An interesting idea to model the collision of Titanic with an iceberg


My Facebook fan and  avid ” shipwreck aficionado”, Doug Kitchener, asked me  if  anyone has thought of building a model of the Titanic in order to study the impact with  an iceberg.  He also asked me to propose this idea for the  International Marine Forensics Symposium of 2012.  I asked my friend and  hydrodynamicist, Sean Kery,  his opinion.  Below is his response.

Mr Kitchener has an interesting idea to model the collision of Titanic with the Iceberg using modern tools.   The following is me talking through the problem as a professional, sizing up the task.   Please feel free to pass this along to him or any of your other readers as you see fit.

We probably have enough data to build such a model with a lot of detective work in various archives.   The problem is that it would take a huge amount of work, probably several man years to build a reasonably accurate structural model.   Probably Solid works or ship constructor would be the software of choice because of the way that the 3D CAD model links to the Finite element strength analysis software that can be linked to those packages.

Titanic was a riveted ship so there are two or three rows of rivets at each plate boundary and tons (litterally as well as figuratively) of rivets.   Each rivet hole is a stress concentration and each metal piece has a range of mechanical properties.   Engineers design to a textbook minimum strength that a plate or other part has to have.   The actual stress that it breaks at will be higher than that but there is often a wide range in the properties that vary from piece to piece, so we really don’t know how strong anything is other than to say it is “at least this much”, and unlikely to fail in normal service.   Some bolts I deliberately broke in a calibrated test were rated at 36000 psi and broke at 120000 to 129000 psi which was quite different that the rated value.

Let that metal rust in seawater for 100 years and the metallic properties are very different, so testing metal samples would probably not give much in the way of useful results.

Ice also has a wide range of mechanical properties and at least 5 different Crystalline forms depending on the temperature and pressure it was formed or was exposed to, for a long time before the iceberg broke off the glacier.   We also have no idea how big the iceberg was and since F=ma, the amount of force acting on the Titanic will depend on both the Mass m of the ship doing 25 knots and the iceberg which has mass plus hydrodynamic mass that was traveling at just over zero knots.   The only way to model that collision would be to assume a size and shape of the the iceberg and run the sumulation, then assume another size and shape and do it again and again until the exact damage pattern was produced.

There is the slight problem that we really don’t know the exact damage pattern that was produced.   Its below the mud line on the wreck and it was probably severely compromised by bottom impact and the nearly 100 years of rust and decay that have happened since.

It sounds like a really fun way to spend maybe half a million dollars to yield a plausible but probably unprovable result.   There are just too many unknowns.   Based upon the survivor testimony which is in unanamous agreement on this singular detail, the waves ranged “flat calm”, all the way up to “too small to care”, so my very fancy ship motions in waves software and analysis techniques cannot be applied to yield meaningful results:=)


Posted in Serving our Seas | Tagged , , , , , | 3 Comments

3 Responses to An interesting idea to model the collision of Titanic with an iceberg

  1. Doug Kitchener says:

    Thanks for checking into this, Pierette, and thank you, Sean, for a really interesting answer. Although I knew there would be many variables, I hadn’t considered the size or any other characteristics of the iceberg and how much of a difference that could make. Funny, what gave me the idea was the passage from the novel The Hunt for Red October where a consultant writes a computer program to model the Russian submarine’s performance, plus the computer simulation of the collision between the Andrea Doria and the Stockholm. The consultant in the novel had access to a Cray computer, which made the whole thing easy – too bad life doesn’t imitate art in this case! The way you summed it up was pretty much what I expected: “It sounds like a really fun way to spend maybe half a million dollars to yield a plausible but probably unprovable result.” – but it’s nice to have that confirmed by a professional! Oh well… it was a fun thought… maybe I can write a novel around it!



  2. Doug Kitchener says:

    PS, Sean, as of now I’m planning to attend the Marine Forensics Symposium; hope to meet you in person there.


Leave a Reply

Your email address will not be published. Required fields are marked *