Short Course: Additive Manufacturing of High-Performance Ceramics

Course Presented by American Ceramic Society

Thomas Henriksen
Thomas Henriksen

Thomas Henriksen Ceramco’s President and Ceramist will be sharing Ceramco’s real-world experiences in additive manufacturing of ceramics with the attendees. Thomas is one of the eight speakers, representing various disciplines, speaking at this short course presented by the American Ceramic Society.

Attendees will learn about the economic and technical aspects of additive manufacturing (AM) ceramics, with an overview of state-of-the-art production methods, gaining insight into its opportunities and limitations. A basic understanding of appropriate applications, systems implementation into their production chain and the pros and cons of different technologies will also be explored.

April 29, 2019—8:30 a.m.–4:30 p.m. | Cleveland Airport Marriott, Cleveland, Ohio | Held in conjunction with Ceramics Expo 2019

Ceramco was the first manufacturer in the United States to have AM (additive manufacturing) capability for ceramics. Recently they doubled their AM production capabilities, adding a second machine.


For more information about Ceramco, please contact Thomas Henriksen:
+1 603-447-2090

Ceramco, Inc.
1467 East Main Street
Center Conway NH USA 03813

Interview with Shawn Allan, Vice President, Lithoz America LLC

Ceramco’s President, Thomas Henriksen, discusses with Shawn Allan the current and future technological advances in lithography-based ceramic manufacturing (LCM). This process is also called: “Additive Manufacturing’ and “3D Printing.”

Mechanical Properties of Certain Ceramic Fasteners


Ceramic materials excel where metallic or plastic substitutes simply cannot provide the necessary performance. High wear situations, electrical or thermal insulating applications, corrosive environments, or instances where magnetism, biocompatibility, or outgassing are concerns for many cutting-edge fields. As the extreme performance demands of oil and gas extraction, chemical detection, or deep space exploration (to name a few) go un-met, engineers and designers are turning to ceramic materials to address the performance gaps. Ceramco’s line of stock solid ceramic fasteners are the coupling solution when the standard falls short.

Ceramco’s solid ceramic fasteners continue to generate considerable interest since they were offered commercially in 1990. With increased interest and exposure has come an increasing desire for mechanical properties (such as recommended installation or service torque and tensile strength) from customers seeking to utilize these fasteners in their applications. To better understand the capabilities of our products and meet the demands and requests of our customers who are becoming more technically advanced in aggregate, Ceramco has obtained unbiased, third-party data based on industry standard test methodologies.  Initial data was used for the below listed values. As testing continues this data will help drive and guide material and process development within the company and expand our product offerings.

Additional preliminary research and prototyping of Ceramic Matrix Composites (CMCs) and Zirconia Toughened Alumina (ZTA) fasteners has already begun.


An initial 800 piece lot of high purity alumina (99.8% Al2O3) fasteners was submitted to an independent laboratory to begin testing, with supplementary lots of varied materials (stabilized ZrO2) and sizes forwarded thereafter. Submitted lots were evenly divided for destructive Tensile and Torsion testing. Sizes and compositions tested to-date are as follows:

SizeLengthTypeMat’l. Tested
5/16"-183.000"Hex BoltAlumina
1/4"-203.125"Hex BoltAlumina/Zirconia
#8-322.000"Hex BoltAlumina
#6-321.500"Hex BoltAlumina
#4-401.063"Hex BoltAlumina
M5 X .840mmHex BoltAlumina/Zirconia
M4 X .740mmHex BoltAlumina/Zirconia
M3 X .525mmPHPAlumina
M2 X .410mmHex BoltAlumina

Test Method

Engineered ceramics of the high-purity oxide type have high relative hardness and correspondingly low ductility. As such, test results would show no definable separation between proportional and elastic limits, thereby limiting the utility of NDT. Acknowledging these inherent issues from the outset, Ceramco elected that all testing would be destructive.

Testing was conducted per the methodologies outlined by ASTM F606/F606M Standard Test Methods for Determining the Mechanical Properties of Externally and Internally Threaded Fasteners, Washers, Direct Tension Indicators and Rivets.

Tested sizes were selected based on historic customer demand (though length was determined by the test facility based on fixturing needs) and tailored to ensure that the range of tested sizes would allow for reasonable extrapolation for uncommon and non-standard sizes. Test specimens were randomly selected from current and former production lots, packaged to maintain randomization, labeled for size and material and delivered to the test facility.

Test Summary

In general, obtained data well represented the normalcy and distribution anticipated of engineered ceramic materials tested in tensile and torsional modes. Similarly, the obtained data had a high degree of conformance with data obtained via internal testing, confirming the validity of historically accepted values and the verification practices employed by Ceramco for both production materials.

Additionally, the material performance was stable and predictable, with good correlation between the tested fastener cross section and the Tensile or Torsional strength.



Tensile strength is not often requested by potential customers since ceramics are not traditionally used for their performance in this area. However, tensile strength is one of the most commonly used measures of material performance which will allow for more appropriate design and use if known, or if retrofitting an existing assembly using other materials. Further, understanding the tensile strength of ceramic fasteners will help Ceramco continue to innovate by establishing a baseline against which new or novel materials may be measured.

As noted above, the test facility recorded and reported the maximum values obtained during destructive testing.

Being sufficient for most ceramic fastener purposes, the high purity alumina (A998) fasteners were tested and analyzed first, returning an average tensile strength of ~15.4 ksi. Tensile strength in this range was anticipated based on known and established alumina ceramic material properties.

Of particular note is the tensile strength of Ceramco’s aluminum oxide ceramic fasteners when compared to other common insulating materials. The Ceramco alumina sacrifices no tensile strength while exhibiting significantly higher maximum service temperature. For example, when compared to Polyether Ether Keytone (PEEK) Ceramco’s alumina has a comparable tensile strength, but an operating temperature over three times higher (16 ksi /480°C vs. ~15.4 ksi /1650°C).

Making use of zirconia (ZrO2) in the manufacture of ceramic fasteners allows Ceramco to enhance fastener performance for more rigorous and/or unique applications. Testing reported an average tensile strength of 21.7 ksi for Ceramco’s stabilized zirconia fasteners. Comparatively, this represents a 29+% increase over the average tensile strength observed in the tested alumina fasteners.

For customers seeking to load ceramic fasteners in a purely tensile mode, or whose application may induce tensile loading during use or operation, Ceramco has delineated the below listed Proof Strengths. Customers should ensure that their application does not exceed the values listed.

High Purity Alumina |High Purity Zirconia 
SizeProof (psi)|SizeProof (psi)
#0-8018.06|#0-80- - -
M2 X 0.426.86|M2 X 0.438.17
M3 X 0.544.94|M3 X 0.563.42
M4 X 0.760.81|M4 X 0.785.55
M5 X 0.878.89|M5 X 0.8110.86
M6 X 1.094.76|M6 X 1.0132.99
5/16-18121.33|5/16-18- - -
M8 X 1.25129.82|M8 X 1.25- - -
3/8-16149.13|3/8-16- - -


As mentioned above, monolithic ceramic materials are inherently brittle. This fact alone makes proper installation torque critical to the function and utility of a ceramic fastener, so it is of little surprise that torque requirements are chief among fielded fastener-related inquiries at Ceramco. Luckily, torsion is the most tangible and easily measured mechanical property an end-user is likely to encounter while working with fasteners. Established limits allow a customer or end-user to simply set a torque wrench and install the fastener without over-tightening and shearing the head off.

As with tensile testing, all torsional testing was destructive. Recorded torsional data on alumina fasteners displayed good correlation with historic data gathered internal to Ceramco, returning torsional strengths between approximately 2 and 44 in*Lbf., respective of fastener size.

As with tensile testing, the results from torsional testing showed increased performance in the zirconia fasteners. On average, zirconia exhibited 28% more torsional strength than a corresponding size alumina fastener. It should be noted that, while there is an apparent inversion point in the comparison of ZrO2 versus Al2O3 around Ø.230, this is due to a slight over-performance in the ¼”-20 tpi alumina fasteners tested.

Ceramco recommends consistent and progressive loading when applying torque to ceramic fasteners, as erratic or off-axis loading may lead to premature fastener failure. It should also be noted that perpendicularity and parallelism between joined members is crucial to mitigate application of bending moments which can have a deleterious effect on ceramic materials. Ceramco has established the below listed (maximum) service torque values for customers whose application(s) may be dependent on specific torque or preload requirements.

High Purity Alumina |High Purity Zirconia 
SizeSCV. Torque (in*Lbf)|SizeSCV. Torque (in*Lbf)
#0-800.13|#0-80- - -
M2 X 0.40.35|M2 X 0.40.5
M3 X 0.50.98|M3 X 0.51.93
M4 X 0.71.97|M4 X 0.75.48
M5 X 0.84|M5 X 0.88.4
M6 X 1.010.95|M6 X 1.015.43
5/16-1825.74|5/16-18- - -
M8 X 1.2527.04|M8 X 1.25- - -
3/8-1640.42|3/8-16- - -

Strength of ceramic fasteners

Ceramco manufactures both custom and stock fasteners out of alumina and zirconia, and although these are high performance materials, they are not going to have strength properties similar to your traditional fasteners.  The tensile strength and/or torque values for these products have not been reported inasmuch we’ve only been able to advise customers on their usage, such as, be careful not to over-tighten them.  And it was up to the customer to be able to translate the typical mechanical properties of these materials to evaluate the material limits.

This is no longer the case!  We now have torque values to report to customers (we did actually provide a small chart with torque-to-failure values before, but those data were from in-house testing, and not to be confused with an installation torque).  We also have tested the tensile strength, which previously had not been reported in any form.  Individual papers on the torque strength and tensile strength will be published in the near future based on the test results.

In general, the fasteners performed as expected and in the case with torque values, were at least in correlation with previously measured values from over 10 years ago.  The strength of the zirconia fasteners compared to the alumina fasteners was lower than expected, and has raised the question as to how fasteners made in ZTA (Zirconia Toughened Alumina) will perform.  ZTA is a composite which combines the strength and toughness of expensive zirconia with the good performance and low cost of alumina.  Testing of a ZTA fastener is underway and to be reported in a later paper.

Other Test-Related video and Images:

A Color Theory for High Purity Aluminum Oxide (Al2O3, A96, A998, High Alumina)

A998 Color Theory

The color of a ceramic is a science in itself, and unless you are you are in the ceramics business you probably wouldn’t find this topic interesting, but this blog is intended to for both ceramic producers and consumers.  We don’t want to get too technical with the details and hopefully you will find this interesting.  First, some background info:

High purity aluminum oxide (Al2O3) in full density, sometimes called “high alumina”,  ranges from about 94% to around 99% in the technical ceramics industry.  It is common to see a 96% and a 99.8% (like Ceramco’s A96 and A998 materials) where the trouble-maker as you might guess is the 99.8% grade.  Color varies with the purity level, as shown below.  By the way, it is very difficult to get good photos of ceramic materials!

From left to right: 96%, 99.7%, and 99.8% alumina samples

From left to right: 96%, 99.7%, and 99.8% alumina samples

The hallmark color of 99.8% purity aluminum oxide is ivory.  No other high purity alumina has this same color, it is unique such that if you go up or down in purity, the color changes.  Color therefore must be a function of purity level?  Ceramic manufacturers struggle with this color pandemic and I’ve heard a lot of theories on what causes the color variation:

  • Furnace plates/setters
  • Furnace elements
  • Contamination of raw materials
  • Dispersion/mixing issues
  • Binder issues
  • Density issues
  • Particle size issues
  • Astrological configuration when the ceramic was made
  • Process moods

Questions remain. Is the problem specific to a process, or should any process be able to make ivory A998 without issue?

The three plates on the left have some white and ivory (mixed shades). The plate on the right is a “control” with uniform ivory color.

Color as a Function of Purity, A998 Still

We may have an explanation for why our A998 is coming out ivory currently.  A customer told us the material was tested and had purity of 99.6% which means we are in the 99.7% neighborhood.  99.7% alumina was made a few decades ago, sort of a gimmick in my opinion, as a high wear type of material – and it’s color was peach.  The test on our material may not be very accurate, however, since our material is not peach at all but it is showing a nice deep ivory.  I predict that we are in the 99.7XXX% purity range still (this can vary with raw material lots) and the test is probably not accurate enough.  We don’t have the data.

Problem solved?

Ceramco has been in business for 30 years now and a year or so (like when this blog was started) isn’t a very long time, at least for a ceramic company, and the issues we’ve had with our A998 material have been numerous but perhaps not severe enough to draw much attention from outsiders.  Except customers, of course.  In most cases, white A998 has been the red headed step child and even caused companies to seek alternate materials to have parts made in, but there is the occasional customer that actually wants the white A998 instead of the ivory… accusing the ivory of being “under fired” and the white being the ideal material.  Ignorance is bliss I guess!

It was intended to blog about every theory of why white A998 material was showing up uninvited and the corresponding inconclusive tests that were run but it’s hard to go back over now that we think we’ve cracked the code.  However a quick review of each couldn’t hurt, right?

Raw Material:

The pedigree of your starting raw materials to make A998 is crucial.  Manufacturers aren’t going to lift their skirt on this topic but most companies do their own powder processing anyway to get the results they need, however the alumina has to come from somewhere.  The less powder processing you do on your own, the more you have to depend on the supplier, and it goes without saying that the quality of your raw material will influence the quality of your end product.  Powder processing includes but isn’t limited to milling (grinding powder), sifting, spray drying, and additives.  At Ceramco we depend a lot on the raw material supplier and have to carefully characterize powders and select the best ones available because we don’t have much powder processing capability.  A theory I’ve heard about alumina with regards to the rogue color variation in it has to do with trace elements in the material, like ceria (cerium oxide), which is a naturally occurring element and well – depending on where the starting material (like bauxite) was mined from could vary just like anything mined from the earth.  Kind of like the character of a beer comes from the water used, alumina may have character depending on the starting material.

Talking about trace elements is almost pointless since can’t really do anything about it and production facilities won’t have the time or energy to nitpick.  We’ve tried using aluminas from many different vendors and the quality and service varies, especially if you are a small player in the game.  Don’t get me started on that, let’s look at the processing:


After we’ve done our batching/raw material prep work specific to our process the A998 is coming out white, we’ve looked atour alumina and can’t find anything wrong, so maybe it’s something we are doing to it (or not doing to it)?  We had a theory about the age/shelf life of our material.  Maybe a batch was sitting in the machine too long and some separating of the mix was happening?  Yupp, we could observe separation happening but even the non-separated mix exhibited the same white end product, so back to the drawing board.  We ended up trying half a dozen things involving the shaping method we use and basically arrived with inconclusive results and more questions.


Moisture, or humidity rather, is an interesting topic.  Anyone lucky enough to have a controlled environment manufacturing space has no idea what a seasonal shift means but for those of us that see cold dry winter weather along with hot muggy summer weather know what I’m talking about.  In an effort to keep humidity from hitting our parts we installed an air dryer… let’s just say it was worth a shot and the jury is still out as to whether it defeated the seasonal affect.

Firing in the Furnace:

We don’t call them kilns.  They are furnaces.  Kilns are in your pottery hobby shop.  Furnaces are the technical ceramic work horses.  Just thought I’d make that clear.

There’s a lot going on in a furnace at high temperature to say the least.  We use electric furnaces exclusively but you can get gas fired ones and there’s a lot to be said about the difference between each.  Let’s just cut to the chase and discuss density with the simplistic view that the function of the furnace is to fully fire (sinter) ceramics.  You have to go to high enough temperature to sinter A998 properly.  We’ve looked at furnace elements, furnace plates, furnace furniture (isn’t it weird how words are so close?  Seems like the word furniture should be exclusively used in furnaces, doesn’t it?), covering the parts, firing pads, “seasoning” the furnace, part location in the furnace, and firing schedules.  Wow have we been busy now that I think back on it!  Every time we opened the furnace door, it was like Christmas morning looking under the tree to see what Santa brought (and then the following curse words when you find white parts, but you asked for ivory parts).


The ivory color comes from the color centers of the crystal structure of the alumina.  Thus we have to make sure the alumina is fired to the proper density.  There’s a trick to doing this, and it’s pretty well known but perhaps not appreciated enough because we knew it all along.  Indeed there’s a very technical explanation and I promised not to go there so I guess we stop here and move on to the many other exciting projects we are working on.

Update – Gas vs. Electric

Thanks to some advice from several experienced ceramic persons we have more insight into the control of color in our A998 material.  It may have been mentioned before that using gas fired kilns versus electric can influence whether ivory color graces us with its presence or not.  What wasn’t understood before is why.  Well, it comes down to oxygen content, actually.  When you run a gas fired furnace, the combustion of the gas fuel produces airflow and reduced oxygen inside the furnace box.  Electrically fired heating elements however are not going to do either of those things (make airflow or reduced oxygen).  It’s apparently common in the industry to use gas fired furnaces to get the right color, i.e. those who have gas firing capability have not found the color of their high purity alumina to be an issue, or put even another way: if you are firing your 99.8% alumina with electricity and having color problems, try gas.

Well, we don’t have any gas firing capacity in-house.  So we fooled around with the air purge function of our electric kilns to try and mock airflow and oxygen content of a gas fired box.  The furnaces have a fan that is already installed which takes air from the outside and purges it through the box and exhausts out the top/chimney that has a mechanical hatch.  Turn the fan on, the hatch opens up, the air goes through the firing area and out the top.  We can control the speed of the fan with a potentiometer type dial on the furnace control panel.

Experiments were run, results are in, and color is good!  While there is room for improvement and better understanding, it would be done best by using an oxygen content monitoring device and we don’t have one, so theory will have to be good enough for now.