Understanding COE

[glazenerd]

COE applications

Coefficient of expansion (COE) is a mathematic equation that measures compression and contraction. The clay body is compressed by the contraction of the cooling glaze. Pure silica glass has a COE of 5.5 x 10(-7) power: glaze calculators just hide the 10(-7) and only show the 5.5. When you add feldspars, alumina, and other oxides: the COE rises.

Clay and glaze both have the basic elements required to make glass (glaze): silica + alumina + fluxes = glass. The difference between them is particle size, ratios of each, and flux levels. Flux levels in clay are much lower than flux levels in glaze. That difference in flux levels is the primary reason there are appreciable differences in clay and glaze.

To understand why COE varies so widely: you need to understand how each glaze ingredient raises or lowers COE. Here is a short list of common glaze ingredients and their individual COE values:

Nep Sy: 9.86 Mahavir potash 8.60. Silica. 3.52. Alumina: 6.47 EPK. 5.06

Any given glaze recipe has a given % of flux, silica, and alumina. The list of ingredients simply supply various amounts of each until a given level of each is reached. There are minimum amounts of each that are required to achieve a complete melt at any given cone: these minimums are known as formula limits. 
**** see " Hesselberth & Roy" for formula limits.

To paint a simple picture of glaze formulation: as the levels of fluxes rise ( sodium, potassium, boron) the Final COE rises. As silica and alumina increases; the final COE lowers. The reason EPK is commonly suggested to fix crazing ( cracking) in the glaze is because kaolin is 37% alumina and 48% silica: both lower COE while maintaining formula limits. The other simple premise: the lower the cone fire; the higher the level of fluxes required to achieve melt. As the cone fire gets higher: then flux levels are lowered: which in turn lowers COE.  Cone 6 recipes typically run 50-55% fluxes in the recipe: more means lower cone fire, less means higher cone fire. This general rule will help you figure out if you have a low, medium, or high fire glaze recipe. Obviously there is wiggle room at any given cone range.

The COE values of cone 6-10 clay bodies run 5.5 up to 6.10 typically. There are variances for highly translucent porcelains, and high talc content low fire recipes. It is best to ask your supplier for the COE of your clay of choice. Glazes typically run 6.95 all the way up to 9.00 (crystalline glaze). Normally they are in the 7.25 to 7.95 range. The trick is to keep the COE of your glaze within 1 to 1.25 of your clay.
Example: clay COE of 5.75 plus 1.25= 7.00 for glaze COE. This is a target range, not an exact number.

The above is the chemistry and science between clay and glaze COE. However, let's throw that in the trash and use something easier. No, this is not accurate but it conveys COE differentials between clay and glaze in digestible terms. Look at COE in terms of percentages.

A clay body with a COE of 6.00 will shrink 6.00% from peak temperature to room temperature. A glaze with a COE of 9.00 will shrink 9.00% from peak to room temperature. The clay is shrinking at 6.00% and the glaze at 9.00%: that is a 50% difference in shrinkage rates. As the differences in shrinkage rates climb: crazing issues increase. The closer you can get these two values together, the chances of glaze issues goes away.  The terms are compression and contraction: but shrinkage conveys both in easy terms. When the clay and glaze are within acceptable percentages of each other: the term is "glaze fit."

Hope this helps.. Tom

 

[glazenerd]

The crazing ( cracking) pattern in your glaze will tell you how close or far away your glaze " fit" is. The crazing pattern on this piece is open and spaced apart.  The farther apart the crazing lines are: the closer you are to having a glaze fit. The actual term is called " low compression." Which means the clay body is experiencing low amounts of compression by the contraction of the glaze.

(Sorry for the blur). As the cracks get closer and closer: the difference in COE between clay and glaze is farther apart.in this case; there is nearly 50% difference between the two. If you noticed: there are long crazing lines that run across the entire piece. These form first as the piece is cooling to compensate for the COE differences. As the piece continues to cool, the crazing pattern continues to get tighter and tighter: reflecting the actual differences.

Tom

 

 

[Min]

Bear in mind that calculate expansion figures, while useful, are not absolute. Glaze calc is helpful to get you in the ballpark but there are limitations.

First off, they don’t work for glazes that aren’t a fully melted gloss. Semi-matte and matte glazes are crystalline structures (as is clay and slip). This means part of the glaze is not fully incorporated in the glaze melt. In the case of semi or matte glazes some of the oxide(s) precipitate out of the glaze matrix. Calcium, barium, strontium, alumina and magnesium matte glazes are examples of this.

Another point is glaze materials don’t always effect the coe in a predictable way. Example would be if you add zircopax to a glaze. With glaze calc the coe will rise with the addition of it. But, the zircon in the zircopax doesn’t actually enter the melt and although the coe raises (and the predicted raise in the likelihood of crazing) just by looking at the figures doesn’t tell you this, adding zircopax in actuality will lower the likelihood of crazing. Colouring oxides can have the same effect, theoretically raising the coe but in practice will often reduce crazing.

There also comes a point where an oxide can be oversupplied, the coe figure might be low enough to assume the glaze won’t craze but it will. Examples of this would be an excess of silica or boron. There are limits to how much silica a glaze can take in without having excess silica not entering the melt. When this happens you will likely will see crazing. With boron, excess of it will negate the elastic effect it has on glazes, and again you will have crazing.

There is also a rather confusing school of thought regarding the combination of oxides and the validity of coe figures altogether. If you take a well fitting glaze and add another oxide the coe can remain the same and yet the glaze can craze. COE figures work more accurately within a "family" or glaze system rather than by comparing two different glazes.

The clay company I buy from, Plainsman, does not publish COE figures,  "The reason is that such numbers often mislead users. First, a body has different thermal expansion characteristics when fired at different temperatures, schedules and atmospheres. Dilatometers are only useful when manufacturers can measure bodies and glazes over time and in the same firing conditions.” They give a very good write of the claybody properties and suggestions for glazes.

Glaze calc and coe figures are a big help but at the end of the day, it comes down to testing, testing, testing. 

 

[glazenerd]

Min: 

you jumped from COE 101 to post grad COE in a hurry!  

Ron Roy and Tony have had a decades old debate over COE values.

[JohnnyK]

And to think that in the 3 college level courses I took COE was never mentioned, much less "glaze fit". My guess would be that the buckets of mixed glazes they provided us with were fits for the clays that they also provided...Hmmm. Hence the lack of the need to provide us with such pertinent info. All of the above is just one of the reasons that I'm a member of this forum and plan to be for a long time into the future. Thanks you all for the education you provide the rest of us!

JohnnyK

[glazenerd]

Delayed crazing.

This rarity occurs after the glaze pieces have cooled and have been removed from the kiln. It can occur within hours or months later; but the underlying cause stems from COE issues. The crazing pattern is distinctive from the patterns commonly associated with COE differentials between the clay and glaze. Typically delayed crazing forms singular cracks, with web cracking extending from it. In the pottery world; this crazing pattern is called " tensile" cracking.

Delayed crazing occurs primarily from the absorption of atmospheric moisture in the clay body, causing expansion. (COE). Low fire earthenware and porous bodies are most susceptible: both can allow absorption. Usually the COE of the glaze and clay are just under the border of fitting; and the added swelling of the clay body from moisture absorption creates enough added stress to cause tensile cracking.

Delayed crazing can also be caused by high viscosity glazes. Sodium has the highest, then potassium; while calcium is lower viscosity. High sodium glazes are often used for the desired effect of glaze run; but can also make the glaze susceptible to delayed crazing. Delayed crazing is a rarity and recipe changes only need occur if cracking actually occurs.

T

[Bill Kielb]

Just food for thought

I like the discussion and often take the position that calculated COE is not really a great design tool. I find that many get lost in the COE and LOI for reasons I cannot begin to imagine.

Interesting that in both discussions above there is no mention of whether a glaze slightly in compression with respect to the body actually strengthens the entire structure. This is relevant to the average Potter in that  incrementally correcting a glaze that crazes until it does not is actually a reasonable approximate assurance that the glaze / body fit is likely beneficial. Significant strength improvements have been measured as a result of this slight compressive force on the claybody.

Often when I mention this folks feel better about making their favorite glaze fit their body.

it is also fascinating that COE, LOI are favored by the same folks layering or using 40,30,20,10 recipes or the ever present 2% bentonite I see in recipe after recipe. Mixing a gloss with a matte to obtain less matte makes my head hurt. All that being said, everyone has their way and most enjoy some success with their methodology.

so testing is always a practical proving ground.

good stuff!

[glazenerd]

Bill:

 I use to post lengthy discussions on various technical aspects of pottery; but these days I just limit them to a starting point. There are plenty of experienced and well versed potters around to fill in any blanks. Calculated COE is the base line: it has to be dialed in by testing as you and Min have correctly pointed out. The joys of pottery; there never seems to be absolutes: only an endless stream of variables.

T

[Min]

7 hours ago, Bill Kielb said:

Mixing a gloss with a matte to obtain less matte makes my head hurt.

Volumetric line blends are a quick and easy way to fine tune the degree of gloss / matte in a glaze. Line blends work and they're simple, especially useful for potters who don't use glaze calc software. Sorry to hear they make your head hurt.

[Bill Kielb]

3 minutes ago, Min said:

Volumetric line blends are a quick and easy way to fine tune the degree of gloss / matte in a glaze. Line blends work and they're simple, especially useful for potters who don't use glaze calc software. Sorry to hear they make your head hurt.

Chemically makes my head hurt! It is a method as I said.

[liambesaw]

6 minutes ago, Bill Kielb said:

Chemically makes my head hurt! It is a method as I said.

The recipes I've seen used like this were chemically designed to be blended, ala Tony Hansen.  Say what you will but that guy knows what he's doing in glaze calculation!

[Hulk]

"...I'm a member of this forum and plan to be for a long time into the future. Thanks you all.."

Boah howdy (emphatic affirmative; it's a colloquialism)!

A glaze I really like crazes badly (the checks are less than an 1/8 inch) on school clay, some improvement after adding silica, however, on my clay (at home), crazing is not nearly at bad, the pattern being much bigger, ~3/8 inch or more.

What I'm not yet understanding is why the glaze crazes when several other glazes I'm using (at school and at home) have coe values in the same ballpark, or even higher. I'm guessing that the difference has to do with elasticity.

The school clay, Laguna wc850, has low coe; home clay manufacturer does not publish coe values...

Any road, more testing and research t'go, and will report back later on.

[Bill Kielb]

Just now, liambesaw said:

The recipes I've seen used like this were chemically designed to be blended, ala Tony Hansen.  Say what you will but that guy knows what he's doing in glaze calculation!

I like his stuff but the average person taking a diluted sample, actually two diluted samples and combining them in some proportion works for him. For me the final chemistry would be impossible for many with any accuracy and I am always interested in durability. I always wonder why not adjust the glaze normally progressively, it just seems easier. Saying all that, that’s just my preference and the why part. Everyone has their method.

[Bill Kielb]

8 minutes ago, Hulk said:

"...I'm a member of this forum and plan to be for a long time into the future. Thanks you all.."

Boah howdy (emphatic affirmative; it's a colloquialism)!

A glaze I really like crazes badly (the checks are less than an 1/8 inch) on school clay, some improvement after adding silica, however, on my clay (at home), crazing is not nearly at bad, the pattern being much bigger, ~3/8 inch or more.

What I'm not yet understanding is why the glaze crazes when several other glazes I'm using (at school and at home) have coe values in the same ballpark, or even higher. I'm guessing that the difference has to do with elasticity.

The school clay, Laguna wc850, has low coe; home clay manufacturer does not publish coe values...

Any road, more testing and research t'go, and will report back later on.

If you are interested in some sort of glaze Calc. Take a look at the work by Anna Burke or Sue McLeod they showed a nice correlation with the Alumina level and Flux ratio within the durable range. From memory 0.25: 0.75 and .55 alumina to 0.2: 0.8 and .5 alumina is a nice starting range for low expansion bodies.

[glazenerd]

Bill:

my chemistry lesson this morning involves thermodynamics. It is 16F out this morning, and my regulated body temperature is 98. After figuring out the heat loss ratio, and factoring body mass: my final calculation was: "I am going home.where it is warm."

These modern chemistry chats make me chuckle a bit ( mostly at myself.) we have the benefit of calculators and nearly 100 years of lab controlled studies as a foundation to stand on. Yet on a daily basis we are still trying to figure it out. I often think about Maria Longworth Storer, Adelaide Robineau, Taxtile Doat: who had no calculators, no theories, no database, no modern kilns, no defined testing protocols : and yet produced stunning pieces. I could not imagine stoking a three story bottle kiln for three days.

The accepted SiAl ratio for clay bodies is 4:1. 

Hulk: obviously there is a COE difference between school/home clay. Adding silica works; but adding silica and alumina works better. ( aka EPK). I will forego the chemistry behind alumina and use a visual: the skeleton in your body is the framework which supports everything else. Alumina in clay and glaze does the same.

T

[Bill Kielb]

1 minute ago, glazenerd said:

Bill:

my chemistry lesson this morning involves thermodynamics. It is 16F out this morning, and my regulated body temperature is 98. After figuring out the heat loss ratio, and factoring body mass: my final calculation was: "I am going home.where it is warm."

These modern chemistry chats make me chuckle a bit ( mostly at myself.) we have the benefit of calculators and nearly 100 years of lab controlled studies as a foundation to stand on. Yet on a daily basis we are still trying to figure it out. I often think about Maria Longworth Storer, Adelaide Robineau, Taxtile Doat: who had no calculators, no theories, no database, no modern kilns, no defined testing protocols : and yet produced stunning pieces. I could not imagine stoking a three story bottle kiln for three days.

The accepted SiAl ratio for clay bodies is 4:1. 

Hulk: obviously there is a COE difference between school/home clay. Adding silica works; but adding silica and alumina works better. ( aka EPK). I will forego the chemistry behind alumina and use a visual: the skeleton in your body is the framework which supports everything else. Alumina in clay and glaze does the same.

T

Like my mother always said: cover your head, wear your scarf and put your gloves on!

[Hulk]

I've run the eight glazes I chose (so far), and five of the school's glaze recipes through GlazeMaster and done some reading - Tony Hansen, Sue Peterson's book (her explanation of limits, imo, the best), some Hesselberth, Britt's book, have seen McLeod' work, aye, good stuff - more research an' testing t'go!

"Adding silica works; but adding silica and alumina works better. ( aka EPK)"

   Will do, thanks 'nerd!

[Min]

@Hulk, in light of the current thread it would be interesting to see what we would all do to adjust your crazing recipe. Don't know if you are at liberty to post the recipe but if so it might be fun. If you do, posting a clear gloss recipe that doesn't craze on your clay and what clay you use would be helpful too. 

edit: Plus the qualities in the glaze you are looking for.

[neilestrick]

2 hours ago, Hulk said:

I've run the eight glazes I chose (so far), and five of the school's glaze recipes through GlazeMaster and done some reading - Tony Hansen, Sue Peterson's book (her explanation of limits, imo, the best), some Hesselberth, Britt's book, have seen McLeod' work, aye, good stuff - more research an' testing t'go!

"Adding silica works; but adding silica and alumina works better. ( aka EPK)"

   Will do, thanks 'nerd!

One of the benefits of adding both EPK and silica is that if you do them in equal amounts, it won't generally affect the si:al ratio, so the look of the glaze won't change significantly. Plus most glazes can use more clay in the mix to aid with suspension and dry durability.

[neilestrick]

6 hours ago, glazenerd said:

 I use to post lengthy discussions on various technical aspects of pottery; but these days I just limit them to a starting point. There are plenty of experienced and well versed potters around to fill in any blanks.

The vast majority of folks who come to the forum looking for help don't have a strong technical background, which is why they come to the forum in the first place. If they had that knowledge base, they wouldn't need us. It's easy to overload them with information that they don't understand- most just want a 'try this and this' response to get them moving toward a solution. It can be tough to diagnose things via the internet, so it's good that we have a lot of people with different backgrounds to offer possible solutions. One of them is bound to work!

[Babs]

Optimist Neil.....but true it depends and testing .....:-))))

And reading I get OMG I understand this or am I just remembering past knowledge.....

So all attempts to solve probs gratefully received.

[glazenerd]

Neil:

bought my domain name a few months back. Now I get to spend next year organizing 500 pages of techno blather into a cohesive website for clay chemistry. That should  cure my habit of making information dumps. I have a new found respect for teachers (which I am not.) Figuring out how to present information gets more challenging than the information itself.

T

[neilestrick]

2 hours ago, glazenerd said:

Neil:

bought my domain name a few months back. Now I get to spend next year organizing 500 pages of techno blather into a cohesive website for clay chemistry. That should  cure my habit of making information dumps. I have a new found respect for teachers (which I am not.) Figuring out how to present information gets more challenging than the information itself.

T

Teaching is difficult. I remember when I was first starting out teaching throwing, I had a really hard time explaining the details of what my hands were doing. It took a lot of work to define my movements and explain them in a way that others could understand. Just because you're good at doing something doesn't automatically make you a good teacher of it. Knowledge and sharing knowledge are two different things. The great thing about teaching, though, is that you can't teach without also learning. I became a better throwing once I analyzed what I was doing and learned to verbalize my techniques.

Looking forward to your info dump!

[C.Banks]

On 12/10/2018 at 8:45 AM, glazenerd said:

The accepted SiAl ratio for clay bodies is 4:1.

I know this can be a complicated subject but I'm curious what the range of acceptance is here.

I'm just a semi-educated bumpkin of sorts so when I see rules of thumbs I want to apply them and see how far off I am. The ^10 clay I've been working with for the last year or so is aproximately 70% silica and 24% alumina. So looking at 70/24 I'm curious how far off this clay is. I inherited a small database of college recipes and using the same simple math most of them are closer to 3:1. I'm wondering if my sometimes dimwit is leading me astray somehow here.

I appreciate some fo the complexities involved - the rest I muddle through as best I can as the need arises. What I would like to understand better is how accepted is a 3:1 ratio and what some of the implications might be for clays closer to, and lower than, 3:1.

[curt]

Nothing magic about 4:1.  Many different ratios will work, including 3:1.  Depends very much on what you need your clay body to do performance-wise.

Much more important in my view is the ratio of silica+alumina to flux, particularly for functional bodies.

And what kind of fluxes you are using - all fluxes are not created equal in terms of melting/vitrification power.

And particle size of the constituent raw materials is also right up there in terms of importance.

And...and...     There is no one size fits all.