I would like to add a couple of things to the conversation:
1. I conducted a little experiment with ice at my home over a year ago and came up with some interesting results. The experiment wasn't very scientific, in that there were no controls or quantitative test sets with calculated variables to plot out on a graph. Simply some crude observations, but still I believe they point toward a better understanding of how icebergs behave.
First I took some different geometric shapes found around the house and filled them with water and froze them in the fridge.
I froze a long narrow cylindrical water bottle.
A shorter, wider, but still cylindrical picnic cup.
A small cubical tupperware dish.
An inverted frisbee, ice was shaped like a pancake.
And a tennis ball with hole cut into it. Formed a sphere.
Next I filled my kitchen sink with cold, very salty water. I added almost a full Morton's Salt container into the water. It actually precipitated to the bottom the water was so saturated with salt. I thought about using 'sea salt' but am much too lazy to make a special trip to the store. (Besides, illegal to drive while drinking...Not that I was drinking when I did this, but when I went to the bathroom I actually precipitated hops and barley!!! But thats another story.)
Cut the 'icebergs' out of their molds and one at a time placed them into the brine water.
The long narrow cylinder floated in a stable fashion on its side longways, like a log in a river. It did NOT float upright like a greek column in the water.
The Picnic cup berg also floated more or less longways like the bottle shape, but was less stable. It rolled and flipped over easier. It did have the greatest surface profile of all the tests. In part due to being slightly flared on one end. This flared end created a sloped ridge rising out of the water.
The cubical berg rolled around a lot. It was very unstable and easy to flip. It had the second highest surface height. And then it didn't, and then it did again...
The frisbee berg, was the most stable. It floated flatly like a life preserver, however it had almost no surface height. That is it appeared to float just under the meniscus of the water, with very little or no visible part breaking above the water surface. It also didn't remain in one piece for very long. Soon it cracked into three pieces, each remained floating flatly though.
Lastly the tennis ball behaved much like the cube, it rolled continually and was the most unstable. Its surface height was also very small.
To me, these admittedly hokey tests illustrated that ice seeks the orientation of greatest stability. In my tests, this meant that each berg tried to put as much of its mass as possible at or near the water surface. It showed that the concept of a long, columnar shaped berg floating upright with 10% of its mass above water and 90% below extending far down into the depths is false. My tests did not exhibit that behavior. My tests lead to the concept of ice trying to flatten itself out at the surface. Presenting the greatest possible surface area, or massive area, closest to the surface.
This ties in perfectly with the idea of 'ice shelfs' surrounding icebergs at sea. The shelf is the broadest face of the most massive side of the berg. Since, as in my tests, you can have a large body of ice floating only slightly submerged, any peaks extending high above water would represent only a tiny portion of the overall mass of the berg. I see it this way, icebergs don't like having big portions of themselves rising out of the water. Nor do they like having large portions of themselves extending deep below the surface. They seek equilibrium at or just shallow below the surface. This makes sense because ice floats due to having lower density than the water around it. The water attempts to push the less dense, air filled ice up and out of the water. Similarly, ice is more dense than air, and the sinks to the bottom of the atmosphere. That equilibrium point between water and air is the ocean surface. Any part of the berg that isn't right at the surface is in conflict with the forces around it.
Since the berg is constantly seeking equilibrium, as it melts, erodes or fractures over time, it will naturally tend to mill itself into the most stable form possible. Based on my tests, and on my reasoning, I predict that the most stable shape for a berg is that of a balanced cylindrical diamond floating vertically. With the majority of its mass located in the center of the body where the cylinder reaches its greatest radius, and two conic ends of equal size and shape. (Imagine two chinese WOKs welded together at their tops, or a spinning top, or two cones placed base to base and fused in that position.)
Of course in real world conditions, this perfect shape likely will never, or rarely, be seen. Icebergs are not stable objects, but are in fact examples of instability. I only propose to describe what is in my opinion how they could theoretically form in perfect conditions. It is this theoretical model I choose to use in my imagination when visualizing Titanic's collision, allision, grounding, or evasive maneuvers. Its not a perfect theory, and its origins are not scientific, but mostly observational. Not empirical.
That is the shape I imagine when I think about the berg that was seen, and contacted by Titanic.
Not a mountain of ice, 100 foot at the widest point, riding 100 feet out of the water, and extending 900 foot under the water. But rather a flatter, shallower object. Still massive, but with the majority of its mass close to the surface and forming a large surrounding shelf of submerged ice perhaps 300 feet or more in radius from the center of the peak. (Presuming the peak is near the center of the body. Who really knows?) Thus Titanic makes contact with the submerged ice as suggested in "Last Log" by Dave Brown, in a grounding situation. In my view, perhaps more of an oblique grounding, impacting first on the starboard-underbelly of the 1st cargo hold then the damage area migrating both foreward toward the forepeak and aft along the firemen's tunnel as the slope of the shelf increases in angle over event time. Until finally momentum and swing of the ship, stop the upward encroachment of the bow along the shelf slope, and the bow begins the slide off of the shelf and back off into the water. In my view, this lifting of the starboard side of the bow along the shelf created a twisting effect on the ship's structure forward of BR5. As is so well put forward by Erik and Dave in terms of structural analysis.
In this way we see that the shape and orientation of the berg is important and speaks to us of its form through the damage pattern on the ship. A berg with a wide, sloped, submerged shelf rising toward a peak which is seen above water fits well with the new theories on the ship's internal damage. And perhaps my kitchen experiments in some way illustrate the legitimacy of that imagined form.
So I caution myself and others to take these ideas with an understanding that they are only ideas and theories. I'm not a scientist and I've never been to sea, nor have I ever seen an actual iceberg in real life.
Which brings me to the second part of my post here:
For some great pictures of icebergs and information on their origins and the currents that move them, check out:
uscg.mil/lantarea/iip/
Always an interesting site to visit.
Yuri