Subject: | |
From: | |
Reply To: | |
Date: | Wed, 12 Sep 2012 09:20:52 +0200 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Thorsten
I assume you talk about solder joints. In these highly plastic materials the temperature gradient determines the deformation mechanism (time dependent creep or time independent plastic deformation) and with that the degradation mechanism. Each of these mechanism has its own mathematical description to estimate the average lifetime. The creep degradation (Low cycle fatigue) can well be calculated with the formula of Coffin and Manson, high cycle fatigue has been modeled by Basquin (Wöhlers law).
Since the two mechanisms of degradation are different metallurgical processes there is no mathematical transformation.
Usually (more than 99% of all cases) heat up and cool down rates in electronic assemblies are in the range of 1°C/min. which activates pure creep with grain boundary sliding (GBS). The maximum acceleration for the test is thus limited by the maximum temperature rate that activates mainly GBS. This rate depends on the temperature itself. This means at high temperature the temperature ramp to activate GBS is steeper than at low temperature.
If you already run the test with 10°C/min everything becomes more difficult. At low temperature (< 20°C)you activate mainly dislocation creep (DC), at temperatures above 50°C you activate maily GBS. By changing the ramp you change how much GBS and how much DC the solder joints will see.
There is no modell which transforms the damage caused by DC to an equivalent damage caused by GBS. Thus, if you change the parameters your result is only valid for this test. This is fine if you want to comparative tests. However, the results will be meaningless for a prediction of the livetime.
Best regards
Günter Grossmann
Scientist
Empa
Swiss Federal Laboratories for Materials Science and Technology
Überlandstrasse 129
8600 Dübendorf
Switzerland
Tel +41 58 765 4279
Fax +41 58 765 4054
[log in to unmask]
www.empa.ch
>-----Ursprüngliche Nachricht-----
>Von: TechNet [mailto:[log in to unmask]] Im Auftrag von Torsten Hagge
>Gesendet: Dienstag, 11. September 2012 11:16
>An: [log in to unmask]
>Betreff: [TN] Reliability
>
>Hi all,
>
>has anybody a simple ALT calculation how to decrease the number of
>thermocycles if the thermo gradient is increased?
>
>e.g. 1000 cycles @ 10°K/min >> xxx cycles @ 15°K/min, yyy cycles @
>20°K/min ; dwell time constant
>
>Torsten
>
>Torsten Hagge
>Gewerbegrund 5-9
>24955 Harrislee
>Telefon +49 (0) 461 7741-624
>Telefax +49 (0) 461 7741-642
>[log in to unmask]
>www.kristronics.de
>Gerichtsstand Flensburg, Handelsregister: HRB 1433 FL Geschäftsführer
>Dipl.-Ing. oec. Thormod Ohm Ust-Id-Nr. DE 811182059
>Bankverbindung: Deutsche Bank AG Flensburg, Konto-Nr. 4216610, BLZ
>21570011
>IBAN: DE32 2157 0011 0421 6610 00, BIC: DEUT DE HH 215
>
>______________________________________________________________________
>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]
______________________________________________________________________
|
|
|