FROM: I. MEMIS, DEPT GGMG/BLDG 257-4, AISLE A, ENDICOTT ENDICOTT ELECTRONIC PACKAGING SUBJECT: CTE Values for PWB The previous postings are atttached. ______________________________________________________________________ The problem mentioned requires accurate local CTE values. For a PWB, CTE changes after each process such as drilling through holes, filling the holes with solder, copper plating etc. The PWB CTE also changes with it's geometry. The entire PWB area of interest must be measured to accurately assess CTE impacts in the area of components and other features such as heat sinks. Advanced computational mechanics tools such as the Finite Element Method (FEM) are tempting for CTE determination. A numerical prediction of CTE, however, requires accurate assessment of material properties such as Youngs modulus, Poissons ratio, CTE etc. and the linear or nonlinear constitutive relationship of each material used in the component. The in-situ material data, especially temperature- dependent material properties and nonlinear constitutive relationships, are not readily available in the literature, which makes prediction uncertain. Recently, moire interferometry was proposed as a tool for CTE measurement of electronic packaging components. The abstract of the paper can be found below. ABSTRACT Moire interferometry is a whole-field displacement measurement technique with a sub-micron sensitivity, which can produce the global CTE and the local CTE across the entire component with high accuracy. The large dynamic range of the method makes it compatible with CTE evaluation of a broad range of components. Two experimental procedures were suggested for (1) CTE over a fixed temperature range and (2) CTE as a function of temperature. The procedures are implemented for various thin small outline packages, a plastic ball grid array package and a stacked memory cube. The results emphasize the importance of a whole-field measurement technique. A supplementary TMA test on an aluminum alloy is conducted and the results are compared with a moire measurement to discuss the accuracy of the proposed measurement procedure. {1} B. Han and Y. Guo, "Determination of Effective Coefficient of Thermal Expansion of Electronic Packaging Components: A Whole-field Approach, IEEE Transactions on Components, Packaging and Manufacturing Technology-Part A, (submitted for publication), 1995. KEY CONTACTS FOR MORE INFORMATION: Bongtae Han, IBM ENDICOTT, 607-757-1077 INTERNET ID - [log in to unmask] C.K. Lim , IBM ENDICOTT, 607-757-1138 INTERNET ID - [log in to unmask] Previous Postings ----------------- We found that using TMA can be misleading, especially in an "unbalanced" design (e.g. amount of copper NOT symetric around the center (thickness) of the board. We found strain gageing both sides of a PWB the only reliable way to get CTE data. Jim Maguire Principal Engineer, E/E M&P (206)657-9063 Boeing Defense & Space Group [log in to unmask] Reza, Might we be able to measure the CTE of this gentleman's boards using TMA? He may have coupons available... It seems he is looking for an analytical means to predict CTE differentials. Just a thought. Phillip ______________________________________________________________________________ _ Subject: CTE Values of Printed Circuit Board and Heatsink combination From: [log in to unmask] at Internet Date: 10/31/95 11:15 AM Help! The Engelmaier equations used to estimate cycles to failure for leadless SMT components has, obviously, as one of it's terms the CTE of the printed wiring board. I have been unable to obtain information on CTE of printed circuit boards, at least with any confidence, and measuring the CTE after build is too late in the product development cycle. This is cirtical to our applications due to the long thermal cycling requirements of some of our programs (7-15 years). The CTE of the board is effected by: Base Material (Kevlar, Polyimide, G10) Resin Materal Lot Layer count Amount of copper (planes) Heatsink bonding Heatsink bonding material (epoxy, acrylic, thermabond) Heatsink material (aluminum, copper) 1. Has anyone done any testing on CTE of boards as a function of the above? 2. Are there any equations which can be used to predict the CTE of the board and/or heatsink combination? 3. Is there any data on the CTE of boards besides the ranges I have seen in the text books. I am considering running an experiment to determine the effects of these parameters. Would anyone be interested in the data and of course help support the effort with boards, heatsink and bonding? Jim Faris AlliedSignal Aerospace Guidance & Control Systems (201) 393-3779 USERID: MEMISIRV, NODEID: ENDVM5, TEL: 855-6022 FAX: 857-1126