BS: how long to design a car

Subject: how long to design a car
From: Charcloth
Date: 28 Jun 02 - 11:14 PM

Johnny Cash sang the song about building a car "one piece at a time" by sneeking out the parts in his lunch box. the song mentions the differences in the year models & it made me wonder, typically, how long does it take the major motor companies to design a car before it is available on the market?

Subject: RE: how long to design a car
From: GUEST
Date: 28 Jun 02 - 11:20 PM

17 minutes

Subject: RE: how long to design a car
From: Charcloth
Date: 28 Jun 02 - 11:27 PM

I wonder which takes longer, the time it takes to design it or the years it takes to pay it off!

Subject: RE: how long to design a car
From: GUEST
Date: 28 Jun 02 - 11:28 PM

the years it takes to pay it off takes longer

On average 32 minutes

Subject: RE: how long to design a car
From: JohnInKansas
Date: 29 Jun 02 - 03:33 AM

If you're looking for a serious answer, the problem is that it's very difficult to pick a specific beginning point for when a design begins and when it's "finished."

The car fanatic magazines fairly often "scoop" the competition with pictures (usually showing nothing meaningful) of new models as much as a year before they appear up on the showroom floor. If you count recall actions and "factory authorized repairs" the design often isn't "finished" until about 3 years after they start selling them.

You may be able to find a copy of the SAE Journal (formerly the Society of Automotive Engineers, but more recently just "SAE International") occasionally on a news stand, which frequently includes "concept car" designs with predicted "future delivery" claims. Most of these are never actually are produced for sale, but you can get a rough idea of how far ahead the stylists are thinking.

A new engine or new suspension geometry that incorporates a significant change may take 3 to 5 years in design and testing, after it has been selected for use in a specific new model - assuming that all of the "concepts" have been proven before you start. Once the design is "final," the engine or axle supplier will need at least a year to solve the production details to be able to get one to the assembly line every few minutes.

If you're just doing a "mix and match" using components that are already "in the pipeline," you might be able to turn out a "new model" with only "cosmetic" changes within as little as 1 or 2 years from "concept."

There are a certain few "radical new concepts" that get announced as "coming out next year" on an annual basis - that have been announced nearly every year for more than the last 30 - and still they haven't shown up in the market because they don't work - - - yet(?). So how long did the design take when/if they do final get there?

John

Subject: RE: how long to design a car
From: Banjer
Date: 29 Jun 02 - 06:04 AM

Anyone having worked on one of the post 1980 or so vehicles will probably tell you there is not much thought given to design anymore! The concept seems to be to take what we had last year and make it worse!

Subject: RE: how long to design a car
From: JohnInKansas
Date: 29 Jun 02 - 08:19 AM

Major model changes may be rolled in every few years, and it does take a couple of years to get from someone's idea to a "firm" concept (lots of committees have to agree). A typical time, once a concept is "committed" would be perhaps 3 years to get to where you have a "list of parts" that's specific enough to start placing orders. Allow 6 months to get enough parts to assemble test vehicles, and a year from the same starting point to place procurement contracts for parts - if you're ready to "commit to production."

If you're lucky, you test for about a year and you're ready to actually start building. (Things usually come out about even - if you're doing the math above - because some critical supplier will always be 6 months late with the first production parts.

A "major rev" design will generally be used for at least 5 years, with "cosmetic changes" to make "new models" so there isn't likely to be anything revolutionary for a while - especially if you buy a new one in the same model line as the one it replaces. A few brand names have carried the same "model base" for as much as 15 years or so.

No one can afford to make an "all new" car. One "new" concept may go on one line, while another "new" concept goes on a different line. The ones that work eventually are "migrated" into all lines.

Watching on a year-to-year basis, it may be difficult to see what is happening, but if you look at all-wheel antilock braking, high-energy ignition, all-passenger air bags, passive restraint belting, stratified charge combustion chambers, computer programmed manifold injection, all-wheel drive, stiffness controlled suspensions, non-$n$o$b$l$e$ catalyst systems, and even GPS navigation - all of which are in a significant number of "current year" vehicles - there has been a lot of change in the last 5 or so years.

At least three major producers will field test fleets of "hybrid" motor/battery vehicles within the next year - and will lose rather large amounts of money on them. At present, this is a "niche" concept, that may someday be ready for general use, but for now it doesn't work well enough for large numbers of people to pay for.

At least four separate and fundamentally different "hydrogen fuel" fleets have been "in the field" for at least 15 years, but if Congress mandated, as has been proposed, that you had to drive one, you'd be very unhappy with the usability of the vehicle.

At least two models from major builders will try out engines based on a new concept called "volume burn direct injection" this year. If you buy one, you won't know it's there - unless your mechanic's answer to everything he can't fix is "it's a new-fangled piece of sh...t" instead of following the manufacturer's service sheets.

Workmanship on individual vehicles? ... well now that's another question.

John

Subject: RE: how long to design a car
From: Charcloth
Date: 29 Jun 02 - 09:21 AM

Thanks Jon, I was kinda guessing about 9 -10 years average. I kinda was basing this using the time laps of first seeing the proto types @ car shows to the time you see their altered product at market. From what you said, it seems my guess may not be far from reality

Subject: RE: how long to design a car
From: Bobert
Date: 29 Jun 02 - 11:46 AM

Well, I'll weigh in on this one since iz not just a hillbilly but a motor head hillbilly. The big three US are always designing and building prototypes most of which never are produced for sale. Now, if you'll remember, GM started the Saturn Division a few years back. The Saturn concept was to kind of start from scratch using GM subcontractors but allowing an indepenent group of designers, engineers, body molders, etc. making Saturn not just another GM model rolling off the same old GM lines but a hybrid. The Saturn program took 3 years. The new Beetle, in spite of being bult on an existing Golf chasis was a 4 to 5 year program. Part of this was marketing and part was engineering and production. Some of the changes in models occur because of safety and emmissions regulations which force the manufacturers to bump up their time tables. One last car that I know about is the Mustang whoes prottype had been in the box for a couple of years when Lee Iacokka pushed Ford to produce it. Had he not been so pursuasive, the Mustang might still be in the box... So, there more to the answer than just thbe engineering and mechanics. Marketing is a major variable.

Bobert

Subject: RE: BS: how long to design a car
From: Grab
Date: 01 Jul 02 - 09:58 AM

I'm writing software now for a new Ford car, a hybrid electric. Pretty major step forward. To take this as an example...

The transaxle (basically gearbox plus 2 electric motors) comes off the Toyota Prius, with some modifications. The engine is a modified 4-cylinder engine using the Atkinson Cycle, which makes it more efficient. The car itself is pretty much a standard Ford Escape, with all the extras bolted on. So the hardware isn't too different, it's the software that's the killer

They started development of this as a prototype 2 years ago. We should have pretty much finished the software by the end of this year. It'll take them 18 months to calibrate it. Fleets will get this in late-2004, and the general public will finally get to buy them in early-2005.

Does that give you an example of the timeframes in really new stuff?

If there's only a minor change to the platform, like using a 3l V6 instead of a 2.8l V6, the software shouldn't have to change, so that saves a lot of time. But there's still all the work to do with calibration to get the engine running right.

Graham.

Subject: RE: BS: how long to design a car
From: Mr Happy
Date: 01 Jul 02 - 10:03 AM

when i first saw this thread, i thought it would be about the dimensions of the vehicle [how long?]picturing extra-stretched limos!

Subject: RE: BS: how long to design a car
From: Ringer
Date: 01 Jul 02 - 12:51 PM

Tell me more about the Atkinson Cycle, please, Grab.

Subject: RE: BS: how long to design a car
From: Grab
Date: 01 Jul 02 - 08:46 PM

I'll see what I can remember - I'm a software engineer, not a mechanic! ;-)

The Atkinson cycle was originally developed with a special crank system - see this link (warning - large animated GIF). But the engine manufacturers have worked out a way of doing roughly the same thing using the valves to set how much air goes in and out. A page about the Toyota Prius drivetrain... and another one about our Ford one. Incidentally, Toyota can give it as many fancy names as they like to try and say they invented it, but in fact they just licensed it off a company called Aisin, the same as Ford have.

The big deal with the Atkinson cycle is that it's very efficient. However, it's got really bad power output at low revs, which means it wouldn't work for normal cars. But low revs on a hybrid-electric is handled by the electric motor, not the engine - the engine doesn't turn on until the car's doing 20-25mph, by which time it can perform just fine. And electric motors put out maximum torque at zero RPM, so that gives it a really good boost. Ford's stated aim for this is 3l V6 performance with a 2.3l engine, and with 35mpg economy instead of sub-20mpg. From latest showings, it looks like they'll get it, too.

Controlling this setup is an absolute nightmare though. The only thing that doesn't have a computer choosing what happens is steering. The brakes have a separate solenoid that can stop the driver's pedal-press engaging the friction brakes to use regenerative braking instead (converting the vehicle's momentum back into electric energy to store in the battery). The powertrain has no gears at all - there's just a 3-way planetary gearbox with the engine and two electric motors on it, so the gear ratio from engine to wheels is entirely controlled by what you do with the electric motors, which basically means it's all software-controlled. The gear lever is just a sensor telling the software what to do, there's no gears physically engaging or anything.

Given how much software there is, and the level of control we have over what the vehicle does, there's always going to be some fun with that. I've personally experienced some really good bugs in this, and I've heard of some even better ones (including an early prototype deciding to engage full forward acceleration when the driver was reversing - new underpants please! ;-). Hopefully none of those bugs make it through to production though! That's why this takes so damn long though - it's so new, and there's so much to get done, and you just can't take the chance on getting it wrong. On something like this, the risks are much more like the risks in an aircraft failing than the risks in a normal car engine going wrong, given that the driver doesn't have the option of braking or even de-clutching if the computer won't let them.

Graham.

Subject: RE: BS: how long to design a car
From: Jeremiah McCaw
Date: 01 Jul 02 - 08:47 PM

Or, as a friend of mine - songwriter/performer Andrew Kerth - observed about the Pontiac Aztec: "Don't you think they should have told the designers they were all working on the same car?"

Subject: RE: BS: how long to design a car
From: DonMeixner
Date: 01 Jul 02 - 09:35 PM

I have a Popular Mechanics from 1947 which lists as concept cars The Jowett Javelin and the Chevrolet Corvair.

Don

Subject: RE: BS: how long to design a car
From: GUEST,ozmacca
Date: 01 Jul 02 - 10:57 PM

How long to design? ... As a disillusioned motorist of may years standing - usually at the side of the road - not long enough if they still have to recall the things for your choice of steering/brake/engine/suspension/etc faults.... We can only hope that some day, they'll build something that's reliable (really reliable), easy/cheap to maintain (back to the fifties when you COULD do a service at home), economical (remember when average 30 mpg in the UK in the 60's was considered good economy - and is STILL quite good after all the recent improvements in technology?), safe, and CHEAP!!! yeah, I know people want slick smooth styling, aircon, comfy seats and power everything, but can't we also have a choice for folks who only want the real essentials like safety and economy?

Subject: RE: BS: how long to design a car
From: Ringer
Date: 02 Jul 02 - 04:39 AM

Thanks, Grab. V interesting. I note from Ford's description of the Atkinson Cycle (your 3rd link) that they leave the inlet valve open late so that the compression stroke pushes some of the charge back into the inlet manifold. Wouldn't the same effect be more efficiently gained by closing it early, rather than late? I'd have thought that with the elimination of reverse-flows in the inlet system you'd get a slight efficiency tweak and also an engine capable of revving higher. But all my i/c-engine experience is 30 years old now (and most of it was on diesels then) so what would I know?

Anyway, I'm grateful for the info you've given.

Subject: RE: BS: how long to design a car
From: Rt Revd Sir jOhn from Hull
Date: 02 Jul 02 - 05:01 AM

dont buy a Lada, they are crap.john

Subject: RE: BS: how long to design a car
From: Rt Revd Sir jOhn from Hull
Date: 02 Jul 02 - 05:06 AM

PS, So are Proton Persona 1.5 GLSi they are crap as well, best car I had was Jaguar Soverign 4.2 Litre (Series 3) I think all foreign cars are crap, I only had 4 foreign cars and they was all crap.

Subject: RE: BS: how long to design a car
From: Rt Revd Sir jOhn from Hull
Date: 02 Jul 02 - 05:18 AM

I just rembered I hadf a forien van as well it was A Simca van (red) french heap of shit, i was going in it from my house to beberley races one day and it blow up and set on fire, so i had to walk there, escort vans are ok I had a few of them but my black one had bad lights and it got nicked, i got it back though and i sold it too the man fron the kebab shop.john

Subject: RE: BS: how long to design a car
From: Grab
Date: 02 Jul 02 - 11:52 AM

Yeah Ringer, it's counter-intuitive. I did once work through it, but I can't remember the exact maths now.

Graham.

Subject: RE: BS: how long to design a car
From: Dave the Gnome
Date: 03 Jul 02 - 11:51 AM

Lada's are OK John. I admit to having 4 of them and they were all fine. Not saying they didn't break down - they did - but when they did they were SOOOO easy to fix. My toolbox comprised of a hammer to knock things off and a roll of Duck tape to stick them back;-)

Seriously though. The 1200 and 1500 (never had a 1600) engines were normaly aspirated analogue ignition models that dated from the 70's. Tons of room to work round them. You could save a lot of fuel (and the odd problems) by putting on a genuine Webber carb and an electronic ignition system but that sort off spoiled the fun...

I've never had motoring that cheap since! Shame they can't pass the emmision tests anymore.

Cheers

DtG