From left to right: Will Willoughby (formerly of the Office of the Secretary of the Navy), Ricardo Vanzetti (Vanzetti Systems), Fred Henley (Texas Instruments), and Jim Raby (formerly of EMPF).
| Created: 3/5/96 | Updated: 4/18/97 |
10 Years as NSF Center of Excellence
In 1986, the National Science Foundation (NSF) awarded a grant to Professors Michael Pecht
and Donald Barker to establish a center of excellence for electronic products research at the
University of Maryland. As a result, the Computer-Aided Life-Cycle Engineering (CALCE)
Center was organized as an Industry-University Cooperative Research Center (IUCRC). The
thrust at that time was to address the life-cycle needs of electronics manufacturers.
In 1990, the name was changed to the CALCE Electronic Packaging Research Center (EPRC), with a focus on cost-effective, reliable electronic packaging. At that time, the CALCE EPRC was advanced to a State-Industry-University Research Center (SIUCRC), receiving a six-fold increase in funding from NSF, with matching funds from the State of Maryland and a tripling in industry sponsorship.
The technical direction concentrated on evaluating widely accepted reliability methods, including allocation, parts selection, reliability prediction, derating, environmental control, screening, and qualification. It became apparent that many manufacturers of electronic hardware had come to rely on the security of government-approved reliability documents such as Mil-Std-785 (Reliability Program for Systems and Equipment) and Mil-Hdbk-217 (Reliability Prediction of Electronic Equipment), even though following them often led to poor part selection, improper derating, high-cost cooling solutions, and long development times. Using these documents, any solution to a reliability question was deceptively simple: select specific devices, derate them, run them cool, and introduce redundancies. Auditing quality was accomplished similarly, with government mandated tests such as Mil-Std-883 (Test Methods and Procedures for Microelectronics), perpetuating the myth that reliability and quality could be tested into a product. The costs for following the mandated guidelines were passed on to the customer, resulting in more expensive products without a commensurate increase in performance or reliability.
Today, the CALCE EPRC has created an environment in which the technical limitations of industrial reliability practices have been identified and new, science based reliability guidance is being implemented. One significant outcome was the development of physics-of-failure reliability assessment software for printed circuit assemblies. This software has been used to evaluate avionics, automotive electronics, missiles, satellites, computers, power supplies, electro-optics, and communications equipment. It has also been commercialized by Texas Instruments and is available in their product called CARMA.
The CALCE EPRC is now implementing a fundamentally new approach to addressing reliability. Based on research into the mechanics of failure processes, knowledge of how failures occur is being gathered in order to gain control over failure mechanisms and manufacturing flaws. Coupling this data with novel simulation techniques, the CALCE EPRC is enabling design for reliability, reliability assessment, and virtual qualification (or qualification by design) of new electronic products.
One CALCE EPRC goal is to help our members develop highly reliable products ten times faster over the next five years
Our approach and recommended procedures will impact all activities in the entire product
development cycle, including basic system architecture, qualification, manufacturing, quality
conformance, requirements for supporting sub-systems (i.e., cooling systems and redundancy),
testing, and repair. The results will include significant cost reductions by eliminating
previously hidden manufacturing costs, increased reliability of electronic systems, and quicker
time to market.
Recognizing the critical competitive advantages of the CALCE EPRC approach, companies in electronics-related industries (avionics, consumer products, telecommunications, automotive) from all around the world are becoming members. With an annual research budget of over $4.5M and a team of over forty researchers, the CALCE EPRC and its members are aggressively preparing for the next ten years.
Laser Inspection System Donation
A Vanzetti Laser Inspection System was donated to the CALCE EPRC by Texas Instruments
(Dallas, Texas). This inspection system for solder joints can significantly increase the
productivity of assembly lines by eliminating visual inspection, and will be used to educate
students, assist industry and government in research, and train factory-floor workers.
The Vanzetti system provides a single laser pulse to individual solder joints, examines its thermal emission, and through pattern-recognition algorithms, compares it to a known thermal signature of a good joint. The system can detect solder defects, such as insufficient and excessive solder, voids, damage, and bridging.
Near-term CALCE EPRC research will utilize the Vanzetti system for multi-beam laser soldering techniques. The advantages of laser soldering over conventional techniques include localized noncontact heating, desirable microstructure and mechanical properties, optimal intermetallic formation, and ease of automation. Application areas include opto-electronics and other micro-joining applications. The proposed investigations are particularly timely in light of the increasing interest in environmentally benign, leadfree soldering materials and fluxless processes.
A donation ceremony with representatives from Texas Instruments, the University of Maryland, and other dignitaries took place on November 20, 1995 (see photo below). Dr. Riccardo Vanzetti, inventor of the system that bears his name, attended the ceremony and gave a public presentation.
From left to right: Will Willoughby (formerly of the Office of the Secretary of the Navy),
Ricardo Vanzetti (Vanzetti Systems), Fred Henley (Texas Instruments), and Jim Raby
(formerly of EMPF).
Ion Mobility In Molding Compounds
Moisture ingress into plastic packages can transport ions to the unpassivated bond pad area
and lead to corrosion-related failures in integrated circuits. There is a high correlation
between the concentration of halide ions in molding compound formulations and corrosion-
related failures in plastic encapsulated microcircuits (PEMs), as well as a correlation between
the concentration of antimony oxide flame retardant and the failure rate in PEMs. In
addition, there is evidence in the literature that ion diffusion rates and failure rates may
increase significantly due to additives in low-stress molding compound formulations. Despite
the correlation between the concentration of ions and the failure rate in PEMs, an evaluation
of ion diffusion rates in molding compound formulations has not been performed.
The CALCE EPRC is now conducting research to obtain a better understanding of the phenomenon of ion diffusion in PEMs and to assess the correlation between ion diffusion rates and failure rates in PEMs. To measure ion diffusion rates in epoxy molding compounds, researchers at the CALCE EPRC have designed an apparatus that incorporates a transfer molded block of molding compound as a semipermeable membrane separating aqueous solutions.
For example, the molded sample may separate an aqueous solution of sodium chloride (0.9% NaCl for salt water) from deionized water, and the diffusion rates of sodium and chloride ions may be measured by ion-specific electrodes. The electrodes, placed in both reservoirs to provide a mass balance for ion transport, are designed to detect ion concentrations in the ppb range. This technique has the advantage of measuring ion diffusion directly, rather than measuring changes in volume resistivity in the molded encapsulant as an indication of ion diffusion rates.
The dependence of the ion diffusion rate on the composition and cure schedule of the molding compound, ion size and charge, concentration of ions in the aqueous solution, and thickness of the molded encapsulant are being investigated. In addition, the dependence of the ion diffusion rate on temperature is also being examined to determine the effect of the high temperatures used for accelerated testing on diffusion rates and failure rates in PEMs.
These studies promise to fill a critical void in the literature on the ion diffusion rates in molding compounds and the relevance of molding compound composition and cure schedule to ion diffusion rates and failure rates in PEMs. With this knowledge, the reliability of PEMs in storage as well as operating conditions will be better understood. For more information, contact Dr. Michael Pecht.
Laser Diode Thermal Management Modeling
Many failure mechanisms in laser diodes are related to the operating temperature of the
device. Consequently, thermal loading during operation has been the subject of many
reliability studies.
The CALCE EPRC has developed a procedure to predict the temperature rise within the active region during operation. To begin, a simple heat generation model is used to allocate the power dissipation to specific regions and in specific quantities, as a function of the operating characteristics of the device.
A finite element analysis (FEA) has been applied to a commercial structure. The FEA was performed on a 2-D cross-section to generate the steady-state temperature profile in the active region during operation and the transient temperature response.
In conjunction, a design-of-experiments (DOE) approach was used to study the effect of a few operating parameters on the temperature rise in the laser structure. The DOE variables studied were thermal conductivity of the heatsink material, laser output power, and spacing between diode laser emitter elements. Statistical constants produced by the DOE analysis and physical relationships derived from application of fundamental heat transfer principles to our structure were then coupled to produce a simple analytical expression.
The expression will accurately reproduce the FEA result in predicting the maximum temperature rise within the laser emitter region. The expression includes the explicit dependence of the DOE variables studied on the temperature rise. The information gained can be used to make practical design choices and to ensure safe and reliable operation of laser devices. This approach can also be used in similar parametric studies and design evaluations. For further information, contact the CALCE Center at (301) 405-5323.
High Temperature Electronics TRP
The development of electronics that can operate at highly elevated temperatures has been
identified as a critical technology for the next century. To address this issue, the CALCE
EPRC is teaming with United Technologies, AlliedSignal, Boeing, Pratt & Whitney,
Honeywell, Moog, Ford, Rockwell, Hamilton Standard, Parker Hannifin, Toranga
Technologies, and others on a government-funded Technology Reinvestment Project (TRP).
The objective is to develop high temperature electronics for integration into military/aerospace
and commercial/industrial systems resulting in improved affordability, reliability, and
performance. Leveraging previous and current efforts, the CALCE EPRC is contributing
reliability assessment methods and software tools to meet the objectives of this high
temperature electronics project. For more information, please contact
Dr. Patrick McCluskey.
Optoelectronic Reliability Assessment
Traditional reliability assessment methods based on statistical curve fitting of failure data have
typically been used in reliability prediction for commercial and military optoelectronic
systems. However, because the optoelectronics industry is experiencing rapid change and
improvement in manufacturer processes, standard failure data is often statistically insignificant
or overly vendor-specific, thus highly inaccurate.
The CALCE EPRC has conducted a study of the reliability of laser diodes (LD), light emitting diodes (LED), and fiber optic cables. As a result, an outline of the major failure mechanisms for LDs/LEDs and optical fibers, and a discussion of the physical principles and environmental factors influencing the degradation process, has been produced. The study also identified some failure mechanisms (bond failure, electrode degradation, electromigration, and electrostatic discharge) that are common between LD/LED and microelectronic devices.
The study identified a great deal of work on the effect of fiber optic reliability on the characterization of silica (or plastic) fiber cable and the processes related to crack growth. However, research on the reliability and performance of assembled fiber connectors remains largely in the infant stages.
The study established that much of the previous work on failure mechanisms unique to optical components has focused almost exclusively on the impact of temperature and, in the case of optical sources, current density. Consideration of factors such as humidity, temperature cycling, mechanical stress, and material properties has typically been given only a qualitative treatment. This particular shortfall in the current research on optoelectronic reliability physics is an area in which the CALCE EPRC plans to have a significant impact.
The CALCE software was recently used to assess the reliability of a laser transmitter module that features a pigtailed optical fiber for output coupling of the laser signal. Algorithms for fiber optic coupling loss were applied following the calculation of vibrational resonances, which yielded component displacement values. Also, an acceleration factor for laser lifetime was determined, based on the calculated steady-state junction temperature under operation.
This software assessment highlights the necessity to treat optoelectronic reliability in a hybrid manner. The knowledge learned from assessing the typically silicon-based laser driver circuitry is quite important. However, the rudimentary treatment for assessment of the laser component and the fiber cable assembly is currently underdeveloped. One future CALCE EPRC goal is to improve the usefulness of our software tools for reliability assessment of optical components and systems typically found in optoelectronic systems. For further information, please contact the CALCE Center at (301) 405-5323.
