Your editorial on odd specifications (“Silliness at 40 Below,” Nov. 20) struck a chord with me. Working on five continents, I’ve seen numerous instances of “silly specifications” in many cultures and languages and I’m always amazed at how, once written, specs seem to acquire biblical respect and authority. When people were struggling with “the Cadillac ashtray” problem, no one seems to have had the wit to take a view from 10,000 feet and ask: “Why are we doing this?” Or perhaps better, “Why are we doing it this way?”
One of the silliest specs I had to deal with involved a fighter plane. The contract mandated that a display needle on an engine instrument could only deviate or wiggle so much. The instrument or display met this specification by a wide margin with one exception: Needle excursions slightly exceeded specified limits when the onboard gun was fired. The fact that the pilot was highly unlikely to be checking his engine instruments while firing weapons was not considered. The spec was the spec. So after a long battle, the company and the Air Force shook hands and mutually agreed to end deliveries of the instrument. The 500 units already in the field were never heard from again.
Richard J. Reilly
I worked for the Tandy Corp. in the early 80s as a quality-control technician. I was told to place computers in the –40°F environmental unit for 4 hours to simulate a 4-hour trucking shipment in the mountains. I then tested the computers after they warmed up. I was also instructed to place software/hardware in the 120/140°F environmental unit for 4 hours to simulate a 4-hour truck ride through the desert.
This kind of testing always made sense to me.
Can we scan the scanners?
Since we are now at the point of not only printing, but also scanning an object in 3D (“Objects Printed from Objects Scanned,” Elisabeth Eitel’s blog, March 11), it begs the question: If you scan and print someone’s patented or copyrighted item, are you required to pay them a fee? It seems that we’re constantly hearing about intellectual property being stolen and cheaply reproduced abroad. I wonder how these companies would feel if somebody scanned their design using their own scanner and then undersold them.
You’ve run through a lot of the available options for freelance work (“How to Become a Freelance Engineer,” Lindsey Frick’s blog, March 7). My hopes for the industry (or the CAD crowd) is that it moves towards crowdtesting rather than crowdsourcing. That could include a small analysis of a buyer’s project before exclusively hiring a freelancer for a full project. Ideally, that would put the emphasis on skilled designers instead of commoditized prices.
Do those figures add up?
In a recent story on a new type of recreational water vehicle (“Amphibian Jet Ski hits 45 mph on Land and Sea, Nov. 20), it says: “Engineers at Gibbs spent more than 18 months and 75,000 engineering man-hours adapting the engine to the Quadski.”
If we take 2,000 hours as a man-year, than 18 months would be 3,000 hr. Dividing that into 75,000 means that 25 engineers worked full time on this one task (adapting the engine) for a year and a half.
That seems like a lot of engineers to work on one task at a small company. If we assume, for argument’s sake, that a man-hour costs the company $100 (which is probably low), then this one task would have cost the company $7.5 million. That seems high for a product that sells for $40k and has a limited market. If the profit per Quadski is 20% ($8k), they would have to sell almost a thousand units just to cover this one cost. Is it possible that the 75,000 figure should really be 7,500?
The estimate of 18 months and 75,000 engineering man hours is actually a conservative one.
The Quadski’s land-and-water powertrain system is much more complex than the average automobile’s powertrain.
The engine and transmission finally selected was significantly different in size from the engine originally under consideration. It had a different output shaft location and drivetrain, along with completely different torque and horsepower curves than the engine originally under consideration. This required a reengineering of the cooling system, the body to accommodate the new engine, engine mounts, and fuel systems, the differential/driveshafts and knuckles, the water jet, and electronics to interface with the engine control module.
The new powerplant also required: EPA engine certification, testing and validation, and a major change to the assembly processes.
The engineering support from purchasing and quality departments is not included in the 75,000-hours estimate but this was needed as well.
There are several additional things you might not have been aware of:
- The program wasn’t handled totally in-house. A number of engineering projects were outsourced to suppliers such as FEV and MBE.
- A high-speed amphibian is not a “simple” product like an automobile. It presents new engineering challenges. An amphibian-powertrain change is certainly not the simplistic “Lego” plug-and-play motor the reader seems to believe exists for major OEMs. It involves developing a dual powertrain for on land and water. The Quadski powertrain is protected by several patents.
- If you are an engineer and believe you could have completed this engine change task in less time and with fewer resources — we want to interview you ASAP!
— Larry Weis, Gibbs Sports Amphibian
Inertia mix up
I believe there was an error in a recent item (“Honda Locks Steel to Aluminum with 3D Design,” March 21). Assuming both doors (or other side panels) are lightened, removing weight on the outer sides of the vehicle body does not move the car’s center of gravity toward the middle of the vehicle. It may well improve handling but that would be because the moment of inertia was reduced.
Think about balancing a bar bell on a fulcrum and then sliding the weights inward equally; the center of gravity doesn’t move.
Andrew J. Brislen
In the article, “The future of additive manufacturing,” March 7, it says: “Currently, few options exist for carbon-fiber parts made with additive manufacturing.” However, in a sense, all composite parts are made by additive manufacturing as they are built up one ply at a time. Automated fiber placement (AFP) automatically places strips of composite materials to build composite structures such as the Boeing 787. In-situ AFP of thermoplastic composites is a truly additive-manufacturing process that builds high-performance composite structures by bonding composite prepreg tapes in a manner similar to FDM. It has been in existence longer than what is now known as additive manufacturing.