a cm in. This probably says something good about their thermal properties, but means that I don't think I should test them yet.
. Now, I'm fairly sure the powder isn't that conductive, but... it might not be as good an additive when powdered as I'd hoped.