Tilt, Soak, Slide: When Does a Model Hillside Fail?

Investigation question: How does the amount of water in sand change the slope angle at which a model hillside fails, and can keeping rain out of a slope or letting it drain away delay the failure?

Scientific purpose

Students build a bed of sand in a lined tray and raise one end slowly until the sand fails, measuring the angle from the height of the raised end. Each group tests two of five water contents, from bone dry to saturated, three times each, and the class pools the results into one graph. In the second session groups rain on a damp slope held at 30 degrees, first bare and then with one mitigation (a plastic cover with a gutter, buried drains, or a toe wall), and count how much rain each slope takes before it fails. The data show that a little water strengthens sand while saturation weakens it, which explains why slopes fail at the end of long wet spells and why engineers fight landslides by managing water.

Why this helps students with this chapter

Students usually think water 'greases' a slope, so any water is bad. Damp sand standing at a steep angle and saturated sand sliding at a gentle one is a surprise they can see and measure, and it forces a better explanation: water films pull grains together, while water filling every gap lightens the grain-to-grain grip and adds weight. Measuring angles turns 'steep' and 'wet' into numbers, the same move scientists make when they map landslide potential. The rain test then links the cause to the cure: if water is the trigger, keeping it out or letting it drain should help, which is exactly how La Conchita's walls and Portuguese Bend's dewatering wells differ.

Standards connections

MS-ESS3-2 (practice, analyzing data on the conditions that trigger slope failure and using them to judge mitigation): Students analyze pooled failure-angle data and rain-to-failure counts to identify the conditions that make slopes fail, then use them to judge which mitigation works and what a landslide watch should monitor. The data come from a model, not from records of real hazards, so this builds toward the PE rather than assessing it.

MS-ESS2-2 (supports, a rapid, small-scale surface change and the threshold that sets it off): The tray shows a slope staying put for a long time and then changing in seconds once a threshold of angle or water is crossed, a small-scale example of the sudden surface change the clarification statement describes.

MS-ETS1-2 (supports, a first, informal comparison of mitigation designs): Session 2 compares three slope mitigations against one criterion, rain taken before failure, using class data. It previews the fuller systematic evaluation in Lab 2 and Lab 3.

Roles (rotate each session)

• Lifter: Raises the upper end of the tray slowly and evenly, about 1 cm every 5 seconds, and holds it still when the Watcher calls.
• Reader: Holds the ruler upright at the raised end and reads the height of the tray's bottom edge at each call, to the nearest 0.5 cm.
• Watcher and timer: Counts the 5-second steps aloud, calls 'crack' at the first crack or rolling grains and 'fail' when sand passes the tape line, and describes the kind of movement.
• Recorder: Writes every reading, converts heights to angles with the table, computes the means and posts them on the class chart.

Materials

Estimated cost: 30 to 55 per class (estimate) (an estimate; Three 50 lb bags of washed play sand at about $6-8 each, a roll of non-slip shelf liner about $6, zip bags about $4, and disposable roasting pans at about $2 each if the school has no trays. Rulers, cylinders, cups, straws, craft sticks and goggles are classroom stock. Prices are typical U.S. retail estimates; the sand is reused for Lab 2.)

Safety

• Wet sand is heavy: a tray with a liter of saturated sand weighs about 2 kg, and the bucket of used sand much more. Keep trays low on the tabletop and carry buckets with two people.
• Wear goggles throughout. Sand can fly when a slope fails, and grit in the eyes is the most likely injury. Never blow on sand to clear it.
• Use washed play sand, not crushed silica, construction sand or sandblasting sand. Keep the sand damp where possible and do not pour dry level A sand from a height; fine dry dust can irritate lungs, so a student with asthma should handle only damp bags.
• Spilled sand and water make floors slippery. Work over paper towels and sweep or wipe spills at once.
• Do not pour sand or sandy water down a sink; it clogs drains. Let it settle in the bucket and pour the clear water off outside.
• Pushpins for the drains are sharp; only the student assigned the drain kit handles them, and they go back in a closed box. Wash hands after the lab; no eating or drinking at the trays.

Teacher preparation and setup

1. The day before, premix the sand in sealed zip bags, 1,000 mL of dry play sand per bag, measured with a measuring jug: A = no water; B = 50 mL; C = 125 mL; D = 250 mL; E = add water slowly until a thin film just appears on the surface after mixing (usually 300-400 mL; write the amount on the bag). Knead each bag to mix and keep it sealed so it does not dry out.
2. Assign two levels per group so each level is tested by at least three groups. For eight groups: G1 A and C, G2 B and D, G3 C and E, G4 A and D, G5 B and E, G6 A and B, G7 C and D, G8 D and E.
3. Cut the shelf liner to fit each tray and put a tape line across the tray 20 cm from the upper end. The sand bed fills the upper 20 cm; the lower part is the runout zone.
4. Measure each tray's bottom length, from the pivot edge on the table to the raised edge under the ruler; flared roasting pans are shorter at the bottom than at the rim. If it is not 30 cm, make a new height column with height = tray length x sine of the angle, using the sine values printed on the recording sheet.
5. Poke the rain cups and check that 50 mL drains in roughly 20-40 seconds; aim for all cups within 10 seconds of each other.
6. Make the class chart: five rows (A-E) with space for each group's mean failure angle.

Investigation design

Student procedure

1. Session 1. Each student writes a prediction for q1. Then put on goggles, lay the liner in the tray, and set the tray flat with its lower end butted against a heavy book so it pivots without sliding.
2. Open your first bag, knead it, and spread the sand in the upper 20 cm of the tray. Level it with a straightedge to about 2.5 cm thick and press it flat with five light pats of an open hand. Build every bed the same way.
3. Tilt test: the Reader holds the ruler upright at the upper end. The Lifter raises that end about 1 cm every 5 seconds as the Watcher counts. The Watcher calls 'crack' at the first crack or rolling grains and 'fail' when sand crosses the tape line by 2 cm. The Reader reads the height at each call.
4. The Recorder converts both heights to angles with the height table and records how the sand moved: grains rolling down the surface, a block sliding, a slump that bulges at the toe, or a flow like wet cement.
5. Lower the tray, scoop the sand back into the zone, rebuild the bed and repeat for three trials. Then do three trials with your second bag. Reseal each bag between trials.
6. Compute the mean failure angle for each of your levels, post both on the class chart, and copy the class means into the class table. Plot the class means on the graph and answer q2.
7. Session 2. Time how long 50 mL takes to drain from your rain cup and record it. Build a bed from a fresh bag of level B sand and prop the upper end on books at 15.0 cm, which is 30 degrees.
8. Rain test, bare slope: hold the rain cup about 10 cm above the upper half of the bed and pour in 50 mL. When it has drained, that is one dose. Keep dosing, up to 10 doses. Record the dose number at the first crack, the first water seeping out at the toe, and failure (sand crossing the tape line by 2 cm).
9. Rebuild with your second bag of level B sand and add your assigned mitigation. Cover and gutter: lay plastic wrap over the upper two-thirds of the bed and fold the index card into a gutter along its uphill edge that carries water off to the side. Drains: poke holes every centimeter along three straws and push them into the base of the bed from the toe, sloping slightly downhill, with their ends sticking out. Toe wall: tape eight craft sticks side by side and push them upright into the sand along the toe.
10. Repeat the rain test on the mitigated slope. Post both results on the class chart, copy the class results, clean up as directed, and answer q3 through q10 individually.

Expected observations

• Expected ranges, not yet tested in a classroom: dry sand (A) fails at about 30-35 degrees as grains avalanche down the surface; damp sand (B and C) stands much steeper, often 50 degrees or more, and fails by cracking or sliding as a block; wetter sand (D) fails lower; saturated sand (E) fails at the lowest angles, roughly 15-25 degrees, by slumping and flowing.
• First cracks usually appear a few degrees before failure in damp sand but almost none appear in dry sand, which fails grain by grain.
• In session 2, bare damp slopes at 30 degrees typically crack and then seep at the toe before failing after several doses; slopes with a cover often survive all 10 doses; drains and toe walls usually delay failure by a smaller amount.

Likely difficulties

• Uneven lifting. A jerk can trigger failure early. Have the Watcher count steadily and the Lifter use two hands; discard and repeat a trial that was jerked.
• Damp sand may hold even at 80-90 degrees. Record 'no failure by 90' and use 90 for the mean, noting it as a lower limit. Above 70 degrees the height barely changes, so readings there are rough.
• Saturated bags separate as water settles. Knead the bag right before each trial, and scoop some of the pooled water in with the sand so the bed stays saturated.
• Groups disagree about when 'fail' happened. The tape line rule (sand crosses it by 2 cm) settles most arguments; the Watcher's call is final.
• Session 2 rain cups drain at different rates. Count doses rather than seconds, and compare each group's mitigated run with its own bare run.

Structured debrief

1. Put the class graph on the board. Ask a group with level B and a group with level E to describe their failures side by side: a damp bed that stood steep and then cracked or slid as a block, against a saturated bed that slumped and flowed at a gentle angle. Then ask the class to state the pattern in one sentence: failure angle rises with a little water, then falls as the sand approaches saturation.
2. Draw two pairs of grains on the board: one pair with a thin curved water film joining them (damp), one pair floating in water that fills the gap (saturated). Explain that the film pulls the grains together, while the water filling every gap pushes outward and lightens the contact between grains, so friction drops; the water also makes the slope heavier. Reject 'lubrication' explicitly.
3. Compare the session 2 mitigations. The cover usually wins in the tray because it keeps nearly all the rain out; drains help by letting water escape at the toe; a toe wall props up the base but does not stop the slope from saturating. Ask which of these real places used which idea: plastic sheeting on Southern California slopes in storms, Portuguese Bend's dewatering wells, La Conchita's wall that was overtopped in 2005.
4. Close on forecasting: if failure depends on how much water is in the slope, then rain totals over days and soil moisture are the data to watch for most slopes, and short-burst rain intensity is the data to watch on burned slopes. That is how landslide and post-fire debris-flow alerts are built.

Recording sheet: what students fill in

Lab 1 recording sheet: Tilt, Soak, Slide

Your bed is a model hillside. The failure angle is the slope at which sand crosses the tape line by 2 cm. Find the angle from the height of the raised end with the table below. Water levels are mL of water mixed into 1,000 mL of dry sand: A 0, B 50, C 125, D 250, E saturated (amount written on the bag).

Analysis questions: answer key

Individual evidence of learning

Each student submits the completed recording sheet, the class graph and answers to q1 through q10. The graph with q4 and q5 is the evidence for explaining how water content changes slope stability; q6 is the calculation; q7, q8 and q10 show the student using the model's data to judge mitigation and to choose what to monitor for a real slope.

Assessment criteria

Accessibility adaptations

• Provide the height table with 5-degree rows highlighted and a calculator for means; let a student with dyscalculia use a pre-made height-to-angle strip taped to the ruler.
• Students with limited strength or fine-motor control can take the Watcher or Recorder role; a partner builds the bed while they direct it.
• For low vision, tape a high-contrast ruler with large centimeter marks upright beside the tray, and have the Reader call every reading aloud.
• Students who cannot handle sand (skin conditions, sensory needs) can run the timing, recording and graphing, and observe with goggles from a step back.
• Accept a labeled photo series of a failure, with spoken commentary recorded on a device, in place of the written description for q2.

Low-cost alternative

If trays or sand are short, run session 1 as a demonstration at the front with two trays (dry and saturated) while groups predict, record and graph, and use clean playground sand dried in the sun. A cardboard box lid lined with a plastic grocery bag can replace the tray for dry and damp levels. Session 2 can use one tray per two groups, each group running one bare and one mitigated slope in turn.

Model limitations

This activity has not been physically tested by the authors; the angles in the key are expected ranges based on how sand behaves, and classes should trust their own data. The tray is a model: real hillsides are hundreds of meters long, made of layered soil, clay and broken rock rather than uniform sand, held together by roots, and fed by groundwater flowing in from uphill over days or weeks. Real slopes often fail along a buried weak layer, which the liner only roughly imitates. Clay-rich soils behave differently from sand, and the model cannot show slow creep or the effect of earthquake shaking. The 50 mL doses fall far faster, relative to the bed's size, than real rain.

Teacher notes

The tilt angles depend on the sand, the liner and the packing, so judge students on correct measurement, clear patterns and explanations rather than matching the expected ranges. If a level B or C bed will not fail before 90 degrees, record 90 as a lower limit; the pattern still holds. Keep q5 central in the debrief: it is the explanation students will reuse for every slope in the chapter. For q6 accept class data. The session 2 cover often 'wins' by a lot; push students in q7 to discuss why real slopes cannot be wrapped in plastic for years.