Video Imagery Video looking straight down onto rubble pile showing: - - PowerPoint PPT Presentation

Video Imagery

• Video looking straight down onto rubble

pile showing:

– Very dynamic conditions within rubble pile – No evidence of circumferential cracking

• Rubble pile edge is about 5 m from edge of cone

– No evidence of ride-up against cone, circulation of rubble pieces is occurring away from cone (stationary ice pieces against cone)

Video Imagery

• Side view video showing rubble pile
• Confirms behaviour seen in previous video

– No ride-up – Stationary ice pieces at top of pile against cone – No circumferential cracks except one of limited extent where the ice is failing against the outer slope of the rubble pile

Video – Ice Under Sheet

• Another video looking down on thinner ice

(about 20 to 30 cm) failing against the cone.

– Very clear evidence of rubble below the ice sheet, extending out further than the rubble pile on top of the ice sheet. Similar to slide 26 but much thicker ice.

Panel Pressures - Cone

0.2 0.4 0.6 0.8 1 1.2 1.4 58:19.2 58:23.5 58:27.8 58:32.2 58:36.5 58:40.8 58:45.1 58:49.4 58:53.8 IFP13Z5c85 IFP15Z1c90 IFP17Z5c93 IFP17Z1c94 IFP13Z6c105 IFP19Z6c117 IFP19Z2c118

Cone Panels – Active Channels

2 4 6 8 10 12

12:58:19 PM 12:58:24 PM 12:58:28 PM 12:58:32 PM 12:58:36 PM 12:58:41 PM 12:58:45 PM 12:58:49 PM 12:58:54 PM

Load – Integrated Cone Panel Pressures

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 58:19 58:28 58:36 58:45 58:54 59:02 59:11 59:20 59:28 59:37 Time (mm:ss) Load (MN)

Event Load: 1.99 MN

Flexural Failure – what did we learn?

• Rubble piles are far more common than ride-up.
• Rubble pile dynamics tend to be complex, and three-

dimensional.

• Sheet failure round the entire circumference never
• ccurs – in hindsight, obvious.
• Sheet failures tend to be spatially distributed around

central portion of cone – again, should be obvious.

• Dynamics are minimal
• So, flexural failure is a three-dimensional process, and

should be modeled as such.

Flexural Failure – what did we learn? – cont.

• Rubble piles are not linear sloped features, but rather bi-linear with a

top slope which is often horizontal.

• There is often stationary ice blocks against the cone, suggesting

that ride-up is not occurring against the cone, but further out.

• Rubble pile porosities can be highly variable, estimated between 0.1

and 0.3..

• There is rubble below the ice sheet, at least for thinner ice (cannot

see with thick ice).

• Circumferential cracks are seldom visible

– Are they under the rubble pile, or – Are they not occurring?

So how do the models do?

• We have made several comparisons between the measured

loads and loads determined from generally accepted models, plus the model developed by Mayne.

• There are several issues that we try to address doing these

comparisons:

– Whether to use average loads or peak (trigger) loads – The uncertainties in the loads and the ice parameters used for the models:

• Thickness
• Strength
• Rubble pile characteristics

– Ensuring that the models are used appropriately for the specific event.

Mayne

• 8
• 6
• 4
• 2

2 4 6 8

• 8
• 6
• 4
• 2

2 4 6 8

Mayne's Model Load_total (MN) Trigger Load at P31 (MN)

Croasdale

Ralston

Nevel

Sidebar – ridge loads

First year ridges

• If you beat your head against a wall long enough,

something might finally sink in (Eric Lemee).

• In our observations of ridge interactions with the

instrumented piers of Confederation Bridge (and we have 10000’s) we continued to observe loads that were much too low according to generally accepted theories.

• And, while we have not yet come close to “design ice

conditions”, we have encountered a credible number of ridges with keel depths of 12m or more.

• So why is this?

Relation between Load and Keel Depth

y = -0.0042x + 1.9385 R2 = 0.0002 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 Keel Depth (m) Load (MN)

Relation between Load and Consolidated Layer Thickness

y = 2.1347x + 0.5815 R2 = 0.2631 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 0.25 0.50 0.75 1.00 1.25 Consolidated Layer Thickness (m) Load (MN)

Pier Panel Pressures - Shaft

5 10 15 20 25 30 35 40 45 50

35:31.2 36:14.4 36:57.6 37:40.8 38:24.0 39:07.2 39:50.4 40:33.6 41:16.8 42:00.0 42:43.2

Time (min:sec) Pressure (kPa) 41 42 53

Panel Pressures - Shaft

• Shaft Pressures:

– Very intermittent – Far fewer pressure records than would be expected given number of keel interactions, and extent of any given interaction – This accounts for the recognized deficiencies of the panels in responding to very localized pressures – At average ice speeds, keel interactions should last from 1 to 3 minutes and result in a large number of “hits” on the panels on the pier shaft.

Pier shaft – Marine Growth

Cone Bottom Edge barnacles stripped off

Ridges

• Very little sustained pressure below the waterline cone
• Rubble Pile volume increases during ridge interaction,

but with only a slight increase in height

• Ridge Load is largely independent of keel depth.
• Failure theory needs to account for geometry of structure
• Waterline cones are extremely effective at minimizing

ridge loads – effectively a knife edge into the rubble of the keel.

• Perhaps first-year ridges are not the “bette noire” that

they are currently taken to be.

Conclusions

• Observations of flexural failure suggest that we need to re-think the

mechanics for modeling purposes

• Rubble piles are far more prevalent than ride-up, and result in the

highest loads.

• Panel pressures seem to confirm the sheet interaction mechanics –
• ne cannot have 180o failure.
• There is not consistent ride-up within rubble piles.
• We see rubble under the ice sheet – suggesting that failure does not
• ccur at the face of the cone.
• We see very little evidence of circumferential cracks – why?
• Despite errors in modeling, some of the models provide reasonably

accurate estimates of the measured loads.

• Ridge keels do not contribute significantly to the load – largely as a

function of pier shape

Acknowledgements

• Those who have funded the program:

– Natural Sciences and Engineering Research Council – Strait Crossing Bridge Limited – Public Works and Government Services Canada – Program for Energy Research and Development (PERD) – Exxon-Mobil – Marathon Oil

Acknowledgements

• And my students:

– Eric Lemee (MSc)– ridge keels – Derek Mayne (PhD)– flexural failure – Dambar Tiwari (MSc)– database – Mohamed El Seify (PhD) – ridge keel loads – Drubha Tripathi (PhD)– local pressures – Ray Peng (BSc)– loads (average files) – Brittany Hope (BSc)– local pressures – Susan Tibbo (MSc)– loads (detailed analysis) – Noorma Shrestha (PhD) – Database and loads – Keely Obert (MSc)– ridges – Chee Wong (MSc)– rubble pile mechanics

• For without them, none of this would be possible

Thank you

Questions?