Designing for high vibration environments

Go To Last Post
10 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi there - I do a lot of work at work on electronics that experience high vibrations (maybe 10-100gs). So far - our company strategy seems to be to mostly ignore the vibrations. I mean we try to zip tie down wires and things like that, but nothing much beyond that. We definitely get periodic wiring failures due to wires shaking around and breaking due to fatigue.

Can anybody suggest good resources/guidelines for running wiring in these sorts of applications? This seems like the kind of thing that an aerospace or maybe an automotive group would be very familiar with, but nobody in my company has any real institutional background in it.

Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You are correct about automotive and aerospace. I really don't have any good references for you. I have seen a very small number of articles over the years in trade magazines (ED, EDN, etc) but nothing I can point to.

One of the things is to prevent flexing where changes in rigidity or mass occur. For example, flexing tends to happen right at the point where wire insulation ends. It also happens where the solder ends (ie, the solder wicking) at soldered connections. Likewise wire entry to terminal strips. I have seen shrink sleeves applied at some of these points to distribute the flexing. The plastic tubes over crimp connections help with this also.

The way I have usually seen this "solved" is to build a vibration table and "shake the hell" out of various test harnesses and assemblies (experimenting with various strategies) until it lasts an appropriate time on the torture machine.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

NASA space flight hardware workmanship standards
http://workmanship.nasa.gov/wkst...

For a start:

Workmanship problems pictorial reference
http://workmanship.nasa.gov/wppr...

Stealing Proteus doesn't make you an engineer.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I still recall one of my first automotive project "failures": I used a socket(s) for an IC chip. I never made that mistake again...

Locking nuts & bolts can be an issue, also. The attached photo, (care of the web, not mine), illustrates using both a locking nut AND a safety cotter pin. The pin goes through the bolt, and prevents the nut from loosening itself.

I think a lot of aviation gear uses either the safety cotter pin or a safety wire. I can recall, but can't find at the moment, an aviation mechanics article on how to properly wrap the wire, (location of hole in bolt, size of wire, number of twists, how the wire lays, etc...). There is either science, or pseudo-science out there on this subject. Perhaps a look at some of the Experimental Aircraft Association books and guides would be another potential source of info, EAA .

JC

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Try to increase the radius of the flexing in the wiring. Spreading the flexing over a longer length of wire will reduce the stress and metal fatigue.

Also, keep in mind that it is the relative movement of the wire wrt the connector that is the killer. If you can fix things so that everything moves together, you'll get much better results.

[edit] Corrected error (I had put decrease instead of increase). :oops: Thanks DocJC. :D

If you think education is expensive, try ignorance.

Last Edited: Thu. Jan 15, 2009 - 07:35 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Big glob of RTV on caps and any other component that can wiggle back and forth. They used to put to5 transistors in little 3 pin holders that looked like the papajohns garlic butter holders. (That stuff is liquid cholesterol. You slop it on your pizza and it goes straight into your arteries and coagulates)

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Good point,Bob. All large/heavy components must be securely anchored. Axial electrolytic capacitors are good for this - they can be laid flat on the PCB and secured using tie wraps.

Transformers are another weak link. Get ones with pins that go into the PCB, not with flying leads. A metal clamp is also recommended.

If you think education is expensive, try ignorance.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
The way I have usually seen this "solved" is to build a vibration table and "shake the hell" out of various test harnesses and assemblies (experimenting with various strategies) until it lasts an appropriate time on the torture machine.

Jim

In the late 70's I had to take my drifting buoy design out of the "lab" and into production. Shock and vibration effects were a concern because some of the buoys would be "launched" (aka thrown overboard) from 20m in rough open ocean weather. All of the electronics pcbs were mounted in a single 3m long aluminium U channel chassis and the flat ribbon cable run under the boards and held with glue clamps.

We built one of Jim's torture machines that held the completed assembly. The very first time it was run I learned how much aluminium U channel could flex when only held at the two extremities ... alarming! The design was altered by gluing two high density rubber strips to the middle of the channel length so that when assembled in the spar hull the strips pressed against the inside of the hull's wall. Worked beautifully.

Shortly later, an "old timer" engineer suggested a cheaper torture machine to me. Just get a long haul transport truck to carry it a few hundred miles and bring what remains back for inspection :lol:

Cheers,

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I worked for a company that did vibration testing. The had a big 1kW valve amp and expensive B&K gear to run it all. I was interesting to see how things hit resonance as the frequency was sweeped. Lots of hot melt in strategic places fixed most of the problems.

I also learn a valuable lesson when the first pcb mount transformers became available. We though these were a grewt idea so we used them in a new product. All was fine until we shipped a few interstate. We got reports from the other office that there was rattly pieces inside the equipment and that they didn't work. When we got the units back we found the transformers had sheared off the circuit board. When solved this by using a cable tie thread throught the board and around the transformer. We never saw the problem again. Had we put the unit through a vibe test, we would've seen the problem immediately. The vibe test in this instance ended up be the 1000km truck ride and rough handling.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
Big glob of RTV on caps and any other component that can wiggle back and forth.

Reading through NASA's workmanship pages, one of the common problems seems to be globs of "spiking" compound which cover the "stress relief" bends in component leads. When you're covering things to make sure they can't wiggle back and forth, be careful not to create new stress points.

Michael