A Work in progress - NanoIndentation

By the end of 2012, (and no Apocalyptic events so far) I am trying to put together results of first set of Nano indentation.

As mentioned earlier, 4 specimens from 60°C testing (3 from original mix design + 1 RILEM standard mix design) are used for analyses. Main problem following indentation process is matching mechanical properties to exact locations on specimens. So far success rate of indent attempts are a bit low (around 55-70%). This was mainly due to some last minute revisions on samples and disturbing the flatness of our samples. Yet there are enough data for our purpose. With our new sample holders I believe this rate will raise drastically.

For the last week or so, I have been working on a Matlab code which gathers thousands of indent data, crunches the numbers and groups curves in to several E Modulus values (with 30GPa intervals). I have been working on a standardised way to identify (hypothesise) a way to identify locations of indents. As they are not (or very hardly visible) through light microscope, we needed to randomly distribute a grid on indent area, then check whether it is statistically comparable with NanoIndenter results.

Here is how it works.

\begin{example}

We had original documentation of our samples where interested indent areas are marked.

Micrograph image under stereo microscopy MagX0.63 - Each grid represents 4mm­² area

After indentations is performed another micrograph is taken with a higher magnification (4X) for detailing indent area alone.

Micrograph image under stereo microscopy MagX4 - Each cross is placed 200microns from each other as indentation points were

It is hardly visible from this image but manual point counting is applied on this image where each 200 microns increment is marked digitally. Finally we end up with two images for matching and comparison purposes. First one representing hypothetical  distribution of indent points and their corresponding material phases. Second one a representation of how E-modulus values vary through indentation grid.

Point count results

Surface map diagram of Indent results for E (GPa) values

This methods allows a standard way to overlay test results with real material properties. This can be further investigated by correlating findings with following subgroups of indent results shown in subplots.

\end{example}

Next set of experiments will start in Jan 3rd 2013. Till then,

Merry Christmas and a happy New year to All.

Nano indentation

Recently I have been working on ways to identify material properties of phases in my concrete samples. As of yesterday we started with our first Nano indentation test. The main aim of this study will be to identify dissolved properties of aggregate and paste in the matrix with known exposure to Accelerated test conditions. Here is a short photo series of how we prepared samples for nano-indenter.

We cut sections from expansion prism (roughly 1.5 cm thick)

Sections are organized named and further selections are marked

Smaller Specimens are prepared to fit the sample holder for Nano Indentation

Mechanical Properties Degradation due to ASR-Swelling: The show has been started!!

After several trails (e.g. Trial Tests 1 ) and a long preparation, this week the first sample have been casted.
Just to give some numbers:
- 425 lt. of concrete will be produced
- in 6 casting along 4 weeks
- 107 sample
- 14 Young's modulus test, 14 splitting tensile test, 6 compressive strength tests will be performed along 52 weeks.
- 5 home made reactor will host the specimens during the storing time.

Here you can find some pictures of our work:

Home Made reactors.

Heater (bottom) and Controller (top) outside view.

Heater (bottom) and Controller (top) inside view.

Small Boxes in which the sample are stored.

Mould 10x10x40 cm for sample used in the Young's modulus test.

Mould 15x15x15 cm for sample used in the splitting tensile strength.

Mechanical Properties Degradation due to ASR-Swelling: Trial Tests 1

Our experimental campaign is becoming larger! We decided to have look also to the degradation of the mechanical properties due to ASR-swelling. 

We will store the specimens in homemade reactors which ensure a temperature of 38C and RH>95%.

Homemade reactors.

The specimens will be tested for Young's modulus and splitting tensile strength.

Just to get familiar with the type of test and the test machines, we had a trail cast at the middle of June and that week we performed the experiments.

Here there are some pictures of the tests.

Preparation specimens for the Young's Modulus test.

Disposition of LDVT for the Young's Modulus test.

Young's Modulus test.

Splitting tensile test.

Crack pattern after splitting tensile test.

Crack pattern after splitting tensile test.

14th ICAAR - Austin, Texas

ICAAR conference series continued in Austin, Texas. Thanks to University of Texas, it has been a remarkable organization from 21st to 25th May. It has been a warm but a knowledgeable week for PAT-ASR team. We had the opportunity to experience interesting researches from around the world. We have share knowledge and connected with researches from our field. Our team also presented 2 papers during the conference and collected useful feedback. one of the highlights of the conference was the visit to Texas University ASR Exposure site (see images). They showing a considerable effort and dedication to run this site.

Over all, it has been a informative week for everyone.

11th ICOLD Benchmark Workshop

The PAT-ASR team has participated to the 11th ICOLD Benchmark Workshop, which took place in Valencia (Spain) during the 3rd International Week on Risk Analysis, Dam Security and Critical Infrastructure Management (http://www.ipresas.upv.es/IIIsemana_E.html).

The numerical problem proposed in the frame of Theme A of the 11th Benchmark Workshop is very representative of those raised to engineers by swelling due to Alkali Aggregate Reaction (AAR). The example selected is that of Kariba dam, a large arch dam built on the Zambezi River between Zambia and Zimbabwe in the late fifties, which has shown evident signs of swelling soon after the starting of its operation.

The Kariba dam.

The aim of the benchmark was to predict the displacement of the crest's dam in 2010, on the basis of measured value given between 1962 and 1995. A very special feature of the Kariba dam case is the apparently strong anisotropy which exist between the vertical and horizontal swelling rate. This has been interpreted as being due to the stress dependence of the swelling process.

We analyzed the dam with a modified version of the Finite Element Software DIANA.
In our model the definition of the volumetric expansion due to AAR is provided through the Larive's kinetic law. The three parameters of this law are calibrated in order to fit the dam's displacements.
The anisotropic effect due to the stress state was studied with different approaches based on the Saouma and Perotti's formulation.
Our model gives one of the best prediction in terms of vertical displacement.

Location of the main instruments.

Vertical displacement of the crest's dam.

Radial displacement of the crest's dam.

At this conference researchers as well as many practical engineers have participated, both of them really interested in AAR!
Here you can find some picture of the event:

Our presentation given by Rita Esposito (left). On the right Christine Noret, the chairman of Theme A.

Some participants with the chairman of the conference Ignacio Escuder Bueno (in the middle).