Acceleration Factors –
How long do we have to wait?
In his recent article, Dr. Bob touched on some of the major factors dealing with acceleration factors. Discussion of this issue is timely.
In our world, the acceleration question has been raised with increasing frequency when test programs are considered. There’s always someone who asks the question, “Do these tests indicate the field-life of this component? And if so, what is it?” Well let’s see if we can shed some light on this issue, the pitfalls that exist, and what we can do about it. How long do we have to wait?
Time is too short and available space is limited to discuss this issue fully, so I will try to be as concise as possible.
There are a number of key environments that can initiate failure mechanisms that impact connector and/or contact systems. Many of these have been mentioned and briefly discussed in past articles. This dissertation will discuss these major contributors: durability, thermal, humidity, atmospheric contamination, and vibration.
This is rather simple and one that Dr. Bob has adequately described. The key is to objectively estimate the expected number of mating cycles the component is expected to endure for the expected life of the product. This may range from one cycle, where it is mated, and will never be unmated during its life cycle. At the other end of the spectrum, it may represent thousands of cycles. If a piece of equipment is designed for five years, and the unit is plugged and unplugged twice a day, this would represent more than 3,600 cycles. This is easy to simulate in testing. The only concern is the type of cycle rate, which should not be more than 500 cycles per hour maximum. This rate has been proven not to generate frictional heating, thus minimizing frictional heat considerations. Durability tests will adequately demonstrate the magnitude of wear that exists for a given contact system. It will vary in intensity contingent on: contact geometry, normal force, plating systems, base metal consideration, and surface conditions.
The base problem is that durability by itself does not guarantee an estimated life. It must be integrated with environments. Again a secondary problem is how to do this. This is the point where rational thought has to be established. Durability generates wear, and hence, debris fields. It’s the interaction between the environment/debris and the displacement of the prime protective material that will allow corrosion to be generated. Durability should be viewed as a conditioning test.
As pointed out in previously written articles, stress relaxation can result in time-dependent failure mechanisms. The key element is loss of normal force. Data has been generated, mainly by the material industry, that predicts what stress relaxation may occur. It also developed data based on different temperature and applied stress levels, allowing an accelerated test setup. One industry standard has three different test temperatures correlated to three-, five-, and 10-year field-life vs. test durations. This sounds very promising, except for two major problems.
Many companies have developed their own proprietary accelerators based on their own proprietary criteria, and the accelerators do not agree with each other.
- There is a fault that exists will all thermal accelerators. Although they accurately can predict thermal relaxation, it’s only valid for defined test strips. They do not take into account residual stresses that exist and are imposed on a contact due to the manufacturing process. These residual stresses may be more significant than the base properties of the material itself. Two different companies with two different tools can produce totally different residual stress patterns. Thus the predictability is very poor as a result of this, and has a low confidence level relative to actual results. The stress relaxation and therefore, the magnitude of the loss of normal force, should be based on:
- The material system used (base metal and plating system)
- The base stress of the material system
- The residual stress from the manufacturing process, the most problematical one to predict
- The operational stress imposed on the system
The last two items have not been integrated into the prediction cycle since the factors involved are not well known, if known at all. However, empirical testing techniques have proven to generate good, accurate information.
There have been attempted accelerators. However, due to the different humidity conditions in the world, these accelerators are wide ranging. It may vary due to the constant high humidity (as in a rain forest) to very sporadic conditions in the desert areas, such as the Sahara or Death Valley, and a whole range in-between.
However there has been data generated that indicates that there are two or possibly three humidity classifications one has to be concerned about:
- Humidity greater than or equal to 70 percent
- Humidity 25 to 70 percent
- Humidity less than 25 percent
This may be significant since humidity is a major catalyst for corrosion. Without it, corrosion would be greatly diminished. Although this is a promising development, additional work is required to define these boundaries. Until that occurs, it’s either pure guesswork, or decisions based on experience (tempered by emotions).
The accelerator factor for this type of concern has by far been the most successful and effective. The test of choice is mixed flowing gas (MFG). The consensus accelerator is two days in the chamber equals one year in the field.
As previously discussed in the many papers I’ve written, I won’t take up time to discuss the details except to say:
- It was based on a 15-year sponsored study.
- Laboratory testing confirmed said accelerator to a tolerance of 16 months.
Vibration is probably the least understood factor affecting product. We tend to evaluate at a constant severity level for a given period of time in a one-dimensional manner (X-axis, Y-axis and Z-axis, one axis at a time). If a product passes the test, one tends to become comfortable, and then becomes completely emotional when it fails in field application situations.
Perhaps the thinking should be to view this attribute as a three-dimensional problem, where the interaction between the three axes is a determining factor. Unfortunately this is a cost-prohibitive approach requiring precise scanning evaluations of actual operating equipment to establish the proper vibration envelope needed for testing.
From an accelerator point of view, this is one of the tests where the accelerators are based on experience and proper feedback. If a connector is prone to failure, it will occur very rapidly. Thus, if a connector performs properly for a one hour/axis (three axes total), it probably will not fail in vibration relative “to the severity levels” tested. Change the severity levels or the fixturing could change the results.
Thermal Cycling/Thermal Shock
There is a significant difference between thermal shock and thermal cycling. Thermal shock involves an instantaneous change in temperature whereby the thermal cycle has specified ramp times that must be met. The thermal shock test generally simulates those applications whereby “rapid” change in temperature is involved with very short durations (five cycles is the norm.). Thermal cycle simulates more of a gradual warming up or cooling down (equipment turning on/off conditions) and long durations are more common (from 500 to 3,500 cycles). The base accelerator for the thermal cycling is one cycle equals one day. Thermal cycle tests require variable monitoring to establish data trends in order to put the projected field life in proper perspective.
A general rule of thumb is if after 1,000 cycles (for both tests) and variable monitoring indicates that stability is maintained, the probability is very low that a problem will occur (as suggested by Jim Whitley, AMP Inc.).
The following is a summary of the accelerators for the various environments used in connector end-of-life predictability.
- Durability : 1 mating cycle = 1 mating cycle in actual operation
- Thermal Shock: None, but can be integrated with thermal cycling
- Thermal Cycling: 1 cycle = 1 turn on/turn off (equipment)
- T-Life: Accelerators exist but vary between user agencies
- Singular Gas Test: None
- Two Gas Tests: None
- Mixed Flowing Gas: 2 days in chamber = 1 year in field (consensus)
- Humidity: Variable depending on field conditions and other interacting factors.
- Salt Spray: None
- Vibration: None
The above is where we stand today. Establishment of accelerators can take a lot of funding, time, and resources, with the commitment to share data with industry at large. So how long do we have to wait? My answer: “More than my lifetime.”
By Max Peel, Senior Fellow, Contech Research
- Nanocrystalline Silver Alloy Contact Finishes in Electronic Applications - April 6, 2015
- Nanocrystalline Silver Contact Platings - March 16, 2015
- Dr. Bob on Gold Flash Contact Finishes (and Max Peel) - September 22, 2014