The early stages of computerized engineering, in the early 1960's to the mid 1980's there was a distinction between designers and stress analysts. Designers designed structures and products and passed those designs to the analysts by way of a set of Engineering drawings, and later electronically with design files. Designers were trained and gained experience in design and using design tools whereas analysts were trained and gained experience using analysis tools.
Analysts were knowledgeable in the limitations and restrictions of element formulations, modeling of boundary conditions, differences in solution techniques, and deformation and stress result interpretations. They were familiar with the need to, and the procedures to, verify analyses. Even with trained and experienced analysts, many analyses were not as optimized as they could have been, but were typically conservative.
Starting in the mid 1980's desktop computer systems and the advance of analysis software began offering Engineering organizations the ability to combine design and analysis capabilities allowing the designer to analyze their designs therefore assuming the role of both the designer and the analyst. While this may have reduced the cost of the designs (by eliminating analysts and using bargain basement analysis software), it shifted the burden of verifying a design to individuals (designers) who are not normally qualified to use analysis software to verify their designs
Companies allowing designers to analyze structures or products, need to be aware of the following:
Liability for a structural or product design lies with the company performing the design and not with the application software used to verify the design.
The designer/analyst needs to assume that all engineering application programs contain errors. He should be confident that the application he is using has been as extensively tested as possible and meets some industrial standard such as NRC software testing standards.
The designer/analyst needs to gain an understanding of the limitations and restrictions of using all parts of the application program. This includes finite element formulation limitations and restrictions and solution procedure limitations and restrictions.
The designer/analyst needs to be able to accurately simulate boundary conditions placed on the finite element model of the structure or product he is analyzing. Accounting for all relevant boundary conditions may require significant time and effort.
The designer/analyst needs to be able to accurately simulate the relevant material properties of the structure or product, in the environment the entity is subjected to, and simulate how they may change under load.
The designer/analyst needs to gage the accuracy of the analysis particularly in regions of high stress (strain) gradients. Changing the mesh density or the order of the elements in those regions may be necessary for an accurate solution.
The designer/analyst needs to know where stress (strain) is computed for any element formulation and how that stress (strain) is distributed across the entity under design. This is necessary to properly interpret the stress (strain) contours that develop in the entity under load.
The designer/analyst needs the resources to independently verify the results of analyses. This requires knowledge, experience, and time to simplify modeling of structures and products and boundary conditions while maintaining a "ball-park" accuracy in the solution process.
The designer/analyst needs to maintain a healthy pessimism about the results of any analysis.
Using a sophisticated stress analysis program requires caution in its use and interpreting the results. In using these types of programs the old adage "Garbage In Garbage Out" is more relevant today than ever before. The user can not expect to use a stress analysis program as a "black box" to obtain accurate solutions. Adequate resources should be allocated for the user to gain an understanding of the strengths and weaknesses of the application program and to gain experience in accurately modeling structures and products and interpreting deformation and stress results.