Constantino,
From the photos, it is very hard to form a conclusion as to what you are seeing.  If it were delamination/blisters/crazing, you would typically see a lighter area in the dent--Ref IPC-A-600G pp18 or pp19, and the anomaly would typically be convex (bulged outwards).  However, you state that it is a dent (concave--bulged inwards) and that it appears on both sides.  The cross section (which would have been invaluable for evaluating the condition) also does not show blister, but does show thinning of the copper? Copper will not become thinner all by itself, nor from thermal excursions, unless it is stretched (tensional thinning), or compressed.  From these clues, my personal opinion would be that the anomaly is probably caused by mechanical considerations rather than thermal.  Are you using some type of clamp (probably with hemispherical ends) during your rework procedure??  This would cause concave dents on both sides, and also the thinning copper.

An alternative hypothesis, is that the thermal excursions involved in reflow/rework are causing a resin shrinkage of uncured laminate/prepregs, especially if the multilayer board (guess based on surface complexity) were laminated at low pressure, and didn't develop its optimum density.  This is easy to check out, by running a Thermal Mechanical Analysis (TMA) of a coupon/board that has not seen the rework/reflow temperatures.  Not just a normal TMA plot, but a hysteresis plot---RT to 150ºC to RT and finally back to 150ºC.  The 150ºC temperature is chosen to be below the typical Tg of most of today's FR-4 materials.  The plots going up, then down, and then back up should be reasonably on top of each other, and reflect the CTE's of the materials. If the material is uncured originally, it will cure during the first "up" cycle of the TMA analysis, but will not overlap during the second "up" cycle.  If a third up/down cycle were added, the second and third plots will overlap, but the first will be offset showing uncured resin.  Your PWB supplier should be able to get the TMA work done by their raw material supplier. (Most laminate vendors have this capacity)

Dennis J. Cantwell
R & D Liaison
Printed Circuits, Inc.
1200 West 96th Street
Minneapolis, MN 55431-2699
952-888-7900
[log in to unmask]
 

-----Original Message-----
From: IPC-600-6012 [mailto:[log in to unmask]] On Behalf Of Constantino J. Gonzalez
Sent: Friday, October 28, 2005 4:50 PM
To: [log in to unmask]
Subject: [IPC-600-6012] Dent

Hello does anybody has a clue what this can be if it is not a blister?
Please read below and see attachments



Attached is the best photo I can get.

Fr top it sinking down, fr bottom it's budging up.

We see this right after reflow on random location.

It is not blister, the cross section photo does not show that.

Only abnormalities is difference in Cu thickness ( slight thinner on dent
area).



Any clue??



Hello Constantino,
How are you ??
I need your advise on recent issue I'm facing in our line.  It pretty new
for me ..
I hv  this PCA which produce "dent mark" during rework.

This dent appeared after performing rework to remove the SRAM (the location
that you see dent) with hot air blower, the temp is set to 350 deg C, blow
time is around 15 seconds (dent will be created - in shape of circle about
10mils diameter ), effective temperature on the comp is 290 deg C.

Every single board which we rework produce this condition.
It's not d/c related as we see similar condition on other d/c as well.
More then one d/c and more one location.
We try to simulate the hot air rework on raw card and the dent appears too.

Today line found 2 more boards with dent after going through IR reflow only
( no rework involved ).

The PCB supplier perform cross section on the dent location,  there is no
abnormalities seen expect for a difference in the copper thickness before
and after the dent.

I hv attached the cross section photo.. I can not get a snap shot of  the
dent due to glossy appearance and light reflection.

I can not figure out the root cause..
1) What actually caused this dent ?? Mechanical or thermal
2) Thermal should expand not contract !!
3) Why the dent appear only after reflow / rework ??

Even the PCB suppliers cracking their head.
This board is not failing   ICT/ FVT however it not as simple as cosmetic
defect !!!

I really hope to hear fr you.