Saturday, June 11, 2005

Logistics

This article appeared in several places in Jan/Feb of 2003. One of them was Winds of Change. A comment of mine from the WoC site has been included at the end here.

==============

Why can't we have the fuel efficient cars we see and hear about in magazines and on television filling the auto company's show rooms in the next model year? Why don't we already have them this year? There's a reason, a one word reason. That word is logistics.

I'd like to discuss here the difference between a prototype built by a school or an auto company and a production auto that you can buy off the show room floor.

I'm going to start out with the very simplest of the new technologies, the Integrated Starter Alternator (ISA). This is a starter motor that's also the alternator (electrical generator) of the car. If this device was made part of the engine we would get a number of valuable improvements:

First, it would be a more efficient electrical generator than the current separate alternator for two reasons. One is that losses from the rubber belts needed to transmit force from the engine to the alternator would be eliminated. Second is that because the ISA would have a larger diameter, it's magnetic structure could be much more efficient than the structure of current belt-driven alternators.

There's a second advantage to a more efficient magnetic structure. In the starter mode the starter motor becomes more powerful and more efficient. Coupled with a higher battery voltage (36 volts nominal, about 42 volts while the engine is running) an engine on demand system becomes viable. That means that when the car is stopped at a stop light the engine can be turned off to save fuel.

A third advantage of an ISA system with a larger battery is that instead of engine braking where the engine absorbs some of the energy needed to slow an auto, the generator/battery system can absorb some of that energy. Better yet, it can return it to the motor on the next start up cycle. In effect, the energy needed to start the engine in stop and go driving is energy that would be otherwise wasted in heating the brakes.

This is a lot of payback from what seems like a simple design change. It has already seen prototypes on the road. So what prevents the car companies from going from a proven design to a million vehicles? Well, we are back to that word: Logistics.

A good place to start is batteries. Twelve volt batteries are easy to find. Twelve volt lamps are manufactured by the billions. Twelve volt accessories like radios and heater fans are commonplace. Where do you get 36/42 volt equipment? Right now, you don't get it anywhere. Now the engineers need a 42V-12V converter and an auxiliary 12 volt battery. Thirty six volt batteries? Fuggedaboutit. In the initial production models they will likely be made from three twelve volt batteries strung together.

The battery design also will have to be different from the current design. A battery called on to make 5-10 starts a day is very different from one that can reliably deliver 500-1,000 starts a day. So the car companies must go to the battery companies with a specification, then ask them to design a battery that will fit in a currently produced 12V case and still do the job. The battery company then asks their battery engineers to take a whack. They come back with a design after a few months of effort. Then the prototype shop comes back with a few copies after a few months more of effort. More back and forth with the engineers.

Now the real fun begins. The battery must be cycle tested. Charged and discharged 1,000 times a day to simulate operation in the new vehicle design. There are problems. There are always problems. So the engineers come up with a new set of compromises (otherwise referred to as the revised design), and the testing begins again. The months tick by. Finally, the design works acceptably at the normal temperature range.

As well all know, however, cars do not operate in a normal temperature range. They must work reliably from a -40 degree arctic environment to a 120+ degrees summer desert environment. More testing. More redesign. More compromises. After 18 to 30 months the battery company finally has a pre-production prototype lot of a few hundred batteries to deliver to the car companies.

Now the serious fun begins. All the suppliers to the car company have been going through a similar drill to get their parts ready for the preproduction prototypes. The parts are all there, and they are then hand assembled into a few hundred pre-production prototype vehicles. Any problems in assembly are noted for further revisions in the delivered parts. Time to live test. The vehicles are first test driven on the auto company's proving ground tracks to shake out the bugs in the hardware and software.

Did I mention software? Most of the function of a vehicle with an ISA system are controlled by software. So now not only do the parts have to work as specified but they also have to work in the way the software commands. If not, either the parts or the software must be modified.

Finally everything passes the track test. The pre-production prototypes can now be parceled out to the auto company executives and their families for a year of test driving in all weather conditions. A few more are parcelled out to writers in the field to test drive and give their opinion. During this year, known modifications are being made and possible changes are being anticipated from the early driving reports. If all goes well, 3-6 months after the year of test driving is done the new component designs are ready for production.

Home free at last? Not by a long shot.

Now the factory designers must come in and design a factory to produce the newly designed and proven components. Orders must be placed for the special tooling required. Orders for plastic cases and plastic grid separators goes out. Orders for lead of a certain chemical composition and thickness. Orders for lead oxide paste.Orders for tank cars of sulfuric acid. Even orders for new software to track the manufacturing process. Orders for machines. Orders for punches and dies. Orders for bins to recycle the scrap produced by the production process. Orders. Orders. Orders.

Now of course, you would like to compress this schedule where possible. It costs money to keep all these testing facilities and engineers and technicians and designers on a project. So it would be good if the design of the factory could be started while final testing was going on. Yet this entails real risk. What if testing reveals that the final design needs to be radically altered? Suppose a light went on in some really bright engineer's head and he found that with a different factory design he could shave ten million dollars off the production of a million batteries for 350,000 cars and reduce the scrap produced by one half million pounds, or because of a serious design flaw it was found that all the batteries built according to the original factory design would be scrap after six months on the road. To accomplish such a change, the battery company would need to adjust it's factory design, take a three month schedule slip, and order two million dollars of new tooling. Ouch. Maybe they'd rather play it safe istead, and start the design of the factory a bit later.

What I've presented here is just one component of a radically new car design, most of whose engineering parameters are reasonably well known from over 100 years of manufacturing experience. I haven't even begun to cover the power electronics required to make all this a going proposition, and we don't have a hundred years or more of experience with multi-kilowatt power electronics on mass produced vehicles. Instead., we have is roughly zero years of experience. This is not a trivial problem.

Especially if these problems lead to designs whose flaws are not immediately obvious. What happens when they have a problem that shows up in one out of every 300 vehicles after two years on the road, and they have made a million of them due to strong demand? Disaster. They now have 3,000 odd dissatisfied customers, and a million vehicles to recall.

It's not easy for an auto company to put a brand new design into the field. The risks are huge. Which is why the changes will not come all at once except in research and development vehicles. So when your favorite greenie asks why we can't have in the show room tomorrow the vehicle he read about in a magazine yesterday you can give him or her the one word answer:

Logistics.

P.S. The good news is that such American made vehicles will be available in the 2004 model year. Happy driving.

***********

The problem with earlier attempts at ISA was the slow speeds of the processors required to control the unit and the low efficiency, high cost, weight, and volume of previous designs.

All these things have come together in the last three or four years. With continued improvements in sight.

As time goes on all auxiliarys (air conditioners, power brakes, power steering etc.) will be electrically powered to eliminate the continuous power rob of belts and pulleys).

In addition valves will be electrically powered. Engine tuning will be changed by changing a chip not grinding a new cam.

The ISA is just a beginning.

No comments: