Book Review: Optimisation of Design of Composite Structures
(Optimisation of Design of Composite Structures is a combination book and software package. In this week's article I review the book. In two weeks, I will take a look at the software.)
If you have worked with composites for any amount of time, you know that optimization is a difficult problem. Not only must you consider many variables (material, ply angles, ply sequence, number of plies, etc.), but you must also restrict many of these variables to discrete values.
Many papers have been written on the optimization problem, but few if any books have been published. I have recently been working on some optimization problems, and progress has been difficult. It was with great anticipation, therefore, that I dove into Optimisation of Design of Composite Structures.
I am sorry to say, though, that the book doesn't live up to its promise. To be blunt, it looks like a PhD dissertation with the important details left out. Even the binding makes it look like a thesis: the book is distributed loose-leaf with a four-ring binder.
As with most composite references, this one starts with the requisite review of lamination theory and failure criteria. The two pages on finite element theory are too brief, even for a review. The most promising section is the description of the optimization method. Unfortunately, the later chapters never show the details of how the method is applied.
And I guess that's my biggest complaint about this book. The hardest part of optimizing composites is selecting the algorithm and applying it to composites. The only times the details of an optimization are shown, though, is when the brute force approach is taken (i.e., run through all cases and pick out the highest or lowest point on the cruve).
In one case, the author states that all optimization techniques which were available failed. Even listing those techniques would be of use, so others would know to avoid them.
For example, I recently spent a few weeks working on a simulated annealing routine for optimizing a particular problem. The algorithm was easy to understand, but applying it to composites took a bit of work. I then spent a week or so tweaking the algorithm parameters, but I never was able to get the program to converge. It is this type of information that Optimisation of Design is missing.
All that being said, the book is not entirely without merits. The discussions of results reveal some general trends, such as stacking sequences to minimize interlaminar shear failures and the best ways to manage ply drop-offs in variable thickness composites.
In fact, the analysis of variable thickness composites is one of the most interesting portions of the book. Results are shown for stress distributions near ply drop-offs, and these are compared to experimental results. This data is then used to optimize plates and beams of variable thickness. Once again, though, there is a frustrating lack of detail.
Despite the book's positive points, it is difficult to justify the US$495 price. Next week, I will take a look at the software which accompanies Optimisation of Design. If it can solve the problems discussed in the book, it will be a useful utility. It is unfortunate, though, that the book does not provide enough details to apply the techniques to other problems.
Details: Optimisation of Design of Composite Structures, by
Antonio Miravete, published by Woodhead
Publishing, 1996, ISBN 1-85573-208-4.
1. Introduction and
preliminaries; 2. Design optimisation of constant thickness composite
structures; 3. Shells; 4. Constant thickness plates; 5. Constant
thickness sandwiches; 6. Design optimisation of variable thickness
composite structures; 7. Variable thickness beams; 8. Variable
thickness plates; 9. Variable thickness sandwiches