(1) you can actually make it dwell = zero to fake the test:  I saw thermal shock with 5 min. dwell that still totally frosted over on the unit in hot chamber.  You actually don't get real temperature.  (of course, that test was disqualified).  You can make the dwell small enough, that units internal temperature is actually no dwell at all, as you wish. 
(2) any temp range and any ramp rate: model as you wish.  Why?  Brittle transition temperature and re-crystallization are important and material dependent.  If you just want to do theoretical study, you can plug diffusion parameter and set boundary condition as you wish plus CTE (make sure you take into consideration of crystallography phase transition and dimensional change of all the materials of your liking).  Assume you know all of the stress parameters.  As for transient you want to simulate: why not get a TE module and ramp up and down (or LED to pulse modulate), you can get any temp and anything you want by control your environment (still air for example).  A thesis for you ;-). 
Good luck.  

Joyce Koo
Materials Researcher - Materials Interconnect Lab
Research In Motion Limited
Office: (519) 888-7465 79945
Mobile: (226) 220-4760

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of John Nieznanski
Sent: Monday, June 25, 2012 12:30 PM
To: [log in to unmask]
Subject: Re: [TN] solder fatigue analysis for continuously ramping temperature swings

Hi Dave,

Thanks for your response, and a couple others I have received. This is certainly appreciated. I think my question is something a bit different. I agree with you that the failure mechanisms are different in the thermal shock vs thermal cycling domains. I don't think I am shocking the parts as the ramp rate (delta temp / delta time) is always 20C / 30 minutes = less than a degree C per minute, during both heating and cooling phases. 

I think what I am asking is there a general predictive SMT leaded/leadless solder fatigue model that applies for cases where both the following conditions apply:

(1) there is no real "dwell time" or temperature plateau, either in operational usage or accelerated testing. So the soak time and the Td (half-cycle dwell time) parameter in E-W is zero.
(2) there is no accelerated testing data available, over any temperature range, or for any ramp rate.

I suspect that the Engelmaier-Wild model (IPC-D-279, Appendix A-3.1) doesn't really apply here per item 1 above. So in the absence of both (1) and (2) above, are there any predictive models that can be used?   

Another way of asking this question is, are there any generally accepted, published and/or standardized methods and analysis models that can be applied here, or is this not possible (e.g., highly application dependent, too many variables, very ad-hoc/custom, or proprietary?) How were these issues addressed when SMT was in it's infancy? I'm trying to track down a few leads on this. 

Thanks again for any helpful thoughts on this topic. 

John Nieznanski


---- "David D. Hillman" <[log in to unmask]> wrote: 
> Hi John - the solder joint fatigue testing a fair majority of the industry 
> conducts is thermal cycle testing - as you detailed in your email, we 
> allow the solder joints to stabilize at a given temperature limit for a 
> specific time period. Your question focuses on a second conditioning 
> protocol - thermal shock testing. The IPC-9701 specification defines 
> thermal cycle testing as temperature transition rates below 20C per minute 
> and thermal shock testing as temperature transition rates greater than 20C 
> per minute. The failure modes for thermal cycle testing and thermal shock 
> testing are very different so any accelerated testing needs to replicate 
> the thermal conditions that your product will experience. The IPC-9701 
> specification can be used for both test methodologies.
> 
> Dave Hillman
> Rockwell Collins
> [log in to unmask]
> 
> 
> 
> 
> John Nieznanski <[log in to unmask]> 
> Sent by: TechNet <[log in to unmask]>
> 06/23/2012 09:13 AM
> Please respond to
> TechNet E-Mail Forum <[log in to unmask]>; Please respond to
> <[log in to unmask]>
> 
> 
> To
> <[log in to unmask]>
> cc
> 
> Subject
> [TN] solder fatigue analysis for continuously ramping temperature swings
> 
> 
> 
> 
> 
> 
> Hello TechNet,
> 
> Has anyone had any success predicting, measuring and correlating solder 
> fatigue wearout for circuits that are continuously ramping up or down in 
> temperature (linearly) between fixed thermal limits over fixed time 
> periods? The fixed time periods are stable and repeatable as is the 
> temperature change over these periods. 
> 
> As an example, a normal ramp-up time is 30 minutes and the temperature 
> ramps linearly from 20C to 40C. Then the temperature immediately ramps 
> down from 40C to 20C over the next 30 minutes. This pattern repeats.
> 
> The Engelmaier method described in IPC-D-279, IPC-SM-785 assumes cyclic 
> temperature swings between two fixed temperature limits.  The solder joint 
> temperatures stabilize at each temperature limit for a fixed interval 
> before periodically switching to the other temperature limit for the same 
> interval.  How can be solder fatigue wearout be quantified in solder 
> joints that instead of ?soaking? or stabilizing at each of the temperature 
> limits, are always either heating up or cooling down between these known 
> limits in a predictable, linear way?  Intuitively, it seems that creep and 
> stress relaxation should be less, but how to quantify when solder joint 
> wearout occurs (N50, mean lifetime)?
> 
> Thanks.
> 
> John Nieznanski
>  
> 
> 
> ______________________________________________________________________
> 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 transmission (including any attachments) may contain confidential information, privileged material (including material protected by the solicitor-client or other applicable privileges), or constitute non-public information. Any use of this information by anyone other than the intended recipient is prohibited. If you have received this transmission in error, please immediately reply to the sender and delete this information from your system. Use, dissemination, distribution, or reproduction of this transmission by unintended recipients is not authorized and may be unlawful.

______________________________________________________________________
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] 
______________________________________________________________________