Is it logical for a PhD in a technical field to take a post-doc if they later decide to go into industry and not pursue a faculty position?

It really depends. Post-docs are mainly geared towards people pursuing careers in academia. They help you build a research base and portfolio of publications, things that most companies care little about. I will admit, however, that I did a post-doc after my PhD and found it was a spring-board to better opportunities. I did a PhD in Mechanical Engineering at Washington University in St. Louis, which, while it's an excellent institution, doesn't have very much name recognition in the engineering industry. I had a few job offers, none of which were very good and ultimately decided to do a post-doc at the University of Wisconsin. After just 1 yr. at UW, I had 6 different job offers for around 25%-50% more than what I got just the year before. Mostly it was due to the network and name recognition of the school. I'd say my case was probably the exception and not the rule though. It basically delayed the start of my career by another year. If it's at all possible, try and avoid it and go straight into industry.

I recommend being very careful and thoughtful as to why you want to do a post-doc.  If you are interested in working in industry, it is best to gain industry experience.  Often there is little reason to delay going into industry and post-doc jobs often only give you incrementally more experience than what you gained in your Ph.D.  Potentially, if you conduct a post-doc in a PI's lab that has industry ties or your research is very relevant to industry, then it may lead to an opportunity.  However this will be 2-6 years down the road. Unfortunately, many Ph.D.'s choose to do Post-docs for the wrong reason - because it is the path of least resistance following your Ph.D.  If you want to work in industry, I recommend trying very hard to get an industry job.   To make this easier, try to gain industry experience during your Ph.D.  This can be done in multiple ways. For instance, you can align your work with industry, possibly working on industry sponsored research in a industry focused lab.  Another effective route is to spend some of your time 'volunteering' your skills at a company or firm of interest to you. This may be in the evenings, weekends, or even during a small part of one of your work days (given that you likely get paid ~20-30k during your Ph.D., your supervisor shouldn't complain too loudly for this last option).  All in all, the shortest route from A to B is a straight line.

No, based that I was a Visiting Fellow at Harvard while working in industry. A Visiting Fellow allows you to do research at a University in collaboration with a professor. It is only necessary when your doctorate is not strong enough to enter a career that requires research. In my case, it was an opportunity to do research with a specific person outside the confines of producing a product.