Well, that was a successful week, I would say. We managed to run all the gypsum dehydration experiments we wanted (varying pore fluid pressure and differential stress), do a number of very nice KBr/KCl replacement reactions with Oliver Plümper and his students and also run the first real-life tests with our new triaxial rig Heitt Mjölnir, which performed extremely well. And it was really good to get out of Edinburgh again, and have a productive, fun time with our team. And, on the team, it was really great to see how much routine we have by now, and how smooth everything runs. There was not a single hick-up that cost us any significant time. Damien and I are working the experimental setups and maintenance quite smoothly, and Roberto is the king of data acquisition and management. And the chaps from Utrecht fit in perfectly, too. And support on the beamline by Federica Marone and Christian Schlepütz was perfect as usual.
We knew that we could run the gypsum dehydration experiments in Mjölnir relatively easily, but for the first time now we also had full control of the pore fluid pressure, having added a new component to Mjölnir that allowed us to feed fluid through the top piston and past the ram. Also, we now have, for the first time, properly used an LVDT on the rig, which measures the piston displacement very accurately. We had some minor issues at the onset when we ran the experiments at 115 °C and the reaction was too slow (probably inhibited to a degree by the pore fluid pressure), but we overcame these by increasing the temperature slightly. After that, everything went really smoothly.
I was particularly pleased with the tests on Heitt Mjölnir. We weren't quite sure whether we could actually see through the rig with the x-ray beam available at TOMCAT (modelling and initial tests with dummy parts left some doubts on how fast we could acquire images). But when we finally had the rig installed at the beamline and the light came on, it turned out that Heitt Mjölnir is definitely transparent enough, even with a copper jacket - excellent news. We did two deformation experiments at room temperature to test practical operations, and eventually settled on an imaging routine that will cover the entire sample (8x18 mm) with 2.75 µm voxel size in eight scans, each of which is 4 seconds. Every 3D dataset is going to be 100 GB, about 4000x4000x8000 pixels. Quite humongous!
Also, henceforth we will not be bringing hard drives anymore, as we will transfer all data with Globus, a grid-ftp service, and which allows us to transfer the incoming data directly to our data store in Edinburgh.
Now we just have to manage the challenges that come with UK immigration these days, and as of tomorrow enjoy our compulsory ten days of self-isolation...
