Eric,

One possibility is that the electro deposited copper ductility is low,
and that's where the highest strain is developed. Surface cracks
initiated in the electro deposited copper create stress concentrations
that lead to catastrophic failure of the traces.

However, please check your math. (The Rogers calculator uses English
units.) 1/2 oz Cu is typically 0.0007 in. thick. The .15 mm overall
thickness would be 0.0059 in. The .005 mm core polyimide thickness you
report is an unlikely 0.0002 in. A better description of the circuit
cross section is required for me to use the Rogers calculator properly.
You didn't mentioned how thick the electro deposited copper over the 1/2
oz copper is. Is there adhesive between the polyimide and the copper?
How thick?

A quick check formula is:
Copper Strain% = 100% * Distance from the outer surface of the copper to
the neutral axis / Bend Radius

With an overall thickness of 0.0059 in. (0.15 mm) I get 18.8% max.
strain at the surface of the coverfilm with a 0.01575 bend radius (0.4
mm). To have only 4.5% strain in the copper the top of the plated copper
would have to be within 0.00071 in. of the neutral axis (not possible as
you described the two layer circuit).

Be careful of over bending or creasing in an attempt to get the flex to
hold an angle as this will increase your peak strain levels.

Many other considerations can also effect flex performance:

Surface finish, edge straightness, and side-wall profile of the traces
can create stress concentrations if nonuniform.
Foreign material trapped under the coverfilm can create local stress
concentrations, and may unbalance the circuit.
Improper selection of adhesive thickness, excessive squeeze-out, or
mixing of adhesive types can unbalance the circuit.
Poor coverfilm bond can lead to delamination, and rapid destruction of
the circuit.
Nicks or cuts in the polyimide along the edges of the part can act as
tear propagation sites.
Handling damage can wrinkle the copper and cause premature failure.
Wrinkles or creases in the polyimide, formed during lamination or
application of a stiffener, can affect the free movement of the flex.
This distorted geometry can lead to high impulse loads during bending.
Solder on the copper in the flex area can unbalance the circuit.

Andy Magee
Flex Guru - Consulting
[log in to unmask]
(937) 435-3629

         Eric Christison wrote,

Perhaps someone can help me diagnose a problem I'm having with a small
flexi
we've just received.

We've been trying to put a fairly sharp bend in the flexi -0.40mm to be
precise. It's a one time bend the PCB doesn't get flexed in operation.
Every
time we try, the tracks that cross the bend crack. Now I did my sums
before we
ordered the board and the Rogers flexi program tells me that the maximum
strain
is ~4.8%. This should be well within the ductility limits of copper as I

understand it - I believe strain to failure is typically 25-30%.

Just to see what I could get out of the PCB I tried increasing the bend
radius
until the cracking stopped.  I had to increase the bend radius to 1.2mm
before
I got no failures.

To give a few more details about the board:

Two layer 1/2oz copper - with electro deposited copper on top.
Layers separated by a polyimide dielectric - 0.005mm thick
Kapton cover top and bottom bringing total thickness to 0.15mm.

It seems to me that the copper has become embrittled would anyone
agree/disagree with me?

Regards,



Eric Christison
Mechanical Engineer
VLSI Vision Ltd
Aviation House
31 Pinkhill
Edinburgh
EH12 7BF

Tel:    +44 (0) 131 539 8165
Fax:    +44 (0) 131 539 7141
A company of the STMicroelectronics Group

################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe:   SUBSCRIBE TechNet <your full name>
To unsubscribe:   SIGNOFF TechNet 
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################