Thin-Skinned Models for Undergraduate Teaching Labs – "Flour Structures"

Dan Davis, Dept. of Geosciences, Stony Brook University Summary Analog modeling of thin-skinned tectonics can be useful in teaching as well as in research. Because ‘sandbox’ models can be difficult and time-consuming to set up and exhume, they are typically not well suited for use in undergraduate teaching laboratories. The lab described here a simple, inexpensive way for an undergraduate class to create model thin-skinned wedges A simple model using flour and marker chalk, pushed across the base of a box using a piece of wood, can be used to demonstrate many aspects of the growth of thin-skinned mountain belts. Depending upon the strength of the décollement, the cross-sectional taper of the deforming wedge will be thin or broad, internal deformation will be mild or intense, and the structures either close to symmetric or strongly forward-vergent, just as in the analogous natural thrust belts. Learning Goals This hand-on lab will allow students to create model thin-skinned wedges and to: • Understand the relationship between boundary conditions and deformation in thrust belts.

-­‐ How basal friction controls wedge taper. -­‐ How basal friction affects the balance between forward- and backward vergent

thrusts. -­‐ How thrust belts can gain along-strike complexity.

• Use those observations to predict boundary conditions based upon surface observations • Directly measure strain and understand its relation to displacement. • Exhume their model mountain belts and observe their internal structure:

-­‐ Observe the relationship between thrusts and their surficial expression. -­‐ Appreciate repetition of strata as indicative of thrusting. -­‐ Map out forward and lateral ramps.

Context for Use This lab requires a full 3-hour lab period. It cannot be interrupted in the middle if the lab room is to be used in the interim by other classes. It is best suited to an upper-division course in structural geology, after the students have had an initial introduction to the concepts of thrust faulting and thin-skinned mountain belts. If the course includes a field trip to sites in a thin-skinned fold-and-thrust belt, the concepts explored in this lab can serve as a useful prelude to the field trip.  

 Teaching  Materials   These experiments require the following easily obtained materials: • Flour: It is best to divide students in groups of 2-4. One 25 lb bag

(inexpensive, and found at stores such as ‘warehouse’ stores such as COSTCO or Sam’s) will be sufficient for 3 groups.

• Sandpaper: About a dozen sheets of 9”x11” 60-grit sandpaper, available in the painting section of most hardware stores. (4 sheets cost ≈ $6, a back of 15 sheets ≈ $11)

• Graphite: A moderately small packet will suffice to be sifted to cover the ≈8 ft2 required between all of the experiments. I recommend however that you buy as large a container as feasible – like sandpaper, graphite is packaged in such a way that its unit cost drops dramatically with quantity (1 oz. costs ≈ $5 but 5 lbs cost ≈ $20)

• Transparency sheets: You’ll want to have about a dozen thin, slightly rigid sheets with which to excavate the models. Photocopier transparency sheets work well, but anything similar in size and equally (or more) rigid will do, such as thin aluminum sheets.

• Drawers/low-wall boxes: One per group (ideally 6 to 8 in total). I use geological sample drawers (typically between 20” and 30” on a side), with any holes in their bases plugged using duct tape. Any similar-sized box with low (2”-4”) walls will suffice.

• Marker chalk: Any flour-like colored material in three colors (say, red, green, and blue) will do. I use Irwin Strait-Line marking chalk, 5 lb. containers of which each cost about $12 and last several years.

• Kitchen sifters: You need 6-8 (one per student station). Each costs just a few dollars. • Brushes: One brush, 3”-4”, per station. Each costs about $5. • Plastic cups: About a dozen small cups for flour and marker chalk. • ‘Hinterland’ blocks: Each student station will need a thin (1/2”-1”) block of wood that the

students will push to cause the deformation. The height of the block should be 3” to 4”, and the width should be just slightly less than the width (shorter horizontal direction) of your box. Thus, a 30”x20” box should have a block about 197/8” wide.

• ‘Basement’ blocks: Two of the models require a thin block to mimic rigid basement rocks to be placed above the base of the box. These should be thin (say, 1/16”). I use Plexiglas, but any rigid material would do. You should have two such blocks with a long dimension just able to fit across the width of the box, and a short dimension considerably smaller than that. For example, if your boxes are 30”x20”, two 1/16” pieces of Plexiglas cut to 20”x6” would be ideal. Other shapes can also be useful (as in Figure 14 below.

• Duct Tape: Of course – duct tape is needed for everything, and this experiment is no exception!

• Clean-up equipment: You  will  want  to  have:  a  mop,  paper  towels,  brooms,  and  garbage  bags.

Teaching  Notes      Introduction Analog modeling of thin-skinned tectonics can be useful in teaching as well as in research. Because ‘sandbox’ models can be difficult and time-consuming to set up and exhume, they are typically not well suited for use in undergraduate teaching laboratories. The lab described here a simple, inexpensive way for an undergraduate class to create model thin-skinned wedges (Figure 1). Experimental Set-up Any large rigid box with low side walls will suffice for these experiments: we use drawers from geologic sample cases (60cm x 75cm, 5 cm deep). Each box is shared by 2-4 students. Each group of students sets up their box with a unique set of boundary conditions. After all boxes are ready, the models are run one at a time, with the entire class observing the running and excavation of each of the experiments.

One key parameter is the strength of basal friction. To study its effects, separate groups can set up experiments with the base as either:

1) the bare base of the box,

2) sheets of ≈60 grit sandpaper duct taped together (Figure 2) and then taped along its edges, face up, onto the base of the box, or

3) a minimal (≈0.1mm) layer of graphite that has been rained in from above using a sifter (Figure 3).

Before introducing flour into the box, a wooden block the width of the box and about 10 cm tall (see Figure 1) should be placed at one far end (the hinterland) of the box. Flour preserves the internal structure of the model mountain belt far better than the granular materials like sand that are better from the standpoint of strength scaling. The total thickness of all the layers of flour (typically about 3 cm) is not critical - it can vary, depending upon the depth of the box and the amount of flour available. It can be used as a separate

Fig. 1

Fig. 2

Fig. 3

variable – the scale of the structures produced depends upon the initial sediment (flour) thickness. Each layer of the flour (typically ≈1cm) is added using a sieve (Figure 4). The model length scale factor (≈105) is such that the cohesion of the flour is higher than ideal, as evident by the fact that after the experiment one can cut stable cross-sectional slices several few cm high. Thus, at each stage the flour should be smoothed very gently, using a soft brush (Figure 5), so as to avoid packing it and increasing its cohesion even further.

At two or three stages in constructing the model, a colored layer (athletic field marker chalk) is sieved across the central 3/4 of the box (Figure 6) to create a marker layer to be viewed when the experiment is eventually excavated. This colored layer should be as thin as possible to insure that it has no significant effect on the deformation. The intervening flour layers are typically no more than about 1 cm thick.

Friction from the side walls distorts the stress field within a couple of cm from the walls, so the structures there are of less interest and it is not necessary to have marker layers there.

It is useful to place thin colored lines at a uniform spacing on the top surface of the experiment (Figure 7), because changes in

that spacing during the experiment can be used as a direct measure of shortening in the model orogenic belt. I suggest a 10 cm spacing between hash marks to make strain calculations easy.

Fig. 5

Fig. 4

Fig. 6

Fig. 7

Any bright, colored granular material will suffice. We sometimes use a green play sand, which can be emplaced using a sieve or pastry bag - anything that makes it possible to place the colored material on the surface of the flour in a controlled manner. Photocopier transparency sheets are useful in excavating the experiment after it is completed (Figure 8). Running the Experiment It is now possible to start the experiment. A thin wooden sheet the width of the box and about 10 cm high is used to provide a push from the hinterland (see Figure 1). It should be pushed slowly and carefully, avoiding sharp accelerations – and despite the image shown, two hands are recommended. Someone else should hold the box from the far (foreland) end to keep it from moving. It is best to stop the experiment well before the frontal thrust reaches the far end of the box, so as to avoid having the deformation affected by that rigid boundary. Once the experiment is complete, it becomes possible to excavate the model and view the internal structure that developed as part of the deformation. Photocopier transparency sheets are very useful for this purpose (Figure 8). There are several things to note:

- A series of thrust scarps appear at the surface, with the youngest thrust toward the foreland.

- Any given location remains almost undeformed (with hash marks remaining 10 cm apart) until the frontal thrust approaches it.

- There is typically a small amount of shortening (compaction) even out in front of the frontal thrust. This can be thought of as analogous to compaction and layer-parallel shortening in front of a natural thrust belt.

- The most rapid strain is around the frontal thrust, though some deformation continues throughout the orogen.

- That deformation produces a wedge-shaped orogen above the décollement, as in the cross-sections in Figures 9 and 10. The taper of a wedge formed over sandpaper (Figure 9) is

Fig. 10

Fig. 9

Fig. 8

quite large and the wedge is tall and intensely deformed. A wedge formed over a weak base in graphite (Figure 10) has a narrower taper and a less pronounced preference for foreland-vergent thrusting. Thrusts in it are more widely spaced, the orogen is wider, and strain (measured using the surface hash marks) is smaller.

With a sufficiently large group of students, it is effective to have multiple stations, each set up by a team of 2-4 students. One group should produce a model with a mixed strong/weak base. The sandpaper and graphite may be arranged in any way, but I find it useful to divide them along strike. In this case, the experiment produces (for example) a steep wedge on the left and a broader but less steep weak-basal one along strike to the right - separated by oblique accommodation structures, as imaged in an along-strike cross-section in Figure 11. In this case, the cross-section has been cut normal to convergence, in order to see along-strike structural variability. It is best if the students assembling the various models keep their boundary conditions secret from their classmates (it helps if a piece of cloth, behind the wooden push-block, covers the newly exposed basal sandpaper and graphite from view). The other students can then try to determine the nature of the detachment on the basis of surface structures and strains. There are other interesting options. One is a ‘forearc’ model, in which a thin, rigid (e.g., Plexiglas) sheet the width of the box is placed on the base, directly in front of the wooden block (Figure 12a). This sheet then acts as a strong, low-angle ‘backstop’, leading to the formation of a two-sided orogen (Figure 12b) Alternatively, the rigid sheet can be placed in the foreland, to mimic basement relief (Figure 13a). Combined with a graphite base and a pre-existing taper, this produces a model orogen reminiscent of the Salt Range and Potwar Plateau (Figure 13b).

Fig. 11

In  this  case,  the  cross-­‐section  has  been  cut  normal  to  convergence,  in  order  to  see  along-­‐strike  structural  variability.  

Fig. 12b

Fig. 12a

The options are nearly limitless, and students enjoy exploring them. Each group should spend the first part of the lab setting up its own experiment. That should take a bit over an hour. It speeds things up if there are enough soft brushes and kitchen sifters for each group. Once all groups are ready, run the experiments in sequence, going from the simplest to the most complex

models. Allow the students time to take photographs and measurements (strain values, fault spacings, vergence, and dips, and wedge taper) at each model. No matter how hard you try, things can get a bit messy, so have mops, paper towels, brooms, and garbage bags handy – and make clean-up a required part of the assignment! Examples of models can be generated include:

- A model with high basal friction (sandpaper), as in Figure 9. Students can measure: the spacing and dips of thrusts, the surface strain (using the 10cm spaced hash marks), and wedge taper.

- A model with low basal friction (a very thin layer of graphite), as in Figure 10. Students can make the same measurements as in the strong base case, and compare results. Note that if the graphite is more than a most minimal coating, some of it is likely to become incorporated into one or two thrusts, lubricating them and allowing them to undergo a great deal of slip.

- A hybrid model, with an along-strike contrast in the strength. This allows direct comparison of the effects of basal strength along strike. This model will produce complex tear structures over the transition from sandpaper to graphite. It is a good idea to excavate this model so as to produce along-strike cross sections (as in Figure 11).

- By including a thin horizontal sheet of wood or Plexiglas (forearc basement rocks) in front of the pushing block allows generation of a model accretionary wedge, outer-are high, and forearc basin. This is illustrated in Figure 12.

- Adding a thin, fixed block in the foreland, as in Figure 13, allows exploration of the effects of basement topography.

- Use a thin basal block like the one in the forearc model, but in this case turn it sideways (Figure 14a) so it covers only a fraction of the widthe of the block. This will produce very pronounced shear and a strike-slip fault.

- Make a triangular basal block and place it on the base of the box against the wooden backsop and the side of the box (Figure 14b). This will produce a convergent margin with oblique convergence and strain partitioning.

Fig. 13a

Fig. 13b

Fig. 3 Fig. 3

Fig. 14a

- Make a basal block that is arcuate. In a 20”x30” box, this could be a thin Plexiglas sheet that is one quadrant of a 20” diameter circle (Figure 14c). If you place it against the wooden backstop block and the side of the box, it will produce a strain partitioned curved forearc.

The options are nearly limitless. Assign each group one model to construct and arrange them in rough order of increasing complexity, to be run in sequence during the last hour of the lab. Assessment The student handout asks the students to make observations, measurements, and interpretations of their own experiment – and to compare/contrast it with two other experiments. This requires a high-level understanding of the underlying physics. Did they calculate tapers correctly and interpret them logically in terms of the basal boundary conditions? Did they measure and interpret strain appropriately? Did they understand the differences in tectonic style, fault vergence, and strain distribution between the models they compared? Did they understand the model scaling and how it may or may not be appropriate?

Fig. 3

Fig. 14b Fig. 14c