help on damping constant 2

[xbmath]

Hi there:
I am working on the structure analysis of the electronic circuit board assembly. Can anybody tell me what the damping constant should I use for the FR4 board, the ceramic components, and the stiffener (aluminium).
Thanks

 

[Drej]

What's your definition of "damping constant" (including units)?


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[xbmath]

I mean damping coefficient, no units.

 

[GregLocock]

Hit it with your fingertip. Does it ring or does it thud?

If it thuds use 0.3, if it rings 0.02


Or measure it properly.


Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Drej]

Nice one, Greg. You're obviously a practical sort of guy.

xbmath - just to clarify, if it is unitless it's (usually) called the damping ratio; the damping coefficient is units of Force*time*length/radian (usually). Like greg says, you probably need to measure it experimentally (give it a whack and measure the response!), unless there are some figures floating around on the web.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[tlee123]

There's been some confusing information in this thread. To clear it up, let's get some definitions fixed.

The damping coefficent (for linear motion) must have units of force/velocity. This is analogous to a spring rate which has units of force/distance. A damping coefficient will have units such as lb/(in/sec), sometimes written as lb-sec/in, and always spoken as "pounds per inch per second."

The damping ratio is the actual damping coefficient divided by the critical damping coefficient, and is therefore unitless.

OK this is funny. I just read Drej's last post and realized he already covered most of this. Oh well, I'll post anyway!

 

[SomptingGuy]

An additional source of confusion is that many design analysts will give you a damping ratio when you really wanted a damping coefficient. What good is 2% when you really wanted a number in Ns/m (or in some antiquated American unit)? Unless you know which mode has 2% damping, you're stuck.

 

[cbrn]

I once read a paper where a strategy was outlined in order to find modal damping coefficients from several damping ratios given at some frequencies, being these quantities, as outlined by SomptingGuy and others, frequency-dependent; the method worked when most of the eigenfrequencies were known (either experimentally or by FE means). I came to this resource simply by Googling "internal+friction+metals" or "internal+damping".
I don't remember this source, since in the company I work in, for "bulk" steel parts we use a Raleygh damping approach, and the values we input for alpha and beta are statistically fixed (probably not extremely precise in absolute, but they work for the parts we handle).
If you think it can be useful, I can post these values.
Bye!

 

[tomirvine]

You can perform a forced response or base excitation analysis as a modal analysis. In this case, you would specify the damping ratio for each mode instead of the damping coefficients. This is the method that I prefer.

As an alternative, the amplification factor Q can be specified.

Q = 1/(2*damping ratio)

Steinberg gives an empirical formula for the circuit board Q in his book "Vibration Analysis for Electronic Equipment."

The best approach is make your measurements, however, as has been previously suggested.

Tom Irvine

http://www.vibrationdata.com