Printer-friendly versionPDF version
Stephen Ferguson

Having successfully sold coolant- jacket simulation work to Ford in the US, Steve MacDonald began to spread his net wider, and struck up a business relationship with Ford’s European Powertrain Division, based in the UK.

One of the engineers at Ford UK was buying some university-type cooperation from this university professor, David Gosman. And of course, it turned out that David Gosman had a CFD code, well a prototype CFD code. Initially when we got into CFD, we were using FLUENT but when we met David, it seemed to me like, well, geez, maybe we ought to cooperate with this guy."

David Gosman remembers: "I’d been asked to go to Ford in the UK to listen to a presentation by a guy called Steve MacDonald on some work his company had on engine cooling using a commercial CFD code, which in those days was based on the castellated grid methodology. I told him that, because he couldn’t properly capture the geometry of the water jacket, the results he was presenting were probably wrong. In retrospect, knowing him as I now do, Steve was remarkably restrained. He listened to what I said, and went away."

In fact, ever the perfectionist, Steve MacDonald was already aware of the problem in employing this "stair-step" representation of the boundaries of his domain, which had become a source of much frustration with his CFD code vendor. His problem was that, until he met David Gosman, he didn’t have much choice in the matter as none of the commercial CFD codes available at the time supported the type of "body-fit ted" meshing methodology that would allow his engineers to accurately capture the actual geometry of the coolant jackets that he was trying to model.

Because of this, adapco quickly became the fir st company to apply David Gosman’s, as yet unnamed, body-fit ted CFD code to industrial type problems.

"Gosman had a pretty good name, and so we leased his code, which was interesting. It had no front end. It had no back end.
It had no manual," remembers Steve MacDonald. "It was just this computer deck. It was basically an early prototype of STAR-CD®."

"I still have the sheet of paper where I was trying to think up a name for our new code," recalls David Gosman. "And in those days, we were thinking in terms of acronyms, something which actually had some CFD context to it, without being too blatantly CFD. And one of the things I wrote down was Simulating Transport in Arbitrary Regions, S-T-A-R! Well, there were other products around called 'STAR', so I had to add something to it to make it unique and that’s where the "CD" got put on the end. So that’s how STAR-CD came into being as a name."

"So we initially had the trials and tribulations of using a prototype CFD code that you had to do a lot of fixing to get to work," says Steve MacDonald.
"And I kept going to Europe to get jobs from other people. I got jobs from Ford in Germany. I got jobs from Daimler- Benz. So I was flying backwards and forwards between New York and Europe, and every time I visited, I would go and see David Gosman."

"And one day, I was riding in the car with David and I said, ‘You know, if you ever think about going into business, adapco would be interested in going into business with you, and we could provide some seed money,' which was interesting because we really didn’t have very much money in those days.

Steve and DavidWe were starving to death. But David said that he and his partner, Raad Issa, were going to go into business and they had been talking to a venture capitalist, and that was very far down the road to forming the company and they couldn’t back out of that. So I said, 'Oh, okay. Well, if anything happens, let me know.'"

"And lo and behold something happened..."

In fact, two very big "somethings" happened. Just as the early history of adapco was influenced by the Three Mile Island Accident, the birth of CD-adapcoTM probably only came about because of not one, but two further "disasters".

The first of these occurred on January 28, 1986, when the Space Shuttle Challenger exploded 73 seconds after launch, tragically killing all seven astronauts on-board, and plunging NASA and the American Space Industry into crisis.

adapco co-founder Bill Wheeler remembers the state of panic at the time: "After the Challenger disaster, NASA began frantically scratching their heads saying: 'What are we going to do about this? How are we going to figure out what caused this?', They kind of knew what caused it, but 'How are we going to know, when we redesign it, that it’s not going to happen again?'."

To answer those questions, NASA turned to adapco, who were rapidly becoming acknowledged as leaders in the simulation of structural engineering problems.

"They called us up, Steve went out there. And we got a job over Christmas time in 1986 to do one of the connections, not the connection that actually blew up, but one that they were actually more concerned about," remembers Bill Wheeler. "It was kind of a demonstration problem. And we had to have this thing done right after the New Year."

VAX 750: adapco's first computer purchase"It was hard work, but we solved it, and they liked the results so much that they gave us a fairly big contract. I think we did over two million dollars’ worth of work for them. In a little over a year, we looked at all of the joints in the rocket motor."
The net result of this is that adapco were suddenly no longer "starving to death". In fact, for the fi rst time in their history, they found themselves with positive cash flow.

The second "something" happened on October 19, 1987, a day which would subsequently be known forever more as "Black Monday", when the Dow Jones stock market lost 12% of its value in a two day period, sending the world’s financial markets in crisis, and David Gosman’s backers scurrying for cover.

"I had decided to form a company and started going out to look for finance and I was introduced to some venture capitalists," recalls David Gosman. "I don’t think they were called that in those days, but that’s effectively what they were. So I made presentations to these finance guys - who understood practically nothing of what I’d said - but who eventually decided that they would provide our start-up equity. And we agreed this on the shake of a hand, because in those days, a gentleman’s word was considered enough. That was in early ’87."

"When the stock market takes a nosedive, venture capitalists disappear," says Steve MacDonald. "So all of the sudden, David’s backers disappeared and Steve MacDonald was standing there talking about all the money that adapco had and what they could do for him, so we joined up with them and formed Computational Dynamics. We put our money into that company, although we were lucky that we didn’t have to put it in one lump sum."

Initially, the partnership operated as two separate companies "Computational Dynamics" and "adapco". While CD concentrated on developing the CFD solver, adapco were busy applying the software to real world engineering problems, and actually trying to sell the software.

"So as I was selling services work, I also began selling software, all over Europe and across the United States," says Steve MacDonald.

"I think in the early days because we were quite different organizations," recalls David Gosman, "CD was a CFD geeky-type organization. We were good at CFD, we were good at writing the basic methodology, codes. We didn’t really know very much about pre- and post-processing and one of the huge benefits we got from joining up with adapco, for which the realization came relatively soon in our joint lives, was that adapco, although they weren’t at that stage writing much in the way of software, had a huge amount of experience in using commercial codes and were therefore aware of what real commercial-level pre- and post-processing should be.

So adapco then took on the job of developing the pre- and post-processing side of our CFD package."

"We began to develop pro-STAR way back then, I think in about 1988, 1989, and we worked on it, and worked on it, and worked on it, and we used it so we could build anything. Nobody else could but we could," remembers Steve MacDonald. 

The first release of STAR-CD was a block-structured, body-fitted code. By its second release in 1991, STAR-CD had become the first truly unstructured commercial CFD code, offering engineers the ability to construct meshes from any combination of hexahedral, tetrahedral and prismatic cells.After selling the first license to Ford, STAR-CD quickly achieved the position as the CFD solution of choice for the automotive industry, also finding application wherever there were difficult problems to solve elsewhere.

"It became increasingly apparent that we really needed to have a more integrated operation," remembers David Gosman. "This compartmentalization meant that we weren’t benefiting enough from the expertise in our respective companies, and this went right across the board from development, sales and marketing to engineering services," says David Gosman.

"Engineering services were something which was clearly benefiting us through the experience that our services people were getting, but we weren’t getting the feedback from that into CD that we would’ve if we’d been working more closely together.

So I think we just gradually realized that we needed to do that. There was a time when a concrete decision was made to become CD-adapco. And I can’t remember the exact circumstances of that, but I do remember that decision being made and from then on, we made a really concerted effort to merge our operations together and I would say that, it was a big, big step forward in our joint history."