agree.  no fun at all ;-)
http://extra.ivf.se/eqs/dokument/7%20pet6005.pdf

> Joyce, I hate to spoil your fun, but even that Neanderthal Rework Process
> will not overly extend the IMF by more than 1 or 2 microns.
> When the original IMF is formed during reflow, any additional rework is
> simply soldering solder to the solder already present. The IMF does not
> increase in thickness appreciably.
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Joyce Koo
> Sent: Thursday, September 17, 2015 8:00 AM
> To: [log in to unmask]
> Subject: Re: [TN] BGA Reworked Intermetallic
>
> Hot plate and soaking the part intend to repair in flux and walk away
> until it was reflowed.  remove the defective part and let the PWA cool
> down.  put back on the hot plate, with additional flux and clean the site,
> remove access old solder.  add additional solder to address the pad.  Cool
> down again by remove from the hot plate.  position the part using
> microscope that is not near the work station with tacker flux, bring back
> to the hot plate and reflow again, using tooling to perform minor
> adjustment to position the part and final cool down... if you check for
> the IMC after such a repair, you get very thick IMC - the overall time
> above TL is 5 to 6 x compare to your SMT the least.... (I was told it is
> not un-common in some places for "repair").  scary.
>                jk
>> But Vlad - those conditions are so extreme that something is clearly
>> incorrect in the process so focusing on the IMC is the  wrong root
>> cause analysis path. The focus should be on how/why such an extreme
>> amount of IMC is being created in the first place.
>>
>> Dave
>>
>> On Thu, Sep 17, 2015 at 7:36 AM, <[log in to unmask]> wrote:
>>
>>> Hi Dave,
>>>
>>> That what exactly my point. It does happen under certain conditions,
>>> so I always advise our customers not to start with a very thick layer
>>> of intermetallics.
>>>
>>> Cheers,
>>>
>>> Vladimir
>>>
>>> Sent from my BlackBerry 10 smartphone on the Rogers network.
>>> *From: *David Hillman
>>> *Sent: *Thursday, September 17, 2015 08:22
>>> *To: *Vladimir Igoshev
>>> *Cc: *TechNet E-Mail Forum
>>> *Subject: *Re: [TN] BGA Reworked Intermetallic
>>>
>>> Hi Vlad - the growth rates of IMCs are pretty well known as they
>>> follow an Arrhenius equation behavior. The growth of IMCs from 25C
>>> -100C is extremely low so unless you have a product service
>>> environment that has a high temperature such as 150C for long time
>>> periods, you just are not going to "grow" enough IMC thickness to
>>> impact solder joint integrity. Having an IMC of 12 microns (400
>>> uinches) is HUGE and clearly not a normal, typical case found in
>>> standard electronics production. The biggest issue with having an IMC
>>> discussion is that the data doesn't support IMC failure as a typical
>>> root cause for solder joints when we start crunching the numbers.  I
>>> working with one of our industry colleagues and we are pulling
>>> together a paper on industry "myths" where we plan on including the
>>> topic of IMCs.
>>> As
>>> you can imagine, its a fun paper to work on and we hope to have it
>>> publish in the upcoming year(ish).
>>>
>>> Dave
>>>
>>> On Wed, Sep 16, 2015 at 9:52 PM, <[log in to unmask]> wrote:
>>>
>>>> Hi Dave,
>>>>
>>>> As far as I remember I said it's a reverie for disaster. As you know
>>>> intermetallics will grow over time and a starting point of 10 or so
>>>> microns doesn't help.
>>>>
>>>> I had several cases of failure (cracked intermetallics) after it
>>>> grew up to 12-15 microns.
>>>>
>>>> Does it happen all the time? I don't know  May be not, but the cases
>>>> I had are good enough "argument" , at least for me.
>>>>
>>>> Regards,
>>>>
>>>> Vladimir
>>>>
>>>> Sent from my BlackBerry 10 smartphone on the Rogers network.
>>>> *From: *David Hillman
>>>> *Sent: *Wednesday, September 16, 2015 22:44
>>>> *To: *TechNet E-Mail Forum; Vladimir Igoshev
>>>> *Subject: *Re: [TN] BGA Reworked Intermetallic
>>>>
>>>> Hi Vlad - what data do you have that 10 micron thick layer of IMC
>>>> results in the failure of the BGA? There is TONS of subjective
>>>> comments in the published literature but no hard data of failures.
>>>> As a basic materials engineering principle, IMCs are brittle but
>>>> that specific material characteristic seems to be the only reason
>>>> folks make statements that a "thick" IMC is bad. Should we work to
>>>> keep IMC layers minimized?
>>>> Absolutely
>>>> but I don't believe, as technologists, we should keep propagating
>>>> the myth about thick IMC layers without having published, reviewable
>>>> data.
>>>>
>>>> Dave Hillman
>>>> Rockwell Collins
>>>> [log in to unmask]
>>>>
>>>> On Wed, Sep 16, 2015 at 9:38 AM, Vladimir Igoshev <
>>>> [log in to unmask]> wrote:
>>>>
>>>>> 10 microns thick layer is a perfect way for a disaster down the
>>>>> road :-).
>>>>>
>>>>> There are no parameters for E-Ni, but the appearance of the
>>>>> interface and a P-enriched layer ‎is important.
>>>>>
>>>>> Vladimir
>>>>>
>>>>> Sent from my BlackBerry 10 smartphone on the Rogers network.
>>>>>   Original Message
>>>>> From: Victor Hernandez
>>>>> Sent: Wednesday, September 16, 2015 10:11
>>>>> To: [log in to unmask]
>>>>> Reply To: TechNet E-Mail Forum
>>>>> Subject: Re: [TN] BGA Reworked Intermetallic
>>>>>
>>>>> On ENIG surface I don't see much of an increase in the IMC
>>>>> formation thickness. However, on Cu it is a different story. I have
>>>>> measured IMC formation greater tham 10 microns. Not sure of the
>>>>> below statement about E-NI parameter. Please explain!!!
>>>>>
>>>>> Victor,
>>>>>
>>>>> -----Original Message-----
>>>>> From: TechNet [mailto:[log in to unmask]] On Behalf Of Vladimir
>>>>> Igoshev
>>>>> Sent: Wednesday, September 16, 2015 8:12 AM
>>>>> To: [log in to unmask]
>>>>> Subject: Re: [TN] BGA Reworked Intermetallic
>>>>>
>>>>> The "magic" Number should stay the same 1-3 micron but you'd also
>>>>> have to keep an eye on what happened to the layer of E-Ni underneath.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Vladimir
>>>>>
>>>>> SENTEC
>>>>>
>>>>> Sent from my BlackBerry 10 smartphone on the Rogers network.
>>>>> Original Message
>>>>> From: Datacom - Juliano Ribeiro
>>>>> Sent: Wednesday, September 16, 2015 09:09
>>>>> To: [log in to unmask]
>>>>> Reply To: TechNet E-Mail Forum
>>>>> Subject: [TN] BGA Reworked Intermetallic
>>>>>
>>>>> Hi to all,
>>>>>
>>>>>
>>>>>
>>>>> When we reworked the BGA, removed the component of the board and
>>>>> replacement another BGA, what's the intermetallic thickness ideal
>>>>> after the rework?
>>>>>
>>>>>
>>>>>
>>>>> p.s: Our pcb is ENIG finished and the solder is Tin Lead.
>>>>>
>>>>>
>>>>>
>>>>> _____________________________
>>>>>
>>>>> Juliano Bettim Ribeiro
>>>>>
>>>>> DATACOM
>>>>>
>>>>> ENGENHARIA DE PROCESSOS
>>>>> Rua América Nº 1000 - Eldorado do Sul - RS CEP: 92990-000
>>>>> +55 (51) 8446-2135
>>>>>
>>>>> +55 (51) 3933-3000
>>>>>
>>>>> Ramal: 3484
>>>>> [log in to unmask] www.datacom.ind.br
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> ___________________________________________________________________
>>>>> ___ This email has been scanned by the Symantec Email
>>>>> Security.cloud service.
>>>>> For more information please contact helpdesk at x2960 or
>>>>> [log in to unmask]
>>>>> ___________________________________________________________________
>>>>> ___
>>>>>
>>>>> ___________________________________________________________________
>>>>> ___ This email has been scanned by the Symantec Email
>>>>> Security.cloud service.
>>>>> For more information please contact helpdesk at x2960 or
>>>>> [log in to unmask]
>>>>> ___________________________________________________________________
>>>>> ___
>>>>>
>>>>> ___________________________________________________________________
>>>>> ___ This email has been scanned by the Symantec Email
>>>>> Security.cloud service.
>>>>> For more information please contact helpdesk at x2960 or
>>>>> [log in to unmask]
>>>>> ___________________________________________________________________
>>>>> ___
>>>>>
>>>>> ___________________________________________________________________
>>>>> ___ This email has been scanned by the Symantec Email
>>>>> Security.cloud service.
>>>>> For more information please contact helpdesk at x2960 or
>>>>> [log in to unmask]
>>>>> ___________________________________________________________________
>>>>> ___
>>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>
>> ______________________________________________________________________
>> This email has been scanned by the Symantec Email Security.cloud
>> service.
>> For more information please contact helpdesk at x2960 or
>> [log in to unmask]
>> ______________________________________________________________________
>>
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> ______________________________________________________________________
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> ______________________________________________________________________
>


______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________