Improving Hard Disk Drive Reliability by Studying Bridge Design
On a desktop PC, the hard disk drive (HDD) is the component with the highest failure rate. While our ThinkCentre desktops have among the lowest HDD failure rates in the industry, our engineering team focuses a lot of effort to make these failure rates even lower. So far the payoff has been great. A ThinkCentre you buy today has about 25% of the chance of experiencing a failed HDD drive during its lifecycle than even systems we shipped just three years ago.
Recently the team was able to make a small but significant change that will improve reliability even further. To understand what they did, consider a bit of history.
One day in 1906, as a group of Russian soldiers went across a bridge in step, the bridge suddenly collapsed. The reason was the soldiers’ walking frequency was the same as the bridge’s natural frequency, a phenomenon known as sympathetic vibration. From that point forward, a rule was set by the army: When walking across a bridge, soldiers must break step.
On November 7th, 1940, the Tacoma Narrows Bridge collapsed suddenly when prevailing winds caused the same phenomenon. Soon after vibration theory became a required course for aspiring engineers.
People familiar with the inside of their ThinkCentre PCs will recognize the original HDD caddy design in the diagram below. It first debuted on our ThinkCentre S50 PC and provided both ease of service plus HDD vibration protection. Yet the team found that under certain conditions the design was susceptible to sympathetic vibrations from its environment much like the bridges mentioned above. This could result in premature HDD failure. Using bridge design theory as a model, our engineers were able to come up with a modified design to eliminate the danger with almost zero change in cost.
You may claim all desktops are the same and even use the “C” word (commodity), but I offer this as evidence that nothing can be further from the truth. There is still plenty of innovation left to be had in the desktop world.
The Lenovo IdeaCentre B510
The Lenovo IdeaPad U160
Lenovo Meet the Modder Dean Liou
Lenovo Meet the modder- Chris Blarsky Diary 2
April 22nd, 2008 3:34 pm
Even though the probability of me purchasing a desktop in the next 12 months is near 0, it is still fantastic to read posts like these and make me proud of the Lenovo name which is now practically synonymous with great engineering.
Thanks! Talk about ThinkPads next time
April 22nd, 2008 5:12 pm
stefan, even though this study in harmonics isn’t specific to a thinkpad, the point to take away from it is that lenovo’s engineering and design staff are aware of the minute details which comprise an overall-better product. before you turn the key in your car’s ignition, think about the harmonics involved with both the firing order and counterbalances in your engine. next time you watch a sports event, consider the harmonics involved in the shutters of high-frame-rate SLR cameras which now have an MTBF of 300,000 actuations. and, next time you press the power button on your thinkpad, think about the rubber vibration dampener surrounding the hard drive that helps keep it reliable just like the example above (assuming, of course, that you have a model made approximately within the last two years).
the battle to mitigate or eliminate harmonics is all around us, even in your thinkpad.
April 22nd, 2008 6:08 pm
You’ve forgotten to emphasize, that beside lowering hard disk failure rates, such HDD caddy shock absorbers do also lower the overall noise level, which is produced via the vibrations. So it offers a sort of double protection, especially for very fast (high speed) rotating hard drives.
April 22nd, 2008 7:23 pm
i should add that many harmonic frequencies which destroy components such as hard drives are at sub-audible levels. encasing something in an elastomer or polymer doesn’t guarantee that destructive harmonics will be lessened in that component — it only means you’ll hear less noise (as in the HDD rails on a thinkpad). elastomers/polymers can carry harmonics, too, even if they successfully cancel noise and vibration. it’s the transfer of these vibrations into a material that is too stiff, the wrong shape, or the wrong material that becomes a concern, as shown by the various colors in the stress model above.
thankfully modern hard drives have internal damping methods in place to start from an inside-out perspective. the caddy improvement might not have been designed to help cancel harmonics from the hard drive but instead to cancel out harmonics from the chassis itself. this may not be an inside-out solution but rather an outside-in solution.
April 22nd, 2008 8:36 pm
It’s good to see this engineering coming out of Lenovo and this keeps me confident when buying your products. This is why I loved IBM systems so much, and I am thrilled you are carrying on the torch.
I hope you consider marketing things like this, because people need to know what they are paying for in a world that they assume is commodity. Most other manufactures focus purely on aesthetics and tech specs, but those of us in IT, business, government, and engineering will quickly find a friend in Lenovo if this kind of detail is stressed.
April 22nd, 2008 11:37 pm
Do you have a reference for the 1906 russian soldiers story?
April 25th, 2008 12:28 am
I wrote a paragraph or two, but decided to forgo it because it’s pretty much what Kevin Bowling said (#5).
Being more on-topic, the ‘original-design’ image is shown warped as if under a stress/strain simulation, whereas the ‘new design’ just appears to be a non-loaded image. I think most people would look at it and conclude that the non-deformed one is better. Of course, that’s what the added braces would do- strengthen the caddy.
On the topic of marketing, I’d like to see more companies explain why some of the design decisions were made. I, for one, would certainly appreciate the perspective and be able to understand why certain compromises were made (example: x300 design).
April 25th, 2008 1:01 pm
Jonathan – explaining design decisions and the facts behind our engineering are certainly some of the things we try to communicate on our blogs. I assume you’ve been over to David Hill’s blog Design Matters. Between these two blogs alone, a number of design decisions and engineering facts have been discussed. As for the X300, search for that term on either blog and you will find several interesting posts. Be sure to read the BusinessWeek article on X300 as well – it has some really good information.
May 2nd, 2008 2:05 pm
Nick — It is embarassing to admit so, but I do not have a citation for the 1906 Russian Soldier story. It was relayed to me, and while I believe it to be true, I cannot find a reference for it either. If you need a reference for a research paper, I suggest you skip this one.
May 2nd, 2008 4:45 pm
wikipedia has a dedicated list of bridge disasters. i didn’t see anything involving russian soldiers in 1906 but we have to keep in mind that this list is incomplete. the tacoma narrows bridge is by far the most popular resonance-induced bridge failure to date thanks to the video footage.
on a similar note, a short article about forced oscillation was written by one of my former physics professors, larry curtis, and is good reading for anyone interested in the physics behind the above caddy example.