Rockets are very loud, and seems to have a lot of vibration caused by turbulence. They're also run by microchips, which I tend to think of as fragile. Do the microchips need to be specially designed or encased to withstand these forces?
$\begingroup$
$\endgroup$
7
-
$\begingroup$ Your mention of "turbulence" make me believe that you are really interested in tolerance of vibration and/or acoustics rather than constant acceleration. $\endgroup$– ErikFeb 18, 2015 at 0:06
-
9$\begingroup$ Microchips themselves are crazy tough. It's their connections to other electronic components that are fragile. $\endgroup$– Russell BorogoveFeb 18, 2015 at 1:50
-
1$\begingroup$ To give some idea of how tough electronics can be, there are guided artillery shells where the GPS unit has to withstand over 15,000 g. I wouldn't be surprised if a rocket is no more arduous for electronics than a car or plane. $\endgroup$– Blake WalshFeb 18, 2015 at 11:59
-
2$\begingroup$ AAA with proximity fuses can take 30,000+ g's. Embedding them in a compound of similar density lowers stress. Of interest, the shells are all x-rayed in manufacture. An void or bubble in the explosive allows a localized collapse and compression that can cause detonation while still in the gun tube! Shock tubes (gas guns) for impact deformation research fire instrumented projectiles at around 5 km/sec and the parts last long enough to get some of the impact data. $\endgroup$– C. Towne SpringerFeb 18, 2015 at 23:27
-
1$\begingroup$ Relevant: en.wikipedia.org/wiki/Potting_(electronics) $\endgroup$– Russell BorogoveFeb 21, 2015 at 0:36
|
Show 2 more comments
2 Answers
$\begingroup$
$\endgroup$
Well, yes and no. The most important feature for modern space-bound CPUs (aside from 100% reliability) is radiation hardening. Radiation hardening is considered at every step of the design process from the materials used to the configuration of the transistors in each internal circuit. These chips are specifically designed to withstand any bombardment of radiation that can be thrown at them.
That being said, there have been some improvements considering susceptibility to vibration in recent history. Traditionally processor circuits for devices meant to live in the harshness of space contain a simple (or in the case of the Voyager 10/11 crafts completely discrete logic) devices that are supported by a large array of discrete logic. This was done, again, in pursuit of radiation hardening. Such circuits were composed of a large number of components with exposed leads. Below is an image one of the main computer boards for the (now retired) space shuttle:
Nowadays the dense CPUs (such as the CPU in the BAE RAD750 single board computer) are designed with some variety of a leadless (the pins are underneath the component) footprint. This document shows the features of the CPU contained within the RAD750, and there is a section on the device's reliability:
The RAD750 will be packaged in a Ceramic Column Grid Array (CGA) package, constructed by attaching extended columns to the original Ball Grid Array (BGA) 360 pin package employed for the 0.25 micron PowerPC 750. The CGA package has been adopted because it is better suited to the stresses of launch and the space environment. The CGA has demonstrated significantly increased reliability during temperature cycling and stress, when compared to BGA packages, as shown in Figure 3. The CGA package has also passed shock and vibration testing for the space launch environment.
As you can imagine, the column grid array components are much more rigid than the leaded components. More information on CGA can be found here.
$\begingroup$
$\endgroup$
1
As Russell said, it's the connections generally speaking that are difficult. The environment for launching is very stressful, and sometimes the connectors come loose. The trick is to make sure the chips are on solidly, and most importantly, to do a vibration test to ensure that the satellites can manage the stress of launching to orbit.
answered Feb 21, 2015 at 0:22