The what and when of upfront CFD training

I know a lot about rolling out upfront CFD for maximum fun and profit. Over the last several months of applying this knowledge with existing customers, a new issue has surfaced. Among the many ways to fail miserably with upfront CFD, I’ve identified another success point that managers should consider:

The what and when of training.

In my experience, companies are either wildly successful with upfront CAE… or the software gathers dust on a shelf. There isn’t much in between. For many companies, the path to apathy starts with their initial training strategy.

If you are lucky, you will have a software stud on the team who just “gets” CAD and CAE. He or she will have success with any tool you drop in their lap. More likely, however, you’ll want to train a larger group of people who generally climb steeper learning curves. If that’s you, then pay attention:

You should seriously consider organizing a customized training class for your team. The “what” should be some of your real models that reflect the kinds of work your team will be doing with upfront CFD after training. If you make valves and half the training course is devoted to electronics cooling, the lessons are unlikely to stick with your people.

Ask your software vendor to prepare a customized course around your real models. Your people will stay more engaged throughout, will learn more about dealing with the initial geometry setup, and will retain their new skills for much longer.

Similarly, you need to pay attention to “when” this training takes place. Don’t schedule the world’s best, customized training class 3-6 months before any of your engineers will have a chance to put their skills to use on a real project. That nearly guarantees failure.

Pick a real live, scheduled project by name.

“Ok, we’re going to schedule our customized training to support the XP12 project. The upfront design work needs to be completed by June 30, so we are scheduling this training for June 10.

I expect to see some preliminary results on at least 1 design iteration at the conclusion of the class. Then, the team will have 2 weeks to complete several more design simulations and prepare a final report.”

The non-software-jockeys on your team will have far more success if you prepare them with highly relevant training in support of a real project. It is essential to give these folks a taste of the candy on their first stab at upfront CFD. Do that, and they are far more likely to willingly come back for more!