Lil4X
Elio Addict
Probably the worst thing you can possibly do when developing a new product is to start with that proverbial "clean sheet of paper" design. It all sounds very nice, to be able to begin again at the beginning and engineer a project from the ground up. I spent quite a few years in product development, both inside a large corporation, then as a private contractor, and some of the things I learned that first year on the job still ring true.
When you start with a clean sheet of paper, you have no data points. If you build a machine with more than a dozen or so interdependent moving parts, when (not IF) you have trouble, you're in for a long siege trying to figure out just which component isn't performing as expected. I worked on some pretty large electro-hydraulic tools that had enough power to literally tear themselves to pieces. Not a good idea if you're trying to make your tool do something other than self-destruct. In the oilfield, for the past 35 years, there's been a dream of developing an automated machine for drilling for oil. One man with a couple of levers - like digging a hole for a backyard pool - well, except the deep end is going to be about five miles down. On paper it's a pretty simple concept - a CNC drill. Ah, but it's not so simple. It's been tried and failed so many times, it's become a legend - and in a long career I got to witness several "totally new" attempts. Everybody wants to chuck all of the good stuff that's been well proven over years of service in favor of a "clean sheet" design. It just makes me want to go hide.
There are simply too many components performing a wide variety of operations to get your head around all at once. Whether it's individual pride or whatever, it seems nobody wants to use anything that's proven - "we can build it better" is the first sign of trouble. If you start with a known, you can proceed into the unknown, developing one small sub-system at a time, then "pivoting" (to use the current Washington vernacular) to the next sub-system.
The use of a proven Geo/Zuke block as a starting point is an excellent example. OK, it's not really the same - the tooling for that engine disappeared years ago, but we can reverse engineer that block as a starting point. If you're going to copy, copy from success. Then you start to tweak. Pistons, bearings, head, valves, manifolds, injection systems, electrics. All have to be developed as sub-systems/components to work together in the finished product. One team makes a change, others may have to rework their designs to accommodate the change . . . but it's not like starting from scratch. Let's build on the success of the past with the tools of today. We can use computer systems to analyze strength of material, heat transfer, flame propagation, vibration, and dozens of other components - even develop the on-board computer that will operate the engine - all in a virtual world. The finished Elio/IMG engine may share a few dimensions with the Geo, but it's going to be a far distant cousin, if any real relation at all.
When you start with a clean sheet of paper, you have no data points. If you build a machine with more than a dozen or so interdependent moving parts, when (not IF) you have trouble, you're in for a long siege trying to figure out just which component isn't performing as expected. I worked on some pretty large electro-hydraulic tools that had enough power to literally tear themselves to pieces. Not a good idea if you're trying to make your tool do something other than self-destruct. In the oilfield, for the past 35 years, there's been a dream of developing an automated machine for drilling for oil. One man with a couple of levers - like digging a hole for a backyard pool - well, except the deep end is going to be about five miles down. On paper it's a pretty simple concept - a CNC drill. Ah, but it's not so simple. It's been tried and failed so many times, it's become a legend - and in a long career I got to witness several "totally new" attempts. Everybody wants to chuck all of the good stuff that's been well proven over years of service in favor of a "clean sheet" design. It just makes me want to go hide.
There are simply too many components performing a wide variety of operations to get your head around all at once. Whether it's individual pride or whatever, it seems nobody wants to use anything that's proven - "we can build it better" is the first sign of trouble. If you start with a known, you can proceed into the unknown, developing one small sub-system at a time, then "pivoting" (to use the current Washington vernacular) to the next sub-system.
The use of a proven Geo/Zuke block as a starting point is an excellent example. OK, it's not really the same - the tooling for that engine disappeared years ago, but we can reverse engineer that block as a starting point. If you're going to copy, copy from success. Then you start to tweak. Pistons, bearings, head, valves, manifolds, injection systems, electrics. All have to be developed as sub-systems/components to work together in the finished product. One team makes a change, others may have to rework their designs to accommodate the change . . . but it's not like starting from scratch. Let's build on the success of the past with the tools of today. We can use computer systems to analyze strength of material, heat transfer, flame propagation, vibration, and dozens of other components - even develop the on-board computer that will operate the engine - all in a virtual world. The finished Elio/IMG engine may share a few dimensions with the Geo, but it's going to be a far distant cousin, if any real relation at all.
.......very nice!
