How did you first become involved with engineering simulation in your career?
After completing a degree in Aeronautical Engineering at the University of La Plata (Argentina) I started to work in the National Institute of Industrial Technology in a group performing stress analysis and design. The year was 1976 and shortly after I joined the group the institute purchased its first finite element code running on a mainframe computer that took up an entire room. I started combining traditional hand calculations with FEA at that time. In the U.S. I completed graduate studies at Washington University in the field of numerical simulation in solid mechanics. I joined ESRD after graduation to develop FEA software as well as participate in R&D projects and consulting work related to simulation.
Why is ESRD participating in the Revolution in Simulation?
ESRD is participating and sponsoring the Revolution in Simulation initiative because we strongly believe that the vision of expanding the use of simulation by non-experts can be safely realized only through the practice of Simulation Governance.
The concept of Simulation Governance (SimGov), first introduced by ESRD in 2011, is a managerial function to provide command and control over all aspects of numerical simulation. In the case of democratization of simulation, SimGov provides the safeguards to ensure that routine analysis in support of design decisions can be performed reliably by engineers without expert training in FEA through the use of Engineering Simulation Apps that meet the technical requirements of SimGov.
Dr. Barna Szabo, one of the founders of ESRD, and myself have volunteered to moderate the SimGov page on the Revolution in Simulation website to increase the awareness and importance of Simulation Governance in Democratization. The development and implementation of a SimGov plan to assure the integrity of engineering analysis requires the attention of managers and education of users. Rev-Sim.org provides a very useful resource center to help industrial users of simulation software get started.
Why is expanding the use of simulation technologies through the democratization of simulation important to your customers?
Proper application of numerical simulation procedures requires expertise in computational engineering that is not widely available. Standardization deployed by means of Simulation Apps (Sim Apps) in the Aerospace & Defense industry leverage this expertise for recurring analysis tasks and process workflows similar to the expertise of specialists in applied mechanics made available through classical engineering handbooks.
To ensure the level of reliability expected in professional use, FEA-based engineering sim apps must incorporate solution verification procedures. This is an essential technical requirement of Simulation Governance. It is important to emphasize that solution verification is possible only when the model definition and the numerical approximation are separated. This rules out the traditional practice of finite element “modeling”, with all its tricks and traps, in the creation of simulation apps. Engineering simulation apps must provide users with objective measures of quality for all quantities of interest.
One argument often used to justify the lack of a-posteriori error estimation is that a-priori rules established by an expert are sufficient, and that the “user can always consult with an expert if the results are wrong”. The problem with this argument is that if the assessment of the quality of a solution depends on the subjective opinion of the expert analyst, the same expert opinion is needed to determine when the solution is wrong. A non-expert user is not qualified to make that determination. Having an automatic solution verification feedback for all the quantities of interest however, tells non-expert users to consult with experts when the computed data does not converge to within a prescribed tolerance.
Engineering simulation apps must not be deployed without objective measures of quality for all quantities of interest. To do otherwise would not be democratization of simulation but a redistribution of risk.
What do you see as the most important benefits of democratizing simulation?
The standardization, automation, and democratization of numerical simulation through Sim Apps subject to the discipline of Simulation Governance offers many benefits to industry at the engineering, product, and business levels. These benefits include: encapsulating complexity, improving productivity, containing cost, and ensuring reliability for the expert simulation analyst and non-expert design engineer alike.
To ensure their reliability, the design, development and deployment of engineering simulation apps must include built-in safeguards to prevent use outside of the range of parameters for which they were designed and must incorporate automatic quality assurance procedures. To meet the technical requirements of Simulation Governance, Simulation Apps must incorporate the proper formulation of idealizations, the adoption of the best available simulation technologies and practices, and solution and data verification procedures.
Can you share with us an example of a benefit?
For the past two decades, ESRD has worked alongside customers from across the A&D industry to create applications designed to streamline engineering analysis and numerical simulation process workflows. The specific nature of many of these applications cannot be disclosed for reasons of confidentiality or national security. However, to see more generic examples of our work visit the applications pages on the ESRD website .
What are some of the challenges that you have also witnessed?
In our view the greatest challenge is the tendency to underestimate the time and effort it takes to develop an application that meets the technical requirements of numerical simulation. A software product that is simple and easy to use is not easy to create and cannot be produced cheaply. Managers should understand that democratization involves substantial up-front costs but also delivers substantial returns on investment. In addition to increased productivity through the use of smart applications, democratization serves to capture and preserve institutional knowledge.
A numerical simulation product created by expert analysts for safe use by engineers must have a clearly defined scope, well-documented methodology, has to be implemented, tested and certified by professionals. It must provide a posteriori error estimation in terms of the quantities of interest and warnings if the error in a quantity of interest is greater than what the user should accept. This is possible only when the mathematical model and the discretization process are treated separately.
What can we expect to see from ESRD in the future to help industry overcome these hurdles?
ESRD will continue to develop applications that meet the technical requirements of simulation governance, train analysts in the art and science of democratization, and will challenge false claims concerning democratization. For example, claims that creation of a user-friendly GUI with tight control of the input parameters passes for democratization.
Are there any other topics you think industrial users of simulation should understand better?
Yes. You know the concept of democratization of expert technologies is quite old, even when we did not call it that! Engineering handbooks and design manuals are examples of democratization practiced in the pre-computer age. Experts, such as S. Timoshenko, solved a variety of problems in mechanics by classical methods in parametric form. These solutions were collected into handbooks and made available to ‘non-experts’. That is, experts in fields other than the theory of elasticity.
This kind of democratization had a serious limitation: There were a number solutions but usually not for the problems engineers actually needed to solve. To get a rough order-of magnitude estimate of their quantities of interest, engineers had to find handbook entries that were close in some sense to their problem on hand.
With the maturing of numerical simulation technology it is now possible to remove the manifold limitations of classical engineering handbooks and provide parametric solutions for the problems engineers actually need to solve. This is the goal of democratization. There is a fundamentally important prerequisite: The exceptionally rare talents of engineer-scientists who populated conventional handbooks have to be democratized, that is, mapped into the world of modern-day analysis. The time has come for democratization to be reinvented.
Where can readers go to learn more?
There are several publications, presentations and other material available in our website. In particular the following pages expand on the topics covered in this interview:
Thank you Dr. Actis!