Desktop Insight 3 - Dealing With Crazing

Learn what crazing is, how it is related to glaze chemistry, how INSIGHT calculates thermal expansion and how to substitute high expansion oxides (e.g. Na2O, K2O) with lower expansion ones (e.g. MgO, Li2O, B2O3).

D. Desktop Insight

Click here to watch this at youtube.com or click here to go to our Youtube channel

This video explains what crazing is, how it is related to glaze chemistry, how INSIGHT calculates thermal expansion and how to substitute some of the amounts of high expansion oxides that often cause crazing (e.g. Na₂O) with lower expansion ones (e.g. MgO, Li₂O, B₂O₃). The side effects of these substitutions are discussed and the reader is shown how to research the purpose of each of the oxides in a glaze formula. Click here for a page having a link to it.

Lesson 3 Transcript: Dealing With the Root Cause of Crazing
What is crazing, why glazes craze, thermal expansion, stress testing, researching, substituting KNaO, adding more SiO₂ and Al₂O₃ via boron, the unique character of Li₂O, adding boron from frits and silica:alumina ratios

Dealing with the root cause of crazing
Welcome to lesson 3. We are going to get to the root of why glazes craze
And I'll talk about thermal expansion, stress testing, researching, substituting KNaO, adding more SiO₂ and Al₂O₃ via boron, the unique character of Li₂O, adding boron from frits and silica:alumina ratios.

Thermal expansion information at Digitalfire
As already noted, you can learn more about crazing by clicking the Reference Database link on the digitalfire.com home page, then Articles. There is a section on Thermal Expansion and if you do an on-page search for the word crazing you will find more.

What is crazing?
There are many listings there in the Articles/Glazes section. This one shows the best solution. There is also a very concise one in the Troubleshooting section that explains what it is and how to fix it.
Crazing occurs when a glaze has a higher thermal expansion
than the clay body it is 'glued' to. Most solids expand on heating, and contract on cooling. If a glaze shrinks more than the ware during cooling in the kiln, it becomes 'stretched' and seeks to relieve the stress by forming a network of cracks.

The key to fixing crazing
While there are many band-aid approaches, the key to fixing crazing is the reduction of glaze thermal expansion (or increasing body expansion). INSIGHT can calculate an approximation of glaze expansion as you can see here with the G1214Z recipe. This is a base matte very similar to the one developed in the last lesson.

Boiling water:Ice water test
It is very important to realize that although a piece might not be crazed out of the kiln, the crack pattern can develop over time as it is repeatedly heated and cooled in use. So to test you need to stress the piece thermally to accelerate time. You can do this using the boiling water:ice water test. It is described in the Tests section in the Digitalfire Reference Library.

Adding silica, a poor option
People often suggest a simple silica addition to stop crazing. But alumina mattes need low silica, adding even a few percent can turn them glossy (notice the SiO₂ is at the minimum in the target formula to achieve the matte effect). Furthermore, mattes are often not well melted, silica is refractory, it will make this worse. Finally, large amounts of silica are required to dilute the expansion-raising effect of relatively small amounts of oxides like K₂O and Na₂O.
So we need a better solution or a solution of which it is a part.

Method 1: MgO for CaO
For my first approach to the problem I have opened G1214Z into recipe 1 and copied it to recipe 2 using this button.
I have set both recipes to calculate RO Unity and am going to work on 2 and compare it with 1.
Notice I have the lessons materials database selected as with other lessons.

Analyzing the formula
Take a look at the formula. CaO is very high compared to the limits. What limits?
I have them open from the last lesson, notice the title bar shows which one.
Can I substitute some of the CaO for another fluxing oxide of similar character and lower expansion? Yes.
Magnesia (MgO) also mattes middle fire glazes and has a much lower expansion. I will double-click the CaO oxide line in the Formula List.

Compare MgO/CaO expansions
The Oxides dialog opens.
Notice the expansion number. Now I will click 'MgO' in the list, notice now much lower it is. I will close the Oxides dialog.

Using the detail report to investigate CaO sources
To see which materials are sourcing the CaO I have selected Detail Recipe Calculation from the Report Type menu (in the File menu).
This report shows how INSIGHT calculates the chemistry. INSIGHT has written it to this text file (in the Insight folder in my documents folder) and asked the operating system to display it. In the Preferences dialog you can configure INSIGHT to ask your operating system to display reports in your web browser or system text editor.
This window is narrowed so we can only see the left most portion but notice the CaO column. Clearly, wollastonite is supplying the major portion of the total. Wollastonite is calcium silicate, so why not source MgO from magnesium silicate, namely talc.

Talc as a source of MgO
I have chosen Edit Materials from the Utility menu, entered Talc here and pressed the go button to search and got three hits. I have selected Nytal Talc. Notice this talc is mainly MgO and SiO₂ but also has some CaO. Talc also has an LOI (like whiting but much lower) and this can contribute to the formation of suspended micro bubbles in the fired glaze. However this is a matte glaze so they wont be as visible (unless of course they break at the surface).

Generic talc
I have clicked generic talc. Notice it only has MgO and SiO₂ and no LOI is specified. Of course such a theoretical material does not exist in nature. The best talcs approach this chemistry better. You might question the selection of talc over a frit to source MgO. But for demonstration purposes using a material like this helps focus on the issues more clearly, we will leave the complexities of incorporating a frit for a later lesson.

Reducing wollastonite for talc
To incorporate some talc I have made sure recipe 2 is selected and changed the amount of wollastonite to 20, that reduces the CaO it contributes to make room for MgO.
I have also checked this box to get the target to weave into the list better, it does not separate K₂O and Na₂O because they are so similar.

Resultant formula changes
Notice that I have checked the .000 check box above the formula list. Most engineers argue that ceramic calculations are not nearly accurate enough for this, they are right. However some like three decimals to see how rounding was done to two. I will turn this back off.
But first, notice how much the B₂O₃, Al₂O₃ and SiO₂ have increased as a result of the wollastonite reduction. I am going to add talc one-part-at-a-time to drive these numbers back down.

Driving CaO down with talc
To do this I have clicked a blank line below the recipe and entered Talc in the Lookup blank and updated the line.
Then I will make sure the increment by field contains a 1 and start clicking the Increment arrow and watch the B₂O₃ and Al₂O₃ amounts come down (in formula 2).
You might think that the CaO has not dropped much as a result in the loss of the wollastonite, but stay tuned to see what happens.

Adjusting silica
I kept clicking till the alumina and boron matched formula 1, this happened at 7 talc.
Now the CaO and MgO total what the CaO alone was before. The addition of the MgO drove the CaO down because fluxes are retotaled to one.
The SiO₂ was off a bit, (wollastonite and talc don't have the same SiO₂ content).
I have clicked the silica line in the recipe and taken it down by one to match them back up.

Assessing the effects of MgO
Take a look at the Calculated Items list. The expansion has dropped from about 7 to 6.6. This is a significant change.
However, before continuing I want to stress that matte glazes have tight limits on their chemistry and do not afford a lot of room to juggle things without losing their surface quality. The addition of MgO might affect stain colors, for example. For now I am going to ignore that.
MgO might also adversely affect melting, so a strong flux might be needed to help. Lithia is really powerful and small amounts are often used for this purpose. I want to demonstrate something about it so I am going to replace all the MgO with it (I would not do this in practice since it would gloss the glaze, but I want to show you something).

Method 2: Trade CaO for Li₂O
To change the talc to lithium carbonate I click the talc line in the recipe list and enter enough of the name in the Lookup for INSIGHT to recognize it.
I have also emptied the label so INSIGHT will change it also.
Then I will update the line.

Comparing talc and lithium
Something interesting happens. 7 parts of lithium carbonate supplies 0.24 Li₂O to the formula whereas 7 parts Talc supplied only 0.16 MgO? Why is this?
7 grams of lithium carbonate powder sources many more molecules of Li₂O because they are alot smaller, remember formulas compare numbers of molecules. Lets look at the weight of Li₂O in the Oxides dialog.

Comparing Li₂O, CaO, MgO
It is only 29.8, whereas CaO is 56.1 and MgO is 40.3.
Amazingly, lithium carbonate has a higher LOI than whiting or talc. Yet still it supplies many more molecules per weight unit.
You can learn more by clicking the Info button. This page at the Digitalfire Reference Library explains that lithia is a very powerful melter and it can have a disproportionately large reducing effect on thermal expansion. Understanding this and how to fire it is actually the secret behind Corning ware.

Second guessing INSIGHT on Li₂O
Notice that even though I added considerable lithium, the calculated thermal expansion is only slightly lower.
This is an example of when we need to second-guess INSIGHT. The relationship between the amount of lithia and its effect on thermal expansion is not linear so it does not 'calculate' as well. From personal experience I would be confident that 5% is enough to fix the crazing problem however it will also affect fired visual character to a greater degree than other fluxes.
To finish I would need to reduce the amount of lithium to match up the boron, alumina and silica.

Method 3: Add B₂O₃
Lets look at another way to reduce thermal expansion:
Boron is a low expansion and low melting glass, its presence makes room for more very low-expansion (but refractory) alumina and silica (I mentioned silica at the beginning of this lesson). Where do we get more boron?
INSIGHT provides a way to search for materials having a certain oxide.

Using the delete line button
To start again I will copy recipe 1 to recipe 2.
Notice the line label is still there. INSIGHT left it there but I will delete it by clicking the delete line button.
Then I will select the Frit 3124 recipe line for recipe 2 and click the Delete button to zero its amount.
Notice again that it zeroed the line but did not remove the name. This time clicking the button again will not remove it because the other recipe still has some of this.
Now I will select a blank line below the recipe.

Choose a line material using the materials dialog
I am going to determine the material for this line by opening the Materials dialog. I have selected B₂O₃ in the Having popup and then I have clicked the Frit 3195 line.
Notice this has 23% boron, that is higher than the vast majority of boron frits so this will be a good source of B₂O₃. Frits are the best source of B₂O₃, the natural ones have LOI, consistency and solubility problems.
Now I will click the Update Recipe Line button, this will insert this material into the blank line I selected.

Adding B₂O₃ from a frit
The Materials dialog closed and here it is.
I have already also keyed 36 for the Amount and Updated. I clicked the frit line again since INSIGHT moved the cursor down after the Update button.

How much B₂O₃?
Notice that the B₂O₃ is up dramatically, the alumina is also up. That's what I wanted. Extra B₂O₃ will make the glaze melt lower and increase its ability to tolerate high melting low expansion SiO₂; but how much do I add? Another number that INSIGHT calculates is useful. Look at the SiB:Al line in the Calculated Items list.
This is simply the total amount of SiO₂ and B₂O₃ divided by the amount of Al₂O₃, or a ratio of glass formers to intermediates/stabilizers. If this ratio is maintained the altered glaze is more likely to have the same degree of fired gloss. Testing will be needed to assess comparative melting.

The SiB:Al ratio
Notice what I have done here: I selected the silica line in recipe 2 and clicked the increment button until the SiB:Al ratio numbers matched. It only took 3 parts of added silica.
Notice that the calculated expansion is again reduced considerably.
I can tell you from experience that this amount of added boron will melt more silica and alumina than I have added so there is room to move the expansion lower using this technique.

The best methods all rely on ceramic chemistry!
In most cases, you will find that the glazes most likely to craze are those having significant Na₂O and K₂O and those lacking Al₂O₃ and SiO₂ (in comparison to limit formulas). Often glazes have both problems. The most effective method of reducing thermal expansion then is obviously reducing KNaO in favor of other fluxes, especially MgO, it has the lowest expansion. Increasing boron to make room for more alumina and silica is also an effective approach.
Think about this: All of the oxides I just talked about come in materials that contribute other oxides also. There is no way you could figure out how to juggle the recipe to effect the changes we have done without ceramic chemistry calculations.

Links

Oxides	MgO - Magnesium Oxide, Magnesia
Troubles	Glaze Crazing Ask the right questions to analyse the real cause of glaze crazing. Do not just treat the symptoms, the real cause is thermal expansion mismatch with the body.
Troubles	Glaze Shivering Ask the right questions to analyse the real cause of glaze shivering. Do not just treat the symptoms, the real cause is thermal expansion mismatch with the body.
URLs	https://digitalfire.com/videos Tutorial Videos at Digitalfire
Tests	Boiling Water:Ice Water Glaze Fit Test Ceramic glazes that do not fit the body often do not craze until later. This test stresses the fit, thus revealing if it is likely to craze later.
Tests	300F:Ice Water Crazing Test Ceramic glazes that do not fit the body often do not craze until later. This progressively stresses the fit until failure point, thus giving it a score
Glossary	Co-efficient of Thermal Expansion The co-efficient of thermal expansion of ceramic bodies and glazes determines how well they fit each other and their ability to survive sudden heating and cooling without cracking.
Glossary	Feldspar Glazes Feldspar is a natural mineral that, by itself, is the most similar to a high temperature stoneware glaze. Thus it is common to see alot of it in glaze recipes. Actually, too much.
Glossary	Digitalfire Insight A downloadable program for Windows, Mac, Linux for doing classic ceramic glaze chemistry. It has been used around the world since the early 1980s.
Articles	Crazing in Stoneware Glazes: Treating the Causes, Not the Symptoms Band-aid solutions to crazing are often recommended by authors, but these do not get at the root cause of the problem, a thermal expansion mismatch between glaze and body.
Articles	Understanding Thermal Expansion in Ceramic Glazes Understanding thermal expansion is the key to dealing with crazing or shivering. There is a rich mans and poor mans way to fit glazes, the latter might be better.
Articles	Bringing Out the Big Guns in Craze Control: MgO (G1215U) MgO is the secret weapon of craze control. If your application can tolerate it you can create a cone 6 glaze of very low thermal expansion that is very resistant to crazing.

Cone 10 mug is crazing after a year. That's OK because it's high-fire, right?

Wrong. This cone 10R grey stoneware, H550, is OK if the glaze fits. But this high feldspar glaze on the inside has begun to craze after a year. The greyer coloration around the craze lines indicates that water is soaking into the slightly porous body. And, this mug has lost the ring it had out of the kiln. Imagine your customers returning these pieces! Could they be refired to be as good as new? Perhaps. But they would return to this condition. The practical solution is to reformulate this glaze to reduce its thermal expansion.

By Tony Hansen
Follow me on

Got a Question?

Buy me a coffee and we can talk