Guenter et al,

Thanks for the input.

The point of environmental testing - in this instance - is to demonstrate that the product is fit for service in the environment the product will operate in for the lifetime of that product.

In my view parallel testing will demonstrate that a product can or cannot withstand a single type of 'stress'. This is useful and of course gives an insight into fundamental properties of some aspect of the design. However in real life a product is subjected to many 'stresses' at the same time and parallel testing will never test for this. Therefore parallel testing may tell you that a product is unfit for service but will never tell you that it definitely is. My view is that it is useful as a precursor to sequential testing.

It would probably be best if one could apply a lot of 'stresses' simultaneously but that is generally impractical. So one is left with sequential testing. Yes, it is probably impossible to devise a test spec that will cover every possible failure mode. However to say it is therefore not worth doing is like saying half a loaf is no better than no loaf at all. It seems to me that careful consideration of standard test specs e.g. EN 60068 which are based on an accumulation of experience combined with one's own specific product knowledge should result in the best qualification programme.

Regards,





[log in to unmask] wrote:

> Eric
>
> Just as a basic thought: It is not the goal of a test to be as severe as possible but to proof that a product/process meets the requirements. To do so, the tests performed have to be chosen carefully and designed to show any week points but not to wreck the specimen on any account.
>
> Now, you talk about different philosophies and different goals:
>
> - Parallel testing:
> The idea of sequential testing is to find the answer of a specimen on a particular stress. There is no sense in starting combined tests is one doesn't know how an item to be tested reacts if it is exposed to let's say high temperature.
>
> - Sequential testing:
> As I get it, you want to evaluate the effects of combined environmental stress on a specimen by exposing the specimen to the stresses one after the other. See, as soon as you get into combined testing the whole thing gets quite complex. Your thought is right, if the stresses do occur one after the other. One example is the degradation of roads due to frost. Firstly water penetrates into small cavities of the surface. When freezing the water expands breaking pieces out of the road. However, if the combined stress occurs simultaneously the attempt to simulate their combined effect by sequential testing seems fruitless to me. Just as an example for this thought we can take diffusion. As you know, diffusion through a membrane is driven by the temperature and by the difference of the concentration of a certain substance on both sides of the membrane. High temperature accelerates diffusion and so does a large difference in concentration. According to your thought one could theoretically
> perform a test where a specimen is firstly exposed to an environment with high temperature and a low difference in concentration dC ( or even worse not caring for the dC) and then making a test with the same specimen where it is exposed to a high dC with low Temperature (or similarly to above not caring for the temperature). As you see clearly, this test is absolutely meaningless when it comes to  find the effect of the combination of temp. and dC. If you want to find that you have to run tests where the dC and the high temperature occur together.
>
> I think I mentioned this earlier. Designing sensible tests is a difficult business and requires a lot of knowledge of physics of failure and incorporates some math. To learn about the mathematical part read Montgomerys "Design and Analysis of Experiments". For the physics of failure part I am afraid you have to go through some literature dealing with physics, failure analysis and material science.
>
> Have a great day
>
> Guenter
>
> .
>
> EMPA
> Swiss Federal Institute for Materials Testing and Research
> Centre for Reliability
> Dipl. Eng. Guenter Grossmann
>
> 8600 Duebendorf
> Switzerland
>
> Phone: xx41 1 823 4279
> Fax :     xx41 1823 4054
> mail:     [log in to unmask]
>
> ---------------------------------------------------------------------------------
> Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
> To unsubscribe, send a message to [log in to unmask] with following text in
> the BODY (NOT the subject field): SIGNOFF Technet
> To temporarily halt delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL
> To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
> Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
> Please visit IPC web site http://www.ipc.org/html/forum.htm for additional
> information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
> ---------------------------------------------------------------------------------

--
Eric Christison
Mechanical Engineer
STMicroelectronics
33 Pinkhill
Edinburgh
EH12 7BF

Tel: (0)131 336 6165
Fax: (0)131 336 6001

---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site http://www.ipc.org/html/forum.htm for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------