Hacker News new | past | comments | ask | show | jobs | submit login
Why the Vasa sank: Lessons learned (up.edu)
81 points by ams6110 on Jan 1, 2012 | hide | past | favorite | 16 comments

This article was clearly written by a non-sailor as his terminology is very odd.

The phrase "heel over" is particularly confusing. All sailboats heel on most points of sail. When he says the boat "heels over" at 10⁰.. I assume that means the boat's Angle of Vanishing Stability (AVS) was a mere 10⁰? If so, the boat was very unseaworthy indeed.

Seaworthy square riggers would have had an AVS of more than 70⁰ with closer to 90⁰ being ideal. Modern externally-ballasted keelboats have an AVS well in excess of 90⁰, many in the 120-130⁰ range, meaning the mast can be completely submerged and the boat will still come back up if it isn't down-flooded.

Another point that stuck out to me: He said that the additional gundeck in the 17th century lead to a change in tactics from boarding with marines and taking the enemy vessel as a prize to simply outright sinking of the vessel.

In fact, prize-taking remained the dominant goal of naval battle until the late 19th century when exploding ordinance and long-range standoff guns rendered it impractical. Consider the battle of Trafalgar in 1805: The British captured 21 French and Spanish vessels and sank only one.

>This article was clearly written by a non-sailor as his terminology is very odd.

This is a parable about modern software development. The author is a compsci professor. From the biography section at the bottom:

Richard E. (Dick) Fairley is a professor of computer science and director of the software engineering program at the Oregon Graduate Institute.

Thanks for the cool facts about ships, though.

This comment was clearly written by a non-Unicode expert ;)

U+2070 SUPERSCRIPT ZERO (⁰) is not the same character as U+00B0 DEGREE SIGN (°). They look rather distinct on my screen (I assume both characters come from Verdana, the font specified in the CSS of this site).

I used Compose-^-0 to generate it. That used to make a degree sign. I guess it makes slightly more sense for it to be a superscript though. I wonder what the new compose sequence for degree is..

> If so, the boat was very unseaworthy indeed.

Yes, indeed. It sank after 1300 m on it's maiden voyage..

For those that haven't seen it, the Vasa museum (http://vasamuseet.se/en/) is really cool. If you happen to be in Stockholm and are interested in these things, a stop is highly recommended.

I can strongly second this, just saw it a month ago. Its hard to appreciate just how big this boat was until you're standing under it. And the 'excavation' process was quite incredible, made possible by the amazing preservation properties of pollution.

It is also very surprising that the ceiling height on the ship is something like 160 cm (5'4") - people were that much shorter only few hundred years ago.

Just FYI, they weren't that much shorter. Below decks everyone was stooped over. Sometimes the captain had a full headroom cabin depending on the size of the ship, and later perhaps the officers in the wardroom as well.

Great museum that's a must visit, and unfortunately it has a time frame on it. When I was there last year, the guides said it's slowly crushing itself, as a boat isn't designed to support itself out of the water. They don't know how long it has left.

I had a chance to visit the Vasa museum in Stockholm, and the Why was explained as follows - the king was an odd fella and builders couldn't question or object his directives. He wanted an extra tall ship and that's exactly what he got. The end.

I just finished Diane Vaughan's excellent Challenger Launch Decision, and it's remarkable how many of the same things that doomed the Vasa were also problems with the Shuttle fleet:

1. Excessive schedule pressure:

The Vasa was launched with an aggressive schedule that did not allow the shipbuilders to adequately redesign the ship to accommodate modifications requested by the king. Likewise the space shuttle fleet was plagued by schedule pressure throughout its history, with Congress repeatedly turning up the pressure on NASA to establish an operational shuttle fleet.

2. Changing needs:

The Vasa was originally designed to be a 108 foot ship, but was completed as a 135 foot vessel. Likewise, the Shuttle, originally was designed as a relatively small spaceplane that would sit on top of the booster stack. Additional scientific and military requirements cause the space shuttle to be scaled up into a much larger system that would sit alongside the boosters. In addition, these changing requirements ensured that testing was conducted on an accelerated schedule, further adding risk to the project.

3. Lack of technical specifications:

This is one place where the Vasa and shuttle narratives diverge. NASA was very diligent about ensuring that technical specifications were up to date and that changes were propagated to the appropriate working groups in a timely manner. These detailed changes and documentation were instrumental in conducting post-disaster analysis.

4. Lack of documented project plan:

The Vasa did not have a documented project plan, or if it did, it didn't survive into the present day. The Shuttle did have a very well documented project plan, but there were so many changes, few knew what the overall plan was, so the effect was remarkably like not having a plan at all. Unlike with technical specifications, mid-level managers and low-level NASA engineers often did not have a good understanding of what the current "big picture" status was with the Shuttle design.

5. Excessive innovation:

The Vasa attempted to incorporate a second gun deck, whilst simultaneously attempting to set a size record for the shipyard and the country. The Shuttle incorporated numerous innovations: reusable heat shields, innovative liquid-fueled motors, new electronic flight control systems. These innovations (particularly the heat shield and the liquid fueled motors) were significant timesinks and drew time and manpower away from other parts of the project that were considered "safe" (in this case, the reusable solid rocket boosters). Likewise, it appears that the shipwrights of the Vasa had so many other concerns, they didn't have time to pay attention to basic seaworthiness concerns.

6. Secondary innovations

This is closely related to the point above. Just like the Vasa required additional innovations to accommodate her second gundeck, the Space Shuttles required numerous secondary innovations to accommodate the new technologies that they were to incorporate. In both the space shuttle disasters, the final, fatal flaw was in a secondary innovation.

In the case of Challenger, the zinc oxide putty that was sealing the gaps between segments of the solid rocket boosters failed, allowing rocket gases to attack and burn holes through the O-rings. In the case of Columbia, the carbon-carbon panels that were shattered by falling ice were there to protect the leading edges of the wings - the original shuttle design didn't have such panels, as the original shuttle design called for the use of heat-resistant titanium alloy.

7. Requirements creep

The Vasa evolved significantly over it's 2.5 years of construction. The space shuttle evolved even more. It went from a relatively modest astronaut transport system to the be-all end-all space launch system for NASA. That, combined with schedule pressure, ensured that there were outstanding design issues when the system was put into production.

8. Lack of scientific methods

Here again, the two narratives diverge. Unlike 16th century shipbuilders, 1970s NASA was highly experienced in designing and building rockets. What sank the shuttle was not lack of scientific measurement and method, but a process of normalization of deviance, where increasingly deviant measurements were redefined as "normal" by virtue of the fact that in this instance nothing untoward resulted from the abnormal parameter.

9. Ignoring the obvious

The Vasa was launched after failing a rudimentary stability test. The space shuttle exhibited hot-gas impingement on it's O-rings on the second test flight. However, in both cases, the solution was to apply a patch and move on as if the problem had been solved. For the Vasa, it was additional ballast. For the shuttle, it was a series of tweaks to putty composition and application technique. In both cases, the tweaks did not fix the underlying problem, and were well-intended best efforts to fix design issues after the system had been built.

10. Possible mendacity

While the Congressional report does indicate mendacity on the part of NASA middle managers, Vaughan's account absolves these managers of mendacity. She says that the Congressional report misinterpreted internal NASA terms (like "waiver", for example) whose meanings had diverged from their common-language usage.

Despite a few points of divergence, it's striking how parallel the courses of the Vasa and the Challenger are. Both were groundbreaking, innovative flagships and symbols of national pride. Both were designed and built under intense schedule pressure and fluctuating requirements. And finally, both met tragic fates due to fundamental flaws in their design. Our knowledge of scientific principles may have improved over the past three hundred years, but our knowledge of sound project management techniques sure hasn't.

Also this: http://brucefwebster.com/2008/04/15/the-wetware-crisis-the-t...

This applies everywhere. Without a culture that encourages everyone at every position of power to be honest and express their true opinions then you just end up with a system built on deception and false beliefs.

Sounds like an early exercise in Agile Shipbuilding.

An improved version of the same document: http://www.it.uu.se/edu/course/homepage/acsd/vt10/SE3.pdf (with pictures and typesetting)

Source, which links to this and other historical accidents: http://faculty.up.edu/lulay/failure/ME421.htm

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact