Ronni Grapenthin - Notes

email:
New Mexico Tech
Dept. of Earth &
Environmental Science
801 Leroy Place
Socorro, NM-87801

Dike Injections with a Twist

Published: 2016/09/08

I have been intrigued by analogue models for a long time. They allow observation of the evolution of processes and can tell us a lot about what might be happening below the surface of the Earth. Given that I am teaching a volcanology class this semester it seems only appropriate to pack the labs with as much analogue models as possible. Here's one that I developed with my TA this last week. I'd be surprised if this was this first time anybody did anything like this, but I didn't find anything in the literature. So, here we go: Dike injections with a twist.


The Idea

Dikes turn into sills when they encounter, for instance, rigidity or density contrasts as they rise to the surface. The sill keeps propagating along the bottom of the obstruction until it can clear it; say once it made it past the edge (or it stalls and the action is over). Here it likely turns into a subvertical dike again (90 degree turns are difficult / unlikely for fluids unless forced). The students should be able to observe the processes in the lab. My TA and I came up with a setup that works quite well to model this using gelatin and Jell-o. If left standing over night, the diffusion between intrusion and host material might be a good analogue to conduction of heat.

The Setup

We used pure gelatin as the crust / host rock and Jell-O in various colors as the intrusion. Now, you may think that it's silly to intrude Jell-o as it's mostly gelatin, but it also contains sugar and food coloring, which changes the properties slightly. Plus, you break it up and suck it into a syringe, which makes it less viscous than the host rock.

The key to this experiment is that we initially make a batch of gelatin (~8 liters) and let it cool enough / almost solidify before we add the cover page of a magazine (EOS, thanks AGU!). In parallel we created a second batch of gelatin (~4 liters) to pour on top. The key is that the second batch should be cool enough to not melt the top layer of the first batch. It didn't quite work, so I stabilized the paper with knives (go me!). Here's a picture (forgot to take a before image, so there's already a volcano that formed):

Arrested Gelatin Dike Annotated
Figure 1: Setup - magazine page (sturdy, coated paper) acts as rigidity contrast; knives hold the thing down, because the second batch of gelatin was still too warm, which threatened to have the paper rise to the top.

The Jell-o is injected from below with a 30 ml syringe (get a bigger one!) after we drilled a hole into the bottom of the container. Note that some of the Jell-o will still drain (fairly slowly though), so it's useful if someone plugs the hole with a finger, while magma is being reloaded.

Arrested Dike / Vent-Clearing Eruption

The first group of students produced what you see below in Figure 2. The dike made it to the obstruction, changed direction and propagated horizontally toward the edge. Here, it came up at about 45 degrees to make it to the surface.

Now, the student didn't just inject Jell-o, but also significant amounts of air. The larger air pockets actually migrated to the dike tip. Clear case of bubbles / volatiles driving the action. Most of the fracturing under and past the obstruction was created by the bubbles. They eventually made it to the surface (there's a fracture going all the way to the top, solid line in Figure 2b) mimicking what might be a 'vent clearing eruption.' At that point, the students had it with Jell-o injection (it's a significant amount!) and moved on to the next experiment. Hence, it's an arrested intrusion. It's notable, that the fracture that links to the surface went up at a 45 degree angle, which is what you'd expect in a homogeneous material when your injection changes from horizontal and is buoyant enough to make it to the surface.

A real crust's heterogeneities would obviously have something to say about this angle. Imagine, for instance, that a vertical / subvertical fault connects the edge of the paper to the surface. This could be a caldera ring fault. The dike would very likely follow this existing unconformity rather than making its own. We could model this by simply cutting into the gelatine. Maybe next time.

Arrested Gelatin Dike Arrested Gelatin Dike Annotated
Figure 2: Dike-sill-dike intrusion below rigidity contrast with a gas created fracture at the tip. Note little sub-sill near the bottom of the dike, which didn't propagate further.

Being the geodesist in the room, I have to point out the magnificent surface deformation that occurred during this intrusion! I marked the original surface (dashed line in Figure 2b) and the uplifted surface (solid line in Figure 2b). The embedded paper also deformed, but it's harder to tell due to its irregular surface. Here, the bulge is of larger amplitude, but smaller wavelength, that is, it's bigger, but more localized.

`Heat conduction' (i.e., diffusion)

Arriving the next morning I found that some of the dike material was moving into the gelatin and the dike edges have become a lot more fuzzy (see Figure 3, below). I will argue that this models heat conduction from the intrusion into the host-rock. How well it models this process is separate question and would require some actual sciencing to compare diffusion and conduction rates, but hey - this is to visualize the processes for education.

We had plugged the hole at the bottom with duct tape (yeah!) to stop the slow leak of Jell-o. Amazingly enough this worked and the surface deformation feature actually held up overnight!

Arrested Gelatin Dike Arrested Gelatin Dike 18 hours later
Figure 3: Comparison of dikes right after the experiment (left) and 18 hours later (right). Note fuzzy-ness of picture to the right due to diffusion of dike materials. The perspectives aren't quite the same. Also - check out that the surface deformation is pretty much the same - it held! Overnight!

Eruption

At last - the red stuff makes it to the surface!

The second group of students used the other end of the same gelatin container to create their intrusion. The syringes didn't have an opening at the tip, but along the needle, that may be the major factor in the different orientation of the dike (perpendicular to group 1; Figures 4 & 5).

Arrested Gelatin Dike Annotated
Figure 4: Dike outline. The group used a differently colored Jell-o (purple) that lost its color quite quickly (higher water content?) and so I added the outline of the initial dike here. In the middle you can see the needle. The magma pools under the obstruction and moves into the 45 degree to the left (their last injection was red Jell-o, better visible)

The procedure here was similar to the first dike, but the students actually pushed the material all the way to the surface. The interesting observation was the significant surface deformation as the intrusion moved shallower toward the surface (see Figure 5 for some remnants of this).

To some extent, we also had a race condition going on given that the dike extended beyond the edges of the paper (Figure 4). Although the needle injects the material below the paper, which determines the front of the dike, I wondered for a bit whether the sill forming and migration might take longer than the edge of the dike making it to the surface right away.

Figure 5 shows quite neatly how the strike of the dike was changed by 90 degrees due to the interference of the paper / stint as a sill. A real crust's heterogeneities would obviously have something to say about this. Imagine, for instance, that a vertical / subvertical fault connects the edge of the paper to the surface. This could be a caldera ring fault. The dike would very likely follow this existing unconformity rather than making its own.

Effusive Eruption
Figure 5: Same dike as in Figure 4, but different angle and it breached the surface. You can see the nice elongated dike below, some of the sill below the paper and then the dike that connects to the surface. Some `crust' still covers the central part with two lava flows symmetrical on either side.

rg <at> nmt <dot> edu | Created: 2016/09/08 | Last modified: September 08 2016 19:42.