How many g's can a typical circuit board withstand?

Hi,
Does anyone know how many g's (acceleration) a typical circuit board (surface mount if it matters) can withstand?
Recently, a rep from the U.S. Military purchased a few of my MAVRIC boards
and mentioned that they will be used for the IMU (inertial measurement unit) on a CKEM (compact kinetic energy missile). Apparently these missiles go from 0 to Mach 6 in just a few thousand meters.
I suspect they will be using my boards to prototype their software while they are in the process of developing their own custom boards to actually reside on the missiles themselves. But they did ask how many g's my boards could withstand. I don't have a clue - any ideas?
-Brian
--
Brian Dean, snipped-for-privacy@bdmicro.com
BDMICRO - Maker of the MAVRIC ATmega128 Dev Board
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
There are a lot of factors to consider. You'd have to have a lab test it for find out for sure. Basically, the heavier an item is the more susceptable it is to G's. It depends on what they put on your boards, crystals, connectors, large parts like voltage regulators, capacitors would likely fail before the surface mount parts. At really high G forces,even the SMT parts would pull loose from the board, usually pulling traces off the substrate as they let go.
I think it could do 50-100 g's face up, maybe more. But the crystals, connectors or wires going to the board would likely fail first.
Unscientific methods, You can get 100's of G's of shock dropping the assembled board flat down onto a hard surface from 3-5 feet up. Components face down would not be able to handle as many G's as face up. Put the board in something streamlined and drop it off a 500 foot tall building onto the concrete below, that's good for maybe a 1000 g's or so. Choose a part and somehow attach something to it, clamp the board down, with a scale attached to the part, steadily pull it away until it lets go. Note how much pressure/weight/pull was used to pull the part off. You can now weight the part, compare it to how much it took to pull it off, and calculate how many G's that would be.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
To add to that. Don't forget that "g's" can also be a part of vibration. If you have vertically mounted components, like TO-220 voltage regulators, they can be susceptible to really nasty cyclical acceleration forces due to vibration (going back and forth). This can make solder joints or the legs themselves fail.
On a component-by-component basis you could estimate forces with the simple F = m x a formula. You need an idea of the mass of each product. The, of course, there's the mounting of the board itself. The board could flex due to inertial forces and traces/components crack. If needed you could build an exo-support mechanism (plates that go on top and bottom of the board to keep components from flying off or getting over-stressed. Also, using epoxy lines under components and conformal coating will add robustness to the overall assembly. Small SMT ceramic components can be suceptible to cracking if the board underneath flexes enough.
Want a cheap (but brutal) test? Zip-tie one of your boards to the suspension (not the body) in your car. The sway bar or one of the A-arms might be good. If you don't have an idea of how to do this safely (for both you, the car and anyone else on the road) don't attempt it, you can get killed or seriously hurt someone else. But, if you know how to do it, go for a ride and see if your board works afterwards. I've done this in the past, it's amazing what vibration can do to a board. Do this at your own risk. I accept no liability for anything that may happen. It is very dangerous.
Lot's to think about.
Sort of an extension of that question about using hobby servos in UAV's in that a lot of the same criteria is applicable.
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Martin Euredjian
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Fri, 01 Aug 2003 23:33:30 -0400, Earl Bollinger wrote:

Well, I just did some quick calculations, and unless I made a mistake, I'm guestimating the acceleration to be around 150 g's. This is assuming that the missile starts off at zero velocity and undergoes constant accleration for 5000 meters, at which time it's velocity is Mach 6, or 3846 m/s. I came up with 2.6 seconds to travel the 5000 meters, which makes the formula for acceleration:
    v = a * t
    a = 3846 / 2.6 = 1479 m/s^2 ~= 150 g's (1 g being 9.8 m/s^2)
That's probably OK for a ballpark working figure. Of course, I have no details on the actual performance of these things - what little I can find on the web is pretty vague. But I suppose if I really knew, I most likely wouldn't be allowed to talk about it anyway :-)
I suppose a board could be encased in an epoxy to keep all the parts in place, assuming the epoxy can withstand the stresses. But aside from physical and structural integrity at high g, what about the actual performance of the electronic components themselves? G-force is not something I'm used to seeing in the spec sheets for electronics components' "absolute maxiumum ratings" sections. Seems like capacitors could distort which could change their behaviour. Crystals may run slower (or faster) under high g. What else?
Maybe it's time for a little destructive testing :-)
Cheers, -Brian
--
Brian Dean, snipped-for-privacy@bdmicro.com
BDMICRO - Maker of the MAVRIC ATmega128 Dev Board
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Pot the boards in epoxy or silicone. General electric makes silicones for this sort of thing, although I suspect that epoxy would be more suitable. < I suppose this would make your nifty screw terminals sort of problematic ;^) >
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Hang on a mo'
before you start potting anything, if it is only 150 g's then the board is probably fine. A 40 pin dip weighs around 5 gms, so at 150 g's 750 gms, it'll take more than that to rip it off.
smt is lighter even if not quite so firmly attached.
I think I agree that the crystal is probably the weakest link, so why not just dump the problem on your crystal supplier? Crystal suppliers usually want you to consider g forces when mounting their parts, just in case some one drops it. They should know.
Maybe you could specify the board is mounted at right angles to the major accelerations or whatever looks most comfortable for it. Try resting 150 times the boards weight on the middle of it, stuff like that.
best regards
Robin G Hewitt
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
While conformal coating like epoxy can help increase the maximum g-load on your electronics board, it is expensive and can also produce heat problems. Conformal coating is a last resort unless corosion resistance is required.
BRW
On Sat, 02 Aug 2003 05:42:00 GMT, "Blueeyedpop"

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

This isn't the whole picture. You cannot ignore vibration, harmonics, resonance, etc. Particularly in something like a missile. Making a board that can endure high, constant, single-axis acceleration is almost a trivial excercise in engineering. The challenge is making something that can deal with real world realities which include acceleration, shock, harmonics, resonance and even metal fatigue.
Interestingly enough these are some of the same issues in making earthquake-safe buildings. Ground acceleration isn't necessarily what destroys a building. Shock and vibration along x,y,z (both translational and rotational) can have devastating effects on a structure that would otherwise endure substantial single-axis acceleration.
An interesting little experiment is to take a flexible metal ruler and clamp it to a desk:
ruler ------------------------ |DESK |
Now, strike various points on the desk with your fist and observe what happens at the far end of the ruler. The tip will begin to oscillate. If the strikes were periodic and at the right frequency you could cause resonance, leading to potential fatigue and severe acceleration loads. You can change the setup and clamp the ruler to the edge of the desk (vertically) and see what that does.
Electronic components and boards are suceptible to this sort of a mechanism, which is a lot more complex than single-axis acceleration. Certainly a missile taking off is not a nice-clean single-axis acceleration environment.
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Martin Euredjian
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
You could possibly get hold of a high speed motor from somewhere, attach a balanced arm to it, and use it as a crude centrifuge to test a board to destruction - the calculations for centripetal acceleration shouldn't be too hard.
Just a thought.
Tom

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Brian -
I'm an electrical engineer in our (large electronics company) government systems group. Our PLC equipment is very similar to your MAVRIC board. We build control systems in electrical panels that have to pass MIL-S-901D (shock) and MIL-STD-167-1 (vibration). Your board will probably have to pass the same test as part of electrical control system.
Our boards are rated at 30 g operating. Looking at the MAVRIC board, it will probably withstand 30 g. The only concern I would have is the heat sink to the right of the "MAVRIC" label.
There are 2 ways that the military mounts control systems - hard mount and using shock isolators. Either way, they will have to fashion a clamp for the header pin connector. That's the weak point of the whole system. Given the header pin connector, your board will easily pass "shock & vib" using shock isolators and will probably also pass hard-mount.
I you want to test your board, I suggest contacting Wyle Laboratories. That's who we use for shock & vib testing.
BRW

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Sat, 02 Aug 2003 08:24:18 -0400, Bennet Williams wrote:

Lots of great info - thanks! You mention the "heat sink" to the right of the MAVRIC label - that's actually a button-cell battery holder for the battery backed real time clock which backs up the Dallas DS1307 I2C RTC. I suspect it could actually just be removed for this application.
Thanks for the info! Lots to think about.
Cheers, -Brian
--
Brian Dean, snipped-for-privacy@bdmicro.com
BDMICRO - Maker of the MAVRIC ATmega128 Dev Board
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Sat, 02 Aug 2003 09:31:27 -0700, Gordon McComb

Actually, it's a new world now - COTS (Commercial Off The Shelf). The military strives to use COTS components from commercial companies to save money. No security clearances are required for COTS work (although U.S. citizenship is usually a requirement).
You are certainly correct about bonding and insurance, though. It's all about contract terms & conditions.
BRW
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Bennet Williams wrote:

I was speaking mainly about Brian becoming a defense contractor providing MIL spec versions of his board. They'd require strict MIL specs for anything on-board a missle, because of the hypercritical nature of the components. And because it's avionics, secrecy is usually the order of the day -- if an enemy knows the system (because it's commercially available), it's easier to figure out ways to jam it.
I sell to the gobment, including the military. I happen to have a security clearance, but not for robotics (contract document automation programming, for the IRS actually). Somehow I don't think the little robot kits they've purchased from me have been for use in Tomahawks! <g>
-- Gordon Robots for Less at Budget Robotics: http://www.budgetrobotics.com/ Author: Robot Builder's Sourcebook & Robot Builder's Bonanza
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Does anyone else here get that depressing feeling of wasted effort when they think about the obscene amounts of expensive components and precision engineering that go into making things that are ultimately designed to destroy themselves, and other expensive things and people in the process?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Yes. When you challenge the designers and manufacturers of such equipment that it results in children having their arms blown off and/or left to die horribly having been given first degree burns over most of their body you meet a callous reaction that makes Saddam Hussein seem to be an absolute saint.
If there could have been any good come of the attack on the World Trade centre, it would have been the shock and horror that results from realising the effects of weaponry upon our fellow humans; the reaction of Yankland, however, was to some extent to justify the attack that was made on them.

they
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

expensive things and people in the process?
Unfortunately, freedom isn't free.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
<snip>

Is this a valid argument? The soviets also had nukes. It's a little one-sided too - for all we know now (and, more importantly, the Russians knew then!), if the USSR hadn't had atomic weapons, the Allies may well have invaded them instead.
It's going to be interesting (and probably rather messy) when we finally reach the point where further advances in military technology are impossible, and everyone's back on an even footing.
Tom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

have
If there had been no nukes, then showing restraint and not invading, given such overwhelming conventional superiority, would have been something really new.
Without nukes the Russians would have been at risk, They had just expanded their borders and were obviously going to be a problem. However, no country with nuclear weapons has ever been attacked.

Whatever happens, we have got The Maxim gun, and they have not.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

it
have
The USSR didn't detonate their first A-bomb until 1949.... There was a span of 5 years where the West still could have invaded...

"Everything that can be invented - has already been invented"
right?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Your point of view is ignorant and offensive.
At least tell us where you are from so that we may have context?
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Martin Euredjian
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.