Printed wire assembly HASS profile development based on HALT
Introduction
Highly accelerated stress screening test (HASS) is an effective mean of screening out infant mortality failure modes of a system before shipment to consumers. Screening tests are generally conducted under accelerated conditions so latent defects are exposed in a short testing time compared to field operation. HASS uses rapid temperature transitions, hot and cold dwells combined with multi-axis vibration to weed out defects without removing significant amount of product life.
Prior to HASS, highly accelerated life testing (HALT) is usually implemented to expose any weaknesses in new designs at early stages of product development cycle. HALT consists of five steps: cold stress, hot stress, rapid thermal cycling, random shock, and finally a combined random shock-thermal cycling profile [1]. The goal of the first three thermal stress tests is to go beyond the specifications and identify failure modes caused by temperature extremes. During the vibration stress step, the device under test (DUT) is subjected to repetitive shocks (RS) as low as 5 gRMS (root mean square g-loads) for a dwell time of 10 min then steadily increasing vibration levels until reaching the destructive limit of the DUT [2]. HALT combined test consists of combined RS and temperature excursions so the DUT is exposed to uncontrolled broadband multiaxial excitation [3]. The thermal profile limits are the rapid thermal transition operational temperatures relaxed by 10 °C. A good description of HALT and HASS tests and how to select test limits is provided in [[4], [5], [6]].
Research suggests that HALT combined test exposes more failure modes than the other individual steps. For example, McLean [4] analyzed 47 HALT tests from 19 industries. The tests resulted in 45 failures during vibration, 4 in thermal cycling, and 20 during the combined test. Similarly, Aoki et al. [7] compared the results of environmental stress testing using HALT, thermal shock, and electro dynamic (ED) vibration stresses for vehicle surface mount and leaded printed circuit boards (PCB). Results suggests that HALT combined environment RS and thermal cycling test exposes multiple failure mechanisms that require separate tests to expose using traditional stresses. It is impossible to correlate structural failure during HALT to life cycle loading [8].
HASS is usually consists of five profiles with each profile consists of five cycles of combined RS and thermal cycling cycles as shown in Fig. 1.
Since all screen tests including HASS remove life from products, the goal of HASS should limit this degradation in life to small amount such as 2–5%. As a result, many organizations require product to survive at least 20–50 repetitive HASS profiles before implementation without causing damage to the outgoing product [9]. This process is referred to as proof of screen POS which is used to demonstrate screening effectiveness and safety. Effectiveness is demonstrated by catching marginal/defective products while safety is demonstrated by not damaging or significantly affecting good parts life [10]. As a result, new products such as printed wire assemblies (PWA) should survive at least 20 HASS profiles. To the best of the authors' knowledge, there are no clear guidelines on how to derive HASS from HALT. Instead, different organizations used different practices. One common practice is to limit the maximum vibration to the 50% of the latest operational vibration level during the HALT combined stress [10,11]. The upper temperature limit is set either at product rated temperature Trating + 5 ° C [10] or 80% of the highest temperature used in HALT combined test [5]. In both cases, the high vibration dwell time is 50% of the high and cold dwell temperature time and ramp rate of thermal cycles is 60 °C/min [2,10]. The 50% vibration level during HASS is a general guideline that needs to be verified by several trial and error testing. In this article, we are proposing a new method to predict HASS life under different RS levels and thermal cycle limits using physics of failure and statistical modeling. Successful implementation of the proposed method will enable practitioners to predict and determine how many profiles PWA can survive under different loading conditions, which will reduce testing time and cost.
Section snippets
Literature review
The majority of HALT and HASS failures are due to either low cycle fatigue (100–5000 cycles) or high cycle fatigue (>100 million cycles). Failure due to thermal cycling is a low cycle while failure due to RS is a high cycle one [12]. While much research has been undertaken on reliability problems of electronic products, less work has been addressed on reliability under combined RS and thermal cycling [13,14]. The two competing failure modes combined will fail PWA's by reducing strength due to
Proposed methodology
In this article, we propose the use of physics of failure approach along with Miner's superposition rule to estimate the cumulative damage impact of both RS and thermal cycling. Since PWA failures during HALT test is proportional to vibration induced damage in SJs which is proportional to board strain [14], we will model accumulated median strain as a power function of vibration and temperature as shown in Eq. (5). It is not uncommon to model product life-data as a function of one or more
Case study
A gauge acquisition PWA used for an oil and gas measurement tool rated at 150 °C is used to demonstrate the effectiveness of the proposed methodology. The PWA uses surface mount technology and 75 × 800 mm FR8 board and 63Sn37Pb soldering material. One of the major components mounted on the board is a QFP 208 pins field programmable gate array FPGA that suffers historically from a relatively high failure rate due to harsh environment of wells and shortcomings of supplier quality. The dominant
Conclusions
HASS proof of screen (POS) for products in general and PWA's in specific is a critical step to ensure screen effectiveness and safety. In this article, a new method is proposed to bridge the gap between HALT and HASS to ensure that HASS stresses are high enough to expose weak parts but not to take >5% of good parts life.
The proposed method is demonstrated using a gauge PWA board used in oil and gas tools. The method requires that HALT and one initial HASS test are performed on PWA under
CRediT authorship contribution statement
Mahmoud I. Awad: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing - original draft, Writing - review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (32)
- et al.
Vibration reliability test and finite element analysis for flip chip solder joints
Microelectron. Reliab.
(2009) - et al.
Coupling damage and reliability modeling for creep and fatigue of solder joint
Microelectron. Reliab.
(2017) - et al.
Evaluation of structural design methodologies for predicting mechanical reliability of solder joint of BGA and TSSOP under launch random vibration excitation
Int. J. Fatigue
(2018) - et al.
A review on thermal cycling and drop impact reliability of SAC solder joint in portable electronic products
Microelectron. Reliab.
(2012) - et al.
Determination of the optimal burn-in time and cost using an environmental stress approach: a case study in switch mode rectifier
Reliab. Eng. Syst. Safe.
(2002) - et al.
Comparison of fatigue life prediction methods for solder joints under random vibration loading
Microelectron. Reliab.
(2019) - et al.
Physics of failure as a basis for solder elements reliability assessment in wind turbines
Reliab. Eng. Syst. Saf.
(2012) Introduction to HALT - making your product robust
- et al.
Efficiency improvement of the highly accelerated life testing system by using multiple hammers
J. Mech. Sci. Technol.
(2014) - et al.
Keynote: simulation and test vibration - nonlinear dynamic effects in vibration durability of electronic systems