Spinning Drives and Movement Question

[sao123]

The California earthquake this morning got me thinking...

If your computer is running (AKA drive spinning) at the moment an earthquake hits...
How much damage to the spinning platters is that going to cause?
Or is the magnetic reader arm secure enough that moving or rocking the drive is not enough force to cause it to touch the face of the platters...

Lets assume the drive/tower only rocks back and forth or up and down, and does not fall off a shelf to impact.

 

[KingFatty]

My last earthquake experience was that it felt more like swaying. So the opposite of an impact. Maybe equivalent to someone picking up a computer and carrying it in their arms as they walk normally, swinging the laptop back and forth with a typical gait.

We know laptop hard drives are designed to easily handle that kind of movement, and I think the 3.5" 7200 RPM drives are pretty robust too. I guess you'd need raw numbers to figure it all out for sure.

But another point: earthquakes can move the ground side-to-side, or up and down. It might make a difference if the movement is along the plane of the spinning disk, or perpendicular to it. But either way, I still think that kind of 'translational' movement is much easier on the drive because it doesn't fight against the gyroscopic effect of the platters, compared to 'angular' movement where you'd be rotating or spinning the drive along an axis that fights against the gyroscopic effect and tweaks the platters into the heads etc.

 

[BrightCandle]

Most hard drives are designed to survive impacts to at least 10G at least while off and parked. A 6.0 earthquake isn't going to even remotely hit that sort of energy into the drive, it needs to be dropped from quite a height or thrown really hard to achieve that sort of force. However if its on and writing data then you will have actually seen a momentary reduction in performance or inability to write or read data and potentially the heads can crash into the disk damaging it.

 

[KingFatty]

Also recall the experiment where a guy literally yelled at a rack of hard drives, and the tiny vibrations of the sound waves from his own voice caused the read/write performance of those drives to momentarily decrease.

So during the earthquake, you'd probably see a decrease in the performance (perhaps due to needing to correct slight misalignment errors) similar to how soundwaves can affect performance, but no damage.

 

[sao123]

Most hard drives are designed to survive impacts to at least 10G at least while off and parked. A 6.0 earthquake isn't going to even remotely hit that sort of energy into the drive, it needs to be dropped from quite a height or thrown really hard to achieve that sort of force. However if its on and writing data then you will have actually seen a momentary reduction in performance or inability to write or read data and potentially the heads can crash into the disk damaging it.

I didnt think drive heads parked until the computer shutdown or went to sleep? Or do they park after each read/write cycle completes?

 

[corkyg]

As for earthquakes, it really depends on the magnitude and type of quake. I have lived most of my life in quake areas, and few would have caused damage to a spinning HDD. More likely is a secondary crash caused by the computer falling off the desk or some part of the house falling on it.

Go SSD and don't worry about it.

 

[sao123]

As for earthquakes, it really depends on the magnitude and type of quake. I have lived most of my life in quake areas, and few would have caused damage to a spinning HDD. More likely is a secondary crash caused by the computer falling off the desk or some part of the house falling on it.

Go SSD and don't worry about it.

lol, no.
I have a 2 bay NAS, soon to be adding another 4 bay NAS.
I need to build or buy a stand of some sort to put them on. Just thinking out loud.

 

[bryanl]

I didnt think drive heads parked until the computer shutdown or went to sleep? Or do they park after each read/write cycle completes?

The heads will retract and the motor will shut off if the drive electronics sense enough shock, either through its piezoelectric shock sensors or a severe anomaly in reading the servo marks or sectors. I don't know how much better shock sensors help protect the drive, but often circuit boards have room for 1-3 sensors, and they're not installed in lower cost or retail models.