Cadillac Club Estonia

* KKK    * Otsi

* Logi sisse   * Registreeru

Tänane kuupäev 29 Mär 2024, 00:40

Kõik kellaajad on UTC + 2 tundi [ DST ]




Tee uus teema Vasta teemale  [ 1 postitus ] 
Autor Sõnum
PostitusPostitatud: 12 Veebr 2009, 17:11 
Eemal
ravi puudub
Kasutaja avatar

Liitunud: 01 Nov 2005, 21:11
Postitusi: 5492
Asukoht: Tallinn
Skype Kasutaja: eldorado67
Pilt

BY FRANK MARKUS
May 2003




Today, untold zillions of molecules of iron, aluminum, and plastic composites awoke at the flick of a human wrist on an ignition key. They were compressed or stretched by the forces of combustion, they transferred heat from cylinder walls to coolant, they transformed vibrations into noise, and they were lubricated against the forces of friction. But before these molecules were crafted into modern engines, virtual facsimiles of them were subjected to vivid, lifelike simulations of these actions, the whole shebang conjured by an artificial intelligence as sophisticated as anything conceived of in The Matrix.

A year ago, the luckiest of these virtual alloy molecules were grooving and grinding to the synthesized machinations of the first V-16 engine of the millennium. The mission to build an actual running engine for the Detroit-show-stealing Cadillac Sixteen concept was launched in mid-March 2002. At that point, Cadillac could have simply FedEx'd a couple of V-8 crate motors to a prototyping shop and subjected them to a welding-torch wedding. That process could have produced a V-16 that would start, idle, and move the concept car onto and off its auto-show runway.

Instead, GM decided to flex its advanced engine-engineering muscle and highlight technology that would soon be appearing on the Gen IV Corvette V-8 by designing and building a real V-16 from scratch—one that could operate under load at redline on a dynamometer. No weld-'em-up motor could ever withstand that.

The engine's refinement and output were supposed to get folks thinking again of Cadillac as the Standard of the World. It's hard to beat a smooth-firing sixteen for refinement, and it's tough to top 1000 horsepower and 1000 pound-feet of torque. But the engine also had to achieve the fuel economy of a competitive V-12 while running on regular-grade fuel.

This daunting mission fell to Larry Cooper and about 10 people from the Powertrain High Performance Vehicle Group. They started with the basic architecture of the next-generation Corvette's V-8, including its variable cam timing—a first among pushrod engines—and a Displacement on Demand system that shuts down cylinders to save fuel under low-load conditions.

Calculations of the cylinder pressures required to hit the output targets suggested the need for a big bump in per-cylinder displacement, so the bore and stroke were both increased by 6mm relative to the current 5.7-liter V-8. Small-block cylinder walls don't have 6mm to spare, so adjacent cylinders ended up touching each other, and coolant could no longer flow in between the cylinders. This complicates the cooling system, sometimes causes the cylinders to go out of round as the engine heats up and cools down, and presents dozens of other problems that don't spring instantly to the human mind at the outset of such a project.

This is where the team engaged the General's considerable computer-engineering might by sending the basic engine design off to be thoroughly modeled, developed, and tested in the digital realm. In this cyberworld, the V-16 would be cold-started, overrevved, bogged down, blown up, and otherwise abused to the point of repeated destruction—all without ever staining a floor.

To get a closer look at this world of artificial abuse, we plugged into GM's engineering matrix, which is presided over by the GM Powertrain Synthesis & Analysis group in Pontiac, Michigan. This crew of 117, most of whose walls are papered with advanced-degree sheepskins, maintains and develops some 300 computerized tools and test procedures used throughout GM. The least refined of these can merely weigh the relative merits of two design concepts; the best developed can guarantee the success of a computer-designed part without physical verification. Most lie somewhere in between. Fourteen of the most highly developed tools are proprietary programs developed in-house. Each tool strives to reduce the time, energy, and cost expended on physical testing and experimentation so that someday every part in a new car will be designed perfectly, so that no prototype (let alone production) part will ever fail.

According to Cooper, every component of the running XV16 is the first prototype piece built. A prime example is the crankshaft, which ended up resembling two V-8 cranks welded together but was designed from scratch using a host of computer tools, including a GM-customized version of LMS International's Powertrain Dynamic Simulator. This software tool helped assess the loads, positions, velocities, and accelerations of all the reciprocating parts in varying operating conditions so engineers could optimize the firing order and balance the engine's crank, rods, pistons, and other rotating parts. In this one case, the math program spit out a design, but the collective human engineering gut felt the counterweights were too small, so a second crank was fabricated. Both were tested, and the computer crank suffered less torsional vibration and weighed 20 pounds less.

The block and heads started out in the computer-aided design department, where electronic models of the V-8 block and heads were compu-welded into thoroughly lifelike V-16 parts. A team of computational fluid dynamics experts took the head design and flowed computerized air molecules through the intake ports, sprayed virtual 87-octane fuel at the intake valves, and then fired digital sparks to ignite cyberfires. The team made a small alteration in the shape of the Gen IV V-8's intake runner, hoping for better flow. In the simulation, the flow rate was unchanged, but the way the air moved inside the cylinder improved the burn rate, allowing a slight advance in spark timing for better power—a benefit that would have gone unnoticed back in the day of simple airflow testing.

The complex shapes of the block and head computer models were then divided into manageable "finite elements"—simple little geometric shapes, the stresses on which can be easily computed and combined to determine the stresses on the whole part. Then all the thermal and mechanical forces acting on the block and head, as computed by the various other teams, were fed into yet another program that simulated the reactionary forces to be expected from the engine mounts and the 4L85-E transmission. The virtual V-16 was then "started," rocked, shook, and twisted by all the forces it would experience in real-world driving, and computer-screen images of the engine were displayed with colors indicating areas of stress and fatigue. Problem spots were strengthened, overengineered areas were thinned, and the program was run again, until the resulting parts were as light as possible and as robust as necessary. The raw XV16 block casting weighs 175 pounds, just 68 more than a machined LS6 V-8 block.

By July, the first castings were ordered from Becker CAD-CAM-CAST in Germany, which can make casting mold cores directly from a computer model. This process, called laser sintering, uses laser light to harden sand in a slurry solution to very quickly form complex core shapes. The sump casting was finished in a week; the block took a month.

There are loads of other tools in GM's computer arsenal that come into play in the development of production engines: one that models airborne oil droplets to aid in the design of oil separator baffles in PCV valves, and myriad others to develop emissions-control systems—an area the XV16 team didn't concern itself with at all.

By October 30, an actual—not virtual—13.6-liter, 32-valve XV16 was assembled and brought to life on a dynamometer. Fully dressed, the all-aluminum engine weighs just 695 pounds—64 pounds less than GM's 8.1-liter production V-8, and its dry-sump lubrication system with eight scavenging pumps makes it compact enough to limbo under the Sixteen's long, ultra-low hood. After a couple weeks of real-world bench tuning, that hapless dyno was groaning under the full onslaught of the 1000 horses and 1000 pound-feet of torque that the computer had prophesied.

The finished engine's performance matched the math model pretty accurately, and if Cadillac decides to build a V-16 like this, Cooper anticipates very little of the basic architecture would change. "We'd upsize the water pump a bit, and we would have to develop the emissions controls, but the basics would probably remain unchanged."

In December, Cooper and company were given about five hours to snuggle the XV16 under that batwing hood and hook up all the controls. It turned few miles in that priceless show car, but it logged some development time under the stretched hood of a Yukon, running with a 45-mph speed governor and a block of wood to limit throttle travel to 45 percent. Confidence in the engine was high, but time to fix anything that broke was nil. "It still managed to smoke the tires in three gears," Cooper says, beaming.

:beer

_________________
„siin on tegemist jällegi ühe toreda riigireetmisega“ (Bretschneider).


Üles
 Profiil  
 
Näita postitusi eelmisest:  Sorteeri  
Tee uus teema Vasta teemale  [ 1 postitus ] 

Kõik kellaajad on UTC + 2 tundi [ DST ]


Kes on foorumil

Kasutajad foorumit lugemas: Registreeritud kasutajaid pole ja 8 külalist


Sa ei saa teha uusi teemasid siin foorumis
Sa ei saa postitustele vastata siin foorumis
Sa ei saa muuta oma postitusi siin foorumis
Sa ei saa kustutada oma postitusi siin foorumis
Sa ei saa postitada siin foorumis manuseid

Otsi...:
cron
Powered by phpBB® Forum Software © phpBB Group
phpbb.ee 3.0.9